cost of environmental degradation training manual[1]

The Cost of Environmental Degradation (COED)

Methodology

July 1‐5, 2008 Crowne Plaza Hotel Beirut, Lebanon

Training Manual on

For further details contact:

Mutasem El Fadel Professor of Environmental Engineering Faculty of Engineering and Architecture

American University of Beirut Bliss Street, PO Box 11‐0236

Riad El Solh 1107 2020 Beirut, Lebanon

Phone: +961 (0)1 350 000 Ext. 3470 Fax: +961 (0)1 744 462

Mobile: +961 (0)3 228 338 Email: [email protected]

i

ACKNOWLEDGEMENTS

Special thanks are extended to Dr. Dahlia Lotayef, Senior Environmental Specialist at the

World Bank and METAP Coordinator for the Middle East and North Africa Region, and

Ms. Saliha Dobardzic METAP Operations Officer at The World Bank, as well as Mr. Fadi

Doumani, Consultant at the World Bank for their support and assistance during the

preparation for this workshop.

The following references have been quoted directly, adapted or used as a primary

source for major parts of this document. Secondary and indirect references are cited

within the document. While the document provides a good introductory summary for

most related topics, it is by no means a complete resource on the subject. The reader is

highly advised to consult relevant references similar to those cited below for in depth

examinations.

1. Asafu-Adjaye, J. 2005. Environmental Economics for Non-economists. Techniques and Policies for Sustainable Development. 2nd Edition. World Scientific Publishing Co., London.

2. Barton, D.N. The transferability of benefit transfer: contingent valuation of water quality improvements in Costa Rica. Ecological Economics, 42, 147–164, 2002.

3. Chapman, D. 1999. Environmental Economics: Theory, Application, and Policy. Addison Wesley, USA.

4. Garrod G. and Willis K.G. 2001. Economic Valuation of the Environment: Methods and Case Studies. Edward Elgar Publishing, UK.

5. Hodge, I. 1995. Environmental Economics: Individual Incentives and Public Choices. MACMILLAN PRESS LTD.

6. Hussen, A. M. 1999. Principles of Environmental Economics, Ecology, and Public Policy. Routledge, UK.

7. King, D.M. and Mazotta, M. Ecosystem valuation. www.ecosystemvaluation.org

8. McComb, G., Lantz, V., Nash, K., and Rittmaster, R., 2006. International valuation databases: Overview, methods and operational issues. Ecological Economics, 60, 461-472.

9. Ministère de l’Aménagement du Territoire et de l’Environnement, 2002. Plan National d’Actions pour l’Environnement et le Développement Durable (PNAE-DD). République Algérienne Démocratique et Populaire

10. Prüss-Üstün, A., Campbell-Lendrum, D., Corvalán, C., Woodward, A. 2003. Introduction and methods: Assessing the environmental burden of disease at national and local levels. Environmental Burden of Disease Series, No. 1. World Health Organization, Geneva.

ii

11. Sarraf, M., Larsen, B., and Owaygen, M. 2004. Cost of environmental degradation: The case of Lebanon and Tunisia. Environment Department Paper No. 97, The World Bank, Washington D.C.

12. Kuchler, F and Kohler, E. 1998. Assigning Values to Life: Comparing Methods for Valuing Health Risks. Agricultural Economic Report No. 784. Food and Rural Economics Division, Economic Research Service, U.S. Department of Agriculture, Washington, D.C.

13. World Bank, 1998. The Effects of Pollution on Health: The Economic Toll. In Pollution Prevention and Abatement Handbook. The World Bank.

14. World Bank, 2002. Cost Assessment of Environmental Degradation in the Arab Republic of Egypt. Sector Note. Report No. 25175 –EGT. Rural Development, Water and Environment Department, Middle East and North Africa Region, The World Bank.

15. World Bank, 2003. Cost Assessment of Environmental Degradation in the Kingdom of Morocco. Report No. 25992-MOR. Water, Environment, Social and Rural Development Department, Middle East and North Africa Region, The World Bank.

16. World Bank, 2004. Cost Assessment of Environmental Degradation in the Syrian Arab Republic. World Bank. METAP.

17. World Bank, 2007. Economic Assessment of Environmental Degradation due to the July 2006 Hostilities in the Republic of Lebanon. Sector Note. Report No. 39787-LB. Sustainable Development Department, Middle East and North Africa Region, The World Bank.

18. Whittington, D. Improving the performance of contingent valuation studies in developing countries. Environmental and Resources Economics, 22, 323-367, 2002.

19. Whittington, D. 1996. Administering Contingent Valuation Surveys in Developing Countries. Economy and Environment Program for Southeast Asia (EEPSEA).

iii

WORKSHOP OVERVIEW

Day Session Time Topic

1 1 08:30‐10:00 Participants’ registration Official opening 1. Introductions and purpose of the workshop

10:00‐10:30 Coffee break

2 10:30‐12:00 2. Brief overview of basic economic principles 3. Introduction to environmental valuation


3 12:30‐14:00 The revealed preference approach 4. The productivity method (Theory, application, advantages, limitations, case‐studies)

5. The market price method (Theory, application, advantages, limitations, case‐studies)

6. The damage cost, replacement cost, and substitution cost methods(Theory, application, advantages, limitations, case‐studies)

14:00‐15:30 Lunch

4 15:30‐17:30 Case‐studies on the productivity method and the market values approach

2 5 08:30‐10:00 The revealed preference approach (cont’d) 7. The travel cost method (Theory, application, advantages, limitations, case‐studies)


6 10:30‐12:00 The revealed preference approach (cont’d) 8. The hedonic pricing method (Theory, application, advantages, limitations, case‐studies)

9. The averting behavior method (Theory, application, advantages, limitations, case‐studies)


7 12:30‐14:00 Group Exercises: Ayubia National Park In Pakistan (Travel Cost) Non‐Priced Forest Recreation Areas In Malaysia (Travel Cost) Valuing Landscape and Amenity Attributes In Central England (Hedonic Pricing)

14:00‐15:00 Lunch

8 15:30‐17:30 Presentation and discussion of group exercises

iv


3 9 08:30‐10:00 The stated preference approach 10. The contingent valuation method (Theory, application, advantages, limitations, case‐studies)


10 10:30‐12:00 The stated preference approach (cont’d) 11. The discrete choice method (Theory, application, advantages, limitations, case‐studies)

12. The benefit transfer method (Theory, application, advantages, limitations, case‐studies)


11 12:30‐14:00 Group exercise: Stated preference approach Air quality in Beijing Ecosystem services in Ejina China Environmental services in the Yaqui River Delta, Mexico Sustainable development in Sweden coastal zone Coastal ecosystems in Phang Nga Bay, Thailand

14:00‐15:00 Lunch

12 15:30‐17:30 Presentation and discussion of selected group exercises Case‐studies: Coastal zone in North Lebanon, Climate Change MENA Region

14:00‐15:00 Lunch

4 13 08:30‐10:00 13. Cost‐benefit analysis Case‐studies: wastewater and solid waste management


14 10:30‐12:00 14. The value of life and health Including the burden of disease (DALY), the human capital approach, the cost of illness approach, and the contingent valuation approach

Case studies: Drinking water quality; Emissions from the cement industry; Particulate matter in urban areas; Lead phase‐out


5 15 8:30‐10:00 Case studies: Drinking water quality; Emissions from the cement industry; Particulate matter in urban areas; Lead phase‐out


16 10:30‐12:00 Group exercises on the value of life and health: Urban air pollution from particulates in selected MENA countries Water, sanitation and hygiene in selected MENA countries



14:15:30 Lunch

18 12:30‐14:00 Wrap‐up case with various concepts: the July 2006 War in Lebanon 15. Policy implications and workshop conclusion Workshop evaluation

14:00‐15:00 Lunch

The highlighted titles are detailed in the following section.

v

ABBREVIATIONS

BCR = Benefit Cost Ratio BoD = Burden of Disease CAC = Command and Control CBA = Cost Benefit Analysis COED = Cost of Environmental Degradation COI = Cost of Illness CS = Consumer Surplus CVM = Contingent Valuation Method DALYs = Disability Adjusted Life Years Dh = Dirham DRRs = Dose‐Response Relations EA = Environment Agency EIA = Environmental Impact Assessment GBD = Global Burden of Disease GDP = Gross Domestic Product GIS = Geographic Information System HCA = Human Capital Approach HPM = Hedonic Price Method IRR = Internal Rate of Return MC = Marginal Cost MBI = Market‐based instruments MENA = Middle East and North Africa METAP = Middle East Technical Assistance Program MR = Marginal Revenue NPV = Net Present Value NSB = Net Social Benefit O&M = Operation and Maintenance OC = Opportunity Cost RAD = Restricted Activity Days SCBA = Social cost benefit analysis TCE = Trichloroethylene TCM = Travel Cost Model UK = United Kingdom US = United States USD = United States Dollar VOSL = Value of Statistical Life W Korean Won WHO = World Health Organization WLD = Work Loss Days WTA = Willingness to Accept WTP = Willingness to Pay WWTP = Wastewater Treatment Plant

vi

[This page was left blank intentionally]

INTRODUCTIONS AND PURPOSE OF THE WORKSHOP

Session 1

Region

al Training Worksho

p on

: Th

e Co

st of E

nviron

men

tal D

egrada

tion

Metho

dology

1

SESSION 1

1 INTRODUCTION AND PURPOSE OF THE WORKSHOP

Environmental degradation has become one of the most prominent adverse phenomena

in today’s world. The scope of environmental problems has grown substantially in the

past decade and will continue to expand and diversify more in the future; no generation

has ever faced a more daunting agenda. The world today confronts a multitude of

environmental problems, more than ever before, over a wider range of spatial and

temporal scales, and requiring various skills for proper control. Within this context, the

Middle East and North Africa (MENA) region is no exception in suffering from serious

environmental problems and natural resource degradation. Environmental pollution is

evident throughout the region which is exhibiting various types of degradation whether

water (coastal and inland surface, and ground), soil, and air (indoor and outdoor).

According to the Middle East Technical Assistance Program (METAP)/World Bank

Country studies1, the cost of environmental degradation in seven countries (Lebanon,

Syria, Jordan, Egypt, Tunisia, Algeria, and Morocco) ranges between US$228 million per

year in Jordan and US$4.2 billion per year in Egypt. Figure 1 illustrates the distribution of

the average estimated annual damage costs of environmental degradation in countries

in the MENA region by country and category in percent of Gross Domestic Product

(GDP).

Figure 1. Average annual damage costs of environmental degradation from studies in MNA countries

(percentage of GDP)

1 METAP website: www.metap.org

2

According to Figure 1, the cost of environmental degradation in MENA countries

constitutes between 2 and 5 percent of the country’s GDP, as compared to 1‐2% of GDP

in OECD countries, 4.5% of GDP in 1991 in India, 3.3% of GDP in Mexico, and 8% of GDP

in China. Yet, it is important to note that these results are underestimates because of

data limitations. As such, they do not include damage stemming from several potential

contributors such as untreated industrial, hazardous, and hospital wastes or losses of

forest cover and biodiversity. Also owing to data constraints, the impact of inadequate

treatment of industrial and municipal wastewater is limited to coastal recreational and

tourism losses.

The Cost of Environmental Degradation (COED) methodology in the MENA region is an

environmental economics tool developed by the METAP/World Bank. This tool enables

key professionals to carry out assessments of the economic cost of environmental

degradation. It has been successfully used in the valuation of environmental degradation

on a macroeconomic and sector levels, in terms of giving an estimate of GDP lost to

environmental degradation. The importance of economic valuation of environmental

degradation is that it allows the quantification of benefits of environmental projects/

policies, thus fostering the incorporation and prioritization of environmental issues in

decision‐making. It can be a powerful means for raising awareness about environmental

issues and fostering progress towards sustainable development. However, a main

obstacle to conducting policy‐relevant and timely studies in environmental economics is

the shortage of human capacity at governmental ministries/organizations as well as

academic institutions.

The current project, which is funded by the World Bank/ METAP and implemented by

the American University of Beirut, aims at enhancing regional capacity in environmental

economics. This document was prepared as part of the training course in COED

methodology. It aims to provide participants a comprehensive and easy to use reference

on this subject and follows the same sequence of the course syllabus.

BRIEF OVERVIEW OF BASIC ECONOMIC PRINCIPLES

& INTRODUCTION TO ENVIRONMENTAL

VALUATION

Session 2

Region

al Training Worksho

p on

: Th

e Co

st of E

nviron

men

tal D

egrada

tion

Metho

dology

3

SESSION 2

2 BRIEF OVERVIEW OF BASIC OF ECONOMIC PRINCIPLES

Activities of economic agents contribute to the generation of pollution. The operation of

the market system is intimately related to the nature and amount of pollution

generated. The purpose of this session is to introduce the participants to the basic

concepts of economic analysis, in terms of how markets work and why markets fail.

2.1 The competitive market

A market can be defined as the coming together of consumers (or buyers) and producers

(or sellers) to exchange goods and services for money. The buyers and sellers do not

have to be physically present to carry out transactions. Usually a market exists for a

single good or service. Markets may be classified according to the numbers of sellers and

buyers. In a perfectly competitive market, there are many buyers and sellers. A

monopoly is a market in which there is a single seller, such as the utilities sector in many

countries, where the government is the sole provider of electricity and water. An

oligopoly is an intermediate case in which there are few sellers. The Australian car

manufacturing market is an example of an oligopoly because there are four main sellers.

A monopsony is a market in which there is a single buyer. For example, a small town

with a single major industry that is the sole buyer of labor.

The competitive market has the following characteristics:

− There are many buyers and sellers and none of them are influential enough to affect

the market price or output

− The buyers and sellers are free to enter and leave the market in response to price

changes

− The goods and services being offered for sale are identical or homogeneous. This

implies that buyers do not care from whom they buy, provided prices are identical

− All the participants in the market have perfect knowledge. That is, consumers know

product prices and producers know input prices.

2.1.1 Consumer behavior and demand

The demand function is a curve that indicates how much of a good a consumer will buy

at various prices (Figure 2). Note the inverse relationship between price and quantity

demanded. This is referred to as the Law of Demand. That is, given income, preferences

and prices of alternative goods, an individual will be willing to purchase decreasing

amounts of a given good (or service) as its price increases.

4

− The points on the curve represent the maximum amount an individual is willing to pay for different quantities of q1

− The individual's demand for good q1 is defined given that all other goods and income remain constant

− The demand curve is defined for a given period of time. Thus, the demand curve in a different period of time will have a different shape and position

Figure 2. Individual demand function for a good q1

2.1.2 The concept of elasticity

The term 'elasticity' refers to the responsiveness of the quantity demanded (or supplied)

to changes in other variables (e.g., price and income). The concept of elasticity is

important because a key factor in the functioning of the economic system is the reaction

of economic agents to price incentives. Own price elasticity of demand is the ratio of the

change in quantity demanded of a given good to the change in its own price. That is:

εD = (percent change in quantity of q1 demanded)/ (percent change in price of q1)

Depending on the magnitude of the elasticity parameter, own‐price elasticity of demand

can be perfectly elastic, relatively elastic, relatively inelastic, or perfectly inelastic, as

illustrated in Table 1.

5

Table 1. Forms of elasticities

Type Graph DescriptionPerfectly elastic | εD | = ∞

A small increase in the price of the good causes the quantity demanded to fall to zero. In practice, no good has perfect price elasticity.

Relatively elastic | εD |>1

A small change in the price of the good causes a relatively large change in quantity demanded. In general, most luxury goods tend to be relatively price elastic.

Relatively inelastic| εD |<1

In this case, a change in the price of the good causes little change in quantity demanded. Necessities such as food and utilities (e.g., water and energy) tend to be relatively price inelastic.

Perfectly inelastic| εD |=0

A change in the price of the good does not lead to a change in quantity demanded

Cross‐price elasticity of demand refers to the responsiveness of the quantity of a

demanded good (q1) as a result of changes in another good (q2).

εD 12 = (percent change in quantity of q1 demanded)/ (percent change in price of p2)

1. εD12>0, implies q1 and q2 are substitutes. That is, an increase in the price of one good

causes consumers to switch to the other, resulting in an increase in the quantity

demanded of the second good. Examples of substitute goods are sugar and

NutraSweet, bus and rail transportation, etc.

2. εD12<0, implies q1 and q2 are complements. Complementary goods are consumed

together and therefore an increase in the price of one good leads to a reduction in

its consumption, and hence a reduction in demand.

Income elasticity of demand refers to the responsiveness of the quantity of a demanded

good (q1) as a result of changes in another good (q2)

6

ηγ = (percent change in quantity of q1 demanded)/ (percent change in income)

1. ηγ > 0, implies the good is a normal good. Most goods are normal goods because an

increase in income leads to an increase in quantity demanded.

2. ηγ < 0, implies the good is an inferior good. That is, an increase in γ leads to a

decrease in q1. There are not many practical examples of inferior goods. However, a

low‐ income family that currently consumes dried vegetables might reduce their

consumption and switch to fresh vegetables in response to an increase in household

income.

2.1.3 Producer behavior and supply

The production function of a good q is a function of various inputs, including labor, land

and capital that are used in producing good q. The producer's aim is to maximize profit

subject to the constraint of the above production function. Given the profit motive, the

producer will increase the output of q if its price rises so as to increase profits. The

production function (Figure 3) is positively sloped because producers are willing to

supply more as price increases. In addition, the curve refers to a given point in time.

− The supply curve is also the marginal cost (MC) curve. That is, it indicates the cost of producing each additional unit of the good. In order to maximize profits, the producer will increase production up to the point where marginal revenue (MR), the price per unit of output in a competition market, just equal marginal cost.

Figure 3. Market supply function

2.1.4 Market Equilibrium in the Competitive Market

The interaction of supply and demand forces in the market determines the equilibrium

or market clearing price, and the equilibrium quantity demanded. The equilibrium price,

in turn, determines the price for each unit of output, that is, marginal revenue. In Figure

4, market equilibrium is achieved at point E. Thus, at price PE, the market demand is

exactly equal to the quantity the market is willing to supply (SE).

7

Figure 4. Equilibrium in the market for a product Q1

Now suppose the market price per unit of Q1 increased to P’, the producer will supply S’,

while the consumer will demand only D’, which will result in demand deficit (Figure 5).

To clear the deficit, the producer will decrease the price and the consumer will increase

the purchase. The producer will decrease the quantities supplied until equilibrium is

reached and demand equals supply. On the other hand, if market price decreases to PA,

the consumer demand increases to DA. However, the producer is not willing to supply at

this price, causing a shortage in the market. This will put upward pressure on prices and

the producer will respond by increasing supply. As the price goes up, consumers will

reduce their purchases until the price decreases, at which point, quantity demanded will

be equal to quantity supplied (Figure 5).

Figure 5. Shift in market equilibrium for product Q1

Factors that can shift the demand curve include income, prices of substitutes or

complements, and consumer tastes and preferences. An income increase causes an

upward or rightward shift in market equilibrium, while an income decrease causes a

downward or leftward shift. For example, an increase in per capita income in a given

population (with all other factors constant) will shift the demand curve for mobile

telephones upwards. However, due to the excess demand, the price of mobile

telephones will rise to re‐establish equilibrium.

8

Decrease in income Increase in income

Figure 6. Shift in market equilibrium with change in income

A decrease in the price of substitutes for a good will cause the demand curve for the

good to shift downward, while a decrease in the price of complements for the good will

cause the demand curve to shift upward. For example, a decrease in the price of oil will

cause a downward shift in the demand for natural gas. The quantity of natural gas falls

and the price also falls to re‐establish equilibrium.

Factors that can cause the supply curve to shift include, price of inputs, taxes, subsidies,

improvements in technology and weather for agricultural products. A decrease in the

price of inputs for making good q will make the supply curve for q to shift outwards,

whereas an increase in the price of inputs will cause an inward shift. Improvement in

technology will cause an outward shift because more output can be produced with the

same level of inputs. For agriculture and other forms of production that are weather

dependent, deterioration in weather conditions will cause a leftward shift in the supply

curve.

Decrease in input price Increase in input price

Figure 7. Shift in market equilibrium with change in inputs

The examples of equilibrium illustrated above are highly simplified in order to convey

the basic concepts. In real life, equilibrium does not tend to be static. The demand curve

9

is constantly shifting due to changes in tastes and incomes, while the supply curve also

shift in response to resource constraints and technological advances. One important

assumption underlying the market equilibrium analysis above is that property rights are

well defined. That is, the seller owns the rights to the good or service and can therefore

appropriate any benefits from the sale.

2.1.5 Consumer and producer surplus

Consumer and producer surplus are illustrated in Figure 8. Consumer surplus is the

maximum amount of money consumers are willing to pay for the good or service minus

the market price. It is a measure of net benefits or welfare (Δ abc). The sole reliance on only the market price could result in an under estimation of benefits. Producer surplus is

the net benefits received by the producer and is given by the difference between the

market price and marginal cost (Δ bcd).

Figure 8. Consumer and producer surplus

2.1.6 Application of the Competitive Model‐ The socially optimal level of forestry

Clear felling of timber has several undesirable effects on society, including loss of forest

cover and associated problems such as increased soil erosion, loss of soil nutrients, loss

of biodiversity, etc. Assume that the stumpage price is currently p. At this price, q logs

will be harvested (Figure 9). Now suppose that the government charges an extra $5 per

log to cover the environment damage. This policy will result in an upward shift in the

supply curve from S to S’ by a vertical distance of $5 (Figure 9). assuming the demand for

logs remains constant during the period of the analysis, the quantity of logs harvested

will decline and equilibrium will be re‐established at q’, the socially optimal level.

10

Figure 9. Socially optimal level of forestry

2.2 Market Failure

Market failure occurs when some costs and/or benefits are not fully reflected in market

prices. The market system fails to function properly for many kinds of environmental

goods because such resources, including the services they provide, are often not traded

in markets. There are many reasons for market failure including:

− Property rights related to ecosystems and their services are often not clearly defined

− Many ecosystems provide services that are public goods

− Many ecosystem services are affected by externalities

− Type of market structure

2.2.1 Property rights

When property rights are not well defined or absent in the economic system, there is no

incentive for an individual to invest in an asset because they cannot appropriate the full

benefits. When one purchases an asset, it comes with a set of well‐defined ownership

rights and responsibilities. These have the following general characteristics:

− Well defined: in the form of registration certificate of purchase receipt. In some

cases the entitlement may be informal and may have been institutionalized by social

or cultural norms.

− Exclusive: the buyer is the only one who has the right to use the asset, although he

may choose who else may use it and under what conditions. It is important to note

that restrictions accompany ownership rights.

− Transferable: the buyer may transfer property rights of his asset to another

individual either permanently by selling it or temporarily by renting it.

− Secure and enforceable: the property rights to an asset are secure and enforceable.

Effective enforcement involves the effective detection and apprehension of violators

and application of an appropriate penalty. To be effective, the penalty must exceed

the actual or anticipated benefits of violation.

11

Property rights regimes

In practice, there are different types of property rights regimes. The complete set of

property rights described above is at one end of the spectrum and is typical of private

goods. At the other end of the spectrum, there are public goods, with congestion goods

in between these two. Public goods can be classified into pure public goods, semi public

goods, and open access or common property goods (Figure 10).

Figure 10. Taxonomy of environmental goods

Most environmental goods fall under the category of pure public goods or open

access/common property goods. In such cases lack of well‐defined property rights

results in market failure. A consequence of market failure is inefficient allocation of

resources (e.g., excessive pollution). For example, a farmer has the right to prevent

someone from polluting his or her land, but cannot prevent anyone from polluting a

nearby river. Furthermore, he or she may have no legal right to receive compensation

from the upstream polluters. The upstream polluters, who do not bear costs of their

activities, have no economic incentives to limit the amount of pollution especially when

they know that the farmer has no property rights. This type of market failure has led to

calls for governments to intervene in the market.

Private vs. public goods

As mentioned earlier, a private good has characteristics such as exclusivity,

transferability, security, and enforceability. In addition, a private good has a positive

marginal cost. That is, the cost of supplying one additional unit is above zero. A private

good is rival in consumption. That is, once someone consumes the good, another person

cannot consume it. On the other hand, a pure public good is non‐exclusive and non‐rival

in consumption, and has zero marginal cost. 'None exclusive' and 'non‐rival' mean the

Goods and

Services

Private goods Congestion goods Public goods

Open access & common property

goods

Pure public goodsSemi-public goods

12

good is available to everyone and that one person's consumption of the good does not

reduce the amount available to others. 'Zero‐marginal costs' means the cost of supplying

an additional unit of the good to any particular individual is zero. Examples of pure

public goods are national defense, biodiversity, clean air, flood protection, noise and

visual amenities. A distinctive characteristic of a pure public good is that consumers do

not have the option of not consuming. As suggested earlier, a pure public good will be

under‐provided because the owner cannot appropriate the full benefits.

Open access goods are rival in consumption, non‐exclusive, non‐transferable, and often

non‐enforceable even when ownership rights exist. Examples of open goods are ocean

fisheries and migratory wildlife. Common property goods (e.g., common grazing land)

are rival in consumption and are exclusive for a group of people (e.g., a local

community). Rights of use may be transferable by individuals or the group. There may

not be legal or formal title to ownership but the group may be able to enforce their

ownership rights by means of social sanctions. Under open access or common property

rights regimes, the resource will be overexploited. However, under some form of

common property systems, resource management is likely to be more efficient because

it is based on communal rules and customs.

Semi‐public (or quasi public) goods are non‐rival in consumption, have a zero marginal

cost of provision and are non‐exclusive although ownership rights exist. An important

distinction of semi‐public goods is that even though the owner or the providers of the

service cannot exclude others form consumption, consumers can choose not to

consume. Examples of this category of goods include radio and TV broadcasts and

lighthouse.

Congestion goods are exclusive and can be either non‐rival or rival in consumption. Such

goods do not fall neatly into any of these categories and may exhibit characteristics of

private goods or public goods at different levels of consumption. An example of this type

of good is a campsite, roads, bridges, an art gallery, fishing and boating sites, and a

historic site.

2.2.2 Externalities

As the word suggests, an externality is an effect that is 'external' to the causing agent.

That is, the person causes an effect that impact on other people. An externality is said to

exist when some agent A (individual or firm) takes an action which has an impact on

another agent B, that B has not chosen to accept. In an externality, agent B cannot

13

choose the level of the impact like in a normal economic transaction and the impact on

B is not a deliberate attempt from A. An externality is often negative. This occurs when

the affected person suffers a loss in utility that is uncompensated. Examples of negative

externalities are air, water and noise pollution. A positive externality (external economy

or external benefit) occurs when the effect is beneficial to the affected person. An

example of a positive externality is immunization. Another example of a positive

externality is where one firm's technological breakthrough benefits other firms in the

industry who have not contributed to the research costs.

The following factors give rise to externalities:

1. Interdependence between economic agents: the market system fails to 'price' this

interdependence, as a result of which the affected party is uncompensated.

2. Lack of or weak property rights: due to lack of or weak property rights, the affected

party is unable to demand that the externality be reduced or ask for compensation.

3. High transactions costs: the cost of negotiating, implementing and enforcing and

agreement between the parties may be high.

Once the affected agent is compensated for his or her loss of welfare, the externality is

said to be 'internalized', and society is better off by the gainer compensating the loser.

Types of externalities

Externalities can be classified into relevant externalities, pareto‐relevant externalities,

static or dynamic externalities, and pecuniary externalities. A relevant externality is

when the affected person is made worse off by the activity and wants the offending

person to reduce the level of the activity. A pareto‐relevant externality exists whenever

its removal results in a pareto improvement. A pareto improvement is a situation where

it is possible to take action such that the affected person is made better off without

making the offending person worse off. A dynamic externality exists when the

externality has adverse impacts for the future. To illustrate static and dynamic

externalities, take the example of two fishers who are operating under an open access

or property rights regime. A static externality is when one creates an externality for the

other by overfishing. However, the externality can become dynamic if the offending

party is harvesting fish that may have some future value. This could happen, for

example, if the offender is harvesting juvenile fish species. In this case, the opportunity

cost of the fish reflects a forgone value in the sense that there will be adverse impacts

for the future. A pecuniary externality is a form of externality that is transmitted through

the price system. An externality is usually an 'unpriced' effect. However, a pecuniary

14

externality occurs when the externality is transmitted through higher prices or reduced

costs. An example is increased rental prices in an area due to a new business opening

there. Pollution is not a pecuniary externality because the effect is not transmitted

through higher prices. Even if penalties exist, they do not reflect the amount of damage

inflicted on the environment.

2.2.3 Type of Market Structure

The type of market structure or organization can also cause market failure. We consider

the following two cases: a perfectly competitive market with external costs and a

monopoly.

Resource allocation in a perfectly competitive market

Consider a gold mining company that dumps mine tailings into a nearby river without

paying for clean up or treating the waste. In this case, production at the mine includes

the production of gold as well as pollution. Or to put it differently, the river water is an

unpriced input in the gold production process. Let us define the following variables:

D = demand curve for gold; MCp = marginal private cost of producing gold (i.e., the

firm's supply curve); MSC= marginal social cost (Figure 11).

Figure 11. Resource allocation in a competitive market with externalities

We assume that MSC is greater than MCp at any level of output because society

considers both the costs of pollution as well as gold production, but the company

considers only its marginal private cost. The marginal social cost of gold production is

therefore given by the marginal external cost, the cost of disutility caused by the

externality, plus the marginal private cost. That is:

C=MCp+ MSC

15

Under a perfectly competitive market structure, the company maximizes its producer

surplus by producing q0 units of gold. However, from society’s point of view, q0 is not

the efficient allocation. Society's net benefits could be maximized by producing less gold,

that is, q’ units. The triangular area Δabc represents a deadweight loss to society. Note the following observations about Figure 11:

1. The socially optimal level of pollution (which is assumed to be proportional to gold

production) is not zero. This implies that it may not be socially optimal to have zero

pollution.

2. In a perfectly competitive market where pollution is unpriced (i.e., there is no

pollution abatement), production results in more output that is socially desirable,

resulting in excessive pollution.

3. If pollution abatement is enforced, the company will reduce pollution but raise the

price per unit of output, resulting in reduced output of gold. However, in this case,

the reduced output is the socially efficient level and the higher price is also the

efficient price.

Resource allocation in a monopoly

Assuming a perfectly competitive market and a system of private property rights, the

price mechanism will combine to result in an efficient allocation of resources. However,

the presence of monopoly rights causes market failure or inefficient allocation of

resources from society's point of view. Take the case of a single monopolistic firm with a

marginal cost curve, MC, facing a market demand curve, D (Figure 12). Under perfect

competition, q units of the good will be supplied by setting the price = marginal revenue

(MR) = marginal cost (MC). Note, however, that in the case of a monopoly, the demand

curve is above the marginal revenue curve and therefore price is not equal to marginal

revenue.

Figure 12. Resource allocation in a monopoly

16

Monopoly profit is maximized by setting MR equal to MC. This results in less output, qm,

and a higher price, pm. Consumer surplus under a monopoly is apmb, which is less than

consumer surplus under perfect competition, Δ ap’c. Recall that the demand curve (D)

represents the marginal benefit of goods to consumers. Figure 12 indicates that for a

monopolist, marginal benefit exceeds marginal cost and therefore the level of output

(qm) is inefficient. Consequently, there is dead weight loss to society represented by

triangle Δbdc.

The monopolist's production decision may be somewhat unexpected because it seems

to suggest that less pollution will be created in a monopoly than in perfect competition.

Furthermore, the monopolist's higher initial price (pm) suggests that, given a fixed stock

of natural capital, the price path will be less steep over time than in perfect competition.

However, caution must be exercised in making such comparisons because other factors

such as elasticity of demand affect the outcome.

2.3 Policy failure

Policy or government failure occurs when the government creates incentives for the

prices of certain goods and services to be lower than the actual cost of production per

unit. An example of policy failure is a government subsidy on pesticides which provides

incentives for farmers to use more pesticide than is socially efficient, resulting in adverse

environmental impacts. Other types of subsidies include guaranteed prices for

agricultural products and subsidies which tend to encourage large scale production and

loss of forest cover. In general, subsidies in the developing countries are on the decline

as most of them have adopted structural adjustment programs over the past two

decades.

2.3.1 Solutions to environmental pollution problems

Two main approaches have been proposed for dealing with externality problems. The

first approach, known as the property rights or market solution, was proposed by Ronald

Coase and involves allowing the free market system to solve the problem through

bargaining between the affected parties. However, it is based on assumptions that may

not apply in the real world, including, zero transaction costs, well‐defined property

rights, perfect competition, and no free‐rider effect. The second approach is by means of

government intervention. There is always a need for government interventions to

correct externality problems. Government pollution control policies can take two main

forms: market based instruments (MBI) and command‐and control (CAC) instruments

(Figure 13).

17

Figure 13. Government pollution control policies

Command and control instruments are the oldest form of pollution control policies. They

require setting the standard and monitoring and enforcing it. They have the advantage

of being a widely understood form of policy and being more pragmatic and socially

acceptable than MBIs. On the other hand, such policies provide no incentive for

pollution reduction beyond standards. In addition, penalties tend to be too low and

enforcement too weak. Governments must know the marginal social cost and marginal

social benefits curves to set an optimal penalty and penalties need to be revised

frequently, which is costly. Furthermore, financial costs for setting and enforcing

standards are high; political costs may arise if standards are stringent; and standards are

uniformly set to all firms and regions.

As for market‐based instruments, they use price or some other economic variables to

provide incentive for economic agents to abate pollution. They have the advantage of

achieving the same objective as CACs at a lower cost, and of generating significant

revenue for the government. However, they cannot be applied where the institutional

framework is weak.

When choosing the right instrument for pollution abatement and control, various

criteria need to be considered including, economic efficiency, effectiveness in achieving

the desired environmental objective, adaptability to changing circumstances, equity in

the distribution of costs and benefits among different groups in the society, and political

acceptability.

18

3 INTRODUCTION TO ENVIRONMENTAL VALUATION

The proper valuation of non‐market environmental commodities has significant policy

implications. In the past, such commodities have been assigned zero to low values due

to difficulties involved in assigning economic values. Failure to properly account for the

values of some environmental resources has resulted in decisions that have had

negative implications for the environment and society.

An environment resource has a range of values that need to be accounted for. These

values can be categorized into use‐values and non‐use values (Figure 14).

Figure 14. The elements of total economic value

Some of the environmental functions are used directly, either contributing towards the

production of marketed outputs or else contributing to consumption directly. For

example, agricultural land provides the medium for the production of crops and timber.

The environment may also be used directly for consumption purposes, for recreation or

as landscape value. The third category of use values is the ecological functions of the

environment, such as flood control, waste assimilation, or carbon storage. Alternatively,

non‐use or intrinsic values are inherent in the good. The satisfaction we derive from the

good is not related to its consumption per se. Non‐use values comprise existence value,

bequest value, and option value. Existence value arises from the benefit that an

individual derives from knowing that a resource exists or will continue to exist regardless

of the fact that he has never seen or used this resource, or intends to see or use it in the

future. An example is the international outcry over the whaling issue. Bequest values

arise from the benefit that individuals derive from knowing that a resource will be

19

available for their children and children’s children. Option value is a little more complex.

It represents the value of potential uses. An individual is prepared to pay now, to retain

options for future uses.

Use values can be readily measured by market prices or other means and are well‐

accounted for in decision‐making processes. However, non‐use values are problematic

because they can constitute a significant component of total economic value, and yet

they are not traded and therefore cannot be valued by market prices. For this purpose,

non‐market valuation techniques have been developed.

3.1 Non‐market valuation techniques

Non‐market valuation methods can be broadly classified into two categories, revealed

preference models and stated preference models (Figure 15).

Figure 15. Non‐market valuation methods

Revealed preference models make use of individual behavior in real and simulated

markets to infer the value of an environmental good or service. For example, wilderness

is valuated from the cost incurred to travel to the area for recreation. Revealed

preference models measure use values only. The choices made are real rather than

hypothetical. The revealed preference models are based on a clear principle but

complicated applications. Examples of revealed preference methods include the travel

cost method, the hedonic pricing methods, the averting behavior method, and the

market values method.

Stated preference models attempt to elicit environmental values directly from

respondents using survey techniques such as questionnaires. They are flexible and

Preferences

Revealed preferences

Stated Preferences

Market Values

Travel Cost Methods

HedonicMarkets

Averting Behavior

Contingent Valuation

Choice Experiments

20

applied to a wide range of goods and they measure both use and non‐use values.

However, these models are subject to many biases. The following chapters discuss in

detail the various non‐market valuation methods.

THE PRODUCTIVITY METHOD &

THE MARKET PRICE METHOD &

THE DAMAGE COST, SUBSTITUTION COST, AND REPLACEMENT COST

Sessions 3 & 4

Region

al Training Worksho

p on

: Th

e Co

st of E

nviron

men

tal D

egrada

tion

Metho

dology

21

SESSIONS 3 & 4

4 THE PRODUCTIVITY METHOD

4.1 Theory

The production function method is one of the most widely used valuation techniques. It

focuses on environmental resources as an input to the production of goods and services.

It is used to estimate the economic value of ecosystem products or services that

contribute to the production of commercially marketed goods. Thus, if a natural

resource is a factor of production, then changes in the quantity or quality of the

resource will result in changes in production costs, and/or productivity of other inputs.

This may affect the price and/or quantity supplied of the final good. It may also affect

the economic returns to other inputs. For example, agricultural production is a function

of soil (S) and other inputs (x). As soil quality declines from S1 to S2 due to soil erosion,

the production function shifts to Q2 (Figure 16). Accordingly, the farmer has the option

of producing at Q2 or to keep production at Q1 by increasing other inputs from X1 to X2.

Figure 16. The production function curve

Two types of benefits (or costs) may be important (Figure 17):

− Changes in the quality or price to consumers of the final good will result in changes

in consumer surplus

− Changes in productivity or production cost changes will result in changes in producer

surplus

22

Figure 17. Market supply and demand functions

Thus, the economic benefits from improvements in the resource can be estimated using

changes in observable data.

The production‐function method is most easily applied in two specific cases:

− Cases where the resource in question is a perfect substitute for other inputs: For

example, improved water quality in a reservoir means that less chlorine is needed

for treating the water. An increase in quantity or quality of the resource will result in

decreased costs for the other inputs. The benefits of improved water quality can be

directly measured by the decreased chlorination costs.

− Cases where only producers of the final good benefit from changes in quantity or

quality of the resource and consumers are not affected: For example, improved

quality of irrigation water may lead to greater agricultural productivity. If the market

price of the crops to consumers does not change, benefits can be estimated from

changes in producer surplus resulting from increased income from the other inputs.

The profits per acre will increase, and this increase can be used to estimate the

benefits of improved irrigation water quality

Selected applications of the production function method are outlined in Figure 18.

Pressure Environmental

Impact Productivity Impact Change in Income

Overgrazing Soil erosion Reduced capacity of soil to sustain crops

Reduced farmers income

Wastewater discharge

Polluted river Reduced capacity to sustain fish stocks

Reduced income of fishermen

Increased vehicle use

Air pollution Increased respiratory problems among workers

Lost workdays

Uncontrolled irrigation

Salinity of cropland

Declining yields Reduced farmers income

Figure 18. Selected applications of the productivity method

23

4.2 Applying the Productivity Method

Steps to be followed when applying this method include:

1. Determine the physical impact solely arising from the driving force or behaviour

under study. Note that it is sometimes difficult to differentiate impacts due to a

series of complex biological interrelationships.

2. Collect data on how changes in the quantity/ quality of the natural resource affect

costs of production for the final good, supply and demand for the final good, and

supply and demand for other factors of production. Sources of data include

experimental data using field trials and statistical data using cross‐section or time‐

series data. Experimental data are usually difficult to extrapolate, while statistical

data are available for short time horizons and are difficult to control for other

factors.

3. Link the impact of changes in the quantity/ quality of the resource to changes in

consumer surplus and/or producer surplus. Problems encountered with this step

include:

− Distorted prices due to government interventions

− Change under study is not large enough to impact market price

− Change in market price is too large

− Change in production alters costs 4. Estimate the economic benefits

The productivity method has the advantage of being a straightforward methodology

that is inexpensive to apply due to limited data requirements and ready availability of

relevant data. However, this methods does not account for non‐use values, hence it

provides only the lower bound estimate. Furthermore, it is limited to valuing resources

that can be used as inputs in production of marketed goods. It also requires information

on the scientific relationships between actions to improve quality or quantity of the

resource and the actual outcomes of those actions. Finally, if the changes in the natural

resource affect the market price of the final good, or the prices of any other production

inputs, the method becomes much more complicated and difficult to apply.

4.3 Illustration 1‐ Polluted municipal drinking water reservoir

A municipal drinking water reservoir is polluted by agricultural runoff. The economic

benefits of implementing measures that eliminate the runoff need to be determined.

Accordingly, the productivity method was selected because environmental quality

directly affects the cost of producing municipal drinking water. In addition, cleaner

water is a direct substitute for other production inputs, such as water treatment

24

chemicals and filtration. Thus, the benefits of improved water quality can be easily

related to reduced purification costs.

To apply the production‐function method, first, specify the production function for

treated drinking water. Inputs include water of a particular quality from the reservoir,

chemicals, and filtration, while outputs include pure drinking water. Second, estimate

how the cost of treatment changes when reservoir water quality changes, using the

production function estimated in the first step. Calculate the quantities of treatment

chemicals and filters needed for different levels of reservoir water quality and multiply

these quantities by their costs. Third, estimate the economic benefits of protecting the

reservoir from runoff, in terms of reduced purification costs. If all runoff is eliminated,

the reservoir water will need very little treatment and the purification costs for drinking

water will be minimal. Compare the outcome to the cost of treating water where runoff

is not controlled. The difference in treatment costs is an estimate of the benefits of

eliminating runoff. The benefits for different levels of runoff reduction can also be

estimated. This requires information about the projected success of actions to reduce

runoff, in terms of the decrease in runoff and the resulting changes in reservoir water

quality.

4.4 Illustration 2‐ Values of Wetlands in the Peconic Estuary, Long Island

The Peconic estuary encompasses productive wetlands of different types, including

eelgrass, salt marsh, and intertidal mudflats. Development and resulting water quality

degradation have reduced the quantity of these wetlands. Various management actions

for the estuary and surrounding land areas need to be considered and assessed using a

productivity study for wetlands.

The study focused on valuing marginal changes in acres of wetlands, in terms of their

contribution to the production of crabs, scallops, clams, birds, and waterfowl, assuming

that wetlands provide both food chain and habitat support for these species. The

productivity of different wetlands types in terms of food chain production was

estimated and linked to production of the different species of fish. The expected yields

of fish and birds per acre of habitat were valued using commercial values for fish,

viewing values for birds, and hunting values for waterfowl.

The study results estimated that:

− An acre of eelgrass is worth $1,065 per year

− An acre of salt marsh is worth $338 per year

− An acre of intertidal mudflat is worth $68 per year, in terms of increased productivity

25

of crabs, scallops, clams, birds, and waterfowl

Based on the results the economic value for productivity services of preserving or

restoring wetlands in the estuary can be calculated. Note that these values are an

understatement of the total economic value for the wetlands, as they only address

values in production of commercially and recreationally valuable species. They overlook

other services, such as erosion and storm protection or aesthetics.

4.5 Case‐study: Degraded agricultural land & rangeland in Morocco

About 93% of Morocco is arid. Fragile soils suffer from water and wind erosion.

Furthermore, overexploitation and unsustainable management is resulting in arable land

loss, decrease in crop yield, silting of dams, loss in biodiversity, and loss in terms of

attenuating emissions of gases causing greenhouse effect. In Morocco, around 65 million

ha of pastureland provide 30% of overall animal food requirements. Erosion, drought,

overgrazing, land clearing and removal of woods are resulting in degraded pastureland.

4.5.1 Methodology

Degradation of agricultural land is estimated by calculating the value of lost agriculture

production due to a decrease in land productivity. Since the majority of agricultural land

is planted with cereals, the cost of degraded agricultural land corresponds to the value

of lost cereal production. As for the degradation of rangeland, it is estimated by

calculating the cost of lost forage production

Step 1: Estimation of degraded agricultural land

FAO classified the degradation of 8.7 million ha in Morocco as “severe”. According to the

FAO, three scenarios are possible:

− 10 to 25% of land is severely degraded

− 25 to 50% of land is moderately degraded

− 50 to 100% of land is slightly degraded

Surveys did not show any case of severe land degradation, thus only moderate and slight

degradation are used.

Step 2: Estimation of the decrease in agricultural yield

Young estimated the decrease in cereal yield as follows:

− Slight degradation corresponds to a 5% decrease in cereal yield

− Moderate degradation corresponds to a 20% decrease in cereal yield

26

Given that the mean yield for cereals in Morocco is 1 Ton/ha, then a slight degradation

will result in a decrease of 50 Kg/ha in cereal production, while a moderate degradation

will result in a 200 Kg/ha decrease.

Step 3: Assessing the cost of degraded agricultural land

The average of the lower bound and the upper bound of moderate and slight

degradation were used. The adopted selling price of cereal was 2,580 Dh/Ton.

Accordingly, the average cost of agricultural land degradation is estimated at Dh 1,263

million, constituting 0.36% of the GDP (Table 2).

Table 2. Assessing the cost of degraded agricultural land

Parameter Lower limit Upper limit

Moderate erosion 25% 50%

Degraded agricultural land (000 ha) 2,175 4,350

Level of decrease 20% 20%

Decrease in yield (Kg/ha) 200 200

Lost production (Tons) 435 870

Lost value (millions of Dh) 1,122 2,244

Slight erosion 50% 100%

Degraded agricultural land (000 ha) 4,350 8,700


Decrease in yield (Kg/ha) 50 50

Lost production (Tons) 217.5 435

Lost value (millions of Dh) 561 1,122

Average (millions of Dh) 842 1,683

Step 4: Estimation of degraded pastureland

When estimating the cost of degraded pastureland, calculations considered only the

areas with dominant steppe and forest covers, excluding the Saharan region. The total

dominant steppe area considered was 12 million ha and the total dominant forest area

considered was 5.1 million ha. According to REEM, 46% of dominant steppe is degraded,

amounting to 5.52 million ha, and 19% of dominant forest area is degraded, amounting

to 0.969 million ha.

Step 5: Estimation of the loss of productivity

The Ministry of Agriculture and Agricultural Development estimated land productivity as

79 FU/ha/year (FU: Forage Unit equivalent to 1Kg Barley) for steppe and 558 FU/ha/year

27

for forest. The Ministry of Agriculture and Agricultural Development adopted 2 levels of

loss, 6% and 10%. Accordingly, the estimated productivity loss was as follows:

Steppe: 6% 26.1 million FU / year

10% 43.6 million FU / year

Forest: 6% 26.1 million FU / year

10% 43.6 million FU / year

Step 6: Assessing the cost of rangeland degradation

Given that the price of barley is 2,270 Dh/Ton, the FU price is 2.27 Dh. Accordingly, the

average cost of rangeland degradation is estimated at 177.4 Dh million, which

constitutes 0.05% of the GDP (Table 3).

Table 3. Assessing the cost of rangeland degradation

Parameter Steppe Forest Total

Pasture area (000ha) 12,000 5,100 17,100

Degraded area (%) 46% 19%

Degraded area (000ha) 5,520 969 6,489

Land productivity (FU/ha/year) 79 558

Loss in yield in degraded area 10%

Lost yield (000FU/year) 43,608 54,070 97,678

Lost value (million Dh) 99.0 122,7 221.7

Loss in yield in degraded area 6%

Lost yield (000FU/year) 26,165 32,442 58,607

Lost value (million Dh) 59.4 73,6 133.0

Average (million Dh) 79.2 98.2 177.4

Based on the above estimations, the total cost of land degradation, including both

agricultural and rangeland, is estimated to range between Dh 975 and 1,895 million,

with an average of 1,440 million Dh, which constitutes 0.41% of the GDP.

5 THE MARKET PRICE METHOD

The market price method makes use of observed market prices for environmental goods

and services. It values changes in quantity and/or quality of a good or service by using

standard economic techniques for measuring the economic benefits from marketed

goods. This method is applied for goods and services with established markets, and

28

which have direct uses, such as plantation timber, commercial fisheries, and tourism;

some indirect uses, such as value of water from protected watersheds; and some option

values, such as gene research and forest conservation.

5.1 Applying the Market Price Method

Market price represents the value of an additional unit of that good or service, assuming

the good is sold through a perfectly competitive market. Applying the method requires

data to estimate consumer surplus and producer surplus. To estimate consumer surplus,

the demand function must be estimated, which requires time series data on the quantity

demanded at different prices, and data on other factors that might affect demand, such

as income or other demographic data. To estimate producer surplus, data is needed on

variable costs of production and revenues received.

5.2 Advantages and limitations

The advantages of the market price method include the following:

− It is relatively simple and straightforward

− It relies on actual market values

− The price, quantity and cost data are easy to obtain for established markets

− The method uses observed data of actual consumer preferences

− The method uses standard, accepted economic techniques

The application of the market price method is associated with several issues and

limitations including:

− Market data may only be available for a limited number of goods and services

provided by a resource

− Available market data may not reflect the value of all productive uses of a resource

− The true economic value of goods or services may not be fully reflected in market

transactions, due to market imperfections and/or policy failures

− Seasonal variations and other effects on price must be considered

− Cannot be easily used to measure the value of larger scale changes that are likely to

affect the supply of or demand for a good or service

− Does not deduct the market value of other resources used to bring ecosystem

products to market, and thus may overstate benefits

5.3 Illustration 1‐ Closure of a commercial fishing area due to water pollution

Water pollution is causing the closure of a commercial fishing area. The benefits of

cleanup need to be evaluated before deciding on their implementation. The market

29

price method was used because the primary resource affected is fish, for which market

data are available.

The objective is to measure the total economic surplus for the increased fish harvest

that would occur if the pollution is cleaned up. The difference between economic

surplus before and after the closure must be estimated. The results of the analysis can

be used to compare the benefits of actions that would allow the area to be reopened, to

the costs of such actions.

Step 1

Use market data to estimate the market demand function and consumer surplus for the

fish before the closure. Assume a linear demand function, where the initial market price

is $5/g and the maximum willingness to pay is $10/g. At $5/g, consumers purchased

10,000 g fish/yr, thus spending $50,000 on fish per year. The shaded area on the graph

represents the total consumer surplus received from the fish before the closure = ($10‐

$5)*10,000/2 = $25,000 (Figure 19).

02468

1012

0 5,000 10,000 15,000 20,000 25,000

Quantity demanded (g)

Pric

e ($

/g)

Figure 19. Demand for fish before closure

Step 2

Estimate the market demand function and consumer surplus for the fish after the

closure. The market price of fish increased from $5/g to $7/g. The total quantity

demanded decreased to 6,000 g/yr. The new consumer surplus is ($10‐$7)*6,000/2 =

$9,000 (Figure 20).

Consumer surplus

30

02468

1012

0 5,000 10,000 15,000 20,000 25,000

Quantity demanded (g)

Pric

e ($

/g)

Figure 20. Demand for fish after closure

Step 3

Estimate the loss in economic benefits to consumers by subtracting benefits after the

closure from benefits before the closure. The loss in benefits to consumers is 25,000 ‐

9,000 = $16,000.

Step 4

Estimate the losses to producers by first measuring the producer surplus before the

closure. Producer surplus is measured by the difference between the total revenues

earned from a good, and the total variable costs of producing it. Before the closure,

10,000 g of fish were caught per year. Fishermen were paid $1/g, with their total

revenues amounting to $10,000 per year. The variable cost to harvest the fish was

$0.50/g (total variable cost = $5,000 per year). The producer surplus before the closure

was $10,000 ‐ $5,000 = $5,000.

Step 5:

Measure the producer surplus after the closure (Table 4).

Step 6:

Calculate the loss in producer surplus due to the closure (Table 4).

Consumer surplus

31

Table 4. Summary of calculations

Before closure After closure

Fish caught per year = 10,000 g Fish caught per year = 6,000 g

Fishermen were paid $1/g Fishermen were paid $1/g

Total revenues = 1 × 10,000 = $10,000 per year Total revenues = 1 × 6,000 = $6,000 per year

Variable cost to harvest fish = $0.50/g Variable cost to harvest fish = $0.60/g

Total variable cost = 0.5 × 10,000 = $5,000 per year Total variable cost = 0.5 × 6,000 = $3,600 per year

The producer surplus = $10,000 ‐ $5,000 = $5,000 The producer surplus = $6,000 ‐ $3,600 = $2,400

Loss in producer surplus due to the closure = $5,000 ‐ $2,400 = $2,600

Step 7:

Calculate the total economic losses due to the closure. Total economic loss = lost

consumer surplus ($16,000) + lost producer surplus ($2,600). Thus, the benefits of

cleaning up pollution in order to reopen the area are equal to $18,600.

Finally, it is important to note that this example is based on assumptions that greatly

simplify the analysis. Some factors might make the analysis complicated. For instance,

some fishermen might switch to another fishery after the closure, and thus losses would

be lower.

6 DAMAGE COST AVOIDED, REPLACEMENT COST, AND

SUBSTITUTE COST METHODS

These methods estimate values of ecosystem services based on the costs of avoiding

damages due to lost services, the cost of replacing ecosystem services, or the cost of

providing substitute services. These methods assume that the costs of avoiding damages

or replacing ecosystems or their services provide useful estimates of the value of these

ecosystems or services. They also assume that if people incur costs to avoid damages

caused by lost ecosystem services, or to replace the services of ecosystems, then those

services must be worth at least what people paid to replace them. The damage cost

avoided, replacement cost, and substitute cost methods are most appropriately applied

in cases where damage avoidance or replacement expenditures have actually been, or

will actually be, made.

The damage cost avoided method uses either the value of property protected or the

cost of actions taken to avoid damages as a measure of the benefits provided. For

example, if a wetland protects adjacent property from flooding, the flood protection

32

benefits may be estimated by the damages avoided if the flooding does not occur or by

the expenditures property owners make to protect their property from flooding. The

replacement cost method uses the cost of replacing an ecosystem or its services as an

estimate of the value of the ecosystem or its services. As for the substitute cost method,

it uses the cost of providing substitutes for an ecosystem or its services as an estimate of

the value of the ecosystem or its services. For example, the flood protection services of a

wetland might be replaced by a retaining wall or levee.

6.1 Applying the methods

Step 1: Assess the environmental service provided

This is done by specifying the relevant services, how they are provided, to whom they

are provided, and the levels provided. For example, in the case of flood protection, this

would involve predictions of flooding occurrences and their levels, as well as the

potential impacts on property.

Steps 2 and 3

For the damage cost avoided method, estimate the potential physical damage to

property, either annually or over some discrete time period. Then calculate either the

dollar value of potential property damage, or the amount that people spend to avoid

such damage.

For the replacement or substitute cost method, identify the least costly alternative

means of providing the service and calculate the cost of the substitute or replacement

service. Then establish public demand for this alternative, which requires gathering

evidence that the public would be willing to accept the substitute or replacement

service in place of the ecosystem service.

Examples of applications of these methods

− Valuing improved water quality by measuring the cost of controlling effluent

emissions

− Valuing erosion protection services of a forest or wetland by measuring the cost of

removing eroded sediment from downstream areas

− Valuing the water purification services of a wetland by measuring the cost of filtering

and chemically treating water

− Valuing storm protection services of coastal wetlands by measuring the cost of

building retaining walls

− Valuing fish habitat and nursery services by measuring the cost of fish breeding and

stocking programs

33


The advantages of these methods include:

− They may provide a rough indicator of economic value, subject to data constraints

and the degree of similarity or substitutability between related goods.

− They are less data and resource‐intensive, whereby it is easier to measure the costs

of producing benefits than the benefits themselves, when goods, services, and

benefits are non‐marketed.

− Data or resource limitations may rule out valuation methods that estimate

willingness to pay

− They provide surrogate measures of value that are as consistent as possible with the

economic concept of use value, for services which may be difficult to value by other

means

Issues and limitations associate with these methods include:

− They do not provide a technically correct measure of economic value, which is

properly measured by the maximum amount of money or other goods that a person

is willing to give up to have a particular good, less the actual cost of the good.

− They assume that expenditures to repair damages or to replace ecosystem services

are valid measures of the benefits provided.

− They do not consider social preferences for ecosystem services, or individuals’

behaviour in the absence of those services.

− They may be inconsistent because few environmental actions and regulations are

based solely on benefit‐cost comparisons, particularly at the national level. For

instance, the cost of a protective action may exceed the benefits to society.

Alternatively, the cost of actions already taken to protect an ecological resource will

underestimate the benefits of a new action to improve or protect the resource.

− The replacement cost method requires information on the degree of substitution

between the market good and the natural resource. Substitute goods are unlikely to

provide the same types of benefits as the natural resource.

− The goods/services being replaced probably represent only a portion of the full

range of services provided by the natural resource. Thus, the benefits of an action to

protect or restore the ecological resource would be understated.

− Without evidence that the public would demand the least cost alternative for the

affected ecosystem, this methodology is not an economically appropriate estimator

of ecosystem service value.

34

6.3 Illustration 1‐ Restoration of degraded wetlands

An agency is considering restoring some degraded wetlands in order to improve their

ability to protect the surrounding area from flooding. Cost‐based methods are used

because the agency is only interested in valuing the flood protection services of the

wetlands and because a limited budget is available for the valuation study.

Step 1

Conduct an ecological assessment of the flood protection services provided by the

wetlands to determine the current level of flood protection and the expected level of

protection after full restoration of the wetlands

Step 2

The Damage Cost Avoided method is applied using two different approaches

− Use the information on flood protection obtained in the first step to estimate

potential damages to property if flooding were to occur. Estimate, in dollars, the

probable damages to property if the wetlands are not restored.

− Determine whether nearby property owners have spent money to protect their

property from the possibility of flood damage by purchasing additional insurance or

by reinforcing their basements. These avoidance expenditures would be summed

over all affected properties to provide an estimate of the benefits from increased

flood protection.

Note that the two approaches are not expected to produce the same estimate.

The replacement cost method cannot be applied since flood protection services cannot

be directly replaced.

The substitute cost method can be applied since a substitute for the affected services

such as a retaining wall or a levee might be built to protect nearby properties from

flooding. In this case, estimate the cost of building and maintaining such a wall or levee.

Also determine whether people would be willing to accept the wall or levee in place of a

restored wetland.

Step 3

Compare the cost of the property damages avoided, or of providing substitute flood

protection services to the restoration costs to determine whether it is worthwhile to

restore the flood protection services of the wetlands.

35

6.4 Illustration 2: Soil Erosion in Korea using the replacement cost method

Urban growth and industrial development in Korea caused farming activities to move

into hilly upland area. Inadequate soil management techniques and errors in field layout

and construction resulted in soil erosion of these upland areas. It is required to evaluate

the benefits of proposed new soil management techniques, including retaining the soil

and nutrients on the upland areas and protecting downslope areas from damage by the

eroded soil.

The researchers measured the cost of physically replacing lost soil, nutrients, and water

in upland areas and the cost of compensating for downstream losses by:

− calculating the annual soil loss per hectare, nutrient loss/hectare, and water

runoff/hectare (Table 5)

− calculating the expected losses, in terms of replacement costs, if the new

management practices were not implemented. The net present value of the losses

amounted to W 2,039,662, using a 15 year time horizon.

Table 5. Cost of replacement activities

Measured parameter Cost (W/ha/yr)

Recovering and replacing eroded soil 80,000

Fertilizer and spreading to replace lost nutrients 31,200

Annual field maintenance and repair 35,000

Damage to downstream fields in lost production 30,000

Supplemental irrigation to replace lost water 92,000

Total cost of soil erosion under existing management 268,200

Net present value using a 15 year time horizon 2,039,662

Then calculate the costs with the new management techniques, including compensation

payments, soil replacement, nutrient replacement, and mulching. The net present value

of the costs of new management techniques was estimated at W 1,076,742.

The cost of new management techniques (W 1,076,742) is about half the replacement

cost (W 2,039,662). Thus, the proposed preventive steps are worth implementing.

6.5 Illustration 3: Oil spill damage in Puerto Rico (replacement cost method)

The Zoe Colocotroni was a ship that spilled oil off the coast of Puerto Rico. The case was

taken to court to determine the monetary damages resulting from the spill’s effects on

the local ecosystem. The replacement cost method was used to estimate monetary

36

damages. This was done by calculating the number of lower trophic organisms killed by

the spill and adding up the cost of purchasing these organisms from a scientific

catalogue. However, the US Court of Appeals rejected the use of the replacement cost

method in this case, as it was not planned to actually purchase the organisms and

restore them to the ocean. By the time such a plan could have been carried out, the

organisms would have restored themselves. Thus the costs of purchasing the organisms

did not accurately measure the actual ecosystem damages.

THE TRAVEL COST METHOD

Session 5

Region

al Training Worksho

p on

: Th

e Co

st of E

nviron

men

tal D

egrada

tion

Metho

dology

37

SESSION 5

7 THE TRAVEL COST METHOD

The travel cost method (TCM) is used to estimate use values associated with ecosystems

or sites that are used for recreation. The concept of using travel costs to value recreation

was first proposed by Hotelling in 1949 and formalised by Clawson (1959). This concept

assumes that the value of the site or its recreational services is reflected in how much

people are willing to pay to get there. This method is useful in planning for the provision

and management of outdoor recreation, such as changes in access costs for a

recreational site, elimination of an existing recreational site, addition of a new

recreational site, and changes in environmental quality at a recreational site.

7.1 Theory

The travel cost method is based on the premises that the cost an individual incurs in

visiting a site reflects his valuation to the site, and that individuals will react to an

increase in entry fees the same way as they would react to an increase in travel cost.

That is, at some high level of entry fee or travel cost, no one would visit the site. By

asking visitors questions relating to where they had travelled from and the costs they

had incurred, and relating this information to the number of visits they make per year, a

trip generation function can be generated for the recreational site under question. An

aggregate demand curve is then derived for visits to the sites per year. The demand

curve shows how many visits people would make at various travel cost prices and is thus

used to estimate the willingness to pay for people to visit the site. The curve is

downward sloping, where the travel cost is inversely related to the number of visits.

That is, people who live farther from the site will visit it less often, because it costs more

in terms of actual travel costs and time to reach the site. Other factors that might affect

the number of visitors to the site include a visitor’s income, the availability of alternative

sites or substitutes, ad factors like personal interest in the type of site, or level of

recreational experience.

Travel cost models can assume a linear functional form or a log‐linear functional form.

Figure 21 illustrates the linear functional form.

38

Figure 21. Linear demand curve

V = α + βC +γS Where: V = number of visits to a site

α = constant β = coefficient of C, usually negative C = cost of travel to gain access to site

γ = coefficient of S, probably negative S = cost of travel to gain access to the respondent’s preferred substitute site

The travel cost model is used to estimate α, β, and γ. The estimated consumer surplus

(CS) for an individual making q visits to the site is as follows:

CS = ‐q2 / 2β

Note that this functional form implies finite visits at zero costs and it has a critical cost

above which negative visits will be demanded.

The Log‐Linear functional form is illustrated in Figure 22.

Figure 22. Log‐linear demand curve

39

lnV = α + βC +γS

Where: V = number of visits to a site

α = constant β = coefficient of C, usually negative C = cost of travel to gain access to site

γ = coefficient of S, probably negative S = cost of travel to gain access to the respondent’s preferred substitute site

The travel cost model is used to estimate α, β, and γ. The estimated consumer surplus

(CS) for an individual making q visits to the site is as follows:

CS = ‐q / β

This functional form has been widely used in TCM models. It implies a finite number of

visits at zero cost, and it never predicts negative visits even at very high costs.

7.2 Forms of TCMs

There are three forms of TCMs: the zonal TCM, the individual TCM, and the random

utility approach.

7.2.1 Zonal TCM approach

In the Zonal TCM approach, concentric zones are defined around each site such that the

cost of travel from all points in a given zone is approximately constant (Figure 23).

Visitors to the site are grouped according to their zone of origin. This approach is the

simplest and least expensive. It can rely on secondary data and it is suitable when

visitors’ origins are relatively evenly distributed and it is unsuitable for linear

recreational sites.

Figure 23. Concentric zones around site S

Steps to apply the Zonal TCM:

1. Identify the site and collect data from visitors on their points of origin, number of

visits from each origin zone, round‐trip mileage from each zone, travel costs per

mile, and demographic information about people from each zone.

2. Define zones of origin and allocate visitors to the appropriate zone. Zones are

commonly defined based on the straight line distance from the site. Geographic

40

Information system (GIS) techniques allow redefining zones based on road distances

or travel times.

3. Calculate zonal visits per household to the site by estimating the number of

households per zone and dividing the number of household visits originating in the

zone by the total number of households in the zone.

4. Calculate the average travel cost from each zone to the site.

5. Use census data to derive variables relating to zonal socio‐economic characteristics.

6. Use data collected above to estimate the trip generation function:

Vh/Nh = f(Ch,Xh,Sh)

where: Vh = # of visits from zone h

Nh = population of zone h

Ch = travel cost from zone h

Xh = a vector of socio‐economic variables that explain changes in V

Sh = a vector of substitute recreational site characteristics for residents of zone h

7. Derive the demand curve

8. Obtain zonal household consumer surplus estimates through integrating under the

demand curve

9. Calculate aggregate zonal consumer surplus by multiplying consumer surplus per

household by the number of households in each zone

10. Aggregate zonal consumer surplus estimates to obtain an estimate of total consumer

surplus or the benefits of the site

7.2.2 Individual TCM methodology

The Individual TCM uses the number of visits made per year by an individual, rather than

the zonal visits, as the basis for generating the demand curve. This method requires

more data collection, as compared to the zonal TCM, and slightly more complicated

analysis. Yet, it is more flexible. It is applicable at a wider range of sites, and it gives

more precise and statistically efficient results.

Steps in applying the Individual TCM:

1. Identify the site.

2. Use an on‐site questionnaire survey to collect data from visitors relating to the cost

of travel to the site, the number of visits to the sites, recreational preferences, and

socio‐economic characteristics.

3. Specify the trip‐generation function:

41

Vij = f(Cij,Tij,Qi,Sj, Yi)

where: Vi = # of visits made by individual i to site j

Cij = travel cost incurred by individual i when visiting site j

Qj = a vector of perceived qualities of the recreation site j

Sj = a vector of available substitute recreational site characteristics

Yi = household income of individual i

4. Estimate the travel cost model taking truncation into account for non‐visitors

behavior

5. Derive demand curve and obtain household consumer surplus estimates through

integrating under the demand curve

6. Calculate aggregate consumer surplus for the site

Note that more complicated and thorough applications of the individual TCM may also

collect information about:

− exact distance that each individual travelled to the site

− exact travel expenses

− the length of the trip

− the amount of time spent at the site

− other locations visited during the same trip, and amount of time spent at each

− substitute sites that the person might visit instead of this site, and the travel distance

to each

− other reasons for the trip (is the trip only to visit the site, or for several purposes)

− quality of the recreational experience at the site, and at other similar sites (e.g.,

fishing success)

− perceptions of environmental quality at the site

− characteristics of the site and other, substitute, sites

7.2.3 The random utility approach

This approach allows for much more flexibility in calculating benefits, yet it is the most

complicated and expensive. It is best suited to estimate benefits for specific

characteristics of sites, rather than for the site as a whole, and it is most appropriate

when there are many substitute sites. The random utility approach focuses attention on

the choice among alternative sites for any given recreational trip and assumes the visitor

is comparing utilities for available destinations. It first models the individual’s decision

on whether or not to participate in recreational activity, and then models the decision

on the number of visits. Models used include probit, tobit, and logit.

42

7.3 Advantages and limitations of TCMs

Advantages of the TCM include the following:

− It is based on real data rather than stated willingness to pay and as such provides

true values

− It is relatively inexpensive to apply

− On‐site surveys provide opportunities for large sample sizes, as visitors tend to be

interested in participating

− The results are relatively easy to interpret and explain

However, the TCM is associated with various issues and limitations that should be taken

into consideration, including the following:

− The method assumes that people perceive and respond to changes in travel costs

the same way that they would respond to changes in admission price.

− The simplest models assume that individuals take a trip for the single purpose of

visiting a specific recreational site. Thus, if a trip has more than one purpose, the

value of the site may be overestimated.

− Defining and measuring the opportunity cost of time, or the value of time spent

travelling, can be problematic. There is no consensus on how to account for time,

whereby travel time may be a benefit if people enjoy the travel itself, leading to an

overestimation of the value of the site.

− The availability of substitute sites will affect values. For example, if two people travel

the same distance, they are assumed to have the same value. However, if one

person has several substitutes available but travels to this site because it is

preferred, this person’s value is actually higher. Some of the more complicated

models account for the availability of substitutes.

− The assumption that travel costs reflect recreational value may not always be true.

Those who value certain sites may choose to live nearby, resulting in low travel

costs, but high values for the site.

− Visits to certain sites could be seasonal and thus survey results could be biased

unless survey is conducted for a long period.

− Interviewing visitors on site can introduce sampling biases to the analysis.

− Measuring recreational quality and relating it to environmental quality can be

difficult.

− Standard travel cost approaches provides information about current conditions, but

not about gains or losses from anticipated changes in resource conditions.

− The demand function requires enough difference between distances travelled to

affect travel costs and for differences in travel costs to affect the number of trips

43

made. Thus, the TCM is not well suited for sites near major population centres

where many visitations may be from "origin zones" that are close to one another.

− The travel cost method is limited in its scope of application because it requires user

participation. Thus, it cannot be used to assign values to on‐site environmental

features and functions that users of the site do not find valuable. It cannot be used

to value off‐site values supported by the site, or to measure non‐use values. It

excludes non‐users who may have significant values for the site

− Certain statistical problems can affect the results, including, choice of the functional

form used to estimate the demand curve, choice of the estimating method, and

choice of variables included in the model

7.4 Illustration‐ Recreational fishing site

A site used mainly for recreational fishing is threatened by development in the

surrounding area. Pollution and other impacts from this development could destroy the

fish habitat at the site, resulting in a serious decline in, or total loss of, the site’s ability

to provide recreational fishing services. Resource agency staff wants to determine the

value of programs or actions to protect fish habitat at the site.

The TCM was selected because the site is primarily valuable to people as a recreational

site. In addition, this site has no endangered species or other highly unique qualities that

would make non‐use values for the site significant. Furthermore, the expenditures for

projects to protect the site are relatively low. Alternative approaches, including the

contingent valuation or contingent choice methods, might produce more precise

estimates of values for specific characteristics of the site and could capture non‐use

values. However, they are considerably more complicated and expensive to apply.

7.4.1 Application of the Zonal Travel Cost Approach

Step 1

Define a set of zones surrounding the site by concentric circles around the site, or by

geographic divisions such as metropolitan areas or counties surrounding the site at

different distances.

Step 2

Collect information on the number of visitors from each zone and the number of visits

made in the last year. For this example, assume that the staff at the site keeps records of

the number of visitors and their zip code, which can be used to calculate total visits per

zone over the last year.

44

Step 3

Calculate the visitation rates per 1,000 population in each zone, which is the total visits

per year from the zone, divided by the zone’s population in thousands (Table 6).

Table 6. Visitation rates for the site

Zone Total Visits/Year Zone Population Visits/1000

0 400 1,000 400

1 400 2,000 200

2 400 4,000 100

3 400 8,000 50

Beyond 3 0

Total visits 1,600

Step 4

Calculate the average round‐trip travel distance and travel time to the site for each

zone. Using average cost per mile and per hour of travel time, calculate the travel cost

per trip, by assuming that this cost per mile is USD 0.30. The cost of time is calculated

using the simplest approach involving the average hourly wage. It is assumed that the

average hourly wage is 9 USD/hour or $0.15 USD/minute for all zones; although in

practice it is likely to differ by zone (Table 7).

Table 7. Travel cost calculation

Zone Round Trip Travel Distance

Round Trip Travel Time

Distance × Cost/mile ($.30)

Travel Time × Cost/minute ($.15)

Total Travel (Cost/Trip)

0 0 0 0 0 0

1 20 30 $6 $4.50 $10.50

2 40 60 $12 $9.00 $21.00

3 80 120 $24 $18.00 $42.00

Step 5

Estimate the trip generation function using regression analysis. This allows the

estimation of the demand function for the average visitor. The analysis might include

demographic variables, such as age, income, gender, and education levels, using the

average values for each zone. To maintain the simplest possible model, calculating the

equation with only the travel cost and visits/1000:

Visits/1000 = 330 – 7.755 * Travel Cost

45

Step 6

Construct the demand function for visits to the site. The first point on the demand curve

is the total visitors to the site at current access costs, or 1,600 visits per year. The other

points are found by estimating the number of visitors with different hypothetical

entrance fees. For example, start by assuming a $10 entrance fee, plugging this into the

estimated regression equation, V = 330 – 7.755C, gives the data in Table 8.

Table 8. Deriving the demand curve

Zone Travel Cost plus $10

Visits/1000 Population Total Visits

0 $10 252 1,000 252

1 $20.50 171 2,000 342

2 $31.00 90 4,000 360

3 $52.00 0 8,000 0

Total Visits 954

This gives the second point on the demand curve: 954 visits at an entry fee of $10. In the

same way, the number of visits for increasing entry fees can be calculated (Table 9).

These points give the demand curve for trips to the site (Figure 24).

Table 9. Number of visits for increasing entry fees

Entry Fee Total Visits

$20 409

$30 129

$40 20

$50 0

0

10

20

30

40

50

60

0 400 800 1200 1600 2000

Total visits

Adde

d co

st p

er tr

ip

Figure 24. Demand curve for the trips to the site

46

Step 7

Estimate the total economic benefit of the site to visitors by calculating the consumer

surplus, or the area under the demand curve. The total estimate of economic benefits

from recreational uses of the site is around $23,000 per year, or around $14.38 per visit.

Thus, if the actions to protect the site cost less than $23,000 per year, the cost will be

less than the benefits provided by the site. If the costs are greater than this, the staff will

have to decide whether other factors make them worthwhile.

7.4.2 Application of the Individual Travel Cost Approach

Step 1

Conduct a survey of visitors on:

− location of the visitor’s home –distance travelled to the site

− how many times they visited the site in the past year or season

− the length of the trip

− the amount of time spent at the site

− travel expenses

− the person’s income or other information on the value of their time

− other socioeconomic characteristics of the visitor

− other locations visited during the same trip, and amount of time spent at each

− other reasons for the trip (is the trip only to visit the site, or for several purposes)

− fishing success at the site (how many fish caught on each trip)

− perceptions of environmental quality or quality of fishing at the site

− substitute sites that the person might visit instead of this site

Step 2

Estimate the relationship between number of visits and travel costs and other relevant

variables using regression analysis. Use individual data rather than data for each zone.

The regression equation gives the demand function for the “average” visitor to the site

and the area below this demand curve gives the average consumer surplus.

Step 3

Multiply the average consumer surplus by the total relevant population in the region

where visitors come from to estimate the total consumer surplus for the site.

Step 4

Value estimates can be improved by adding other factors to the statistical model,

including additional data about visitors, substitute sites, and quality of the site. Including

47

information about the quality of the site allows the researcher to estimate the change in

value of the site if its quality changes. In this case, two different demand curves would

be estimated; one for each level of quality. The area between these two curves is the

estimate of the change in consumer surplus when quality changes

7.4.3 Application of the Random Utility Approach

The agency might want to value the economic losses from a decrease in fish

populations, rather than from loss of the entire fish stock. The random utility approach

focuses on choices among alternative sites which have different quality characteristics. It

assumes that individuals will pick the site that they prefer, out of all possible fishing

sites. This model requires information on all possible sites that a visitor might select,

their quality characteristics, and the travel costs to each site.

Step 1

Conduct a telephone survey of randomly selected residents of the state, asking residents

if they go fishing or not. If they do, then ask a series of questions: how many fishing trips

they took over the last year (or season), where they went, the distance to each site, and

other information similar to the information collected in the individual travel cost

survey. One might also ask questions about fish species targeted on each trip, and how

many fish were caught.

Step 2

Estimate a statistical model that can predict the choice to go fishing or not and the

factors that determine which site is selected. If quality characteristics of sites are

included, the model can estimate values for changes in site quality, for example the

economic losses caused by a decrease in catch rates at the site.

7.5 Case application‐ Hell Canyon preservation

The following is a case application of the Travel Cost Method as part of the efforts to

preserve the.

The Hell Canyon is situated on the Snake River separating Oregon and Idaho. It offers

spectacular vistas and outdoor amenities to visitors and supports important fish and

wildlife habitat. It also has economic potential as a site to develop hydropower. Yet,

generating hydropower there would require building a dam. The dam and resulting lake

would significantly and permanently alter the ecological and aesthetic characteristics of

Hell Canyon. During the 1970’s, there were controversies regarding the future of Hell

48

Canyon. Thus, environmental economists were asked to develop an economic analysis

justifying the preservation of Hell Canyon in its natural state.

Accordingly, the net economic value (cost savings) of producing hydropower at Hell

Canyon was $80,000 higher than at the "next best" site which was not environmentally

sensitive. The recreational value of Hell Canyon was estimated via a low‐cost/low

precision travel‐cost survey at about $900,000. Even if the "true value" of recreation at

Hell Canyon was ten times less than their estimate, it would still be greater than the

$80,000 economic payoff from generating power there as opposed to the other site.

Congress voted to prohibit further development of Hell Canyon, based largely on the

results of this non‐market valuation study.

7.6 Case‐study: The value of forestry in Britain

The travel cost method was used to estimate the total recreational value of Forestry

Commission Woodland in Great Britain, by Willis. In order to do this it was first

necessary to define some representative forest types. Using a statistical technique called

cluster analysis, 14 similar groups of forests were identified and travel cost studies were

undertaken of sample forests from each of these groups. Interviews were undertaken

with visitors in 15 forests. Visitors to sites were allocated into 20 concentric distance

zones, at five‐mile intervals. Those from further away were allocated together into a

single further zone. Willis then estimated relationships between the visit rate from each

zone and the transport costs, taking account of the socio‐economic characteristics of the

zones.

These relationships, referred to as trip generating functions, were then used in order to

estimate the consumer surplus, or the total value of each visit, represented by the

maximum willingness to pay, less the cost of each trip. Summing across all visitors to

each site produced estimates of the total value for each site. Some of the results are

shown in Table 10. This shows the estimated consumer surplus per visit, the consumer

surplus per hectare of forest at the survey site, the total annual number of visitors at all

of the forests within this group of forests and the total consumer surplus generated by

this group of forests.

49

Table 10. Consumer surplus estimates for non‐priced recreation for forest districts

Forest Consumer surplus per recreational

visitor (£)

Consumer surplus per hectare of

forest (£/ha)

Annual number of visitors to the cluster of forests (1000s)

Total consumer surplus (£million)

New Forest 1.43 425 8,000 11,440

Loch Awe 3.31 <1 34 0.114

Brecon 2.26 27 2,117 4.784

South Lakes 1.34 31 1,968 2.637

Thetford 2.66 14 4,742 10.718

By summing all of the estimates of consumer surplus, it is possible to obtain an estimate

of the value of non‐priced recreation for Forestry Commission forests as a whole. This

gives a figure of £53 million. This compares with a figure of £71 million income to

Forestry Enterprise, the timber production arm of the Forestry Commission, from the

sale of timber 1988. In this year, there was an annual net subsidy of £8.5 million paid on

forestry recreation and amenity. This amount represents a net cost to the Forestry

Commission of providing facilities, such as visitor centers, forest walks, wildlife

conservation and amenity tree planting. The figures indicated the importance of

recreation in the role of the Forestry Commission and suggested that its provision for

non‐priced recreation represents good value.

50


THE HEDONIC PRICING METHOD

&

THE AVERTING BEHAVIOR METHOD

Session 6

Region

al Training Worksho

p on

: Th

e Co

st of E

nviron

men

tal D

egrada

tion

Metho

dology

51

SESSION 6

8 THE HEDONIC PRICING METHOD

The Hedonic Pricing Method (HPM) is used to estimate the value or price of an

environmental feature by looking at actual markets where the attributes are traded. It is

most commonly applied in relation to the public’s willingness to pay for housing/

property and in labour markets for health economic valuation.

8.1 Theory

The HPM is based on the assumption that people value the characteristics of a good, or

the services it provides, rather than the good itself. Thus, prices will reflect the value of a

set of characteristics, including environmental characteristics that people consider

important when purchasing the good. For example, the price of a car reflects the

characteristics of that car, in terms of transportation, comfort, style, luxury, fuel

economy, etc. One can value the individual characteristics of a car or other good by

looking at how the price people are willing to pay for it changes when the characteristics

change.

The HPM assumes that the price of a product is a function of its characteristics; the

range of product choices is continuous; the choice is based on perfect information and

with no mobility restrictions; and the amount of a particular characteristic can be varied

independently.

The HPM is relatively straightforward and uncontroversial to apply, because it is based

on actual market prices and fairly easily measured data. If data are readily available, it

can be relatively inexpensive to apply. However, if data must be gathered and compiled,

cost can increase substantially.

The HPM is usually applied ex post, to examine the effects of developments and policies

after implementation. It can be used to estimate economic benefits or costs associated

with:

− Environmental risk. For example, the effect of information of different levels of

earthquake damage on property values

− Environmental quality, including water pollution, such as the impact on waterfront

property; air pollution; noise, such as the impact of highway noise and aircraft noise;

soil quality and erosion

52

− Environmental amenities, such as aesthetic views, proximity to recreational sites,

hazardous sites, waste management sites, etc.

The hedonic technique may also be applied to wage rates. It is based on the assumption

that an individual choice of job may be influenced by the job location if it improves

access to desirable services. The main issue with this technique is high unemployment,

where individuals cannot satisfy their demand for environmental improvement due to

unavailability of suitable jobs in areas of higher environmental quality.

8.2 Applying the Hedonic Pricing Method using housing prices

The price of a house is related to the structural characteristics of the house (plot size,

number of rooms, garage space, structural integrity, etc.), local socio‐economic and

public sector characteristics (unemployment rate, social conditions, quality of schools,

etc.) and local amenity (environmental quality, access to services, communications, etc.).

Upon controlling for non‐environmental factors, any remaining differences in price can

be attributed to differences in environmental quality. For example, if all characteristics

of houses and neighbourhoods in a given area were the same, except for the level of air

pollution, then houses with better air quality would cost more. This higher price reflects

the value of cleaner air to people who purchase houses in the area.

Step 1

Collect the needed information. Data requirements fall into two broad categories:

− Specific data: cross‐section and/or time‐series data on property values and property

and household characteristics for a well‐defined market area including, structural

and locational information, and details of purchase or tenancy (price, date, personal

and financial particulars of the purchasers).

− Local data: pertaining to the area where transaction occurred including,

neighbourhood, amenity, environmental, and socio‐economic factors, and a

measure or index of the environmental amenity of interest.

Sources of data depend on country/ state involved. These may include government

agencies, estate agents and realtors, mortgage granting institutions, GIS, postcode

classification of neighbourhood types, etc.

Step 2

Analyze the data using regression analysis, relating the price of the property to its

characteristics and the environmental characteristics of interest. The analysis will

indicate how much property values will change for a small change in each characteristic,

53

holding all other characteristics constant. This analysis may be complicated by a number

of factors. For example, the relationship between price and characteristics of the

property may not be linear, whereby prices may increase at an increasing or decreasing

rate when characteristics change. Another factor is multicollinearity, where many of the

variables are likely to be correlated, so that their values change in similar ways. This can

lead to understating the significance of some variables in the analysis.

Different functional forms and model specifications for the analysis must be considered.

Restrictive functional forms include: linear, semi‐log, log‐linear, and linear Box‐Cox.

These involve simple and transparent relationships between variables. Other more

flexible functional forms may also be used.


The advantages of the HPM include:

− It can be used to estimate values based on actual behaviour and choices.

− Property markets are relatively efficient in responding to information, so can be

good indications of value.

− Property records are typically very reliable.

− Data on property sales and characteristics are readily available through many

sources, and can be related to other secondary data sources to obtain descriptive

variables for the analysis.

− It is a versatile method that can be adapted to consider several possible interactions

between market goods and environmental quality.

The issues and limitations that are associated with the HPM include:

− The scope of environmental benefits that can be measured is limited to things that

are related to property values.

− It will only capture people’s willingness to pay for perceived differences in

environmental attributes, and their direct consequences. Thus, if people aren’t

aware of the linkages between the environmental attribute and benefits to them or

their property, the value will not be reflected in home prices.

− It assumes that people are free to select the combination of characteristics satisfying

their preferences, given their income. However, the housing market may be affected

by outside influences, like taxes, interest rates, or other factors.

− It is relatively complex to implement and interpret, requiring a high degree of

statistical expertise.

− The results depend heavily on model specifications.

54

− It is susceptible to multicollinearity i.e. a high degree of correlation among the

variables under study which makes it difficult to estimate their individual effect. For

example, air pollution measures where the levels of one form of pollution (PM) is

closely related to levels of another (NO2)

− Large amounts of data must be gathered and manipulated.

− The time and expense to carry out an application depends on the availability and

accessibility of data.

8.4 Illustration‐ Open space preservation program

Agency staff wants to measure the benefits of an open space preservation program in a

region where open land is rapidly being developed. The Hedonic Pricing Method is used

because housing prices in the area appear to be related to proximity to open space, and

because data on real estate transactions and open space parcels are readily available.

Alternative approaches include the travel cost method, if the open space of concern is

used mainly for recreation. Survey‐based methods, like contingent valuation or

contingent choice may be used, but these are more difficult and expensive to apply.

Step 1

Collect and compile data on residential property sales in the region for a specific time

period including

− Selling prices and locations of residential properties

− Structural characteristics (lot size, number and size of rooms, number of bathrooms)

− Local socio‐economic characteristics

− Local amenity including the environmental characteristic of concern‐ the proximity

to open space

Collect data on the amount and type of open space within a given radius of each

property, noting the direct proximity of a property to open space. Data may be obtained

from computer‐based GIS maps.

Step 2

Statistically estimate a function that relates property values to the property

characteristics, including the distance to open space. The resulting function measures

the portion of the property price that is attributable to each characteristic. Estimate the

value of preserving open space by looking at how the value of the average home

changes when the amount of open space nearby changes

55

Step 3

Evaluate agency investments in open space preservation and determine the benefits of

preserving each parcel, which can then be compared to the cost

8.5 Case Study 1: Values of Environmental Amenities in Marickville and Rockdale,

Sydney

An early study of hedonic prices was undertaken in two municipalities in Sydney,

Marrickville and Rockdale by Peter Abelson. Data were collected from house sales in

these areas in 1972 and 1973, giving a total sample of 1,414 observations. Information

was included on about twenty characteristics of each property and its local

environment. These included the size, age and construction of the house, the size of the

plot of land, the type and amount of traffic on the road outside the house, access to

public transport and shops, aircraft noise, zoning and whether there were any plans for

road widening. Some of these variables could be measured directly, such as the number

of rooms. Some were measured on a subjective scale. For example, road traffic levels

were described on a three point scale (noisy, normal and quiet). For a third group of

variables it was only possible to define whether or not an item was present, such as

whether or not the house had a double garage.

Various different types of statistical relationships between house prices and

characteristics were tested. These were able to explain about two thirds of the variation

in house prices. The major determinants of house prices were found to be house quality

and size and plot size. Aircraft noise was found to be a significant determinant of house

prices in Marrickville. The results indicated that the price of a very noisy house would be

about $1,250 less than that of a quiet house. For higher priced houses in Rockdale the

difference was about $3250. The price difference between a house on a noisy road in

Marrickville and one on a quiet road was about $1,400 or 5.6 per cent of the house

price. The value of a good view (assessed subjectively) in Rockdale was valued at about

$440 compared with an average view, which in turn was worth $440 more than a poor

view.

Abelson recognizes the limitations of his analysis. There were difficulties in measuring

several of the housing attributes. The hedonic prices which were estimated may not

represent the willingness to pay for amenities because buyers may not have been well

informed when making decisions about house purchase or the housing market may not

have been in equilibrium, i.e. residents may not have been able to select the houses and

characteristics which would best meet their preferences. Abelson did not undertake the

56

second stage to the analysis, so that the estimates may only relate to small changes in

environmental characteristics.

8.6 Case‐study 2: Quarries in Mount Lebanon

Quarries can cause various environmental impacts, including destruction of natural

vegetation and habitats, air pollution from dust, and a reduction in aesthetic value in

and around such localities. There are more than 700 quarries in Lebanon, of which more

than half are in Mount Lebanon province. Many of the quarries are abandoned with

minimal or no rehabilitation and many have been established with little consideration

for the environment and surrounding human settlements. While it would be a significant

undertaking to assess the damage cost of all the quarries, the impact of four quarries on

surrounding settlements in Mount Lebanon was assessed in this study.

8.6.1 The methodology

The cost of degradation due to quarrying was estimated by measuring the loss in land

and apartment values associated with a reduction in aesthetic value. A survey of impacts

on surrounding areas around four quarries in Mount Lebanon (Shnanaayer, Abou‐Mizan,

Antelias, and Nahr Ibrahim) was conducted. Additional impacts recorded to occur during

quarries operation include, structural damage to buildings and infrastructures from

explosives used, dust pollution, and traffic congestion due to quarry transport activities.

These impacts represent a fraction of the losses in land and apartment values and were

not included in this assessment.

The degradation cost associated with the surveyed quarries cannot be extrapolated to

the other quarries in Lebanon (>700), due to differentials in property prices and

locations

Step 1: Estimate the loss in land value around surveyed quarries

First, determine the area of land affected. Then, multiply the area of land by the decline

in land price, based on information from municipality officials and real estate agents

(Table 11). Note that the area around Nahr Ibrahim quarry experienced a reduction in

land prices during quarry operation due to traffic congestion and dust. However, after

quarry closure land prices were no longer affected anymore (contrary to other quarries).

57

Table 11. Calculation of loss in land prices due to quarrying

Quarry Areas affected Land area affected (m2)

Decline in land price

(US $/m2)

Loss in land value

(US $ million)

Shnanaayer Shnanaayer municipality 600,000 125 75.0

Abou‐Mizan Shirine, Bteghrine, and other villages

175,000 7.5 1.3

Antelias Raboueh and Qornet Chehouane municipality

100,000 50 5.0

Total 875,000 93 81.3

Annualized loss (“low”)* 8.1

Annualized loss (“High”)* 9.6

Step 2: Estimate the loss in apartment values around surveyed quarries

First, determine the number of apartments affected and their surface areas. Then,

multiply the estimated are by the decline in apartment price, based on information from

municipality officials and real estate agents. Only the quarries in Shnanaayer and

Antelias are overlooked by residential buildings (Table 12).

Table 12. Calculation of loss in apartment prices due to quarrying

Quarry Areas affected Apartments affected (m2)

Decline in apartment

price (US $/m2)

Loss in apartment value(US $ million)

Shnanaayer Shnanaayer municipality 36,000 225 8.1

Antelias Raboueh and Qornet Chehouane municipality

8,000 100 0.8

Total 44,000 202 8.9

Annualized loss (“low”)* 0.9

Annualized loss (“High”)* 1.0

Step 3: Estimate land value occupied by other quarries (not surveyed)

Caution is warranted before extrapolating costs to other quarries in Lebanon due to

differentials in property prices and location. As a conservative estimate, the cost of

degradation associated with the more than 700 other quarries is calculated as the value

of the land that the quarries occupy. Estimated land value ranges between 3 and 5

US$/m2. Average size of quarry ranges between 15,000 and 20,000 m2 (Table 13).

58

Table 13. Calculation of land value occupied by quarries

Lower bound Higher bound

Number of quarries 710

Average area of quarry (m2) 15,000 20,000

Average land value (US $/m2) 3 5

Total land value (US $ million) 207.6 415.3

Annualized loss (US $ million/year)* 5.0 5.9

Percent of GDP (%) 0.03 0.04

Total losses in land value were annualized at a discount rate f 10% over 20 to 100 years (high and low estimates)

Step 4: Assess the total cost of degradation due to quarries

The total cost of degradation due to quarries is assessed by summing up the loss in land

value around surveyed quarries, the loss in apartment value around surveyed quarries,

and the annual land value occupied by all quarries (Table 14). Accordingly, the average

total annual degradation cost due to quarries in Lebanon is 15.25 USD million or 0.10

percent of the GDP.

Table 14. Calculation of the total cost of degradation due to quarries

Parameter Lower bound

Higher bound

Loss in land value around surveyed quarries (US $ million) 8.1 9.6

Loss in apartment value around surveyed quarries (US $ million) 0.9 1.0

Annual land value occupied by all quarries (US $ million) 5.0 5.9

Total loss (US $ million) 14.0 16.5


This is estimated at USD 48 million. As an annual damage cost, this corresponds to USD

5‐6 million per year. In total, the annual damage cost of quarries is conservatively

estimated at USD 14‐16 million, or about 0.1 percent of GDP.

9 THE AVERTING BEHAVIOR METHOD

Actions are taken to reduce or avoid the consequences and costs of environmental

damage. The costs incurred due to these actions are considered equivalent to the costs

of environmental degradation. Averting behaviors may include, drinking bottled water

or purchasing water filters due to polluted water, frequent painting of dwellings due to

59

smoke emissions from a nearby factory, moving away from a polluted location, installing

air purifiers, staying indoors, installing soundproof walling to reduce noise, etc. In many

cases, several types of aversive expenditures are undertaken simultaneously. For

example, possible action in response to a noisy road may include installing double

glazing and moving to another area. Thus the total benefits are estimated by summing

up all expenditures.

The application of this technique differs with the type of pollution and the aversive

behavior adopted. A general methodology includes:

Step 1‐ Identification of the environmental hazard and the affected population

Monitoring equipment may be used to measure variables indicative of the

environmental hazard. Common sense should be adopted in defining the population at

risk.

Step 2‐ Observation of the responses of individuals

Survey design should avoid biased sample, strategic bias, and self‐selection. Public

expenditures should be indentified and included with the Willingness to Pay (WTP)

estimation.

Step 3‐ Measurement of the cost of taking action

One should understand why the individual is taking a certain action and if the chosen

course is enough to avoid the hazard.

Issues to consider when applying this methodology include:

− Some actions are difficult to monetize, such as moving house and leaving a familiar

neighborhood, thus the cost of the action is a minimum estimate.

− Some impacts have consequences with no possible averting actions, such as the

impact of air pollution on reduced visibility or the impact of air pollution on lake

acidification. Hence the cost of the action is not accurate or complete.

− Some goods provide additional non‐environmental benefits, for example, bottled

water tastes better, or air conditioning ameliorates room temperature. This needs to

be accounted for to avoid overestimation of benefits.

− Some are only partial substitutes for the environment. For example, double glazing

partially reduces noise, this discomfort may still occur. This should be included in the

analysis.

60

9.1 Case Study 1: The Cost of Pesticide Contamination of Drinking Water

Consider the cost of pesticide contamination of drinking water. Households may take a

variety of possible actions in order to reduce the risks faced by pollutants. Abdalla and

his colleagues have studied the responses of residents in Perkasie, a small town in

Pennsylvania, USA, to the chemical contamination of water supplies. In late 1987,

Trichloroethylene (TCE) was detected in well supplying water to the area. Levels of TCE

were as high as 35 parts per billion, well in excess of the Environmental Protection

Agency's limit 5 parts per billion. As there were no means of reducing the

contamination, water consumers were notified of the contamination in June 1988. No

solution had been implemented as of December 1989. A postal survey was undertaken

in September 1989 of a sample of 1,733 households in the town. Replies were received

from 761 respondents, a response rate of about 45 per cent. The questionnaire asked

for information on actions taken to avoid exposure to the chemical. These included

increased purchases of bottled water by those who had previously purchased it,

purchases by those who had not purchased it before, installation of home water

treatment systems, bringing in water from other sources and boiling water. On the basis

of the responses received, the costs of the actions were estimated. Because the water

treatment systems would last for longer than the expected period of chemical

contamination, only a proportion of this cost was included. The results are shown in

Table 15. As there is no clear logic for choosing the value to attach to the time spent on

averting behavior, the table shows two possible approaches.

Table 15. Costs of averting actions undertaken

Actions undertaken Low estimatea (USD) High estimateb (USD)

Increased purchase of bottled water 11,135 11,135

New purchases of bottled water 17,342 17,342

Home water treatment systemsc 4,691 4,691

Hauling water 12,513 34,013

Boiling water 15,633 64,135

Total cost 61,313 133,334 a Time valued at minimum wage rate (3.35 USD/hr) b Time valued at estimated hourly wage c Because such a system would last for longer than the contamination period, a proportion of the cost was included

The results should be regarded as a minimum estimate of the costs of the chemical

contamination. It is notable that, despite the requirement that households should be

notified, only 43 per cent of respondents were aware of the presence of TCE in their

water. It must be assumed therefore that expenditure would have been higher had

61

more people known about the contamination. It must also be assumed that the averting

behavior by households did not remove all consequences of the chemical

contamination. No allowance has been made for any possible ecological impacts which

may not have any impact on local consumers or which would not be affected by actions

taken within the home. The analysis does suggest that if the contamination could be

avoided for an expenditure of USD 60,000, this should be undertaken. However, this is

not to say that preventative measures should not be undertaken if they cost more than

this figure. The full costs of contamination may well by substantially higher than the

costs identified in this survey.

9.2 Case‐study 2: Consumption of Bottled Water in Lebanon

Lebanon’s population consumes a large quantity of bottled water mostly due to the

perception that municipal water is of a low quality. Water pollution and possible

contamination of municipal water in the distribution system has a cost to society.

According to the State of the Environment Report (SOER), bottled water expenditure

represent 0.60% of total per capita expenditure. The average price of one liter of bottled

water is 0.23 US$. Thus, bottled water consumption is about 115 liters per capita per

year (Table 16).

Table 16. Bottled water consumption in Lebanon

Parameter Unit Value

Per capita expenditures in Lebanon US$/capita/yr 4,465

Per capita bottled water expenditures in Lebanon % 0.60

Bottled water expenditure in Lebanon US$/capita/yr 26.8

Average price of bottled water in Lebanon US$/liter 0.23

Actual bottled water consumption in Lebanon Liter/capita/yr 115

Some consumption is due to taste and lifestyle preferences. This estimate is based on

bottled water consumption in Europe and the United States in the 1970s (prior to the

large increase in bottled water consumption in the 1980s and 1990s, widely believed to

be due to perceptions of inferior municipal water quality). The expected bottled water

consumption associated with preference is estimated in the Table 17. The expected

consumption is adjusted for GDP per capita differentials and price differentials between

several European countries (in the 1970) and Lebanon.

62

Table 17. Estimate of expected bottled water consumption in Lebanon if consumers perceived no health risk of potable municipal water

Unit Value

GDP per capita 2000 (Western Europe and USA) US$/capita 17,253 ‐ 27,750

GDP per capita 2000 in Lebanon US$/capita 3,875

Bottled water consumption in several European countries in 1970’s

Liter/capita/yr 30

Income elasticity of bottled water demand 0.25 to 0.4

Price elasticity of bottled water demand (“low”) ‐1.5 to ‐1.5

Price elasticity of bottled water demand (“High”) ‐2 to ‐2

Average price of bottled water in European countries US$/liter 0.3 to ‐0.3

Average price of bottled water in Lebanon US$/liter 0.23

Expected bottled water consumption in Lebanon “Low” Liter/capita/yr 30 ‐ 24

Expected bottled water consumption in Lebanon “High” Liter/capita/yr 34 ‐ 27

The cost of municipal water of inferior quality (in terms of bottled water consumption) is

the difference between actual bottled water consumption and the estimated

consumption associated with taste and lifestyle preferences. The cost is estimated at

US$82 ‐89 million per year, or around 0.5 percent of GDP (Table 18).

Table 18. Estimation of bottled water consumption to protect against risk

Parameter Value

Low High

Actual bottled water consumption in Lebanon (Liter/capita/yr) 115 115

Expected bottled water consumption in Lebanon “Low” (Liter/capita/yr) 30 24

Expected bottled water consumption in Lebanon “High” (Liter/capita/yr) 34 27

Average expected bottled water consumption in Lebanon (Liter/capita/yr) 32 26

Bottled water consumption to protect against risk (Liter/capita/yr) 83 89

Lebanese population in 2000 (million capita) 4.2

Total bottled water consumption to protect against risk (million liter/yr) 356 383

Total cost of bottled water consumption to protect against risk (million US$/yr) 82 88

% GDP 0.49 0.53

THE REVEALED PREFERENCE APPROACH:

GROUP EXERCISES

Sessions 7 & 8

Region

al Training Worksho

p on

: Th

e Co

st of E

nviron

men

tal D

egrada

tion

Metho

dology

63

SESSIONS 7 & 8

GROUP EXERCISES

The group exercise comprised of the following case‐studies: Ayubia National Park in Pakistan‐ The Travel Cost Method Himayatullah, 2003. Economic Valuation of the Environment and the Travel Cost Approach: The Case of

Ayubia National Park. The Pakistan Development Review 42: 4 Part II (Winter 2003) pp. 537–551 Non‐priced Forest Recreation Areas in Malaysia‐ The Travel Cost Method Garrod G. and Willis K.G. 2001. Economic Valuation of the Environment: Methods and Case Studies.

Edward Elgar Publishing, UK. Valuing Landscape and Amenity Attributes in Central England‐ Hedonic Pricing Garrod G. and Willis K.G. 2001. Economic Valuation of the Environment: Methods and Case Studies.

Edward Elgar Publishing, UK.

64


THE CONTINGENT VALUATION METHOD

Session 9

Region

al Training Worksho

p on

: Th

e Co

st of E

nviron

men

tal D

egrada

tion

Metho

dology

65

SESSION 9

10 THE CONTINGENT VALUATION METHOD

The Contingent Valuation Method (CVM) is the most widely used method for estimating

non‐use values. It is called “contingent” valuation, because it is contingent on simulating

a hypothetical market for the good in question. It involves directly asking individuals

how much they would be willing to pay (WTP) to preserve or use a given good or service

or the amount of compensation they would be willing to accept (WTA) to forgo specific

environmental services. The CVM can be used to estimate economic values for all kinds

of ecosystem and environmental services, for both use and non use values. The CVM has

been applied to estimate the values of landscape, recreation, beaches, water quality,

nature conservation, endangered species, visibility and air quality, etc. Yet, the CVM is

the most controversial of the non‐market valuation methods, whereby many

economists, psychologists and sociologists, for many different reasons, do not believe

that the dollar estimates that result from CV are valid. In addition, many jurists and

policy‐makers will not accept the results of CV. However, studies have shown that a

carefully composed and tested study, where the circumstances are not too distant from

the experience of the respondent and the issue is not too emotive, can produce answers

of value.

10.1 Steps in a CVM Procedure

10.1.1 Setting up the hypothetical market

This step involves devising a convincing contingent valuation scenario to demonstrate

that respondents are actually stating their values for these services when they answer

the valuation questions. A reason for the good or service needs to be established and

pictorial aids could be of use.

10.1.2 Obtaining bids Bids, or the people’s WTP values, are obtained through a questionnaire survey. Possible

bid vehicles include income taxes, property taxes, value added or sales tax, utility bills,

entry fees, and payments into a trust fund. Yet, not all bid vehicles are viable options in a

given situation. The chosen bid vehicle should have a plausible connection with the

valued amenity and should be perceived as fair and equitable. People have different

views on the acceptability of different types of taxes.

Focus groups should precede surveys. They provide insight on the respondents’ likely

understanding of and attitude towards the issue being investigated. They also provide

66

valuable information in framing and designing a CV study and questionnaire survey.

Focus groups are usually drawn from a cross‐section of the population, stratified by

social class. Around 8 to 10 participants in a focus group meet for one to two hours to

discuss their understanding of the context of the good, the good itself, its value, who

should provide it, how it should be paid for, whether they should contribute, and how

much they are willing to pay. However, care should be taken when handling focus

groups due to many reasons. For instance, responses may be influenced by the person

conducting the focus group. In addition, focus group participants have a longer time to

think about the issue than in a typical CV survey and they have more information to base

their judgment on. Furthermore, individuals behave differently in group situations

compared to situations when they are alone.

As mentioned previously, bids are obtained through a questionnaire survey and an

elicitation format where respondents are asked to state their maximum WTP to increase

quantity/ prevent quantity decrease of an environmental good or their minimum WTA

compensation to forgo an increase in the quantity/ accept less of the good. Various bid

elicitation methods may be used including, bidding games, payment cards, open‐ended

questions, and close‐ended questions. Bidding games are when respondents are given

progressively higher bids until they reach their maximum WTP. The payment card

method involves providing a range of values to the respondent on a card, and asking him

to choose from them. In open‐ended questions, no value is specified, and respondents

are asked to report their maximum WTP. However, this method is not recommended as

respondents that have no prior experience in purchasing the good in question may find

it difficult to respond. Close‐ended questions can be asked under various formats. One is

the dichotomous choice referendum, where a single amount is offered and respondents

are asked to agree or disagree. In the double‐bounded referendum, respondents who

disagree are offered a lower amount and those who agree are offered a higher amount.

This method is highly recommended. In the trichotomous choice, respondents are

offered three choice: ‘yes’, ‘no’, and ‘indifferent’.

In the questionnaire survey, three sets of information are obtained from respondents

(Figure 25).

67

Figure 25. Information obtained in questionnaire surveys

Note that data on use, preferences and substitutes should be collected at the beginning

of the questionnaire. Respondents must be reminded of their budget constraints when

eliciting their bids.

Questionnaires are administered in a number of ways, including face‐to‐face interviews,

self‐filled questionnaires, telephone interviews, and mail shots. Face‐to‐face

interviewing is the usual method adopted. It allows the definition and explanation of the

good more thoroughly and it minimizes non‐response. However, it is expensive to

conduct. Self‐fill questionnaires involve questionnaires left at recreation sites for visitors

or in public places. While this is a cheap data collection method, it can only be used with

questions that can be easily comprehended and it often yields a low response rate. Mail

shots are used where the hypothetical market is easily explained to respondents. This

method is most appropriate when respondents are widely scattered over space, and

when they have expert knowledge and interest in the good. Only questions that are

easily comprehended may be used in mail shots. Finally, telephone interviewing is often

faced with problems due to the absence of visual cues and due to the difficulty in

maintaining respondents’ attention. A combined telephone interview and mail shot is

usually recommended as it can be cost‐effective and it can increase the response rate.

The telephone secures the respondent’s interest and the mail follow‐up provides visual

and questionnaire material.

Deciding on a sample size for the questionnaire is a crucial step, as it determines the

precision of the sample statistics used as mean WTP/ WTA. The larger the sample, the

smaller the variation in mean WTP measured by standard error and confidence intervals.

Mitchell and Carson (1989) devised a system to determine sample size based on choice

of acceptable deviation between the ‘true’ and estimated WTPs. For a deviation of 5%,

95% of the time, a sample of 6,000 is needed; while for a deviation of 20%, 90% of the

time, a sample of 286 is needed Mitchell and Carson argue that a sample greater than

68

600 is needed for applications seeking to evaluate policy. This ensures a deviation of

15%, 95% of the time.

10.1.3 Estimating mean and median WTP/WTA

WTP means, medians, modes, trimmed and modified estimators, and standards of

deviation can be found from individual bids. Mean WTP, or trimmed or modified

estimators based on mean WTP are the most appropriate, as they represent cardinal

measures of the utility that individuals derive from the good. Median WTP is

recommended because it is unaffected by large bids and because it is lower than the

mean WTP and may underestimate the value. As for the trimmed estimator, it involves

trimming the top and bottom 5% or 10% of the distribution of WTP observations.

However, this may result in the omission of some true estimates of WTP. The modified

estimator is considered to provide the truest value, as it identifies and excludes biased

and illegitimate responses by a series of questions included in the questionnaire.

Probit, logit and random utility models can be used for close‐ended referendum bids.

Bid curves can also be estimated by regressing WTP against socio‐economic variables.

Differentiating bid curves (dWTP/dV) provides the demand curve for the good and the

consumer surplus can thus be calculated as the area under the curve.

WTPi = f(Yi, Vi, Pi, Si, Ei)

Where, Y = income level; V = visits; P = preferences = S = substitutes; E = socio‐economic

variables (age, education, etc.)

10.1.4 Aggregating WTP or WTA amounts

Mean WTP/ WTA from the sample survey are aggregated across the total population.

TOTAL VALUE of the good/ service = (mean WTP) × (# of population units)

While mean WTP/ WTA may be modest for non‐use benefits, the populations over

which they are aggregated can be large.

10.1.5 Assessing the validity of CV studies The validity of the CV study is assessed by examining three aspects: content validity,

criterion validity, and construct validity. Content validity examines the appropriate

framing of the study and questions asked in relation to the good being valued. Criterion

validity involves the comparison of CV estimates with actual market or simulated market

69

experience. As for construct validity, it examines the convergence between a CV

measure and other methods such as travel cost and hedonic price measures of the value

of the same good. It measures the extent to which the findings of the CV study are

consistent with theoretical expectations.

10.2 Considerations and biases When applying a CV study, the following need to be considered:

− Before designing the survey, learn as much as possible about how people think

about the good or service in question.

− Determine the extent of the affected populations or markets for the good or service

in question, and choose the survey sample based on the appropriate population.

− The choice scenario must provide an accurate and clear description of the change in

environmental services associated with the event, program, investment, or policy

choice under consideration. Convey this information using photographs, videos, or

other multi‐media techniques, as well as written and verbal descriptions.

− Specify whether comparable services are available from other sources, when the

good is going to be provided, and whether the losses or gains are temporary or

permanent.

− The respondent must believe that if the money was paid, whoever was collecting it

could effect the specified environmental change.

− Respondents should understand the frequency of payments required, for example

monthly or annually, whether or not the payments will be required over a long

period of time in order to maintain the quantity or quality change, and who would

have access to the good and who else will pay for it, if it is provided.

− Thoroughly pre‐test the valuation questionnaire for potential biases.

− Include validation questions in the survey to verify comprehension and acceptance

of the scenario and to elicit socioeconomic and attitudinal characteristics of

respondents.

− Make sure that survey results are analyzed and interpreted by professionals before

making any claims about the resulting dollar values.

The CVM is associated with many biases that need to avoided or minimized. These

include:

− Hypothetical bias: occurs since individuals do not have to pay their stated amounts.

This will cause them to overstate their true WTP. This bias is not very significant.

− Embedding effect: occurs when WTP is lower when it is valued as part of a more

inclusive good. This is attributed by some to the existence of substitutes.

70

− Strategic bias: occurs when an individual deliberately overstates/understates the bid

to influence a particular outcome. This bias is difficult to detect and test but, it is not

significant.

− Bid vehicle bias: occurs when respondents give different WTP amounts, depending

on the specific payment vehicle chosen. For example, an individual disliking taxation

might understate his WTP. In such cases, a neutral vehicle such as a trust fund is

recommended.

− Starting point bias: occurs when a start‐off amount is misinterpreted by the

respondent as a cue for an appropriate WTP range. Extensive pre‐tests may minimize

this bias.

− Information bias: occurs when insufficient information makes it difficult on the

respondent to give a proper valuation, especially if the issue is new to him. Too much

information will be a definite source of bias

− Part‐whole bias: occurs when respondents asked to valuate a given asset and then to

valuate a part of it tend to give a similar answer. This is minimized by reminding

respondents of their budget constraints and by restricting valuation to whole goods

rather than parts of goods.

− Non‐response bias: occurs because people with interest in the subject are more

likely to respond. Non‐respondents are likely to have, on average, different values

than individuals who do respond. This is minimized by using questions that are easy

to answer.

Biases may be minimized by including certain questions as part of the survey. The

following are show cards prepared to elicit legitimate and illegitimate reason for WTP

answers.

71

Show card to elicit legitimate and illegitimate reason for NOT being WTP towards low flow alleviation in rivers

Question Bias

a. I cannot afford to pay more water charges at present

b. I have no interest in having different flow levels in rivers

c. I would not pay anymore in water charges but I would be prepared to pay by some other means of payment

Payment vehicle

d. Someone else should pay rather than me Strategic

e. The water company should pay not customers Bid vehicle

f. Low levels in rivers are not a problem

g. I require more information to answer this question

h. Other reasons. Please specify

i. Don’t know

j. Refused to answer

Show card to elicit legitimate and illegitimate reason for being WTP towards low water quality

Question Bias

a. It was the most I could afford to pay

b. Rivers and beaches are important for recreation and I am happy to pay to ensure that they are well looked after

c. I would pay this much each year to ensure that rivers and beaches are protected for future generations

d. Rivers and beaches are important for wildlife and ecology and I am happy to pay to ensure that they are well looked after.

e. I wanted to show my support for environmental improvement in general Strategic

f. It’s an important issue and by saying I’d pay such a large sum each year I hope to get something done about it

Strategic

g. I’m very concerned about this issue and although I’m not sure I could afford to pay this much each year I wish I could

Hypothetical

h. Rivers and beaches are important for a number of reasons and I am happy to pay to ensure that they are well looked after

i. Other reason. Please specify;

j. Don’t know

k. Refuse to answer

72


Advantages associated with the CVM include:

− It is the most widely accepted method for estimating total economic value including

use values and all types of non‐use values

− It is straightforward and highly flexible, whereby it can used to estimate the

economic value of virtually anything

− It requires few theoretical assumptions

− The nature and results of CV studies are easy to analyze and describe. Dollar values

can be presented in terms of a mean or median value per capita or per household, or

as an aggregate value for the affected population.

− A great deal of research is being conducted to improve the methodology, make

results more valid and reliable, and better understand its strengths and limitations.

Issues and limitations associated with the CVM include:

− There is considerable controversy over whether CVM adequately measures people's

willingness to pay for environmental quality. CV assumes that people understand the

good in question and will reveal their preferences in the contingent market just as

they would in a real market. However, most people are unfamiliar with placing dollar

values on environmental goods and services and may not have an adequate basis for

stating their true value.

− Expressed answers to a willingness to pay question may be biased.

− Respondents may make associations among environmental goods that the

researcher had not intended. For example, if asked for willingness to pay for

improved visibility (through reduced pollution), the respondent may actually answer

based on the health risks that he or she associates with dirty air.

− WTA very significantly exceeds WTP. This result may invalidate the CVM approach,

showing responses to be expressions of what individuals would like to have happen

rather than true valuations.

− The “ordering problem”: in some cases, people’s expressed willingness to pay for

something has been found to depend on where it is placed on a list of things being

valued.

− Difficulty to validate externally the estimates of non‐use values.

− When conducted appropriately, contingent valuation methods can be very expensive

and time‐consuming, because of the extensive pre‐testing and survey work.

− Many people, including jurists policy‐makers, economists, and others, do not believe

the results of CV.

73

10.4 Illustration‐Mining on public land

A remote site on public land provides important habitat for several species of wildlife.

The management agency in charge of the area must decide whether to issue a lease for

mining at the site. For this purpose, they must weigh the value of the mining lease

against the wildlife habitat benefits that may be lost if the site is developed. Non‐use

values are the largest component of the value for preserving the site because few

people actually visit it, or view the animals that rely on it for habitat. This necessitates

the use of the CVM.

Step 1

Define the valuation problem by determining what services are being valued and who

the relevant population is. In this case, the resource to be valued is a specific site and

the services it provides are primarily wildlife habitat. Because the land is federally

owned public land, the relevant population would be all citizens of the U.S.

Step 2

Make preliminary decisions about the survey itself: whether it will be conducted by mail,

phone or in person; how large the sample size will be; who will be surveyed, etc. The

answers will depend, among other things, on the importance of the valuation issue, the

complexity of the question being asked, and the size of the budget. The researchers

decided to conduct a mail survey, as they want to survey a large sample, over a large

geographical area. They are asking questions about a specific site and its benefits, which

should be relatively easy to describe in writing in a relatively short survey

Step 3

The actual survey design may take six months or more to complete. It is accomplished in

several steps. It starts with initial interviews and/or focus groups with the types of

people who will be receiving the final survey, in this case the general public. The

researchers would ask general questions about peoples’ understanding of the issues

related to the site, whether they are familiar with the site and its wildlife, and whether

and how they value this site and the habitat services it provides. In later focus groups,

the questions would get more detailed and specific in order to help develop specific

questions for the survey and to decide what kind of background information is needed

and how to present it. People might need information on the location and

characteristics of the site, the uniqueness of species that have important habitat there,

and whether there are any substitute sites that provide similar habitat. The researchers

would also want to learn about peoples’ knowledge of mining and its impacts, and

whether mining is a controversial use of the site. If people are opposed to mining, they

74

may answer the valuation questions with this in mind, rather than expressing their value

for the services of the site. At this stage, different approaches to the valuation question

and different payment mechanisms would be tested. Questions that can identify any

“protest” bids or other answers that do not reveal peoples’ values for the services of

interest would also be developed and tested at this stage.

After a number of focus groups, pretesting of the survey is started. The survey should be

pretested with as little interaction with the researchers as possible. Pre‐testing will

continue until a survey is developed that people seem to understand and answer in a

way that makes sense and reveals their values for the services of the site.

Step 4

At this stage, actual survey implementation takes place. The survey sample is selected.

The sample should be a randomly selected sample of the relevant population, using

standard statistical sampling methods. For instance, a mailing list of randomly sampled

U.S. Citizens may be obtained and a standard repeat‐mailing and reminder method may

be used to get the greatest possible response rate for the survey.

Step 5

The results are compiled, analyzed and reported. Data must be entered and analyzed

using statistical techniques appropriate for the type of question. The researchers also

attempt to identify any responses that may not express the respondent’s value for the

services of the site. They can deal with possible non‐response bias in a number of ways.

The most conservative way is to assume that those who did not respond have zero

value.

Step 6

The final step involves estimating the average value for an individual or household in the

sample, and extrapolating this to the relevant population in order to calculate the total

benefits from the site. If the mean willingness to pay is $.10 per capita, the total benefits

to all citizens would be $26 million.

10.5 Sample application 1‐ Mono Lake

Reduced water flows to Mono Lake affect food supplies for nesting and migratory birds.

The State of California Water Resources Control board has to decide about the water

quantity to be allocated to Los Angeles from sources flowing into Mono Lake. A

contingent valuation study was conducted to measure the use and non‐use values of

citizens in California households for increased water flows in Mono Lake. An initial mail

75

survey was conducted where residents of California were told that, according to

biologists, the higher flows to the lake were needed to maintain food supplies for

nesting and migratory birds. Residents were then asked whether they would pay more

on their water bill for higher cost replacement water supplies, so that natural flows

could once again go into Mono Lake. According to the respondents, average WTP per

household was $13 per month or $156 per year. The total benefits exceeded the $26

million cost of replacing the water supply by a factor of 50. As follow‐up to this survey,

the State of California hired a consulting firm to perform a more detailed CV survey. The

new survey involved the use of photo‐simulations showing what the lake would look like

at alternative water levels. It gave detailed information about effects of changing lake

levels on different bird species. The survey was conducted over the telephone with

people who had been mailed information booklets with maps and photo‐simulations.

Survey respondents were asked how they would vote in a hypothetical referendum

regarding Mono Lake. This study showed that the benefits of a moderately high (but not

the highest) lake level were greater than the costs. Accordingly, the California Water

Resources Control Board reduced Los Angeles’ water rights by half, from 100,000 acre

feet to about 50,000 acre feet, to allow more flows into Mono Lake.

10.6 Case‐study‐ Exxon Valdez Oil Spill At 12.04 a.m. on 24 March 1989, the oil tanker Exxon Valdez ran aground on Bligh Reef

in Prince William Sound, Alaska. It was carrying crude oil from wells on Alaska's North

Slope, brought to Valdez through the Tans‐Alaska pipeline. From Valdez, the crude is

carried by tankers to refineries in the southern United States. The Exxon Valdez was

carrying more than 50 million gallons of crude oil, of which approximately 11 million

gallons poured into the Sound. This was the largest spill in United States history. By

August, the oil had moved across nearly 10,000 square miles of water and about 1,600

miles of the Sound's convoluted shoreline was heavily oiled.

The oil had a massive impact on wildlife in the area, killing many birds and marine mammals. Over 20,000 dead birds were recovered, mostly murres but also many other

species including 100 bald eagles. The total numbers killed were probably three to six

times the numbers recovered. About 2,650 sea otters died, probably about 40 per cent

of the population in the affected area. Seals and many other species were also killed or

damaged by the spill, including plants and microorganisms. However, none of these

losses threatened the extinction of the species involved. It was expected that bird and

mammal populations would recover to their pre‐spill levels in about three to five years.

76

The spill had a variety of impacts on human values. The Sound is important for

commercial and recreational fishing and tourism, and these uses were severely damages

by the oil spill. These types of impact could be valued relatively easily. However, we may

anticipate that the area also has significant non‐use values: existence, option and

bequest values. As part of an assessment of the total damages arising from the oil spill, it

would be important to find some means of estimating the extent of these non‐use

values. In connection with legal action taken by the State of Alaska against Exxon

Corporation and related companies, a contingent valuation study was commissioned

from a number of well‐known economists working in the field of valuation.

The study involved a survey of residents across the United States. Alaska was excluded

from this as the aim was to focus on non‐use values. In principle, the survey should seek

to estimate the population's willingness to accept compensation for the damage arising

from the Exxon Valdez oil spill. However, because of the difficulty involved in designing

surveys of this sort, it was decided to adopt a willingness to pay approach. The valuation

question was based on a hypothetical proposal for a scheme to prevent future oil spills

of the sort which had just been experienced. Respondents in the survey were asked to

indicate whether or not they would vote for a proposal to provide escort ships to

accompany oil tankers through the Sound. The ships would carry special booms which

could be put into place immediately any oil spill occurred so as to hold the oil within a

confined area. The spilt oil could then be skimmed off the surface and taken away for

safe disposal. This system has been used successfully by the Norwegians in the North

Sea. Without this scheme, respondents were told that over the next ten years, another

large oil spill can be expected to occur in Prince William Sound. The scheme would be

paid for from a special tax on oil company profits and from a single tax on all

households.

In carrying out the survey, after investigating respondents' prior knowledge of the issue,

interviewers provided respondents with information on the Exxon Valdez oil spill and its

impacts. The basic valuation question put to each respondent was whether or not he or

she would vote for a proposal to implement the scheme, given a specified level of a

single one‐time tax on each household. The initial tax values were set at $10, $30, $60

and $120 for different households. If respondents answered that they would be

prepared to vote for the scheme at this level of tax, the amount was raised and the

question asked just once more. If they refused to vote for the scheme, the amount was

lowered and the question was asked just once more. The survey also collected

information on interests in environmental issues, the composition of the respondent's

household, education and incomes.

77

Excerpts of the administered questionnaire: The only mammals killed by the spill were sea otters and harbour seals. This card shows information about what happened to Prince William Sound. According to scientific studies, about 580 otters and 100 seals in the Sound were killed by the spill. Scientists expect the population size of these two species to return to normal within a couple of years after the spill. Many species of fish live in these waters. Because most of the oil floated on the surface of the water, the spill harmed few fish. Scientific studies indicate there will be no long‐term harm to any of the fish populations. #2. Of course, whether people would vote for or against the escort ship program depends on how much it will cost their household At present, government officials estimate the program will cost your household a total of $______. You would pay this in a special one‐time charge in addition to your regular federal taxes. This money would only be used for the program to prevent damage from another oil spill in Prince William Sound. If the program cost your household a total of $______, would you vote for or against it?” #3. What if the final cost estimate showed that the program would cost your household a total of $_______. Would you vote for or against the program? #4. What is it about the program that made you willing to pay something for it? #5. Before the survey, did you think the damage caused by the Valdez oil spill was more serious than was described to you, less serious, or about the same as described? #6. Is anyone in your household an angler, birdwatcher, backpacker, or environmentalist? #7. This card shows amounts of yearly incomes. Which category best describes the total income from all members of your family before?

A total of 1,043 interviews were successfully completed, achieving a response rate of 75

per cent. The proportions of respondents indicating that they would vote for the

proposed scheme at the alternative tax levels are shown in Table 19. As would be

expected, the proportion falls as the specified cost of the scheme increases, although

there is little difference between the $30 and $60 questions.

78

Table 19. Positive response to alternative tax levels

Questionnaire version

Initial tax level per household (USD)

Percent of respondents willing to

pay taxes

A 10 67

B 30 52

C 60 51

D 120 34

From these results, it is possible to estimate statistically a figure for the median

willingness to pay. This is midpoint in the distribution, so that there is an equal number

of a figure above and below the median. The preferred estimate was for $31 per

household. A figure of $94 was estimated as the mean willingness to pay, but the nature

of the questions asked meant that this figure is unreliable. Just over one‐third of

respondents were not willing to vote for the scheme at either of the prices offered to

them and of these one third indicated that a reason for this negative response was that

they felt that the oil companies should pay. Varying the assumptions used in the analysis

either tended to increase the estimated median willingness to pay or else had little

effect on it.

It is possible to extrapolate the results of the survey in order to estimate the total value

for the non‐use values lost in the United States as a whole. This figure is obtained by

multiplying the total number of households by $31 and produces a figure of $2.8 billion,

with a confidence interval of $2.4 billion to $3.2 billion.

This analysis was at the centre of a fierce debate. If Exxon and the other companies were

found to be liable for the damage caused, they would face an enormous bill for the non‐

use values alone. The debate focused on the role of contingent valuation in estimating

these sorts of values. In this context a panel of distinguished economists was established

by the National Oceanic and Atmospheric Administration, including two Nobel Prize

winners, to examine the role of contingent valuation in the valuation of non‐use values

for the purposes of assessing the damages from oil spills. The panel concluded that

contingent valuation studies can convey useful information that is sufficiently reliable

from a starting point for the process of assessing damages in the courts. However, they

had reservations about the ways in which contingent valuations have often been

undertaken and set out a series of guidelines for the way in which they believed that

they should be undertaken.

THE DISCRETE CHOICE METHOD &

THE BENEFIT TRANSFER METHOD

Session 10

Region

al Training Worksho

p on

: Th

e Co

st of E

nviron

men

tal D

egrada

tion

Metho

dology

79

SESSION 10

11 DISCRETE CHOICE METHOD

Contingent choice, also referred to as conjoint analysis, was developed in the fields of

marketing and psychology to measure preferences for different characteristics or

attributes of a multi‐attribute choice. It asks the respondent to state a preference

between one group of environmental services or characteristics, at a given price or cost

to the individual, and another group of environmental characteristics at a different price

or cost. It is especially suited to policy decisions where a set of possible actions might

result in different impacts. For example, improved water quality in a lake will improve

the quality of several services provided by the lake, such as drinking water supply,

fishing, swimming, and biodiversity. While contingent choice can be used to estimate

dollar values, the results may also be used to simply rank options, without focusing on

dollar values.

The contingent choice method is similar to contingent valuation, whereby it involves

asking people to make choices based on a hypothetical scenario. Furthermore, it can be

used to estimate economic values for any ecosystem or environmental service and it can

be used to estimate non‐use as well as use values. Yet, the contingent choice method

differs from contingent valuation in that it requires people to evaluate several

alternatives separately, it does not directly ask people to state their values in dollars,

and the values are inferred from the hypothetical choices or tradeoffs that people make.

There are a variety of formats for applying contingent choice methods. Contingent

rating, contingent ranking and paired rating are summarized in Table 20, and choice

modeling is detailed in the section below.

80

Table 20. Types of contingent choice formats Contingent Rating − Respondents presented with a set of attributes associated with each alternative

− Respondents requested to rate their preference for several alternatives on a scale

− Ratings regressed against attributes − Marginal rate of substitution between an attribute and its price provides an estimate of the value of the attribute

− Summing up all values provides the aggregate WTP for the environmental value Contingent Ranking

− Respondents asked to rank all alternatives from least preferred to most preferred− Analysis similar to contingent rating − Rankings converted to rating scale and analyzed with multiple regression techniques

− Other measures such as probit, or logit analysis may be used Paired Rating − Respondents presented with successive sets of two choices and asked to rate the

difference between them in terms of preference on a scale − Data analyzed using multiple regression, probit, or logit models

Choice modeling − Respondents presented with a series of alternatives, each defined by a set of attributes and containing three or more resource use options

− Attributes varied across alternatives − Respondents to choose preferred alternative − More flexible and versatile but requires complex survey designs

11.1 Choice experiments

Choice experiments are used to examine the response of the individual to changes in the

attributes of the scenario as well as the scenario as a whole. They allow breaking down

the relevant attributes of the situation and determining preferences over attributes and

they allow for more flexibility than CVM. Choice experiments attempt to identify the

utility the individuals have for the attributes of the goods and services by examining the

tradeoffs that they make between them when making choice decisions. Steps in the

choice modeling experiment are summarized in and detailed below:

1. Identify the good or service to be investigated

2. Identify key attributes and determine the attribute levels to be used

The initial screening of the attributes is a crucial stage in study design. The attributes

should be familiar and relevant to respondents and the attribute levels should be

measureable using quantitative or qualitative scales. Attributes may be portrayed

verbally or pictorially, etc. Defining an appropriate number of attributes is important

whereby too many attributes burden the respondents and too few cause problems

with estimation and reliability. Pre‐testing and focus groups are helpful in defining

attributes and determining their numbers.

3. Develop an appropriate experimental design for profiles

This involves the specification of a factorial or fractional factorial experimental

design to estimate the utility for the good in question. An orthogonal main‐effects

plan sampled from the complete factorial design is used to select the profiles to be

81

used in the choice experiment. Procedures in computer packages such as SAS and

SPSS may be used to create an orthogonal matrix based on the attribute levels

specified by the researcher

4. Design questionnaire survey and incorporate choice experiments. The types of

questionnaires that may be adopted along with a description of each are presented

in (Figure 26).

Face‐to‐face interviews − The usual method adopted − Allow the definition and explanation of the good

more thoroughly − Minimize non‐response − Expensive to conduct

Self‐fill questionnaires − Questionnaires left at recreation sites for

visitors or in public places − Low response rate − Cheap data collection − Only questions that can be easily comprehended

may be used Telephone interviews − Problems due to absence of visual cues − Problems due to difficulty in maintaining

respondents attention − A combined telephone interview and mail shot

can be cost‐effective and increase response rate − Telephone secures respondent’s interest − Mail follow‐up provides visual and questionnaire

material

Mail shots − Used where hypothetical market easily

explained to respondents − Most appropriate when respondents − Are widely scattered over space − Have expert knowledge, interest in the good,

etc. − Only questions that are easily comprehended

may be used

Figure 26. Types of survey questionnaires

5. Perform pre‐tests and undertake survey

6. Analyze choices made and determine trade‐offs made by respondents. Random

utility theory is used to model the choices as a function of attribute levels, based on

the hypothesis that individuals make choices based on the attributes of the

alternatives along with some degree of randomness. Following repeated

observations of choice, one can examine how the levels of various attributes affect

the probability of choice. An assumption of normality leads to the binary probit

model, while an assumption of a Gumbel distribution means that the multinomial or

Mother Logit can be employed.

7. Calculate welfare measures

8. Aggregate over population of relevance

Whatever format is selected, choices that respondents make are statistically analyzed

using discrete choice statistical techniques, to determine the relative values for the

different characteristics or attributes. If one of the characteristics is a monetary price,

then it is possible to compute the respondent’s willingness to pay for the other

characteristics.

82

A good contingent choice study will consider the following:

− Before designing the survey, learn as much as possible about how people think

about the good or service in question. Consider people’s familiarity with the good or

service, as well as the importance of such factors as quality, quantity, accessibility,

the availability of substitutes, and the reversibility of the change.

− Determine the extent of the affected populations or markets for the good or service

in question, and choose the survey sample based on the appropriate population.

− The choice scenario must provide an accurate and clear description of the change in

environmental services associated with the event, program, investment, or policy

choice under consideration. If possible, convey this information using photographs,

videos, or other multi‐media techniques, as well as written and verbal descriptions.

− The nature of the good and the changes to be valued must be specified in detail, and

it is important to make sure that respondents do not inadvertently assume that one

or more related improvements are included.

− The respondent must believe that if the money was paid, whoever was collecting it

could effect the specified environmental change.

− Respondents should be reminded to consider their budget constraints.

− Specify whether comparable services are available from other sources, when the

good is going to be provided, and whether the losses or gains are temporary or

permanent.

− Respondents should understand the frequency of payments required, for example

monthly or annually, whether or not the payments will be required over a long

period of time in order to maintain the quantity or quality change, and who would

have access to the good and who else will pay for it, if it is provided.

− In the case of collectively held goods, respondents should understand that they are

currently paying for a given level of supply. The scenario should clearly indicate

whether the levels being valued are improvements over the status quo, or potential

declines in the absence of sufficient payments.

− If the household is the unit of analysis, the reference income should be the

household’s, rather than the respondent’s, income.

− Thoroughly pre‐test the questionnaire for potential biases. Test different ways of

asking the same question and test whether the question is sensitive to changes in

the description of the good or resource being valued.

− Conduct post‐survey interviews to determine whether respondents are stating their

values as expected.

83

− Include validation questions in the survey to verify comprehension and acceptance

of the scenario and to elicit socioeconomic and attitudinal characteristics of

respondents.


The main advantages of the contingent choice method include:

− It can be used to value the outcomes of an action as a whole, as well as the various

attributes or effects of the action.

− It does not ask the respondent to make a tradeoff directly between environmental

quality and money. The tradeoff process may encourage respondent introspection

and make it easier to check for consistency of responses. Respondents may be able

to give more meaningful answers to questions about their behavior (i.e. they prefer

one alternative over another), than to questions that ask them directly about the

dollar value of a good or service or the value of changes in environmental quality.

− Respondents are generally more comfortable providing qualitative rankings or

ratings of attribute bundles that include prices, rather than dollar valuation of the

same bundles without prices.

− Even if the absolute dollar values estimated are not precise, the relative values or

priorities elicited by a contingent choice survey are likely to be valid and useful for

policy decisions.

− It minimizes many of the biases that can arise in open‐ended contingent valuation

studies where respondents are presented with the unfamiliar and often unrealistic

task of putting prices on non‐market amenities.

− It has the potential to reduce problems such as expressions of symbolic values,

protest bids, and some of the other sources of potential bias associated with

contingent valuation.

The main issues and limitations that are associated with a contingent valuation process

include:

− Respondents may find some tradeoffs difficult to evaluate, because they are

unfamiliar.

− The respondents’ behavior underlying the results of a contingent choice study is not

well understood. Respondents may resort to simplified decision rules if the choices

are too complicated, which can bias the results of the statistical analysis.

− If the number of attributes or levels of attributes is increased, the sample size and/or

number of comparisons each respondent makes must be increased.

− When presented with a large number of tradeoff questions, respondents may lose

interest or become frustrated.

84

− Contingent choice may extract preferences in the form of attitudes instead of

behavior intentions.

− By only providing a limited number of options, it may force respondents to make

choices that they would not voluntarily make.

− Contingent ranking requires more sophisticated statistical techniques to estimate

willingness to pay.

− Translating the answers into dollar values, may lead to greater uncertainty in the

actual value that is placed on the good or service of interest.

− Validity and reliability for valuing non‐market commodities is largely untested.

11.3 Illustration‐ Mining in public land

There is a remote site on public land that provides important habitat for several species

of wildlife. The management agency in charge must decide whether to issue a lease for

mining at the site. Suppose that there are several possible options for preserving and/or

using the site, including allowing no mining and preserving the site as a wilderness

habitat area, or specifying various levels and locations for the mining operation, each of

which would have different impacts on the site. The contingent choice method was

selected because the outcomes of several policy options need to be valued and because

non‐use values are the largest component of the value for preserving the site. Thus,

TCM will underestimate the benefits of preserving the site. The CVM might also be used,

but the survey questions might become very complicated.

Contingent choice and contingent valuation have very similar application. The main

differences are in the design of the valuation question(s), and the data analysis.

Step 1

Define the valuation problem by determining exactly what services are being valued,

and who the relevant population is. In this case, the resource to be valued is a specific

site and the services it provides is wildlife habitat. Because it is federally owned public

land, the relevant population would be all citizens of the U.S.

Step 2

Make preliminary decisions about the survey, including whether it will be conducted by

mail, phone or in person, how large the sample size will be, who will be surveyed, and

other related questions. In this case, the researchers decided to conduct a mail survey

since it will be administered to a large sample over a large geographical area. Questions

about a specific site and its benefits should be relatively easy to describe in writing.

85

Step 3

The actual survey design is accomplished in several steps. It starts with initial interviews

and/or focus groups with the types of people who will be receiving the final survey, in

this case the general public. In the initial focus group, the researchers would ask general

questions about peoples’ understanding of the issues related to the site, whether they

are familiar with the site and its wildlife, and whether and how they value this site and

the habitat services it provides. In later focus groups, the questions would get more

detailed and specific. Different approaches to the choice questions are tested. Each

choice might be described in terms of the site’s ability to support each of the important

wildlife species. People will be making tradeoffs among the different species that might

be affected in different ways by each possible choice of scenario. After a number of

focus groups, pretesting of the survey is started. The survey should be pretested with as

little interaction with the researchers as possible. People would be asked to assume that

they’ve received the survey in the mail and to fill it out. Then the researchers would ask

respondents about how they filled it out, and let them ask questions about anything

they found confusing. A mail pretest might be conducted. This process is continued until

a survey is developed that people seem to understand and answer in a way that makes

sense and reveals their values for the services of the site.

Step 4

At this stage, actual survey implementation takes place. The survey sample is selected.

The sample should be a randomly selected sample of the relevant population, using

standard statistical sampling methods. For instance, a mailing list of randomly sampled

U.S. Citizens may be obtained and a standard repeat‐mailing and reminder method may

be used to get the greatest possible response rate for the survey.

Step 5

The results are compiled, analyzed and reported. The statistical analysis for contingent

choice is often more complicated than that for contingent valuation requiring the use of

discrete choice analysis methods to infer willingness to pay from the tradeoffs made by

respondents. The researchers need to estimate the average value for each of the

services of the site, for an individual or household, and then extrapolate to the relevant

population in order to calculate the total benefits from the site under different policy

scenarios. The average value for a specific action and its outcomes can also be

estimated, or the different policy options can be ranked in terms of peoples’ preferences

The results of the survey might show that the economic benefits of preserving the site

by not allowing mining are greater than the benefits received from allowing mining. The

86

mining lease might not be issued, unless other factors override these results. The results

might indicate that some mining scenarios are acceptable, in terms of economic costs

and benefits. The different options should be ranked and the most preferred option

selected.

11.4 Case‐application: Landfill Siting in Rhode Island With its primary landfill nearing capacity, the State of Rhode Island was faced with the

need to choose locations for new landfills. Besides technical considerations, the State

wanted to address the social and economic tradeoffs and values related to the location

of a landfill to avoid some of the controversy associated with landfill siting. A contingent

choice, paired comparison, survey was conducted. The survey asked Rhode Island

residents to choose between pairs of hypothetical sites and locations for a new landfill,

described in terms of their characteristics. The site comparisons described the natural

resources that would be lost on a hypothetical 500 acre landfill site area surrounding the

landfill. Each comparison gave the cost per household for locating a landfill at each

hypothetical site or location. The results were used by the State to predict how residents

would vote in a referendum on different possible landfill locations. First, 59 possible

sites were selected, based on geological and public health criteria. Sites were ranked

using the contingent choice survey results, in order to come up with a short list of

potential sites. The final decision, based on geological, public health, public preferences,

and political considerations, was to expand the existing landfill site.

11.5 Case‐study: The Environmental Costs of Low River Flows

11.5.1 Background River flows may be reduced to sub‐optimal levels by natural phenomenon such as low

rainfall, or may be caused by the abstraction of water either from the river itself or from

the underlying aquifer. In the United Kingdom (UK) when river flows are seriously low,

the Environment Agency (EA) is responsible for the design and implementation of

schemes to alleviate this problem. A number of options are available to the EA to

alleviate low flows in these rivers. All of the available options involve a cost, and before

any decisions are made regarding which particular solutions to the low flow problem are

adopted, careful consideration has to be given to the question of whether or not the

additional benefits from increasing the flow to some environmentally acceptable flow

regime outweigh the costs involved.

The case study focuses on the south west of England, an area encompassing 176

beaches designated as Euro‐beaches by the European Community, with other smaller

beaches and coves, and approximately 4,000 miles of rivers. The particular focus was on

87

seven rivers that were identified as being seriously affected by low flows at certain times

of the year. Given that the benefits of increasing river flows were likely to comprise both

use values and non‐use values, it was decided that some form of expressed preference

method would be most appropriate to elicit willingness to pay. The issue of non‐use

values was an important consideration, and any survey of the general public was likely

to result in a high proportion of non‐users being samples. This would include who do not

visit rivers at all and, more commonly, individuals who not visit any of the designated

low‐flow rivers in the south west of England. This lack of familiarity would have made it

difficult for respondents to give meaningful answers to an open‐ended WTP question or

to a bidding game relating to EA activities. This consideration led to the concentration on

the choice experiment approach with the aim of estimating the marginal WTP of the

general public for unit improvements in low flow alleviation in rivers in the south west of

England. The choice experiment would also be used to estimate the public's marginal

WTP for unit improvements in the numbers of clean beaches and miles of unpolluted

rivers in the area.

This concentration on marginal WTP confronts the issue of scale and should provide

more robust welfare estimates for decision making. Furthermore, the more holistic

choice experiment approach is less vulnerable to other sequencing effects such as

embedding.

11.5.2 The benefits of low flow alleviation The total economic value of low flow alleviation in a given river is the sum of all use

values derived from it, plus any non‐use values which this activity may generate. Use

values are benefits arising either directly or indirectly from the improvement in flows,

while non‐use values are generated by the consumption of the flow of information

about the good which is consumed as a preservation benefit, i.e. a value arising from the

knowledge that the river remains healthy and viable and will persist. A survey of the

general public survey in the south west of England was carried out in the summer of

1996, including both users and non‐users. Non‐users were identified as those

respondents who did not visit any of the low flow rivers in the south west specified in

this project.

11.5.3 Questionnaire design A series of focus groups was undertaken prior to the design of the questionnaire with

the intention of informing the design process and suggesting the levels of information

that respondents would require. The focus groups suggested that the public considered

problems of coastal and river pollution to be the most pressing water quality issues in

88

the south west; however, when shown a portfolio of photographs illustrating the varying

effects of low flows most participants agreed that this issue was also important (though

not as pressing as pollution, to which it was seen as a contributory factor). The design of

the questionnaire was intended to provoke respondents into thinking more deeply

about the consequences of their response to the choice experiments and to make those

responses as realistic as possible given the artificial context. The effect of the

questionnaire design on welfare estimates in this case should have been to limit them

rather than inflate them, thus leading to conservation welfare estimates that can be

interpreted as lower‐bound figures. The magnitude of welfare estimates was limited by

using a sequence of questions and statements designed to remind respondents of other

environmental quality issues that they might wish to support, rather than allowing them

to focus on the water quality issues with which this study was concerned. The notion of

a multi‐good environment was introduced early in the questionnaire, when respondents

were asked questions regarding their donations to good causes and their willingness and

ability to contribute more to such causes in the future. This helped to establish a context

within which respondents could begin to determine how much they would be willing to

pay towards water quality improvements. This approach was tested and refined over

two separate pilot surveys. The notion of a multi‐good environment was introduced

early in the questionnaire, when respondents were asked questions regarding their

donations to good causes and their willingness and ability to contribute more to such

causes in the future. This helped to establish a context within which respondents could

begin to determine how much they would be willing to pay towards water quality

improvements. This approach was tested and refined over two separate pilot surveys.

Respondents were then presented with a brochure describing the EA's activities under

three headings: reducing river pollution; monitoring marine pollution in coastal waters;

and improving flows in low flow rivers. Text was kept to a minimum and illustrations

were used wherever possible. Information was limited to bullet points describing the

problem being tackled, its causes and consequences and what the EA was doing to

tackle the problem across the south‐west. Rather than just emphasize the scale of the

problem, the brochure also attempted to show how much had in fact been achieved in

tackling these problems. This had the twin effects of demonstrating that additional

spending could bring about further positive environmental benefits, but that current

levels of spending (£90 million per year in 1996) were already achieving significant

improvements in each of the categories shown in the brochure.

The focus on activities in the south west was thought most appropriate in a survey that

would cover a wide variety of households in that region. Specifically, it was felt that

89

respondents would relate better to more familiar local issues than to national ones and

this would promote more considered response to the choice experiment questions. In

the choice experiments, respondents were given a sheet similar to that shown in Figure

27 and asked to choose one of the three choices. Having made their choice respondents

were shown three more cards and asked to choose their preferred choice from each.

Cards were chosen at random form an orthogonal set of 64 choice cards. In the choice

experiments the issue of low flows is embedded in a broader set of water quality goods

discouraging a single focus on the issue of low flows. This confronts the issue of

embedding by ensuring that respondents act consistently and make choices based on

the same set of related goods.

CARD 06

Please choose one column

CHOICE 1(current situation)

CHOICE 2 CHOICE 3

Increase in water charges needed to achieve targets

No increase £5 increase £10 increase

Beaches in the South West NOT MEETING European standards on cleanliness

9 beaches 5 beaches 3 beaches

Rivers in the South West WITHOUT good quality water

990 miles 350 miles 350 miles

Rivers in the South West WITHOUT acceptable flow levels


Figure 27. Example of a choice experiment card

11.5.4 River usage by the general public Nearly half of the households interviewed claimed to live one mile or less from a river;

and more than two‐thirds lived within two miles of a river. Most households regularly

undertook recreational activities along rivers, with only 23% claiming not to undertake

some form of regular recreation along rivers. Around 88% of households had visited

more than one river the 12 months preceding the survey, but only 45% had visited one

of the low flow rivers in the south west. The frequency of visits to beaches over the

summer months had much the same distribution as visits to rivers, though greater

differences began to emerge during the winter.

11.5.5 Empirical Results

Preliminary questions demonstrated public perceptions of the abstraction problem.

Nearly three‐quarters of respondents thought that rivers were an important source of

water; but just over half thought too much water was being abstracted from rivers by

90

water companies and other users. Investigations into the public's WTP for good causes

showed that 80% of respondents would prefer to see additional public expenditure on

the natural environment (a lower rate than for health or education). However, only 40%

of respondents indicated that they would be willing to contribute more than they

currently contribute towards that they considered 'good causes'.

Following Adamowicz et al. (1996) the responses from choice experiments were used to

estimate a discrete choice model of the probability Pr(i) of choosing a given alternative i:

Pr (i) =exp (sVi)/∑exp (sVj)

Models were estimated using linear and quadratic functional forms. Under the quadratic

specification some attribute coefficients were not statistically significant; therefore the

linear functional form was used for benefit estimation. Table 21 reports WTP for the

marginal improvements in water quality defined by unit reductions in the number of

polluted beaches and the lengths of river affected by low flows and poor water quality.

Respondents were willing to pay between £1.31 and £1.43 to ensure that one additional

beach meets EC standards on cleanliness and £0.02 to clean up a mile of polluted river.

Similarly, respondents were willing to pay up to £0.06 per mile to improve conditions on

low flow rivers. These estimates were used for the purposes of aggregation, but this

relies on the presumption that it is reasonable to assume constant marginal WTP for

water quality improvement measures across the south west. This may be the case for

beaches but it is possible that following substantial reductions in the length of rivers

affected by low flows and pollution that WTP for additional lengths to be improved.

Table 21. WTP for marginal improvement in water quality Reduction Extended specification

1 polluted beach £1.307 £1.431

1 mile of polluted river £0.017 £0.019

1 mile of low flow river £0.052 £0.058

11.5.6 Estimating the number of users and non‐users The population of users for this study was defined to encompass all households who had

visited a given low flow river in the two‐year period immediately preceding the survey.

Respondents were shown a list of the seven low flow rivers that were the main focus of

this study and asked whether or not they had visited these rivers at any time during the

last two years. This data can be used to estimate the number of households that visit

each low flow river. Nearly 45% of households in the south‐west had visited at least one

91

of the low flow rivers during the last two years. Table 22 reports results for each river to

give an estimate of the bi‐annual number of visitors to each river. The population of the

south west is divided into user and non‐user households for each of the seven low flow

rivers. It also reports the length of each river affected by low flows. This latter figure is

important in the estimation of the aggregate benefits of low flow alleviation. Although

quite high, these estimates seemed reasonable. Rivers are linear features and instead of

offering a single point of access provide many and varied opportunities for individuals to

encounter them. Added to this is the fact that all of the rivers are located in areas with

considerable scenic attractions where local people as well as those from further afield

would be expected to enjoy the considerable recreational opportunities on offer.

Table 22. User and non‐user populations for low‐flow rivers from the south west River User Households Non‐user households Miles affected by low flows

Allen 85,897 1,562,533 20

Upper Avon 230,717 1,417,713 35

Meavy 166,760 1,481,670 7

Otter 157,982 1,490,448 5

Piddle 157,854 1,490,576 16

Tavvy 240,155 1,408,275 16

Wylye 162,481 1,485,949 30

11.5.7 Aggregate benefit estimates The choice experiments reported in the previous section can be used to derive random

utility models based on the subsets of users and non‐users defined previously. These

models yield estimates of the marginal WTP for a unit decrease in the length of rivers

affected by low flows. The estimate for users will comprise a combination of use values

and nonuse values, while for non‐users the estimate is made up entirely of non‐use

values. The validity of this latter estimate is highly suspect because the coefficient on the

LOWFLOW variable in the non‐users' model was not statistically significant at any

reasonable level. This implies that when selecting their preferred card in the choice

experiments, the non‐user population did not give much weight to improvements in low

flow rivers. Rather, their choices were based upon cost and the improvements that

could be made to polluted rivers and beaches.

The population of low flow river users was estimated at 734,161. For each low flow river,

the maximum aggregate annual benefit for user households is calculated by multiplying

this figure, first by the length of river affected by low flows, and then by £ 0.076 per mile

(see Table 23). The proportion of this amount contributed by visitors can also be

92

calculated using the same procedure and substituting the estimated number of visitors

for the user population. The aggregate benefits for non‐user households are calculated

in a similar fashion, using the marginal value of £0.0435.

Table 23. Approximate aggregate annual benefits for improving low flows

River Aggregate benefits for user Households

Aggregate benefits for visitor households

Aggregate benefits for non‐user households

Allen £1,115,925 £130,563 £795,414

Upper Avon £1,952,868 £613,707 £1,391,975

Meavy £390,574 £88,716 £278,395

Otter £278,981 £60,033 £198,854

Piddle £892,740 £191,950 £636,331

Tavvy £892,740 £292,028 £636,331

Wylye £1,673,887 £370,457 £1,193,121

11.5.8 The costs of low flow alleviation The costs of various options to alleviate low flows on rivers in the south west were

calculated by Environmental Resources Management. Table 24 summarizes the present

values of the costs of the cheapest available option for low flow alleviation on six of the

seven rivers: no solution has yet been put forward for low flow alleviation on the Tavy.

Table 24 also reports the present value of the benefits of low flow alleviation for the

user sub‐sample. Present values were calculated by assuming a constant flow of benefits

across the period 1997 to 2017 and discounting at 6%. Net benefits and benefit cost

ratios were calculated using these figures. These figures can be used to carry out partial

benefit‐cost analysis in order to identify which options require further investigation

before they can be implemented (Table 24).

Table 24. Net present value of aggregate benefits for improving low flows across the length of the rivers

River Present value of costs

Present value of aggregate user

benefits

Net present value

Benefit‐cost ratio

Allen 11,867,000 13,915,000 2,048,000 1.17

Upper Avon 763,000 24,252,000 23,589,000 31.92

Meavy 80,000 4,870,000 4,790,000 60.88

Otter 34,430,000 3,480,000 ‐30,950,000 0.10

Piddle 5,471,000 11,132,000 5,661,000 2.03

Tavvy Unknown 11,132,000 ‐ ‐

Wylye 224,000 20,873,000 20,649,000 93.18

93

Benefits exceed costs by a wide margin for the rivers Avon, Meavy and Wylye, and to a

lesser extent for the Piddle. Costs are prohibitive only on the Otter, while the benefit‐

cost ratio for the Allen suggests that while benefits probably exceed costs there is

relatively little difference between them. As benefits are based only on the user sample,

these benefit‐cost ratios ignore any benefits that might accrue to non‐users but include

non‐use benefits for low flow river visitors who do not visit the river in question. If only

the present value of the aggregate benefits accruing to visitors is considered then the

least cost solutions for the Piddle and Allen fall to levels substantially below that of the

associated costs, while the benefits for the A von, Meavy and Wylye still outstrip costs

by a ratio of at least 10 to 1 .

The results of the study strongly suggest that there are considerable welfare gains to be

made from implementing the least cost low flow alleviation options on the Rivers Wylye,

Meavy and Upper Avon. Similarly, there is strong evidence to suggest that only a low

flow alleviation option costing considerably less than the one costed in this study would

be justified on the Otter. The benefits of the least‐cost low flow alleviation options for

the Piddle and Allen can only be justified on the basis of non‐use values, and then only

tentatively. In these cases we would recommend that investigation of the non‐

recreational benefits of low flow alleviation be carried out to provide a clearer picture in

a fuller benefit‐cost analysis.

12 THE BENEFIT TRANSFER METHOD

The benefit transfer method involves transferring values that have been estimated for a

similar good or service from another location/context to the current location/context. It

represents a useful method under budget and time constraints. This method has been

applied to value the impact of improved water quality on recreation values and public

health. It has also been the normal procedure adopted in regulatory command and

control mechanisms in which common standards are applied. For instance, the EU

assumes that benefits of environmental improvement are of equal value in different

areas of the EU. Yet, it should be noted that benefit transfers can only be as accurate as

the initial study.

The simplest type of benefit transfer is the unit day approach, where existing values for

activity days are used to value the same activity at other sites. The estimates are based

on expert judgment in combining and averaging benefit estimates from a number of

existing studies and “unit day values” may be adjusted for characteristics of the study

94

site when they are applied. A more rigorous approach involves transferring a benefit

function from another study. The benefit function statistically relates peoples’

willingness to pay to characteristics of the ecosystem and the people whose values were

elicited. Adjustments can then be made for differences in these characteristics, thus

allowing for more precision in transferring benefit estimates between contexts.

12.1 Approaches for applying benefit transfer and assessing the validity of the attempts

12.1.1 Unit day values The ‘unit day value’ was applied by the US Forest Service in the 70s and 80s. Federal

guidelines in 1982 recommended a value of $6.10‐$17.90 per day for specialized

recreation (wilderness use, trout fishing, big‐game hunting, white water boating) and

$1.50‐$4.50 per day for general recreation (picnicking, swimming, small game hunting,

camping, boating). When applied to a new site, unit day values are adjusted on the basis

of the demand functions of site‐visitors. Demand depends on site attributes such as

congestion, accessibility and parking conditions, environmental quality; scenery, pests,

water, air, climate, socio‐economic characteristics of recreationalists, preferences, price,

and availability of substitute sites. None of these factors will be identical across different

sites. Expert judgment is required to assess what the benefits of a new site might be

from a range of possible values. The unit day values can be updated to account for

inflation and observed changes in price and income elasticities for recreation over time.

12.1.2 Transfer from HPM models

Benefit transfer from studies using hedonic price models may be applied by relying on

judgments of real‐estate agents’ to adjust the results. Yet, while some research

suggested close correlation between estate agents’ estimates of total house price and

estimates derived from a hedonic price model, other research revealed discrepancies.

12.1.3 Transfer from TCM models

Benefit transfer via the Travel Cost Model may be applied by transferring demand

functions from existing facilities, resembling closely the prospective facility in the type of

recreation provided. If the catchment areas of the two sites are mutually exclusive, then

multiplying the existing site coefficients by the values for independent variables of the

new site will give estimates of the number of visits and benefits attributable to the new

site. This approach is expected to yield more accurate results than simply applying an

average value of benefit per visitor day to the site. However, if the proposed facility is

situated within the catchment area of an existing facility, the existing demand function

should be applied to the new site as if unique. If the new consumer surplus exceeds the

95

existing one, the net gain from having the new facility is the difference between the two.

The lack of homogeneity in product mix may be remedied by valuing the different

recreational activities separately and then aggregating, rather than developing a

demand curve for the site as a whole. Errors in Benefit Transfer via TCM may occur due

to several reasons, including, choosing the wrong functional form, selecting an

incomplete or inappropriate set of arguments, measuring arguments incorrectly such as

value of time, income, cost of access, measuring the dependent variable with error, and

the presence of substitute sites. The latter could be cancelled out if sites are randomly

distributed via a simulation model.

12.1.4 Transfer from CVM models

Benefit transfer from studies using contingent valuation models can be affected by ex

ante‐ ex post valuation perspective, whereby some estimates elicited after the

uncertainty about the good is removed are employed in an ex ante project appraisal.

Application may also be affected by scale or quantity value. For instance, if the new

good or policy is identical to the old and lies within the same market area, then it

represents an additional quantity of the good and should be valued less than the existing

good at the site. The sequential position of the supply of the good may also affect

application, particularly where goods are complements or substitutes. In this case, the

sequence in which a particular good is provided in relation to others determines its

value. Other factors influencing application of benefit transfer via CVM include

differences in attributes, as well as compositional effects, or when respondents have

difficulty in disentangling the structure of the substitution and complementary

interrelationships among attributes within the same holistic set.

12.1.5 Transfer from meta‐analysis

Benefit transfer may also be applied via meta‐analysis, where data‐based aids are used

to explain variations in estimated benefits across different studies with the aim of

applying past results to future resource policy decisions. This approach attempts to

assess environmental values by investigating the relationship between benefit estimates

(WTP), the features of the goods, and the assumptions of the models. It entails the

systematic application of statistical methods to assess common features and variations

across a wide range of prior studies. It is undertaken using a variety of techniques

encompassing qualitative and quantitative econometric methods. Yet, meta‐analysis is

relatively underdeveloped in the field of benefit transfer. However, it is important as a

means of investigating the factors and issues involved in the derivation and construction

of value. For example, Walsh et al. (1989) conducted a study to explain variations in net

economic benefits per activity day in terms of site, location, and methodological

96

variables. For this purpose, 287 benefit estimates were compared: 156 based on TCM,

129 based on CVMs, and 2 based on HPMs. Some main findings included that omitting

travel time in TCM studies reduced benefit estimates by 34% and that Individual TCM

estimates were about 46% greater than Zonal TCM estimates using the same functional

form. Another finding was that if TCM was accepted as the standard for benefit

estimation, then CV estimates needed to be adjusted upwards by 20‐30%.

12.2 Standards of benefit transfer studies Different standards for benefit transfer may be applied in different contexts A higher

standard of accuracy may be required when the costs of making a poor decision are

higher. On the other hand, a lower standard of accuracy may be acceptable when costs

are lower, or when the information from the benefit transfer is only one of a number of

sources of information, or when it is used as a screening tool for the early stages of a

policy analysis.

The benefit transfer method is most reliable when the original site and the study site are

very similar in terms of quality, location, and population characteristics, when the

goods/services in both sites have similar characteristics, when the original valuation

study has been carefully conducted and used sound valuation techniques, and when

values in original study have not been valuated a long time ago since preferences change

over time.

Three tests have been suggested to determine the accuracy of benefit transfer:

− Comparing benefit transfer values with primary data values obtained from the policy

site

− Determining whether different populations have the same preferences for the same

non‐market good, after controlling for differences in socio‐economic characteristics

− Determining whether transfers are stable over time

When applying the benefit transfer methodology, the following steps should be

followed:

1. Identify existing studies or values that can be used for the transfer. There are a

number of valuation databases that can be useful.

2. Evaluate the existing values to determine whether they are appropriately

transferable. Consider whether the service being valued is comparable to the service

valued in the existing studies (site features, site qualities, availability of substitutes)

97

and whether the characteristics of the relevant population are comparable in terms

of demographics and people’s preferences.

3. Evaluate the quality of studies to be transferred. The better the quality of the initial

study, the more accurate and useful the transferred value will be. This requires the

professional judgment of the researcher.

4. Adjust the existing values to better reflect the values for the site under

consideration, using available and relevant information. Supplemental data may

need to be collected through survey key informants, by talking to the investigators of

the original studies, getting the original data sets, or collecting some primary data at

the study site to use to make adjustments.

5. Estimate the total value by multiplying the transferred values by the number of

affected people.


Advantages of benefit transfer include:

− The BT method is less costly than conducting an original valuation study

− The economic benefits are estimated faster than when undertaking an original

valuation study

− The BT method can be used as a screening technique to determine if a more

detailed, original valuation study should be conducted

− The method can easily and quickly be applied for making gross estimates of

recreational values

− The more similar the sites and the recreational experiences, the fewer biases will

result

Issues and Limitations associated with benefit transfer:

− Lack of accuracy, except for making gross estimates of recreational values, unless the

sites share all of the site, location, and user specific characteristics

− Unavailability of good studies for the policy or issue in question

− Difficulty in finding appropriate studies, since many are not published

− Reporting of existing studies may be inadequate to make the needed adjustments

− Difficulty in assessing the adequacy of existing studies

− Extrapolation beyond the range of characteristics of the initial study is not

recommended

− Benefit transfers can only be as accurate as the initial value estimate

− Unit value estimates can quickly become dated

98

12.4 Illustration A park is being upgraded to provide additional recreational opportunities. A proposal is

to add a swimming beach to the lake in the park. The benefits of the new beach needs to

be explored, however, there is limited budget for the valuation study. For this purpose,

the Benefit Transfer Method is preferred because of the lack of a large budget for site‐

specific benefits studies and because values for recreational uses are relatively easy to

transfer.

The applied methodology is as follows:

Step 1

Identify existing studies or values that can be used for the transfer. Look for studies that

value beach use, specifically for lake beaches if possible. Assume that the researcher has

found two travel cost studies that estimated values for swimming at lake beaches.

Step 2

Decide whether the existing values are transferable by examining various criteria. The

existing values or studies would be evaluated based on several criteria, including:

− Is the service being valued comparable to the service valued in the existing

studies in terms of similar types of sites (lake beaches in a park), similar quality of

sites (water quality and facilities), similar availability of substitutes (the number

of other lake beaches nearby)

− Are characteristics of the relevant population comparable? Are demographics

similar? If not, are data available to make adjustments?

In the example, the first study is for a similar lake beach. The beach is also in a park, has

comparable water quality and facilities, and a similar number of substitute sites in the

area. It is located in an urban area, while the beach being valued is in a rural area. The

characteristics of visitors can be expected to be different for the two sites. The second

study is in a rural area with similar types of visitors, but the lake has many more

available substitutes.

Step 3

Evaluate the quality of studies to be transferred. In this example, the researcher has

decided that both studies are acceptable in terms of quality.

99

Step 4

Adjust the existing values to better reflect the values for the site under consideration. In

this case, the sites valued in each of the existing studies differ from the site of interest.

The researcher might adjust the values from the first study by applying demographic

data to adjust for the differences in users. If the second study has a benefit function that

includes the number of substitute sites, the function could be adjusted to reflect the

different number of substitutes available at the site of interest. Because the beach will

be new, the researcher will need to estimate how many people will use the beach. They

may conduct a survey of park visitors, asking whether they would use a beach on the

lake, and how many times they would use it. Then, these visitation estimates can be

multiplied by the value per day for beach use (adjusted for differences in population and

site characteristics), to get an estimate of the economic benefits for the new beach.

12.5 Case‐application: Wetlands Restoration in Saginaw Bay, Michigan

The State of Michigan is considering plans to protect and restore coastal wetlands along

the southern shore of Saginaw Bay. The State must estimate the potential benefits from

protecting and restoring the wetlands. A survey asked people about their support for

restoring wetlands, but did not include a valuation question. The researchers used

benefit transfer methods to estimate the value of protecting and restoring wetlands

around the Bay.

A valuation study for proposed wetlands protection and restoration of Ohio’s Lake Erie

coastal wetlands was used for the benefit transfer. Researchers assumed that the values

estimated for Ohio were similar enough to be transferable to Michigan. The study

valued similar programs and quantities of wetlands to those proposed in Michigan.

However, coastal residents were not surveyed. The transfer of values from the Ohio

study to coastal residents in Michigan requires the assumption that coastal residents

have values similar to those of residents of other areas of the drainage basin.

Estimates of wetland values for Michigan, based on the Ohio study were $500 ‐ $9,000

per acre for residents of the drainage basin and $7,200 ‐ $61,000 per acre for residents

of the State of Michigan.

100


THE STATED PREFERENCE APPROACH:

GROUP EXERCISES &

CASE‐STUDIES

Sessions 11 & 12

Region

al Training Worksho

p on

: Th

e Co

st of E

nviron

men

tal D

egrada

tion

Metho

dology

101

SESSIONS 11 & 12

GROUP EXERCISES & CASE‐STUDIES

Air quality in Beijing

Wang, X.J.,Zhang, W., Li, Y., Yang, K. Z., and Bai, M. Air quality improvement estimation and

assessment using contingent valuation method: A case study in Beijing. Environmental

Monitoring and Assessment, 120, 153‐168, 2006.

Ecosystem services in Ejina China

Zhongmin, X. Guodong, C. Zhiqiang, Z., Zhiyong, S. and Loomis, J. Applying contingent valuation

in China to measure the total economic value of restoring ecosystem services in Ejina region.

Ecological Economics, 44, 345‐358, 2003

Environmental services in the Yaqui River Delta, Mexico

Ojedaa, M.I., Mayerb, A.S., and Solomon, B.D. Economic valuation of environmental services

sustained by water flows in the Yaqui River Delta. Ecological Economics, 2007. (In Press).

Sustainable development in Swedish coastal zone

Söderqvista, T., Eggertb, H., Olssonb, B., Soutukorvac, A., 2004. Economic valuation for

sustainable development in the Swedish coastal zone. SUCOZOMA research program.

Case‐studies

Doumani, F. 2007. Economic valuation of the coastal zone of the Mohafaza of North Lebanon:

Coastal zone municipal assessment, Coastal zone direct and indirect use value, Coastal zone

economic activity. Short and Medium Term Priority Action Program III. Integrated

Management of East Mediterranean Coastlines: Northern Lebanon, funded by The European

Commission. University of Balamand, Lebanon.

World Bank/ METAP, 2008. Climate Change Adaptation in the Water Sector in the Middle East &

North Africa Region: A Review of Main Issues. Presented at the Mediterranean Workshop on

Integrated Coastal Zone Management (ICZM) Policy, Alghero, Sardinia (Italy), May 19‐21,

2008.

World Bank/METAP, 2008. Carbon Finance Instrument to Improve Coastal Zone Solid Waste

Management. Presented at the Mediterranean Workshop on Integrated Coastal Zone

Management (ICZM) Policy, Alghero, Sardinia (Italy), May 19‐21, 2008.

102


COST‐BENEFIT ANALYSIS

Session 13

Region

al Training Worksho

p on

: Th

e Co

st of E

nviron

men

tal D

egrada

tion

Metho

dology

103

SESSION 13

13 COST BENEFIT ANALYSIS

Cost benefit analysis (CBA) is one of the most widely used techniques for project

appraisal in the public sector. It represents a framework for policy decision‐making. Its

first formal application was in 1768, to evaluate the net benefits of the Forth‐Clyde

Canal in Scotland.

13.1 Measures of benefit

The demand curve, also referred to as the marginal benefit curve, indicates the cost of

consuming one extra unit of good and provides an idea of changes in ‘utility’ or level of

satisfaction. The price one is willing to pay for a good depends on the satisfaction one

derives from consuming it, which is taken as a measure of benefits. For environmental

goods, the benefit or WTP exceeds the market price, if it exists (Figure 28). Valuation

methods discussed earlier are used to obtain estimates of WTP.

Total benefits = Total Revenue + Consumer Surplus

= Area of 0ECD + Area of ΔACE

Figure 28. Demand curve or marginal benefit curve

13.2 The concept of costs The cost in a cost benefit analysis refers to the opportunity cost (OC) to carrying out the

investment. Under perfect competition, the OC of a good is the same as the market

price of that good. For environmental goods, there is no market‐price; alternative

methods are to be used to measure the OC.

13.3 The concept of Net Social Benefits It is important to distinguish between a social CBA and a private CBA. A social cost

benefit analysis (SCBA) is conducted from a society’s perspective and is referred to as

economic analysis, while a private CBA is carried out from an individual investor’s view

point and is referred to as financial analysis. A project may be financially viable but

104

socially undesirable. The objectives of a SCBA is to determine whether a project is

socially beneficial, whereby:

Net Social Benefit (NSB) = Benefits – Costs = WTP – OC > 0

If NSB > 0, then the state can use the surplus to compensate the losers.

13.4 Steps in Conducting an SCBA The steps in conducting a SCBA are outlined in Figure 29. The following sections will

detail the steps of a cost benefit analysis process along with a case‐study illustrating its

step by step application.

1. Define objectives and scope of project

2. Identify and screen alternatives

3. Identify and value the costs and benefits for the remaining alternatives

4. Calculate discounted cash flows and project performance criteria for each alternative

5. Rank the alternatives in order of preference

6. Conduct a sensitivity analysis and/or risk analysis for the preferred alternative(s)

7. Make a final recommendation Figure 29. Steps in a social cost‐benefit analysis

The case‐study involves the Bintuli Wastewater Treatment Project. The city of Bintuli in

the Republic of Kabastan is a center of commerce and industry. Main industries include,

metal manufacturing, coal extraction, chemical manufacturing, construction, paper

making and food processing. The value of industrial output was estimated at 200 million

USD in 1990 as compared to 16 million USD for agriculture. Quantities of domestic and

industrial effluents in water bodies increased significantly, whereby the total industrial

effluent amounted to 163,700 m3/day, with the total effluent including 271,700 m3/day

of domestic wastewater. Only 30% of industrial effluent was treated by the existing

Bintuli Wastewater Treatment Plant (WWTP), while no domestic effluent was being

treated. River courses in the city turned black and emitted unpleasant odors. Thus, it

was proposed to build a wastewater treatment facility with pumping stations and

drainage networks to treat 28% of industrial waste and the remaining domestic waste.

The treated effluent discharged in river to be used by industries and agriculture.

105

13.4.1 Defining objectives and project scope The objectives are often specified by decision‐makers in the bureaucracy. Objectives

should be clear and unambiguous. In the case of the Bintuli WWT Project, the objectives

are to improve the health of the community and to increase economic activity by

improving wastewater treatment facilities in the city.

13.4.2 Identifying and screening alternatives All possible options for reaching objectives need to be listed, including the ‘do nothing’

option. A preliminary screening of alternatives should be conducted. In the case of the

Bintuli Wastewater Treatment Project, possible alternatives include:

− Maintaining the status quo

− Expanding the existing WWT facilities. This alternative was ruled out because it

uses outdated technology and would be difficult to maintain

− Building a new WWT facility

− Various locations and site options. Only one potential site considered in this

application

13.4.3 Identifying and valuing benefits and costs As mentioned earlier, costs and benefits differ for an SCBA as compared to private

investors. A benefit in an SCBA refers to an outcome resulting in an increase in an

individual’s utility and a cost in an SCBA refers to an outcome resulting in a decrease in

an individual’s utility. In identifying costs and benefits, it is important to note the

following:

− An incremental approach is adopted in assessing costs and benefits, which involves

first identifying and valuing costs and benefits of the project, and then compare

them with the situation to prevail without the project. The difference is the net

incremental benefit arising from the project. Only additional changes in costs and

benefits are considered, and not total costs and benefits.

− Sunk costs and benefits incurred before project commencement must be excluded.

Previous costs are not an opportunity cost as they do not represent a loss of future

income from an alternative use of resources.

− Transfer payments must be excluded. These include taxes, subsidies, loans, and debt

services, which do not result in an increase in net benefits. Taxes by foreign investors

should be included.

− Depreciation and interest are excluded from the cost in a SCBA. SCBA involves

discounting values of capital items at their opportunity costs. Thus, including

depreciation as a cost will result in double counting. The discount rate in an SCBA

106

already takes into account the interest. Including interest as a cost will result in

double counting.

Costs and benefits are normally classified into primary costs and benefits related directly

to the project, and secondary costs and benefits. The latter arise from events and

activities triggered by the project and should be handled with care as they could

exaggerate estimates. The opportunity cost must be used as a guideline. Resources are

sometimes merely transferred from one part of the economy to another. Costs and

benefits may also be classified into market costs and benefits non‐market costs and

benefits. In the case of the Bintuli WWT Project, the identified costs and benefits are

presented in Table 25.

Table 25. Costs and benefits associated with the Bintuli WWT project

Costs Benefits Primary Investment

− Construction of a pumping station, office building, WWT facilities

− Purchase of equipment Operation and maintenance (O&M)

− Wages and salaries − Fuel and chemical costs − Other costs (project management,

preparation, training and commissioning

Primary Economic User charges Reduction in health costs and mortality

rates Reduction in costs of treating increasingly

polluted water supplies Increase in labor productivity due to

reduction in absence from work due to illness

Secondary Benefits to industry and agriculture from

using recycled water Additional revenues from re‐afforestation Increase in reed harvesting for the paper

mill industry

Valuing benefits and costs allows comparison between alternatives. The valuation

should be done according to the opportunity cost principle, whereby prices of inputs

that do not reflect their true value to the society are adjusted (shadow pricing).

Comparison of costs and benefits should focus on the “with vs. without” the project,

rather than “before vs. after” the project.

Valuing the costs involves first finding the market prices for the inputs and outputs. All

costs must be in present day or constant prices, whereby costs incurred over the project

lifetime must be valued at prices prevailing at the time of the project appraisal and the

annual costs must be assumed to increase at the inflation rate. The residual value should

be considered for assets with an economic life that exceeds the planning horizon or

project life. The economic life is the estimate of the duration of the operation of an asset

107

before it requires refurbishment. The residual or salvage value of the asset must be

included as a cash inflow at the end of the planning horizon. This is calculated using

either a linear method or a diminishing value method. The linear method assumes that

the asset value declines linearly over time:

Residual value at time t is (1‐t/n)P

Where t = time; n = economic life; P = initial price

For example, consider an asset purchased at $100,000 and has an economic life of 20

years. At the end of the planning period of 15 years, its residual value is (1‐

15/20)*100,000 = $25,000

As for the diminishing value method, it assumes that the asset value declines by a fixed

proportion of the beginning of year value per annum:

Residual value at time t is (1‐1/n)t P

Where t = time; n = economic life; P = initial price

Other costs that need to be considered include:

− Land and pre‐existing building and plant, property that is already owned by

operating authority must be valued at its opportunity costs. Opportunity costs

should be current variations based on the most profitable alternative uses.

− When a project is to be constructed in stages, only the portion of investment and

operating costs to satisfy demand in the current planning horizon must be attributed

to the project.

− Working capital, which often constitutes 2% of the total capital outlays, must be

considered as cash outflow at the time when capital expenditures are made and cash

inflow at the end of the project.

− Operating costs, which include labor, utilities, supplies, repairs and maintenance,

equipment hiring and leasing, insurance and administrative overheads, are to be

estimated on an annual basis.

− Implicit costs should also be considered. These include opportunity costs and social

costs, such as the use of land, buildings, plants, already purchased by the local

authority, and time spent on the project by agency staff.

The costs incurred in the case of the Bintuli WWT Project, are summarized in Table 26.

108

Table 26. Costs of the Bintuli WWT Project

Item Cost (million USD)

Investment costs Buildings and structures Equipment and supplies Total investment cost

3.42 13.15 16.57

O&M costs Electricity Salaries Chemicals Maintenance Other Total O&M costs

0.68 0.09 0.06 0.58 0.21 1.62

They are based on the following:

− All equipment and construction materials are imported and valued in US$

− Fuel and chemical supplies are adjusted by subtracting the government subsidies on

these items

− Because of high unemployment in the area, unskilled labor is shadow priced at 50%

of the wage rate

− Skilled labor valued based on annual salaries

− Construction is to take 3 years

Benefits of the Bintuli WWT project encompass revenues from user charges, and

economic benefits derived from the WWT plant, including, reduced mortality,

productivity gained from reduced morbidity, water treatment cost savings, sale of

recycled water, afforestation benefits, and reed harvesting.

User charges

User charges include both new charges and existing charges. New charges are based on

the principle of full cost recovery and estimated at 6.9 cents/m3. For a total of 54.75

million m3/year of effluent treated, the annual revenue is estimated at 3.78 million USD

per year. As for the existing charges, they are estimated at 0.61 million USD per year (for

a user charged of 6.9 cents/m3 and a total effluent of 11.4 million m3/year already being

treated). Thus, the net incremental sales revenue is estimated at 3.17 million USD by

year 6, when the new plant is at full capacity.

Recycled water

Regarding recycled water benefits, about 60% of the treated wastewater will be reused

for irrigation and industrial purposes. The opportunity cost of this recycled water is

109

estimated at 10 cents/m3. The economic benefits from recycled water were therefore

estimated at 66,000 USD at year 4, rising to 3.29 million USD per year by year 8.

Afforestation

As for afforestation benefits, pine and hard wood species will be planted on 142.8 ha.

The net return per hectare was taken from estimates provided by Kabastani authorities

for experimental plots. These were reported at 689 USD/ha. Thus, the net benefits are

estimated at 10,000 USD in year 8 and a maximum of 100,000 USD by year 17.

Reed harvesting

Economic benefits are also expected from reed harvesting for the paper mill industry.

The Kabastani authorities have estimated the net returns to be 258.4 USD per ha.

Applying this figure to the projected area of 95.25 ha results in net annual benefits of

about 20,000 USD starting Year 6.

Reduced mortality benefits

Using World Bank estimates for Kabastan, mortality reduction from the project was

taken to be 0.005, 0.008, and 0.024 percent respectively, for the age categories of 15‐24

years, 25‐29 years, and over 60 years. On the basis of estimates for the number of

people in each category, the total number of deaths per year was calculated. Using the

estimated proportion of people employed in each age category and the mortality

reduction rates, an estimate of both employed and unemployed deaths was made.

Given the local annual wage of 620 USD (which includes housing subsidies and other

government payments) and assuming average working lives of between 5 and 45 years

for the three age categories, annual income losses avoided were estimated. For the

unemployed, a leisure value of half the annual wage was assumed. On the basis of this

estimate, the annual gains in leisure from saving deaths were estimated. Given that the

project will treat about half of Bintuli’s wastewater, only 50 percent of the potential

mortality and morbidity benefits were attributed to the project. The income benefits

from reduced mortality were therefore estimated at 10,000 USD in year 4, rising to

110,000 USD by the end of the project.

Productivity gains from reduced morbidity

A major social impact of the project is the reduction of the incidence of pollution‐related

illnesses and hence a reduction in worker absenteeism. These benefits were estimated

as follows. First, it was assumed that the current average number of days lost per

worker per year as a result of illness is 3 days. Next, using the employment statistics,

110

potential productivity losses avoided per worker per annum were estimated to be about

180,000 USD in year 4, rising to about 1.8 million USD by the end of the project.

Water treatment cost savings

As indicated above, the benefits of the project include the avoided cost of treating

polluted water. An estimate of 0.002/m3 of water treatment cost savings estimated by

the World Bank was assumed. The estimated benefits of water treatment cost savings

was estimated to be 110,000 USD per year,

The incremental economic benefits derived from the project were estimated at 96,000

USD in year 4, rising to 5.38 million USD by the end of the project (Table 27).

Table 27. Incremental economic benefits of Bintuli WWT project (million USD)

Year Recycled water

Afforestation Reed harvesting

Reduced mortality

Reduced morbidity

Water treatment cost savings

Incremental economic benefits

1 2 3 4 0.66 0.01 0.18 0.11 0.965 0.99 0.02 0.38 0.11 1.506 1.64 0.02 0.03 0.57 0.11 2.377 2.63 0.02 0.05 0.76 0.11 3.578 3.29 0.01 0.02 0.06 0.95 0.11 4.449 3.29 0.02 0.02 0.06 1.00 0.11 4.5010 3.29 0.03 0.02 0.06 1.06 0.11 4.5711 3.29 0.04 0.02 0.07 1.11 0.11 4.6412 3.29 0.05 0.02 0.07 1.17 0.11 4.7113 3.29 0.06 0.02 0.07 1.23 0.11 4.7814 3.29 0.07 0.02 0.08 1.30 0.11 4.8715 3.29 0.08 0.02 0.08 1.37 0.11 4.9516 3.29 0.09 0.02 0.09 1.44 0.11 5.0417 3.29 0.10 0.02 0.09 1.51 0.11 5.1218 3.29 0.10 0.02 0.10 1.59 0.11 5.2119 3.29 0.10 0.02 0.10 1.68 0.11 5.3020 3.29 0.10 0.02 0.11 1.76 0.11 5.39

13.4.4 Calculating discounted cash flows and project performance criteria

Once the costs and benefits with and without the project have been identified and

valued in monetary terms, the analyst is now ready to make a decision as to which

alternative to accept or reject. This requires reducing future streams of benefits and

costs to their present values to enable comparisons to be made between alternatives.

Given a stream of benefits (B0, B1…Bn) and a stream of costs (C0, C1…Cn), the Net Present

Value (NPV) is calculated using the following equation:

111

∑= +

−=

+

−+

+

−+

+

−+−=

n

t nrnCnB

nrnCnB

r

CB

r

CBCBNPV

0 )1()1(...2)1(

221

1100

where r = discount rate

The discount rate in a SCBA reflects society’s preferences between present and future

consumption. A high discount rate implies that society has a stronger preference for

present consumption over future consumption, while a low discount rate implies that

society has a stronger preference for future consumption over present consumption.

The choice of a discount rate is often controversial. Environmentalists argue against high

discount rates, which they believe are associated with environmental degradation.

Economists tended to use long‐term interest rates on government bonds as a measure

of opportunity cost of capital. For example, in the US, a rate of 10 percent is used and in

Australia, a rate of 8 percent is used. The discount rate must be a real rate. That is, the

nominal interest rate minus the inflation rate.

Another critical factor in computing the discounted cash flows is the planning period or

horizon. The planning period varies with nature of project. It should be determined by a

period within which estimates are made with a certain degree of confidence and it

should correspond to the economic life of the project.

Project performance criteria include the following, NPV, benefit‐cost ratio (BCR), internal

rate of return (IRR), and payback period. The BCR is the ratio of the present value of

project benefits to the present value of the project costs. It is calculated as follows:

∑=

+

∑=

+=

++

++

++

++

++

++

= n

tnrnC

n

tnrnB

nrC

r

Cr

CC

nrB

r

Br

BBBCR

n

n

0)1(

0)1(

)1(...2)1(10

)1(...2)1(10

21

21

The payback period is defined as the number of years required for a project to recover

its costs. In general, it discriminates against projects with high capital expenditures and

long‐term benefits. It is not recommended as a measure of project worth.

The IRR is the discount rate at which the present value of project benefits equals the

present value of project costs. It represents the maximum interest rate at which a

project could recover the investment and operating cost and still break even. It is

difficult to calculate and may not exist or may not be unique. A trial and error method

112

must be used. The IRR can be found by finding the discount rate at which the following

equation holds:

0)1(

...2)1(22

111

00 =+

−+

+

−+

+

−+−

ninCnB

i

CB

i

CBCB

The rule is to accept a project when NPV ≥ 0, BCR ≥ 1, and IRR > the social opportunity

cost of capital. The NPV is the most preferred criterion because it provides an estimate

of the size of the Pareto improvement. If two or more projects have NPVs > 0, then IRR

can be used to rank them.

A real rate of 12% was chosen as a discount rate to produce the discounted cash flows.

This rate is the average of the published World Bank discount rates for the past 10 years.

A planning period of 20 years was used based on advice received from engineers. Based

on the calculations shown in Table 28, the NPV at the 12% discount rate is estimated at

12.08 million USD and the IRR is 21 percent, which is above the opportunity cost of

capital of 12 percent. Therefore, it can be concluded that the Bintuli Wastewater

Treatment Project is economically viable.

Table 28. Incremental net benefits of Bintuli WWT project (million USD)

Year Incremental

economic costs Incremental sales

revenue Incremental

economic benefitsIncremental net

benefits 1 2.01 ‐2.01 2 8.45 ‐8.45 3 6.11 ‐6.11 4 2.42 1.91 0.96 0.45 5 1.62 1.91 1.5 1.79 6 1.62 3.17 2.37 3.92 7 1.62 3.17 3.57 5.12 8 1.62 3.17 4.44 5.99 9 1.62 3.17 4.50 6.05 10 1.62 3.17 4.57 6.12 11 1.62 3.17 4.64 6.19 12 1.62 3.17 4.71 6.26 13 1.62 3.17 4.78 6.33 14 1.62 3.17 4.87 6.42 15 1.62 3.17 4.95 6.50 16 1.62 4.43 5.04 7.85 17 1.62 4.43 5.12 7.93 18 1.62 4.43 5.21 8.02 19 1.62 4.43 5.30 8.11 20 ‐0.02 4.43 5.39 9.84

113

13.4.5 Conduct a sensitivity analysis and/or risk analysis In economics and business, a distinction is made between risk and uncertainty. Risk is

the potential outcome whose magnitude and probability of occurrence are known or

can be determined. Uncertainty refers to a situation where the magnitude of the

outcome may or may not be known and the probability of occurrence is unknown. In

practical situations, however, it is difficult to define precisely the probability of an

occurrence. Therefore, the distinction between the risk and uncertainty may not be

clear‐cut. Common methods for accounting for risk and uncertainty include, sensitivity

analysis, break‐even analysis, cross‐over values, and risk analysis.

Sensitivity analysis is used to assess the possible impact of uncertainty by posing ‘what

if’ questions. It highlights the critical factors affecting the project’s viability. Parameters

usually subjected to sensitivity analysis include discount rate, length of project planning

horizon, different timing of the project’s operation, changes in the capital outlays,

changes in the price of non‐market goods, and changes in social and environmental

benefits and costs. Sensitivity analysis is carried out by recalculating project

performance criteria using a range of values for the uncertain parameter. The project

performance criteria commonly used are NPV and IRR. Sensitivity analysis is conducted

by first determining a realistic range of values for the variables that are subject to

uncertainty. For example, capital cost ± 30 percent, O&M costs ± 30 percent, and

product prices ± 30 percent. Then, the effect of possible changes on the project

selection criteria are calculated while varying one variable and holding the others

constant. Finally, the economic viability of the project is reconsidered in light of the

performed calculations.

The break even value of a given project is the value of the discount rate at which the

NPV is zero or the value at which the entire costs will be recovered. On the benefit side,

if a variable appears to be higher than the break‐even level, this increases confidence in

the project’s viability. On the cost side, an estimate that is lower than the break‐even

level means that the project is likely to be economically viable.

The switching or cross‐over value of a project performance criterion (e.g. NPV) is the

discount rate at which the ranking of two projects changes. This method is

recommended when considering only one uncertain variable.

In the case of the Bintuli WWT project, a sensitivity analysis was conducted. For this

purpose, critical, uncertain variables were chosen for analysis, including changes in

114

capital and O&M costs, and changes in the net incremental economic benefits (Table

29).

Table 29. Sensitivity analysis for the Bintuli WWT project

Change in net economic benefit

Changes in capital costs

‐30% ‐15% 0% +15% +30%

‐30% 23 25 27 29 31

‐15% 20 22 24 25 27

0% 17 19 21 23 24

+15% 15 17 19 21 22

+30% 14 16 17 19 20

Changes in O&M costs

‐30% 19 21 23 25 26

‐15% 18 20 22 24 25

0% 17 19 21 23 24

+15% 16 18 20 22 23

+30% 15 17 19 21 22

The sensitivity analysis indicated that IRR is robust whereby a 30% decline in economic

benefits reduced IRR to 17% assuming no change in capital and O&M costs. Similarly, a

30% increase in capital costs assuming no change in economic benefits reduced the IRR

to 17%. Furthermore, a 30% increase in operating costs reduced IRR to 19%. Thus, it can

be concluded that the estimate is insensitive to large changes in the projected economic

costs and benefits.

Risk analysis is suitable in the cases where the values of several parameters are

uncertain. Risk analysis involves the use of the probabilities of occurrence of the key

variables as weights to recompute the project performance criteria. This is carried out

using special purpose computer packages such as ‘@RISK’, which generates probability

distributions for NPV and IRR. A major difficulty in risk analysis is obtaining probability

estimates. Common probability distributions include uniform distributions, which

require minimum and maximum estimates, and triangular distributions, which require

most pessimistic (minimum), most likely (mode), and most optimistic (maximum)

estimates. Beta distributions may also be used.

Finally, based on the cost‐benefit analysis conducted for the proposed Bintuli WWT

project, it can be concluded that water pollution in Bintuli is a serious problem and that

project implementation is urgently required to protect the health of the community and

115

reduce environmental degradation. The proposed project is expected to yield

substantial economic benefits. The IRR was estimated at 21% and sensitivity and risk

analysis indicated that the estimate is insensitive to costs and benefits. Therefore it is

recommended to implement the project with the institution of a good monitoring

program.

13.5 Cost Cost Effectiveness Analysis

Although CBA is a useful tool to assist decision‐making, it may not be a suitable

approach in all situations. When it is not possible to value a project's major benefits in

dollar terms, or when two projects have similar economic benefits, then a cost

effectiveness analysis (CEA) may be used. For example, if the decision problem is to

choose between building two hospitals, a CEA would be appropriate since the social

benefits in either case would be similar. Both CBA and CEA are based on the principle of

economic efficiency and therefore do not consider equity or distributional issues. That

is, a project is deemed to be socially desirable if the gainers can potentially compensate

the losers. They both do not deal with the issue of who the losers are or how they

should be compensated. Cost effectiveness analysis looks only at financial costs. A CEA takes the objective as

given, and then works out the costs of the alternative ways of achieving that objective.

The decision on whether to use CEA instead of CBA will depend on a number of factors

including the following: − The size and complexity of the project;

− The extent to which there are quantifiable benefits; and

− The extent to which the benefits can be valued in monetary terms.

For large‐scale projects CBA is the preferred approach because it enables the major

items of costs and benefits to be identified and valued and discounted cash flow

performance criteria to be computed. However, in cases where the major benefits

cannot be quantified in dollar terms, CEA is the preferred approach. CEA is also

appropriate in a case where the choice is between, say, two wastewater treatment

options with the same outputs or service levels but the difference is in, say, location.

Most of the foregoing discussion on CBA applies generally to CEA. Unlike CBA, CEA does

not have absolute criteria by which to judge the economic viability of projects. CEA is

therefore not recommended when a decision about the level of output or service to be

provided is at issue.

116

13.5.1 Conducting a Cost‐Effectiveness Analysis The steps involved in carrying out a CEA are similar to those for a CBA. They include:

1. Project definition

2. Choice of method of analysis

3. Identification and valuation of costs and benefits

4. Discounted cash flow analysis

5. Calculation of measures of effectiveness, and

6. Sensitivity analysis

THE VALUE OF LIFE AND HEALTH

Session 14

Region

al Training Worksho

p on

: Th

e Co

st of E

nviron

men

tal D

egrada

tion

Metho

dology

117

SESSION 14

14 THE VALUE OF LIFE AND HEALTH

Environmental changes, particularly increased pollution, often result in adverse impacts

on human health, which can be translated into monetary values. In estimating monetary

values of changes in human health associated with environmental changes, two links

need to be established (Figure 30).

Figure 30. The process of health impact valuation

The first link is between environmental change and change in health status. This involves

measuring health impacts and establishing dose‐response relations and calculating the

burden of disease (BoD) through disability adjusted life years (DALYs). The second link is

between the change in health status and its monetary equivalent, which involves

establishing willingness to pay values.

14.1 Measuring health impacts

Health impacts of pollution may be well recognized. Air pollution affects human health

in a variety of ways, from itchy eyes and chest discomfort, to chronic bronchitis and

asthma attacks. Inadequate water supply and sanitation affects human health through

diarrhea, intestinal nematodes, and other diseases. Health impacts are measured

through various types of studies including, epidemiology and field studies, human

clinical studies, and laboratory and toxicology studies.

14.1.1 Epidemiology and field studies

Epidemiological and field studies involve estimating a statistical relationship between

the frequencies of specific health effects observed in a study population and measured

levels of pollutants. There are two main types of epidemiological studies: the cohort

studies and the population studies. Cohort studies analyze the incidence of health

effects in a sample of identified individuals usually selected specifically for the study.

They allow better control of risk factors since characteristics of individuals are well

known. Population studies rely on the data available for the population as a whole

rather than tracking the effects on specific individuals. These studies are readily

available and cost‐effective.

118

Advantages of epidemiology and field studies include:

− Providing sufficient information to infer a concentration‐response function used to

predict a change in the number of cases of a given health effect and pollutant

concentration

− Defining, health effects in terms of factors that can be directly related to perceived

welfare, including risks of premature death and days with noticeable symptoms.

The main limitation associated with such a type of studies is the uncertainty about

whether the causal factors for the observed association with health effects has been

fully and accurately specified.

14.1.2 Human clinical studies

Human clinical studies examine the response of human subjects to pollutant exposure in

a controlled laboratory setting. Such studies can provide evidence of causation because

confounding variables are well controlled. They have the advantage of providing more

accurate dose‐response information. However, the application of human clinical studies

is limited to considerations of short‐term reversible health effects that can be induced

on purpose in human subjects and it requires assumptions to link human exposure in

real life to health effect observed in a laboratory setting.

14.1.3 Laboratory and toxicology studies Laboratory and toxicology studies use animal subjects and human tissue or cells to study

biological responses to pollutants in a controlled laboratory setting. They provide

important information about specific biological pathways and mechanisms by which

pollutants cause harm to living organisms. Laboratory and toxicological studies has the

advantage of well‐controlled pollutant exposures and reduced variations in confounding

factors. In addition, they can consider both long term and short term exposures.

However, analysis and assumptions are required to link human exposure in real‐life to

laboratory exposure. Furthermore, these times of studies are associated with

uncertainty in extrapolating data from animal subjects to human populations. They

sometimes focus on health effects that are difficult to interpret in terms of specific

symptoms

The aim of these types of studies is to establish dose‐response relations (DRR) linking

environmental variables with observable health effects, particularly in the case of air

pollution. DRRs correlate mortality and morbidity outcomes for susceptible population

groups with ambient concentration of a given air pollutant. Most of the conducted

studies have focused on mortality effects. For instance, chronic exposure to PM results

119

in exacerbated respiratory illness, pulmonary disease and cardiovascular disease,

culminating in premature mortality. Similarly, acute exposure to PM affects individuals

in a weakened state or especially susceptible, resulting in premature mortality. An

example of estimated DRRs for air pollutants are presented in Table 30.

Table 30. Estimated increments in annual health effects associated with increments in air pollutants

(IDIEN, 1998)

Outcome PM10

(10μg/m3

)

SO2

(10μg/m3) Ozone(pphm)

Lead (1 mg/m3)

NO2

(pphm)

Premature mortality (% change) 0.96 0.48

Premature mortality/ 100,000 6.72

Respiratory hospital admissions/100,000 12 7.7

Emergency room visits/100,000 235.4

Restricted activity days/person 0.575

Lower respiratory illness/child 0.016

Asthma symptoms/asthmatic 0.326 0.68

Respiratory symptoms/person 1.83 0.55

Chronic bronchitis/100,000 61.2

Minor restricted activity days/person 0.34

Respiratory symptoms/1,000 children 0.18

Respiratory symptoms per adults 0.1 0.1

Eye irritations/person 0.266

14.2 Burden of Disease

A BoD study aims to quantify the burden of premature mortality and disability for major

diseases or disease groups. It uses a summary measure of population health (DALY) to

combine estimates of the years of life lost and years lived with disability. Data are

broken down by age, sex, and region. The Global Burden of Disease (GBD) constituted

the most comprehensive set of estimates of mortality and morbidity yet produced

(Murray and Lopez, 1996). The World Health Organization (WHO) now regularly

develops BoD estimates at regional and global level for a set of more than 135 causes of

disease and injury. National BoD studies involve obtaining country‐specific estimates for

input to national policy. The following section discusses the calculation of DALYs, the

unit measures of BoD.

120

14.2.1 Disability Adjusted Life Years The DALY measures health gaps as opposed to health expectancies, using time

measures. It estimates the difference between a current situation and an ideal situation

where everyone lives up to the age of standard life expectancy. DALY is based on two

key value choices:

− How long should people in good health expect to live?

− How should we compare years of life lost through deaths with years lived with poor

health or disability of various levels of severity?

DALY combines in one measure the time lived with disability and the time lost due to

premature mortality:

DALY = YLL + YLD

Where:

YLL = years of life lost due to premature mortality

YLD = years of life lost due to disability

YLL corresponds to the number of deaths multiplied by the standard life expectancy at

the age at which death occurs:

YLL = N × L Where:

N = number of deaths

L = standard life expectancy at age of death in years

YLD is estimated by measuring the incidence of disability and the average duration of

each disability. The number of disabilities is multiplied by the average weight factor that

reflects the severity of the disease on a scale from 0 (perfect health) to 1(dead). Thus

the Years of Life with Disability (without applying social preferences):

YLD = I × DW × L

Where:

I = number of incident cases

DW = disability weight

L = average duration of disability (years)

121

Disability weights quantify societal preferences for different health states. It is important

to note that disability weights DO NOT represent the lived experience of any disability or

health state and they DO NOT imply any societal value for the person in the disability.

For example, a weight for paraplegia of 0.57 does NOT mean that the person in this

health state is half dead, or that the person experiences life as half way between life and

death, nor that society values them less as a person compared to healthy people. It

rather means that society judges a year with blindness (0.43) to be preferable than a

year with paraplegia. It also means that society would prefer living for 3 years followed

by death (1.7 lost healthy years) than have one year of good health followed by death (2

lost healthy years). Disability weights for various diseases calculated by Murray and

Lopez (1996) are presented in Table 31.

Table 31. Disability weights (Murray and Lopez, 1996)

Disease Mean disability weight


AIDS 0.50 Asthma, cases 0.10

Infertility 0.18 Deafness 0.22

Diarrhea disease, episode 0.11 Brain injury, long term 0.41

Measles episode 0.15 Spinal cord injury 0.73

Tuberculosis 0.27 Sprains 0.06

Malaria episode 0.20 Burns (> 60%) long term 0.25

Cancer, terminal stages 0.81 Congestive heart failure 0.32

Parkinson disease cases 0.39 Benign prostatic hypertrophy 0.04

Alzheimer disease cases 0.64

Other social values that could be taken into account when calculating DALYs are age and

time. Age weights are sometimes used whereby a year of healthy life lived at younger

and older ages was weighted lower than for other ages. This was based on the fact that

various studies have shown a broad social preference to value a year lived by a young

adult more than a year lived by a young child or lived at older ages. However, age

weights in DALYs are controversial. Time discounting may also be used whereby the net

present value of lives lost is estimated using a 3% discount rate. This is based on studies

showing that people prefer a healthy year of life immediately, rather than in the future.

However, it is important to note that BoD studies may or may not include time

discounting and age weights depending on local preference

122

Calculating DALYs with a 3% discount rate:

)1( e rL

rNYLL −−=

Where:

N = number of deaths

L = standard life expectancy at age of death

r = discount rate (0.03)

)1( e rL

rLDWIYLD −−

××=

Where:

I = number of incident cases

DW = disability weight

L = duration of disability in years


Calculating DALYs with age weight and a 3% discount rate:

)1(1]]1)([]1))(([[)(

)())((2 e rLaraLr

ra

rKareaLre

rKCeYLL −+−++− −

−+−+−−−++−

+= ββ

βββ

Where:

a = age of death (years)


β = age weighting constant (Ex: β = 0.04) K = age weighting modulation constant (Ex: K =1)

C= adjustment constant for age‐weights (Ex: C = 0.1658)

L = standard life expectancy at age of death (years)

)}1(1]]1)([]1))(([[)(

{ )())((2 e rLaraLr

ra

rKareaLre

rKCeDWYLD −+−++− −

−+−+−−−++−

+= ββ

βββ

Usually DALYs adjusted for age and time are calculated using MS Excel spreadsheets,

Examples of DALY spreadsheets are presented below.

123

Figure 31. Calculating YLL for diarrhea

124

Figure 32. Calculating YLD for Alzheimer

14.3 Valuating Health Impacts

After measuring the health impacts of pollution, the established DRRs and/or DALYs

need to be monetized. Several methods may be used to value health impacts using the

WTP approach (Figure 33). Many of these methods have been discussed in detail in

previous chapters and thus will only be mentioned briefly.

125

Figure 33. Methods for valuing health impacts

14.3.1 The Human Capital Approach (HCA)

The HCA considers individuals as units of human capital that produce goods and services

for society. It values human life and time spent ill or recovering using forgone earnings.

As such, it measures loss of productivity resulting from an individual’s death (Work Loss

Days‐WLD) and injury (Restricted Activity Days‐RAD)

HCA = (# of Life Years Lost due to premature death or due to illness) × (Average Wage Rate)

WLD and/or RAD are either estimated for specific individuals in a detailed study or for

average individuals. The latter is most commonly applied. HCA usually provides a lower‐

bound estimate, as it does not account for pain and discomfort accompanying a certain

illness.

HCA values calculated are dependent on income, skill level, and country of residence.

Accordingly, this method is considered as the most difficult and controversial aspect of

valuing health effects associated with environmental changes. Table 32 presents the

human capital and mortality costs by age in the US for the year 1992. Cost estimates are

based on life‐expectancy at the time of death and include labor‐force participation rates,

average earnings, the value of home‐making services, and a 6% discount rate.

126

Table 32. Human capital and mortality cost by age in the US

Age group (yrs) Life years lost Mortality cost (1992 US$)

< 5 75 502,421

5‐14 68 671,889

15‐24 57 873,096

25‐44 42 785,580

45‐64 25 278,350

65+ 10 22,977

The following steps need to be followed when applying the HCA

1. Specify the type of economy for the population of interest

2. Specify the characteristics of the economy for the population of interest

3. Specify the family and community structure

4. Specify the unit of analysis

5. Specify the desired measure of productivity changes

6. Estimate the maximum loss in productive time as a result of the health outcome.

This requires information as to the groups of patients that are working and requires

decisions about value of time of children and retired people

There are various problems associated with the HCA. This approach faces difficulty in

accurately estimating forgone earnings, since employee’s compensation includes

pension plans, health insurance, flexible hours, and not just wages. Furthermore, the

HCA does not provide information about the individual’s WTP to reduce probability of

loss of life. It also does not measure net contribution to society. It assumes full

employment and no substitutability of labor. It also assumes a dominant cash economy

where market prices exist, which is not the case in developing countries. HCA also

ignores non‐market activities important to individuals. It undervalues retired people,

children, and home‐makers, and it does not value pain and suffering, the individual’s

own well‐being and preferences, and the sentiments of the society. Finally, the

estimated value using the HCA highly depends on the discount rate used. The higher the

discount rate, the lower the economic value of children.

Several issues need to be considered when applying the HCA:

− Uncertainty about the number of days or years an individual actually takes off work

which requires an assumption about life‐expectancy

− Productivity estimations do not consider the declining economic value as people get

older

127

− WLD and RAD depend on the individual and how he/she responds to symptoms and

illness

− Lack of labor market data in many developing countries

− Data cannot be generalized between populations and countries as values are highly

dependent on local factors

14.3.2 Cost of Illness (COI) Approach The cost of illness approach involves measuring two types of costs, (1) the direct costs or

the costs of medication, hospitalization, and doctors’ visits, and (2) the indirect costs or

the forgone labor earnings due to days spent in bed, days missed from work, and days

when activity was restricted due to illness. The latter are calculated following the HCA

approach mentioned earlier.

The COI approach is considered a useful economic tool as it indicates the direction and

magnitude of the economic flows resulting from health shocks to the economy. It is

easily understood and often readily available being based on available market and

expenditure data. However, COI provides an estimate of an individual welfare loss.

Direct expenditures do not correspond to a drop in income or consumption for the

economy as a whole, but constitute a redirection of economic activity, with some

sectors benefiting from increased activity. Furthermore, COI does not provide a direct

measure of disease severity. Direct medical expenditures are influenced by income

distribution, whereby increased income is accompanied with increased consumption of

health care. Thus direct medical expenditures reflect the ability of current medical

techniques to treat the disease under consideration. For example, treatment of malaria

is expected to generate less expenditure than treatment of cold because the former has

few remedies as compared to the latter. The COI not only measures disease severity but

also the population’s education, skill level, income, insurance coverage, types of medical

interventions currently available, etc.

There are various issues pertaining to its application, including:

− Difficulty to disaggregate hospital payments, including drugs administered on the

premise and salaries paid to health professionals and staff

− Inaccuracies in hospital diagnostic data and the fact that expenses might not be

attributed to the correct illness

− A number of illnesses may be grouped under one diagnostic code making it hard to

decipher individual expenses

− Large data sets assume the same charge for all types of physician services. For

example, a visit for a routine checkup does not cost the same as a visit for cancer

128

− Treatment of multiple conditions where all expenses are allocated to the patient’s

primary condition

The following steps are recommended when estimating the direct cost of medical care

(WASH, 1991):

− Estimate the proportion of those affected at each level of severity of the disease

− Estimate the proportion of those desiring treatment and who have access to

treatment

− Specify the process of treatment for each level of severity of the disease

o Resource use

o Number of inpatient days

o Outpatient visits

− Estimate the unit costs of resources used for treatment and the side effects for each

level of severity of the disease taking into account that many fixed costs are not

affected by reductions in the use of the health service

− Estimate total treatment costs for each level of severity of the disease without

intervention

− Determine the proportion of the costs that can be avoided in the short‐ and long‐run

− Determine the direct costs that would have been avoided

14.3.3 Hedonic Pricing Hedonic pricing involves the valuation of incremental morbidity or mortality by

identifying wage differentials due to risk differences. It is based on the theory that

workers have to be paid a premium to undertake jobs that are inherently risky, which

can be used to estimate the implicit value individuals place on sickness or premature

death. It assumes that there is a fixed supply of jobs and a freely functioning job market

where individuals choose jobs based on perfect information and with no mobility

restrictions. The value of a statistical life in the US, estimated using the hedonic pricing

method ranges between 1.9‐10.7 million USD (1990 dollars).

In the HPM, calculation are based on the assumptions that

− The only difference between two jobs is the level of risk

− The attitudes to risk are identical between individuals

− Labor markets are competitive

− Individuals only take risky jobs because they pay more

− Individuals correctly perceive risk

129

The general hedonic wage equation is:

P = P(J,R,S)

Where:

P = payment rate for a given job

S = vector of skills required to do the job

J = vector of other job‐related attributes (working hours, holiday, sickness benefits)

R = risk of death

The partial differential of this function with respect to R gives an estimate of the

additional payment required by individuals to accept a marginal increase in the chance

of death.

Issues and limitations associated with HPM:

− Faces difficulty in assessing an objective measure of the risk of death

− Contains a high degree of uncertainty

− Requires considerable data sets for regression analysis, containing data on all

relevant and confounding variables

− Results are not transferable between countries due to differences in attitudes to risk

and incomes

14.3.4 The Contingent Valuation Method

The CVM has a great potential for eliciting WTP for environmental health interventions,

including developing countries. One main advantage of the CVM is that questions can be

structured so that respondents can value only the benefits of interest. As such, health,

amenity, and non‐use benefits can be separated for the same environmental health

intervention. Accordingly,

Total Benefit/Cost

= ∑WTP of all concerned members of the society

= Value of Statistical Life (VOSL)

Main steps in a CVM process include:

1. Define the sample of respondents

2. Give respondents a detailed description of the hypothetical market and the good

being evaluated

3. Ask respondents the price they are willing to pay to receive the amenity

130

4. Collect information on demographics and socio‐economic characteristics of

respondents

5. Estimate aggregate WTP

Advantages of a CVM

− Can take into account non‐use values

− Can be designed to include only the variables or characteristics of the market

relevant to the objective of the study

− Allows individuals to consider the true costs to themselves of a particular injury or

illness

− CVM results are repeatable in terms of similarity in results across different settings

and using a test‐retest methodology

Problems associated with CVM

− Does not require cash transactions

− Biases: Strategic, design, hypothetical, etc.

− Survey responses cannot be verified except through comparison with actual

behavior following survey

− WTP vs. WTA

− Short time given to respondents to think about the answer

− In developing countries, questionnaires need to be adapted carefully and trained

researchers are required to administer the surveys

Issues to consider when conducting a CVM study:

− WTP questions should be clear and unambiguous

− Respondents must be familiar with the valued commodity

− Health risk studies involving common, mild illnesses have a greater chance of being

understandable, meaningful, plausible, than severe, rare diseases

− Respondents should have prior valuation/ choice experience with respect to

consumption levels of the commodity in order to give it well‐formed values

14.3.5 Benefit transfer Values may be adopted from the other countries by adjusting for per capita income as

follows:

131

Pe r capita incom e of country i = X i

⇒ Income ra tio Xj/X i

Pe r capita incom e of country j = Xj

⇓

V alue of mortality or morbi dity outcome in country i = Y i

⇒ Multip ly Y i by Xj /X i ⇒ Value of morta lity or morbid ity outcome in country j = Yj

14.3.6 Disability Adjusted Life Years The VOSL obtained from wage differential and contingent valuation studies may be

linked with the corresponding number of DALYs lost in a specific study and so estimate

the implicit value per DALY.

The cost of a DALY lost may be valued by two approaches:

1. DALY (yrs) × GDP/capita (USD/year)

This is based on the rationale that the economic value of a year lost to illness or early

death is the productive value of that year, which is approximated by GDP per capita. It

usually represents the lower bound estimate and has nothing to do with the non‐

economic value of life in general

2. DALY (yrs) × WTP for mortality reduction

This is based on the WTP by an individual to reduce the risk of death. Valuations arrived

at, in studies in the United States and Europe that apply WTP, are substantially higher

than the GDP per capita approach (at least for adults).

Finally recommended methods of valuation for health related benefits are summarized

in Table 33.

132

Table 33. Recommended methods for valuation of health related benefits

Types of benefits Market value(COI, HCA)

Avertive expenditure

Hedonic pricing

Contingent valuation

Improved health‐related quality of life

Improved life expectancy

Medical cost avoided ( )

Reduced time spent in care ( )

Reduced travel expenses to care ( )

Reduced avertive expenditure ( ) ( )

Increased productivity ( )

Reduced sick leave ( )

THE VALUE OF LIFE AND HEALTH: CASE‐STUDIES

Session 15

Region

al Training Worksho

p on

: Th

e Co

st of E

nviron

men

tal D

egrada

tion

Metho

dology

133

SESSION 15

CASE STUDIES

These case‐studies will be distributed as handouts to the workshop participants.

Drinking water quality in Lebanon

El‐Fadel, M., Maroun, R., Semerjian, L., and Harajli, H. A health‐based socio‐economic

assessment of drinking water quality, Management of Environmental Quality, 14, 3, 353‐

368, 2003. (Literati Awards of Excellence 2004)

Emissions from the cement industry, Lebanon

El‐Fadel, M., Kobrossi, R., and Metni, M. Economic benefits of reducing SO2 emissions from the

cement industry, Journal of Environmental Assessment Policy and Management, 5, 1, 99‐

120, 2003.

Particulate matter in urban areas, Lebanon

El‐Fadel, M., Massoud, M. Particulate matter in urban areas: Health based economic

assessment, The Science of Total Environment, 257, (2‐3) pp. 133‐146, 2000.

Lead phase‐out in Lebanon

Hashisho, Z. and El‐Fadel, M. A case study in socio‐economic benefits of the phase‐out of leaded

gasoline, Environmental Management and Health, 12, 4, 389‐406, 2001.

134


THE VALUE OF LIFE AND HEALTH GROUP EXERCISES

Sessions 16 & 17

Region

al Training Worksho

p on

: Th

e Co

st of E

nviron

men

tal D

egrada

tion

Metho

dology

135

SESSIONS 16 & 17

GROUP EXERCISES

Urban air pollution from particulates in selected MENA countries

Sarraf, M., Larsen, B., and Owaygen, M. 2004. Cost of environmental degradation: The case of

Lebanon and Tunisia. Environment Department Paper No. 97, The World Bank, Washington

D.C.

World Bank, 2002. Cost Assessment of Environmental Degradation in the Arab Republic of Egypt.

Sector Note. Report No. 25175 –EGT. Rural Development, Water and Environment

Department, Middle East and North Africa Region, The World Bank.

World Bank, 2003. Cost Assessment of Environmental Degradation in the Kingdom of Morocco.

Report No. 25992‐MOR. Water, Environment, Social and Rural Development Department,

Middle East and North Africa Region, The World Bank.

World Bank, 2004. Cost Assessment of Environmental Degradation in the Syrian Arab Republic.

World Bank. METAP.

Water, sanitation and hygiene in selected MENA countries

Sarraf, M., Larsen, B., and Owaygen, M. 2004. Cost of environmental degradation: The case of

Lebanon and Tunisia. Environment Department Paper No. 97, The World Bank, Washington

D.C.

World Bank, 2002. Cost Assessment of Environmental Degradation in the Arab Republic of Egypt.

Sector Note. Report No. 25175 –EGT. Rural Development, Water and Environment

Department, Middle East and North Africa Region, The World Bank.

World Bank, 2003. Cost Assessment of Environmental Degradation in the Kingdom of Morocco.

Report No. 25992‐MOR. Water, Environment, Social and Rural Development Department,

Middle East and North Africa Region, The World Bank.

World Bank, 2004. Cost Assessment of Environmental Degradation in the Syrian Arab Republic.

World Bank. METAP.

136


WRAP‐UP CASE: THE JULY 2006 WAR IN LEBANON

& POLICY IMPLICATIONS AND

CONCLUSIONS

Session 18

Region

al Training Worksho

p on

: Th

e Co

st of E

nviron

men

tal D

egrada

tion

Metho

dology

137

SESSION 18

15 POLICY IMPLICATIONS AND CONCLUSIONS

Environmental values are used in policy and project appraisal in a number of ways. They

are less routinely incorporated into policy and project appraisal in a systematic way.

Environmental changes tend to be assessed through Environmental Impact Assessment

(EIA) rather than through the estimation of changes in environmental values and cost‐

benefit analysis. There are many instances where the environmental impacts of projects

are only described or enumerated in physical terms with no monetary values attached to

them. This leaves the decision‐makers to make intuitive judgments on whether the

welfare gains from the project will outweigh the ensuing environmental degradation.

The environmental impact assessment study quantifies and describes the physical

impact of projects and policies and documents complexity of an environmental issue.

However, it fails to help the decision‐maker who has little knowledge of how

environmental changes affect the utility of the individual. As such, environmental

valuation gives the ‘true’ value of environmental resources to the society and tends to

remove ambiguity and vagueness in the decision‐making process. However, care should

be taken in order not to apply environmental valuation in order to maximize benefits in

order to justify a policy or in order to minimize the estimated externality values of a

project to ensure its approval.

Environmental impacts should be valued in monetary terms in order that they are given

due and proper weight in the decision‐making process. The non‐monetization of

environmental impacts may mean that either they are under‐valued or over‐valued in

the intuitive decision‐making process. Monetization will permit the comparison of

various environmental management proposals. Many studies revealed the inconsistency

of intuitive decision‐making compared with a more structured approach of

environmental valuation. The numerous cognitive psychological biases in intuitive

decisions render rational choice problematic.

Environmental values have been used limitedly in decision‐making due to many factors,

including:

− Skepticism towards environmental valuation methods

− Lack of environmental economists within government agencies

− Absence of a legal requirement to undertake a CBA of projects or policies

− Uncritical acceptance of other methods such as

138

− Effect on production

− Dose response

− Opportunity cost approaches

− Suspicion of non‐use values

− Distorted perceptions of the valuation methods by non‐economist

− Large variance associated with mean WTP and WTA values

Currently, environmental values are less routinely incorporated into policy and project

appraisal in a systematic way. Environmental changes tend to be assessed through EIAs

in the US and EU rather than through economic valuation and CBA. The World Bank and

the Asian Development Bank advocate the use of valuation methods to estimate the

welfare effects of environmental changes. Environmental valuation studies in different

European countries were undertaken spasmodically with varying degrees of influence on

decisions and with marked variations between countries. For instance, Switzerland has

produced a number of academic/ scientific employing TCMs, HPMs, CVMs, all being

applied, and researchers receiving funds from a variety of institutions. Studies in

Germany have been proportionately fewer and more policy oriented. In the UK, a shift

away from TCMs to HPMs and CVMs occurred in the 1990s, which was mostly attributed

to the nature of the goods being valued. In the Netherlands, while academic interest in

environmental economics is strong, demand for valuation studies by governments and

organizations is low. In Norway, benefit estimation studies provided support for

environmental decision‐making but had not played a crucial role in the process.

In some cases, the environmental valuation process is formalized and fairly explicit and

institutionally incorporated in the decision‐making process. For example, at the US

Forest Service, there is an explicit inclusion of environmental values in the application of

‘unit day values’ of recreational opportunities and resources. Environmental values are

also explicitly included in Type A assessment of natural resource damage from pollution

spills under CERCLA legislation. Economic damages are calculated from an economic

database in which injuries and losses to particular species of fish, water fowls, etc. are

measured as reductions in harvesting or in recreational use values. For major pollution

incidents, a Type B assessment under CERCLA requires a site specific investigation.

Furthermore, the US Department of Interior regulations authorized methods for

environmental valuation. Where a reasonably competitive market exists for a resource,

market price is used to estimate economic damage. If market prices are not appropriate,

appraisal can be based on Uniform Appraisal Standards for Federal Law Acquisition.

Where neither of these approaches is appropriate, environmental valuation methods

are adopted. Use values may be measured via TCM, HPM, unit values, CVM, and stated

139

preference techniques. Non‐use values may only be measured via CVM and stated

preference techniques.

The desire to establish formal benefit transfer methods by governments and agencies

and the advocacy for the use of benefit transfer by organizations will result in

environmental valuation methods that are more institutionalized and more routinely

included in CBAs. In the near future, environmental valuation will witness a search for

more accurate and robust semi‐ and non‐parametric estimators, improved

understanding of the psychology of making choices and decisions, the analysis of the

non‐stationarity of environmental values, and the application of other theories and

techniques from other branches of economics, such as the Bayesian Perspectives and

the Game Theory.

In the MENA region, effort should be direct to increase awareness on environmental

valuation, to build capacity on COED methodology, to institutionalize COED

methodology in decision‐making process, and to establish a database for environmental

valuation studies in the region. Various databases on environmental valuation could be

of help, as listed in Table 34. Other useful websites are listed in Table 35.

140

Table 34. Main features of selected valuation databases (McComb et al, 2006)

Name of database Web host Purpose of the database Number of studies

Regions covered

Available languages

Environmental Valuation Reference Inventory

Environment Canada on behalf of the EVRI Club1 http://www.evri.ca

To help policy analysts using the benefits transfer approach to estimate economic values for changes in environmental goods and services or human health

1,500 International English, French

Envalue New South Wales Environment Protection Authority http://www.epa.nsw.gov.au/envalue

To help stakeholders value changes in environmental quality

400 International English

Ecosystem Services Database

Gund Institute for Ecological Economics, University of Vermont http://esd.uvm.edu

To provide a data and analysis portal to assist in the informed estimation of the economic values of ecosystem services


Review of Externality Data

European Commissionhttp://www.red‐externalities.net

To assist policy makers in capturing the effects of externalities from new policies that have sustainable development as their core concern


New Zealand Non‐market Valuation Database

Lincoln University, Canterbury, New Zealandhttp://oldlearn.lincoln.ac.nz.markval

To help researchers identify nonmarket valuation studies undertaken in New Zealand

Searchable database with 100 primary studies from New Zealand

ValuebaseSwe Beijer International Institute of Ecological Economics, and the Swedish Environmental. Protection Agency http://www.beijer.kva.se/valuebase.htm

To provide a survey of empirical economic valuation studies on environmental change in Sweden

Database with 200 primary studies from Sweden

Beneficial Use Values database

Department of Agricultural and Resource Economics, University of California, Davis http://buvd.ucdavis.edu/

A guide for decision makers, policy analysts, and others interested in valuation of water resources

Database of economic values for beneficial uses of water. Variety of sources

Sportfishing Values database

Industrial Economics, Incorporated under contract to the U.S. Fish and Wildlife Service http://www.indecon.com/fish/default.asp

To provide a detailed account of the contents of numerous recent non‐market valuation studies

One hundred non‐market valuation studies of sports fishing activity

141

Table 35. Useful web resources for environmental economists

Name Description Website

Association of Environmental and Resource Economists (AERE)

AERE was established as a means of exchanging ideas, stimulating research, and promoting graduate training in resource and environmental economics. AERE provides many forums for exchanging ideas relevant to the allocation and management of natural and environmental resources and has two journals, the Journal of Environmental Economics and Management (JEEM), and the Review of Environmental Economics and Policy (REEP), and a newsletter issued to members twice a year.

www.aere.org

European Association of Environmental and Resource Economics (EAERE)

EAERE is an international scientific association which aims to contribute to the development and application of environmental and resource economics as a science in Europe, to improve communication and contacts between teachers, researchers and students in environmental and resource economics in different European countries, and to develop and encourage cooperation between university level teaching institutions and research institutions in Europe.

http://www.eaere.org

South Asian Network for Development and Environmental Economics (SANDEE)

SANDEE is a regional network that seeks to bring together analysts from different countries in South Asia to address its development‐environment problems. SANDEE's mission is to strengthen the capacity of individuals and institutions in South Asia to undertake research on the inter‐linkages among economic development, poverty, and environmental change and to disseminate practical information that can be applied to development policies.

http://www.sandeeonline.org

The Economy and Environment Program for Southeast Asia (EEPSEA)

EEPSEA is similar to SANDEE and supports training and research in environmental & resource economics in South East Asia. This web site offers downloadable Research reports on issues relevant to developing countries.

http://www.idrc.org/eepsea

Latin American and Caribbean Environmental Economics Program (LACEEP)

LACEEP was launched with a grant from IDRC. The program will operate in ways similar to EEPSEA and its South Asian counterpart SANDEE, offering research awards, short courses, workshops and mentoring.

http://www.laceep.org

Middle East and North Africa Network for Environmental Economists (MENANEE)

MENANEE is a joint venture between the Beijer Institute of Resource Economics and the Library of Alexandria. It is considered as a regional network that aims at strengthening the capacity of individuals and institutions in the region in the field of environmental and resources economics. It also intends to highlight to policy and decisions‐makers the linkages between economic development and environmental changes.

http://www.bibalex.com/MENANEE/Home/Home.aspx

142

Name Description Website

Center for Environmental Economics and Policy in Africa (CEEPA)

The mission of CEEPA is to enhance the capacity of African researchers to conduct environmental economics and policy inquiry of relevance to African problems and increase the awareness of environmental and economic managers and policy makers of the role of environmental economics in sustainable development.

http://www.ceepa.co.za/mo.html

Society for Environmental Economics and Policy Studies (SEEPS)

SEEPS is a scientific association which aims to contribute to the theoretical and empirical research of environmental economics and policy studies; to improve communication and contacts between teachers, researchers and students in environmental economics and policy studies; to promote international scientific cooperation in environmental economics and policy studies.

http://wwwsoc.nii.ac.jp/seeps/eng/index.html

Environmental Economics Unit at Göteborg University, Sweden (EEU)

EEU is a research and teaching unit within the Göteborg University with graduate students working on various projects related to natural resources and environmental economics. EEU specializes in environmental economics research and training. The research is focused on natural resource management in developing countries, the choice of policy instruments for transport, industrial environmental problems and welfare related issues.

http://www.hgu.gu.se/item.aspx?id=2496

The UK Network of Environmental Economists (UKNEE)

UKNEE aims to bring together environmental economists from academia, consultancy and public and private sectors to foster closer relationships, follow recent developments and share experience. UKNEE organizes regular seminars on topical subjects in environmental economics followed by social evenings.

http://www.eftec.co.uk/UKNEE/index.htm

Asociacion Hispano Portuguesa de Economia de los Recursos Naturales y Ambientales (AERNA)

AERNA was founded in 2002 as a response to the motivation of a group of academics and researchers in the Iberian Peninsula, for exchanging ideas and knowledge, stimulating research and supporting the government and other interested groups decisions on the field of the multiple relations between economics and environment.

http://www.aerna.org/paginas.asp?id_pagina=1

International Society for Ecological Economics (ISEE)

ISEE facilitates understanding between economists and ecologist and the integration of their thinking into a trans‐discipline aimed at developing a sustainable world

http://www.ecoeco.org/

REGIONAL WORKSHOP REGIONAL WORKSHOP ONON

THE COST OF THE COST OF ENVIRONMENTAL ENVIRONMENTAL

DEGRADATION (COED) DEGRADATION (COED) METHODOLOGYMETHODOLOGY

July 1-5, 2008Crowne Plaza Hotel

Beirut, Lebanon

REGIONAL WORKSHOP ONREGIONAL WORKSHOP ON

THE COST OF ENVIRONMENTAL THE COST OF ENVIRONMENTAL DEGRADATION (COED) DEGRADATION (COED)

METHODOLOGYMETHODOLOGYJuly 1-5, 2008

Crowne Plaza HotelBeirut, Lebanon

INTRODUCTIONINTRODUCTION

• Environmental degradation one of the most prominent adverse phenomena in today’s world

• The MENA region is suffering from serious environmental problems– air pollution– water pollution– land degradation– forest and biodiversity loss– waste pollution– coastal zone degradation

Air pollutionAir pollution Land and coast degradationLand and coast degradation

Liquid Liquid waste waste

pollutionpollution

Solid waste pollutionSolid waste pollution

Economic valuation of environmental degradation

Quantification of benefits of environmental projects/

policies

Incorporating and prioritizing environmental issues in

decision-making

Environmental ValuationEnvironmental Valuation

Raising Raising awareness onawareness onenvironmental environmental

issuesissues

Progress Progress towards towards

sustainable sustainable developmentdevelopment

Cost of Environmental DegradationCost of Environmental Degradation• According to the METAP/World Bank Country studies

– US$228 million per year in Jordan and US$4.2 billion per year in Egypt

• In comparison with other countries:– OECD countries: 1-2% of GDP– India: average annual 4.5% of GDP in 1991– Mexico: average annual 3.3% of GDP– China: 8% of GDP

0123456

Tunisia(1999)

Jordan(2000)

Lebanon(2000)

Syria(2001)

Algeria(1999)

Morocco(2000)

Egypt(1999)

Perc

ent G

DP

Water Soil Air Coastal Zone Waste

Cost of Environmental DegradationCost of Environmental Degradation

• The COED in the MENA region is an environmental economics tool – developed by the METAP/World Bank– enables professionals to carry out assessments of the

economic cost of environmental degradation– successfully used in the valuation of environmental

degradation on a macroeconomic and sector levels

A main obstacle to conducting research in environmental economics is the shortage of human capacity at governmental ministries/organizations and local

universities

Capacity Building for COEDCapacity Building for COED

• Aims to enhance regional capacity in environmental economics

• Funded by the World Bank/ METAP• Implemented by AUB• Main tasks:

Environmental Economic Unit at

AUB

Training course in COED

methodology

Policy Papers on COED

•Maghreb •Mashreq

COURSE COURSE OUTLINEOUTLINE


1 1 08:30‐10:00 Participants registration Official opening Introductions and purpose of the workshop


2 10:30‐12:00 Brief overview of basic economic principles

Introduction to environmental valuation and policy implications


3 12:30‐14:00 The revealed preference approach a) The productivity method (Theory, application, advantages, limitations, case‐studies)

b) The market values approach including damage cost, replacement cost, and substitution cost methods (Theory, application, advantages, limitations, case‐studies)

14:00‐15:30 Lunch

4 15:30‐17:30 Case‐studies on the productivity method and the market values approach

2 5 08:30‐10:00 The revealed preference approach (cont’d) c) The travel cost method (Theory, application, advantages, limitations, case‐studies)


6 10:30‐12:00 The revealed preference approach (cont’d) d) The hedonic pricing method (Theory, application, advantages, limitations, case‐studies)

e) The aversive behavior method (Theory, application, advantages, limitations, case‐studies)


7 12:30‐14:00 Group Exercises: Ayubia National Park In Pakistan (Travel Cost) Non‐Priced Forest Recreation Areas In Malaysia (Travel Cost) Valuing Landscape and Amenity Attributes In Central England (Hedonic Pricing)

14:00‐15:30 Lunch


COURSE COURSE OUTLINEOUTLINE


3 9 08:30‐10:00 The stated preference approach a) The contingent valuation method (Theory, application, advantages, limitations, case‐studies)


10 10:30‐12:00 The stated preference approach (cont’d) b) The discrete choice method (Theory, application, advantages, limitations, case‐studies)

The benefit transfer method (Theory, application, advantages, limitations, case‐studies)


11 12:30‐14:00 Group exercise: Stated preference approach Air quality in Beijing Ecosystem services in Ejina China Environmental services in the Yaqui River Delta, Mexico Sustainable development in Sweden coastal zone Coastal ecosystems in Phang Nga Bay, Thailand

14:00‐15:30 Lunch


Case‐studies: Coastal zone in North Lebanon, Climate Change MENA region

4 13 08:30‐10:00 Cost‐benefit analysis Case‐studies: wastewater and solid waste management


14 10:30‐12:00 The value of life and health Including the burden of disease (DALY), the human capital approach, the cost of illness approach, and the contingent valuation approach

Case studies: Drinking water quality, Emissions from the cement industry Particulate matter in urban areas, Lead phase‐out


5 15 08:30‐10:00 Case studies: Drinking water quality, Emissions from the cement industry Particulate matter in urban areas, Lead phase‐out


16 10:30‐12:00 Group exercise on the value of life and health: Urban air pollution from particulates in selected MENA countries Water, sanitation and hygiene in selected MENA countries



14:00‐15:30 Lunch

18 15:30‐17:30 Wrap‐up case with various concepts: The July 2006 War in Lebanon

Policy implications and workshop conclusion

Workshop evaluation

OUTLINEOUTLINEBasic Economic

Principles Market Pricing

Production Function Travel Cost

Averting Behavior

Hedonic Pricing


Contingent Choice

Value of Life and Health

Benefit Transfer

Cost-Benefit Analysis

Decision-making and Policy

Group exercise

Group exercise

Case-studies

Group exercise

July 2006 War

Lebanon

Revealed Preference

Stated Preference

Valuation Methods EEnd of nd of SSession ession 11

Thank YouThank You



DEGRADATION DEGRADATION METHODOLOGYMETHODOLOGY

Session 2OVERVIEW OF BASIC ECONOMIC

PRINCIPLES&

INTRODUCTION TO ENVIRONMENTAL VALUATION AND POLICY IMPLICATIONS


THE COST OF ENVIRONMENTAL THE COST OF ENVIRONMENTAL DEGRADATION METHODOLOGYDEGRADATION METHODOLOGY

Session 2aOVERVIEW OF BASIC ECONOMIC PRINCIPLES

&INTRODUCTION TO ENVIRONMENTAL

VALUATION

OUTLINEOUTLINE• The competitive market

– Consumer behavior and demand– Producer behavior and supply– Market equilibrium

• Market failure– Property rights– Types of goods– Externalities– Types of market structure

• Policy failure• Pollution control approaches• Introduction to environmental valuation

The Competitive MarketThe Competitive Market

• A market is defined as the coming together of consumers and producers to exchange goods and services for money– It exists for a single good or service– It has many buyers and sellers (perfectly

competitive)– Buyers and sellers do not have to be

physically present to carry out operations


• Classification of markets according to buyers and sellers– Monopoly: a single seller

• Ex: the utilities sector in many countries where the government is the only provider of water or electricity

– Oligopoly: few sellers• Ex: the domestic car manufacturing industry in Australia

– Monopsony: a single buyer• Ex: a mine as a single major industry in a town and a sole

buyer of labor and other goods and services


• Characteristics– Many sellers and buyers with none of them highly

influential to affect the price– Buyers and sellers are free to enter or leave the

market in response to price changes– Goods and services offered for sale are

homogeneous • i.e. choice of buyer is only affected by price

– All market participants have perfect knowledge • i.e. consumers know product prices and producers know

input prices

Consumer Behavior and DemandConsumer Behavior and Demand

• Demand function– A curve indicating how

much of a good a consumer is willing to buy at various prices

• Inverse relation between price and quantity demanded

• The demand for a good is defined given that all other goods and income remain constant

• The demand curve is defined for a given period of time

B0

A

Consumer’s Demand

Quantity

Pric

e pe

r un

it C

D

The points on the curve represent the maximum amount of money a consumer is willing to pay

(WTP) for different quantities of the good


• Elasticity– The responsiveness of quantity demanded/ supplied

to changes in variables such as price and income

– Own-price elasticity of demand can be:

Perfectly elastic (εD=∞) Relatively elastic (εD >1)

Perfectly inelastic (εD =0) Relatively inelastic (εD <1)

Own-price elasticity of demand:εD = % change in quantity of q1 demanded

% change in price of q1

Price

Quantity

DPrice

Quantity

D

Price

Quantity

D

Quantity

D

Price


• Cross-price elasticity of demand– The responsiveness of the quantity of a

demanded good (q1) as a result of changes in another good (q2)

– If εD12 > 0, then q1 and q2 are substitutes

– If εD12 < 0, then q1 and q2 are complements

Cross-price elasticity of demand:εD

12 = % change in quantity of q1 demanded% change in price of p2


• Income elasticity of demand– The responsiveness of the quantity of a

demanded good (q1) given a changes in income

– If ηϒ > 0, then the good is a normal good– If ηϒ < 0, then the good is an inferior good

Income elasticity of demand:ηϒ = % change in quantity of q1 demanded

% change in income

Producer Behavior and SupplyProducer Behavior and Supply

• The production function is a function of various inputs– Labor, land, capital, etc.

• The producer’s aim is to maximize profit– The producer will increase

the production output if its price rises

• Main observations– The curve is positively

sloped• Producers supply more as

price increases– The supply curve refers to

a given point in time

0

A

Market supply curve(Marginal cost curve)

Quantity

Pric

e pe

r un

it

C

D

The marginal cost curve indicates the cost of producing each additional unit of the good. To

maximize profit, the producer increases production up to a point where marginal revenue, the price per unit of output in a

competitive market, just equals marginal cost.

Market Equilibrium in the Market Equilibrium in the Competitive MarketCompetitive Market

• The interaction of supply and demand forces in the market determine– Equilibrium price/ Market revenue (PE)– Equilibrium quantity demanded (SE)

Quantity produced0

PE

Market demand

Market supply

E

SE P

Price per unit

At price PE, market demand is exactly equal to the quantity the market is willing to supply (SE)


• If price ↑ to P’– Producer will supply S’– Consumers will demand only

D’– Demand Deficit

• To clear deficit– Producer will ↓ price– Consumers ↑ purchase– Producers ↓ supply– Demand = Supply

• If market price ↓ to PA– Consumers demand ↑ to DA– Producers not willing to supply– Surplus demand

• Shortage of supply– Upward pressure on price– Producer increases supply– Demand = Supply

Market demand

Quantity produced0

PE

Market supply

SE

Price per unit

E

PA

DAS’D’

P’


• Factors that can shift the demand curve– Income

• Income increase causes an upward (rightward shift)

• Income decrease causes a downward (leftward shift)

– Prices of substitutes/ complements

• Decrease in price of substitute causes a downward shift

• Decrease in price of complement causes an upward shift

– Consumer tastes and preferences

Quantity produced0

P1

Market demand

Market supply

S1 P

Price per unit

S2

P2

Quantity produced0

P2

Market demand

Market supply

S2 P

Price per unit

S1

P1

Incr

ease

in in

com

eD

ecre

ase

in in

com

e


• Factors that can shift the supply curve– Price of inputs, taxes,

subsidies• Input price increase

causes an inward shift• Input price decrease

causes an outward shift– Technology improvement

• Technology improvement causes an outward shift

– More output produced at same input level

– Weather improvement• For weather dependent

productions, weather deterioration causes a leftward shift Quantity produced

0

P2

D1

S1

S2 P

Price per unit

S1

P1

Incr

ease

in in

put p

rice

S2

Dec

reas

e in

inpu

t pric

e

Quantity produced0

P1

S1 P

Price per unit

S2

P2

S1

S2

D1


• Notes– Equilibrium examples simplified– Equilibrium does not tend to be static

• Demand function shifts due to changes in taste and income

• Supply function shifts due to resource constraints and technological advances

– It is assumed that property rights are well defined

• The seller has the rights to the goods and to any benefits from sale

Consumer and Producer SurplusConsumer and Producer Surplus

• Consumer surplus (Δabc)– is the maximum amount of

money consumers are willing to pay for the good or service MINUS the market price

– is a measure of net benefits or welfareThe sole reliance on the market price could result in an underestimation of benefits

• Producer surplus (Δbcd)• the net benefit received by

the producer• the difference between the

market price and marginal cost

Quantity produced0

a

Market demand

Market supply

c

SE P

Price per unit

b

d

Producer surplus

Consumer surplus

Application of the Competitive Model:Application of the Competitive Model:The socially optimal level of forestryThe socially optimal level of forestry• Clear felling of timber causes

– Loss of forest cover– Increased soil erosion– Loss of soil nutrients– Loss of biodiversity, etc.

• Effect of policy to include environmental cost to stumpage price

Application of the Competitive Model:Application of the Competitive Model:The socially optimal level of forestryThe socially optimal level of forestry• At current stumpage price

p q logs are harvested• An extra 5 USD

government charge per log– Upward shift of supply

curve from S to S’– Assuming constant

demand, • Equilibrium established at

q’• the quantity of harvested

logs decline (q’ < q)Number of harvested logs

0

p

S

q

Stumpage price ($)

5 $

5 $

S’

p’

q’

MARKET FAILUREMARKET FAILURE

• It occurs when some costs and/or benefits are not fully reflected in market prices– Common for many kinds of environmental goods

which are not usually traded in markets• Reasons for market failure

– property rights related to ecosystems and their services are often not clearly defined.

– many ecosystems provide services that are public goods

– many ecosystem services are affected by externalities

– type of market structure

Lack of or Weak Property Lack of or Weak Property RightsRights

When property rights are weak or lacking in an environmental system, there is no incentive for an individual to invest in an asset because he cannot appropriate the full benefits

• Characteristics of ownership/ property rights– Well-defined

• Formal (certificate, receipt,…)• Informal (institutionalized by social and cultural norms)

– Exclusive– Transferable– Secure and enforceable

Lack or Weak Property RightsLack or Weak Property Rights

Most environmental goods

Pure public goods or open access/common property goods

Lack of well defined property rights results in market failure

Inefficient allocation of resources

Taxonomy of Environmental GoodsTaxonomy of Environmental Goods

Goodsand

Services

Private goods Congestion goods Public goods

Open accessand common

property goodsPure public goodsSemi-public goods

Private vs Pure Public GoodsPrivate vs Pure Public Goods

• Private goods– Are exclusive– Have a positive

marginal cost• Positive cost for

supplying additional goods

– Are rival in consumption

• If one person consumes it, another cannot

• Public goods– Are non-exclusive

• Goods available to everyone

– Have zero marginal cost• zero cost for supplying

additional goods– Are non-rival in

consumption• Goods consumption does

not affect goods availability– Consumers do not have

the option of not consuming

Open Access and Common Open Access and Common Property GoodsProperty Goods

• Open access goods– Rival in consumption– Non-exclusive– Non-transferable– Non-enforceable even

when ownership rights exist

Ex: ocean fisheries and migratory wildlife

• Common property goods– Rival in consumption– Exclusive for a group of

people– Rights of use may be

transferable by individual or group

– May be enforceable through social sanctions

Ex: common grazing land

Under these regimes resources may be exploited (Hardin’s tragedy of the commons).However, under some form of common property systems, resource management islikely to be more efficient because it is based on communal rules and customs

SemiSemi--public and Congestion Goodspublic and Congestion Goods

• Semi-public goods– Non-rival in consumption– Non-exclusive– Zero marginal cost– Ownership rights exist– Consumers can choose not

to consume

Ex: TV broadcast and lighthouse

• Congestion goods– Rival or non-rival

consumption– Exclusive – May exhibit characteristics

of private or public goods at different levels of consumption

– May be enforceable through social sanctions

Ex: roads, boating sites, historic sites

ExternalitiesExternalities

• An externality exists when some agent A(individual or firm) takes an action which has an impact on another agent B, that B has not chosen to accept– It is negative when the affected person suffers a loss

in utility that is uncompensated• Ex: air, water, and noise pollution

– It is positive when the effect is beneficial• This is very rare• Ex: immunization

• Some features of externalities:– Agent B cannot choose the level of the impact like in

a normal economic transaction– The impact on B is not a result of a deliberate attempt

from A


• Causes of externalities– Interdependence between economic agents

• The market system fails to account for the interdependence, resulting in an uncompensated affected party

– Lack of or weak property rights• The affected party is unable to demand a reduction of the

externality or ask for compensation

– High transaction costs• Cost of negotiating, implementing and enforcing an

agreement

Once the affected party is compensated, the externality is Once the affected party is compensated, the externality is ‘‘internalizedinternalized’’and the society is better of by the gainer compensating the loseand the society is better of by the gainer compensating the loserr


• Types of externalities– Relevant externalities

• When the affected person is made worse off by the activity and wants the offender to reduce it

– Pareto-relevant externalities• When it is possible to take action such that the affected person is

made better off without making the offender worse off– Static vs dynamic externalities

• When the externality has adverse impacts for the future, it becomes dynamic

– Pecuniary externalities• Transmitted through the price system and is not a result of market

failure• Ex: increased rental prices in an area due to a new business

opening there• Pollution is not pecuniary because even if penalties exist

Types of Market StructureTypes of Market Structure

• The type of market structure can cause market failure– Perfectly competitive market with external

costs– Monopoly

Resource allocation in a perfectly Resource allocation in a perfectly competitive marketcompetitive market

• Illustration:– A gold mining company

dumping mine tailings in a nearby river without paying for cleanup or waste treatment

• D: Demand curve for gold

• MCp: Marginal private cost of producing gold

• MSC: Marginal social cost

– Assume: MSC > MCp since MSC = MCp + external cost of pollution

Quantity (q)0

p0

MCp

q0

Price/Cost ($)MSC

p’

q’

D

b

c

a


• Illustration (cont’d):– The company

maximizes producer surplus by producing q0

– For society, • q0 is not an efficient

allocation• q’ (less gold) will

maximize society’s benefits

– Δabc is a deadweight loss to society

Quantity (q)0

p0

MCp

q0

Price/Cost ($)MSC

p’

q’

D

b

c

a


• Illustration (cont’d):– The socially optimal

pollution level is NOT zero– When pollution is unpriced,

• Production results in more output than is socially desired

• Excessive pollution results– If pollution abatement is

enforced• Company will raise price

per unit of good• Company will reduce

output• The reduced quantity is

socially sufficient and the price is efficient

Quantity (q)0

p0

MCp

q0

Price/Cost ($)MSC

p’

q’

D

b

c

a

Resource allocation in a monopolyResource allocation in a monopoly

• Monopoly rights cause market failure from society’s point of view– A single monopolistic

curve • marginal cost curve MC• Demand curve D

– Under perfect competition

• Price = MR = MC• q’ units will be supplied

Quantity (q)0

P’

MC

q’

Price/Cost ($)

MR

pm

qm

D

c

d

b

a

Resource allocation in a monopolyResource allocation in a monopoly

• In the case of a monopoly– Demand curve above

marginal revenue curve– Price ≠ MR– Monopoly profit maximized

by setting MR = MC• Less output (qm)• Higher price (pm)• Consumer surplus = Δ apmb instead of Δ ap’c

– Marginal benefit exceeds marginal cost

– The level of output is inefficient

– Deadweight loss to society = Δbdc

Quantity (q)0

P’

MC

q’

Price/Cost ($)

MR

pm

qm

D

c

d

b

a

POLICY FAILUREPOLICY FAILURE

• Policy failure occurs when the government creates incentives for the prices of certain goods to be lower than the actual cost of production per unit– Ex: Government subsidy on pesticides

• In general, subsidies in developing countries are declining due to the adoption of structural adjustment programs

POLLUTION CONTROL POLLUTION CONTROL APPROACHESAPPROACHES

• Two main approaches– Property rights or market (Coasian) solution

• Allowing the market system to solve the problem through bargaining between affected parties

• Based on assumptions that may not apply in the real world

– Zero transaction costs– Well defined property rights– Perfect competition– No free-rider effect

– Government intervention

Government PoliciesGovernment Policies• There is always a need for government

interventions to correct the externality problem

Other InstrumentsMarket Based(MBIs) Instruments

Subsidies MarketablePermitsCharges

— Emission charges— User charges— Product charges— Administrative charges

— Deposit refundSchemes

— Ecolabelling— Performance bonds— Traditional

property rights

Other MBIs

Command and Control (CAC)

Instrument

AmbientStandards

EmissionStandards

— Performance-basedstandards

— Technology-basedstandards

— Voluntary incentives

— Liabilitylegislation

— Education— Zoning— Fines— Bans

Command and Control Command and Control InstrumentsInstruments

• Oldest form of pollution control policies• Require setting the standard and monitoring and

enforcing it• Advantages

– A widely understood form of policy– More pragmatic and socially acceptable than MBIs

• Disadvantages– Provides no incentive for pollution reduction beyond standards– Penalties tend to be too low and enforcement too weak– Governments must know the marginal social cost and marginal

social benefits curves to set an optimal penalty– Penalties need to be revised frequently which is costly– Financial costs for setting and enforcing standards are high– Political costs may arise if standards are stringent– Standards are uniformly set to all firms and regions

MarketMarket--based Instrumentsbased Instruments

• Use price or some other economic variables to provide incentive for economic agents to abate pollution

• Advantages– Achieve the same objective as CACs at a lower cost– Generate significant revenue for the government

• Disadvantage– Cannot be applied where the institutional framework

is weak

Choosing the Right InstrumentChoosing the Right Instrument

• Criteria to consider– Economic efficiency– Effectiveness in achieving the desired

environmental objective– Adaptability to changing circumstances– Equity in the distribution of costs and benefits

among different groups in the society– Political acceptability

INTRODUCTION TO INTRODUCTION TO ENVIRONMENTAL VALUATIONENVIRONMENTAL VALUATION

Assigning zero/ low values to non-market environmental commodities

Failure to account for the values of environmental resources

Decisions/ Policies with negative environmental and social implications

The Elements of Total Economic The Elements of Total Economic ValueValue

Benefits

Direct Use Values

Ecological Function Values

USE VALUESUSE VALUES(consumptive and non-

consumptive)

Unpriced Benefits

Marketed Outputs

• Recreation• Landscape• Local culture

• Crops• Meat• Timber• Renewable

energy

• Flood control• Carbon storage• Water catchment• Waste

assimilation

NONNON--USE VALUESUSE VALUES(inherent in the good)

Benefits

Option Values

• Future drugs• Potential gene

pool• Recreation

options

Benefits

Existence Values

• Satisfaction from knowledge of existence

Benefits

Bequest Values

• Passing benefits to future generations

NonNon--use valuesuse values

• Can constitute a significant component of total economic value

• Not traded cannot be valued by market prices

• Non-market valuation methods were developed for this purpose

NonNon--market Valuation Methodsmarket Valuation Methods

• Revealed preference models– Make use of individual

behavior in real and simulated markets to infer the value of an environmental good or service

– Measures use values only– Choices made are real

rather than hypothetical– Clear principle but

complicated applications– Example

• Wilderness valuated from the cost incurred to travel to the area for recreation

• Stated preference models– Elicits environmental values

directly from respondents using survey techniques such as questionnaires

– Flexible and applied to a wide range of goods

– Measure use and non-use values

– Are subject to many biases discussed later

Preferences

Revealedpreferences

StatedPreferences

MarketValues

Travel Cost Methods

HedonicMarkets

Averting Behavior

ContingentValuation

ChoiceExperiments

USE VALUES USE + NON-USE VALUES

Dose Response Functions

NonNon--Market Valuation MethodsMarket Valuation Methods

EEnd of nd of SSession ession 22

Thank YouThank You



Session 2bMacroeconomics and Policy Implications

Macroeconomics and the Macroeconomics and the EnvironmentEnvironment

• Macroeconomic instability is not good for the environment.

• Macroeconomic reforms may exacerbate existing policy and market failures; this is an argument for reforming the environmental sector, not abandoning macroeconomic adjustment

• Many macroeconomic reforms are positive for the environment – royalty collection, user fees, pricing reforms, subsidy

reductions, pollution taxes, trade liberalization and consequent access to new technology

The Policy Matrix: What The Policy Matrix: What Instruments?Instruments?

Lessons learned fromLessons learned fromsuccessful financing initiatives:successful financing initiatives:

Successful sustainable development and environmental management initiatives usually have the following characteristics:

• Financial sustainability• Administrative sustainability

Lessons learned from Lessons learned from successful financing initiativessuccessful financing initiatives

• Public/private consensus• Policy integration

Financing sustainable development is not only about having the financial resources but also about these other dimensions. Investing time in ‘consensus’ building can reduce the financial needs of a project.

Financial sustainabilityFinancial sustainability

• Governments often face strict fiscal regimes –effective policies help generate resources

• The environment is usually a second (or third) priority

• Society is willing to pay for a better environment– Defining property rights– who is responsible??– Economic instruments can help in internalizing

external costs and raising revenues• When the externality is global, resources can be

captured internationally– GEF; Carbon Funding

Administrative sustainabilityAdministrative sustainability

• The creation of markets and the imposition of new taxes require environmental management bureaucracies/ institutions

• Countries can often use existing fiscal systems

• Need to invest in training and capacity building: building institutions is however slow

Building a public/privateBuilding a public/privateconsensusconsensus

• Important to identify ‘winners’ and ‘losers’and clearly communicate this information

• Subsidy removal will be opposed by established interests (that benefit from the subsidies)

• Need to promote a public demand for change; governments very rarely lead in environmental policy reform

Achieving policy integrationAchieving policy integration

• Governments need to generate financial resources, often via taxes

• Revenue raising environmental instruments need to be compatible with existing fiscal regime and take into account equity considerations

Achieving policy integrationAchieving policy integration

• A mix of financing mechanisms is needed to• At the macroeconomic level, consider the

links between environmental management and– Liberalization and privatization– Fiscal and monetary instability / Exchange rate

instability– Growth– Poverty alleviation

ConclusionConclusion

• Very little/ no new Official Development Aid for the environment

• Large present flows of private and public capital offer much promise

• Financing tools exist that can be used to increase financing for the environment

• Both “polluter pays” and “beneficiary pays” approaches can be used

• Subsidy reduction – politically hard to do/ often unpopular

• Polluter pays – especially for new developments

ConclusionsConclusions

• User fees – show large promise for recreational/ amenity values

• Both Command and Control and Economic-based Instruments

• Institutions and monitoring needed for both types of policies –Command and Control and Economic-based Instruments

• Vested interests powerful (and clever!)

• Political will needed for effective financing reform

• Public awareness and involvement essential to create political will

EEnd of nd of SSession ession 2b2b

Thank YouThank You




Sessions 3 & 4ENVIRONMENTAL VALUATION

USINGMARKET VALUE METHODS

WORKSHOP ONWORKSHOP ON

COST OF ENVIRONMENTAL COST OF ENVIRONMENTAL DEGRADATION METHODOLOGYDEGRADATION METHODOLOGY

Sessions 3 & 4ENVIRONMENTAL VALUATION USING

MARKET VALUE METHODS

Preferences

Revealedpreferences

StatedPreferences

MarketValues

Travel Cost Methods

HedonicMethod

Averting Behavior

ContingentValuation

ChoiceExperiments



Environmental Valuation MethodsEnvironmental Valuation Methods

MARKET VALUESMARKET VALUES

• Production function method

• Market price method

• Damage cost, Replacement cost, and substitution cost methods

Revealed Preference ApproachRevealed Preference Approach

The Production Function The Production Function MethodMethod

Production Function MethodProduction Function Method

• One of the most widely used valuation techniques

• Focuses on environmental resources as an input to the production of goods and services

• Used to estimate the economic value of ecosystem products or services that contribute to the production of commercially marketed goods

Production Function MethodProduction Function Method• If a natural resource is a factor of production, then

– changes in the quantity or quality of the resource will result in changes in production costs, and/or productivity of other inputs

– this may affect the price and/or quantity supplied of the final good– it may also affect the economic returns to other inputs.

Production

Other inputs

Q1

Q2

X1 X2

Due to soil erosion

Q1 = f(S1, X)

Q2 = f(S2, X)

Production Function

• Production is a function of soil (S) and other inputs (x)

• As soil quality declines from S1to S2 the production function shifts to Q2

• Options for farmer• Produce at Q2• Keep production at Q1 by increasing other inputs from X1 to X2


Two types of benefits (or costs) may be important

– Changes in the quality or price to consumers of the final good changes will result in changes in consumer surplus

– Changes in productivityor production cost changes will result in changes in producer surplus

Quantity produced0

a

Market demand

Market supply

c

SE P

Price per unit

b

d

Producer surplus

Consumer surplus

Thus, the economic benefits from improvements in the resource can be estimated using changes in observable market data


• Selected applicationsPressure Environmental

ImpactProductivity Impact Change in Income

Overgrazing Soil erosion Reduced capacity of soil to sustain crops

Reduced farmers income

Wastewater discharge

Polluted river Reduced capacity to sustain fish stocks

Reduced income of fishermen

Increased vehicle use

Air pollution Increased respiratory problems among

workers

Lost workdays

Uncontrolled irrigation

Salinity of cropland

Declining yields Reduced farmers income


• The method is most easily applied in two specific cases: 1. Cases where the resource in question is a perfect substitute for

other inputs• Ex: increased water quality in a reservoir means that less chlorine

is needed for treating the water.– An increase in quantity or quality of the resource will result in

decreased costs for the other inputs. – The benefits of increased water quality can be directly measured by

the decreased chlorination costs2. Cases where only producers of the final good benefit from

changes in quantity or quality of the resource and consumers are not affected• Ex: improved quality of irrigation water may lead to greater

agricultural productivity– If the market price of the crops to consumers does not change,

benefits can be estimated from changes in producer surplus resulting from increased income from the other inputs.

– The profits per acre will increase, and this increase can be used to estimate the benefits of improved irrigation water quality


Applying the Productivity Method 1. Determine the physical impact solely arising from

the driving force or behaviour under study• Sometimes difficult to differentiate impacts due to a series

of complex biological interrelationships2. Collect data on how changes in the quantity/ quality

of the natural resource affect• costs of production for the final good • supply and demand for the final good • supply and demand for other factors of production

– Sources of data• Experimental using field trials

– Difficult to extrapolate• Statistical using cross-section or time-series data

– Available for short time horizons– Difficult to control for other factors


Applying the Productivity Method 3. Link the impact of changes in the quantity/

quality of the resource to changes in consumer surplus and/or producer surplus• Problems include

– Distorted prices due to government interventions– Change under study is not large enough to impact

market price– Change in market price is too large– Change in production alters costs

4. Estimate the economic benefits


Illustration 1*– A municipal drinking water reservoir is

polluted by agricultural runoff – The economic benefits of measures to

eliminate the runoff need to be determined• Productivity Method selected because

– Environmental quality directly affects the cost of producing municipal drinking water

– Cleaner water is a direct substitute for other production inputs, such as water purification chemicals and filtration

*Adapted from www.ecosystemvaluation.org


• Step 1– Specify the production function for purified drinking water

• Inputs:– water of a particular quality from the reservoir– Chemicals– Filtration

• Output: – pure drinking water

• Step 2– Estimate how the cost of purification changes when reservoir

water quality changes, using the production function estimated in the first step

– Calculate the quantities of purification chemicals and filters needed for different levels of reservoir water quality

– Multiply these quantities by their costs


• Step 3:– Estimate the economic benefits of protecting the

reservoir from runoff, in terms of reduced purification costs

• If all runoff is eliminated, the reservoir water will need very little treatment and the purification costs for drinking water will be minimal

• Compare this to the cost of purifying water where runoff is not controlled

• The difference in purification costs is an estimate of the benefits of eliminating runoff.

– The benefits for different levels of runoff reduction can also be estimated

• requires information about the projected success of actions to reduce runoff, in terms of the decrease in runoff and the resulting changes in reservoir water quality.


• Illustration 2*Values of Wetlands in the PeconicEstuary, Long Island

Cornell university library



• Illustration 2*Values of Wetlands in the PeconicEstuary, Long Island– The estuary includes productive wetlands of

different types• eelgrass, salt marsh, and intertidal mudflats



• Illustration 2*Values of Wetlands in the Peconic Estuary, Long Island– The estuary includes productive wetlands of different

types• eelgrass, salt marsh, and intertidal mudflats

– Development and resulting water quality degradation have reduced the quantity of these wetlands

• Challenge– Considering various management actions for the

Estuary and surrounding land areas– Assessing these management actions using a

productivity study for wetlands*Adapted from www.ecosystemvaluation.org


• Analysis – Valuing marginal changes in acres of wetlands, in terms of their

contribution to the production of crabs, scallops, clams, birds,and waterfowl

– It was assumed that wetlands provide both food chain and habitat support for these species

– The productivity of different wetlands types in terms of food chain production was estimated and linked to production of the different species of fish

– The expected yields of fish and birds per acre of habitat were valued using

• commercial values for fish• viewing values for birds• hunting values for waterfowl


• Results– The study results estimated that

• An acre of eelgrass is worth $1,065 per year• An acre of salt marsh is worth $338 per year• An acre of intertidal mudflat is worth $68 per year, in terms of

increased productivity of crabs, scallops, clams, birds, and waterfowl

– Based on the results the economic value for productivity services of preserving or restoring wetlands in the Estuary can be calculated

– These values are an understatement of the total economic value for the wetlands

• They only address values in production of commercially and recreationally valuable species

• They overlook other services, such as erosion and storm protection or aesthetics

ProductionProduction Function MethodFunction Method

• Advantages– Straightforward methodology – Inexpensive to apply due to

• Limited data requirements• Ready availability of relevant data

ProductionProduction FunctionFunction MethodMethod

• Issues and Limitations– Does not account for non-use values hence it

provides only the lower bound estimate– Limited to valuing resources that can be used as

inputs in production of marketed goods– Information is needed on the scientific relationships

between actions to improve quality or quantity of the resource and the actual outcomes of those actions

– If the changes in the natural resource affect the market price of the final good, or the prices of any other production inputs, the method becomes much more complicated and difficult to apply

The situation• 93% of Morocco is arid• Fragile soils suffer from water and wind erosion• Overexploitation and unsustainable management

–– arable land loss arable land loss –– decrease in crop yielddecrease in crop yield– silting of dams– loss in biodiversity– loss in terms of attenuating emissions of

gases causing greenhouse effect

The Production Function MethodThe Production Function MethodCaseCase--study 1: Degraded agricultural landstudy 1: Degraded agricultural land

& rangeland in Morocco& rangeland in Morocco

The situation• 65 million ha of pastureland providing 30% of

overall animal food requirements• Erosion, drought, overgrazing, land clearing and

removal of woods

Degraded pasturelandDegraded pastureland





The methodology• Degradation of agricultural land

–– value of lost agriculture production due to a value of lost agriculture production due to a decrease in land productivitydecrease in land productivity

– The majority of agricultural land is planted with cereals

Cost of degraded agricultural land Cost of degraded agricultural land corresponds to the value of lost cereal corresponds to the value of lost cereal

productionproduction• Degradation of rangeland

Cost corresponds to the value of lost Cost corresponds to the value of lost forage productionforage production

The methodology (Agricultural Land)• Step 1: Estimation of degraded agricultural land

– FAO* classified the degradation of 8.7 million ha in Morocco as “severe”

– According to the FAO method 3 scenarios are possible:

• 10 – 25% of land is severely degraded• 25 – 50% of land is moderately degraded• 50 – 100% of land is slightly degraded

– Surveys did not show any case of severe land degradation, only moderate and slight degradationare used

* FAO, Land Resources Potential and Constraints at Regional and Country Level, World Soil Resources Report 90, Rome, 2000



The methodology (Agricultural Land)• Step 2: Estimation of the decrease in agricultural yield

– Young* estimated the decrease in cereal yield

• Slight degradation 5% decrease in cereal yield

• Moderate degradation 20% decrease in cereal yield

– The mean yield for cereals in Morocco is 1 Ton/ha• 50 Kg/ha for slight degradation• 200 Kg/ha for moderate degradation

* Young, A. Land degradation in South Asia: its severity, causes and effects upon the people, 1994* Young, A., Land resources: now and for the future, Cambridge University Press, Cambridge, U.K.1998



The methodology (Agricultural Land)• Step 3: Assessing the cost of

degraded agricultural land– The average of the lower bound

and the upper bound of moderate and slight degradation were used

– Selling price: 2,580 dirham/Ton

Lower limit Upper limitModerate erosion 25% 50%

Degraded agricultural land (000ha)

2,175=25% ×8,700 4,350=50% ×8,70


Decrease in yield (Kg/ha)

200=20% ×1Ton/ha 200

Lost production (Kg) 435,000=2,175×200

870,000=4,350 ×200

Lost value (millions of Dh)

1,122=435,000 ×2,580

2,244=870,000 ×2,58

Slight erosion 50% 100%

Degraded agricultural land (000ha)

4,350=50% ×8,700 8,700=100% ×8,700


Decrease in yield (Kg/ha)

50=5% ×1Ton/ha 50

L t d ti (K ) 217 500 435 000

The average cost of The average cost of agricultural land agricultural land

degradation: degradation: Dh 1,263 millionDh 1,263 million=(842+1,683)/2=(842+1,683)/2

0 36% f th GDP0 36% f th GDP



The methodology (Rangeland)• Step 1: Estimation of degraded pastureland

– Calculations considered only the areas with dominant steppe and forest covers (excluding the Saharianregion)

• Dominant steppe area: 12 million ha• Dominant forest area: 5.1 million ha

– REEM calculated the average percentage of degraded rangeland as:

• Dominant steppe 46% 5.52 million ha degraded• Dominant forest 19% 0.969 million ha degraded

* REEM: Rapport sur l’Etat de l’Environnement du Maroc, Ministry of Land Use Planning, Environment, Urbanism and Habitat, 2001.



The methodology (Rangeland)• Step 2: Estimation of the loss of productivity

– MAMVA* estimated land productivity as:• Steppe: 79 FU / ha / year (FU: Forage Unit 1Kg Barely)• Forest: 558 FU / ha / year

– MAMVA adopted 2 levels of loss 6% and 10%:• Steppe loss: 6% 26.1 million FU / year

10% 43.6 million FU / year• Forest loss: 6% 26.1 million FU / year

10% 43.6 million FU / year* MAMVA: Ministry of Agriculture and Agricultural Development, Plan National d’Aménagement des Bassins Versants,

Phase II, Volume 1, 1994.* MAMVA: Ministry of Agriculture and Agricultural Development, Plan National d’Aménagement des Bassins Versants

Priorités régionales, Phase II Rapport de synthèse, 1995.



The methodology (Rangeland)

• Step 3: Assessing the cost of rangeland degradation– Barley price: 2,270 Dh/Ton– FU price: 2.27 Dh

The average cost of The average cost of rangeland degradation: rangeland degradation:

177.4 Dh million177.4 Dh million0.05% of the GDP0.05% of the GDP

Steppe Forest TotalPasture area (000ha) 12,000 5,100 17,100

Degraded area (%) 46% 19%

Degraded area (000ha) 5,520=46%

×12,000

969=19%

×5,100

6,489=5,520+

969

Land productivity (FU/ha/year) 79 558

10% lossLoss in yield in degraded area 10%

Lost yield (000 FU/year) 43,608=5,520

×79 ×10%

54,070=969 ×558

×10%

97,678

Lost value (million Dh) 99.0=43,608

×2.27

122,7=54,070

×2.27

221.7

6% loss



The Results

Average estimate: Average estimate: 1,440 million1,440 million

0.41% of the GDP0.41% of the GDP

Cost of degradation of agricultural land

Cost of degradation of rangeland

Cost of land degradation = +



The Production Function MethodThe Production Function MethodCaseCase--study 2: Beach degradation in Lebanonstudy 2: Beach degradation in Lebanon

The situation• The coastal zones of Lebanon represent unique

economic and recreational assets• Coast line

– > 240 km long– Inhabited by > 50% of population

• Untreated municipal wastewater disposal, seafront solid waste dumps, uncontrolled development of resorts and vacation homes, etc.

Coastal zone/Beach degradationCoastal zone/Beach degradation(Reduced recreational and tourism value)(Reduced recreational and tourism value)


The methodology• Annual cost of coastal degradation

– Domestic recreational losses– Losses of ecological and non-use value– Fishery losses due to pollution–– International tourism lossesInternational tourism losses

• Lebanon would likely have attracted a number of beach tourists if the coast was not degraded

Value of international beach touristValue of international beach tourist--nights nights unrealized due to beach degradationunrealized due to beach degradation


The methodology• Step 1: Estimation of international beach-tourist nights

lost– Calculated based on a comparison with Tunisia and adjusting for

• Differences in Kms of beaches/tourist zones, and• Domestic prices in Lebanon

Extrapolation of Western European (WE) and Northern American (NA) beach tourism in Tunisia for 1999

TunisiaTunisiaCoastline (km) 1,300

Beaches (km) 575

Tourist zones (km) 80

International tourist nights from WE and NA (in 1999) 28,500,000

Beach tourism 90%

International beach-tourist nights 25,650,000

International beach-tourist nights per km beaches 44,609

International beach-tourist nights per km tourist zones 320,625

International beach-tourist nights per km tourist zones in Tunisia:Tunisia:320,625 nights320,625 nights


The methodology• Step 1: Estimation of international beach-tourist nights

lost (cont’d)– Adjustment for differences in kms of tourist zones between Lebanon and

TunisiaLebanonLebanon

Low HighCoastline (km) 243 243Beaches (km) 36 36High potential beaches (if not degraded) (km) 5 10International beach-tourist nights per km tourist zones in Tunisia 320,625

Potential international beach-tourist nights lost in Lebanon*

1,603,125=320,625

×53,206,250

* Unadjusted for domestic prices


The methodology• Step 1: Estimation of international beach-tourist nights lost

(cont’d)– Adjustment for domestic price differentials between Lebanon and Tunisia

• Prices are generally higher in Lebanon• The adjustment was conducted by applying a price elasticity of international

tourism demand*– Price elasticity of demand: -2 and -2.25

*Papatheodorou, A. 1999. The Demand for International Tourism in the Mediterranean Region. Applied Economics, 31, (5), 619-630.*Syriopoulos, T. and Sinclair, M.T. 1993. An Economic Study of Tourism Demand: The AIDS Model of US and European Tourism in

Mediterranean Countries. Applied Economics, 25, (12), 1541-1552.

Potential international beach-tourist nights lost in Lebanon: Lebanon: 726,942 726,942 –– 849,769 nights849,769 nights


The methodology• Step 2: Assessing the value of international

beach-tourism loss– Average tourist expenditure in Tunisia was about 50 US$ in 1999– A range of 75-100 US$ (average 87.5 US$) was used for Lebanon

Low HighPotential international beach-tourist nights lost in Lebanon 726,942 849,769Expenditure per tourist per night (US$) 75 100Average expenditure per tourist per night (US$) 87.5

Total international beach tourism revenue losses (US$/year)63,607,388=726,942

×87.574,354,794


The average international beachThe average international beach--tourism revenue tourism revenue losses:losses:

69 US$ million, 0.42 % of the GDP69 US$ million, 0.42 % of the GDP


The Market Price MethodThe Market Price Method

Market Price MethodMarket Price Method

• Makes use of observed market prices for environmental goods and services

• Values changes in quantity and/or quality of a good or service

• Uses standard economic techniques for measuring the economic benefits from marketed goods

• Applied for goods and services with established markets, and which have– Direct uses

• Ex: Plantation timber; commercial fisheries; tourism– Some indirect uses

• Ex: value of water from protected watersheds– Some option values

• Ex: gene research; forest conservation


• Applying the Market Price Method– Market price represents the value of an additional unit

of that good or service, assuming the good is sold through a perfectly competitive market

– Applying the method requires data to estimate consumer surplus and producer surplus.

• To estimate consumer surplus, the demand function must be estimated

– time series data on the quantity demanded at different prices– data on other factors that might affect demand, such as income

or other demographic data• To estimate producer surplus

– data on variable costs of production and revenues received


Illustration 1– Water pollution causing the closure of a commercial fishing area– The benefits of cleanup need to be evaluated

• This method was used because– The primary resource affected is fish, for which market data are

available• Application of the Market Price Method

– The objective is to measure total economic surplus for the increased fish harvest that would occur if the pollution is cleaned up

– the difference between economic surplus before and after the closure must be estimated

– The results of the analysis can be used to compare the benefits of actions that would allow the area to be reopened, to the costs of such actions


Market Price MethodMarket Price MethodStep 1• Use market data to estimate the market demand function

and consumer surplus for the fish before the closure.– Assume a linear

demand function• the initial market price

= $5/g• the maximum willingness

to pay = $10/g– At $5/g

• consumers purchased10,000 g fish/yr

• consumers spent $50,000on fish per year

– The shaded area on the graph represents the total consumer surplus received from the fish before the closure = $25,000

Consumer Surplus = ($10-$5)*10,000/2 = $25,000

Consumer Surplus

Demand for fish before closure

Market Price MethodMarket Price MethodStep 2• Estimate the market demand function and consumer

surplus for the fish after the closure– the market price of fish increased

from $5/Kg to $7/Kg– the total quantity demanded

decreased to 6,000 Kg/yr – The new consumer surplus

is $9,000Step 3

– Estimate the loss in economic benefits to consumers

• Subtract benefits after theclosure from benefits beforethe closure

– The loss in benefits to consumers is• 25,000 - 9,000 = $16,000.

Consumer Surplus = ($10-$7)*6,000/2 = $9,000

02468

1012

0 5,000 10,000 15,000 20,000 25,000

Quantity demanded (Kg)

Pric

e ($

/Kg) Consumer Surplus after closure

Demand for fish after closure

Step 4• Estimate the losses to producers by first measuring the

producer surplus before the closure – Producer surplus is measured by the difference between the total

revenues earned from a good, and the total variable costs of producing it

– Before the closure• 10,000 Kg of fish were caught per year• Fishermen were paid $1/Kg

their total revenues = $10,000 per year• The variable cost to harvest the fish was $0.50/Kg

total variable cost = $5,000 per year• The producer surplus before the closure was

$10,000 - $5,000 = $5,000


Market Price MethodMarket Price MethodStep 5:• Measure the producer surplus after the closure

Step 6:• Calculate the loss in producer surplus due to the

closure

Market Price MethodMarket Price MethodSteps 4, 5, and 6

Before closure After closureFish caught per year = 10,000 Kg Fish caught per year = 6,000 KgFishermen were paid $1/Kg Fishermen were paid $1/Kg Total revenues = 1 × 10,000 =

$10,000 per yearTotal revenues = 1 × 6,000 =

$6,000 per yearVariable cost to harvest fish = $0.50/Kg

Variable cost to harvest fish = $0.60/Kg

Total variable cost = 0.5 × 10,000 $5,000 per year

Total variable cost = 0.6 × 6,000$3,600 per year

The producer surplus = $10,000 - $5,000= $5,000

The producer surplus = $6,000 - $3,600= $2,400

Loss in producer surplus due to the closure$5 000 - $2 400 = $2 600

Market Price MethodMarket Price MethodStep 7:• Calculate the total economic losses due to the closure

The benefits of cleaning up pollution in order to reopen the area are equal to $18,600

• Notes– This example is based on assumptions that greatly simplify the

analysis– Some factors might make the analysis complicated

• Some fishermen might switch to another fishery after the closure, and thus losses would be lower

Lost consumersurplus$16,000

Lost producersurplus$2,600

Total economicloss

$18,600+ =


• Advantages– Relatively simple and straightforward– Relies on actual market values– Price, quantity and cost data are easy to

obtain for established markets – The method uses observed data of actual

consumer preferences– The method uses standard, accepted

economic techniques


• Issues and limitations– Market data may only be available for a limited number of goods

and services provided by a resource– Available market data may not reflect the value of all productive

uses of a resource– The true economic value of goods or services may not be fully

reflected in market transactions, due to market imperfections and/or policy failures

– Seasonal variations and other effects on price must be considered

– Cannot be easily used to measure the value of larger scale changes that are likely to affect the supply of or demand for a good or service

– Does not deduct the market value of other resources used to bring ecosystem products to market, and thus may overstate benefits


Damage Cost AvoidedDamage Cost AvoidedReplacement CostReplacement Cost

Substitute Cost MethodsSubstitute Cost Methods

Damage Cost Avoided, Replacement Damage Cost Avoided, Replacement Cost, and Substitute Cost Methods Cost, and Substitute Cost Methods

• Estimate values of ecosystem services based on – the costs of avoiding damages due to lost services – the cost of replacing ecosystem services, or– the cost of providing substitute services

• Assume that– the costs of avoiding damages or replacing ecosystems or their

services provide useful estimates of the value of these ecosystems or services

– if people incur costs to avoid damages caused by lost ecosystem services, or to replace the services of ecosystems, then those services must be worth at least what people paid to replace them.

• Are most appropriately applied in cases where damage avoidance or replacement expenditures have actually been, or will actually be, made


• The damage cost avoided method – uses either the value of property protected or

the cost of actions taken to avoid damages as a measure of the benefits provided

• Ex: if a wetland protects adjacent property from flooding, the flood protection benefits may be estimated by

– the damages avoided if the flooding does not occur – or the expenditures property owners make to protect their

property from flooding


• The replacement cost method– uses the cost of replacing an ecosystem or its

services as an estimate of the value of the ecosystem or its services

• The substitute cost method – uses the cost of providing substitutes for an

ecosystem or its services as an estimate of the value of the ecosystem or its services.

• Example, the flood protection services of a wetland might be replaced by a retaining wall or levee


Applying the methods• Step 1

– Assessing the environmental service provided• specifying the relevant services

– how they are provided, to whom they are provided, and the levels provided.

• Example: in the case of flood protection, this would involve predictions of flooding occurrences and their levels, as well as the potential impacts on property


Applying the methods• Steps 2 and 3

– for the damage cost avoided method• estimate the potential physical damage to property, either

annually or over some discrete time period• calculate either the dollar value of potential property damage,

or the amount that people spend to avoid such damage– for the replacement or substitute cost method

• identify the least costly alternative means of providing the service

• calculate the cost of the substitute or replacement service• Establish public demand for this alternative

– This requires gathering evidence that the public would be willing to accept the substitute or replacement service in placeof the ecosystem service

Damage Cost Avoided, Replacement Cost, Damage Cost Avoided, Replacement Cost, and Substitute Cost Methods and Substitute Cost Methods

• Examples of applications – Valuing improved water quality

• by measuring the cost of controlling effluent emissions. – Valuing erosion protection services of a forest or wetland

• by measuring the cost of removing eroded sediment from downstream areas.

– Valuing the water purification services of a wetland• by measuring the cost of filtering and chemically treating water

– Valuing storm protection services of coastal wetlands• by measuring the cost of building retaining walls.

– Valuing fish habitat and nursery services• by measuring the cost of fish breeding and stocking programs


Illustration 1* (damage and substitute cost methods)

– An agency is considering restoring some degraded wetlands in order to improve their ability to protect the surrounding area from flooding

– Cost-Based Methods are used because• Agency only interested in valuing the flood protection

services of the wetlands• Limited budget available for valuation study• the easiest and least costly method to apply in this case



Step 1• Conduct an ecological assessment of the flood protection services provided

by the wetlands to determine– the current level of flood protection– the expected level of protection after full restoration of the wetlands

Step 2• The Damage Cost Avoided applied using two different approaches

– use the information on flood protection obtained in the first step to estimate potential damages to property if flooding were to occur

• estimate, in dollars, the probable damages to property if the wetlands are not restored.– determine whether nearby property owners have spent money to protect their

property from the possibility of flood damage• by purchasing additional insurance or by reinforcing their basements.• These avoidance expenditures would be summed over all affected properties to provide

an estimate of the benefits from increased flood protection– the two approaches are not expected to produce the same estimate


Step 2 (cont’d)• The replacement cost method

– flood protection services cannot be directly replaced, so this cannot be applied

• The substitute cost method – a substitute for the affected services such as a retaining wall or a levee

might be built to protect nearby properties from flooding• estimate the cost of building and maintaining such a wall or levee• also determine whether people would be willing to accept the wall or levee in

place of a restored wetland.

Step 3– Compare the cost of the property damages avoided, or of providing

substitute flood protection services to the restoration costs to determine whether it is worthwhile to restore the flood protection services of the wetlands


Illustration 2*: Soil Erosion in Korea (replacement cost method)

• Background

• The challenge– Evaluate the benefits of proposed new soil management techniques

• retaining the soil and nutrients on the upland areas• protecting downslope areas from damage by the eroded soil


urban growth and industrial development

farming moved into hilly upland areas

Inadequate soil management techniques and errors in field layout and construction

Heavy soil erosion on upland areas


• Analysis – Researchers measured the cost of physically replacing lost soil,

nutrients, and water in upland areas the cost of compensating for downstream losses

• calculate the annual soil loss per hectare, nutrient loss/hectare, and water runoff/hectare

• calculate the expected losses, in terms of replacement costs, if the new management practices were not implemented

Measured parameter Cost (W/ha/yr)Recovering and replacing eroded soil 80,000Fertilizer and spreading to replace lost nutrients 31,200 Annual field maintenance and repair 35,000Damage to downstream fields in lost production 30,000Supplemental irrigation to replace lost water 92,000Total cost of soil erosion under existing management 268,200Net present value using a 15 year time horizon 2,039,662


• Analysis (cont’d)– calculate costs with the new management techniques

• compensation payments• soil replacement, nutrient replacement, and mulching

– The net present value of the costs of new management techniques was estimated at W1,076,742

• Results– the cost of new management techniques (W1,076,742) is

about half the replacement cost (2,039,662)

– The proposed preventive steps are worth implementing


Illustration 3*: Oil spill damages in Puerto Rico (replacement cost method)• Background

– The Zoe Colocotroni was a ship that spilled oil off the coast of Puerto Rico– The case was taken to court to determine the monetary damages resulting from the

spill’s effects on the local ecosystem

• Analysis– The replacement cost method was used to estimate monetary damages

• Calculating the number of lower trophic organisms killed by the spill• Adding up the cost of purchasing these organisms from a scientific catalogue

• Results– The US Court of Appeals rejected the use of the replacement cost method in this

case• It was not plannned to actually purchase the organisms and restore them to the ocean• By the time such a plan could have been carried out, the organisms would have restored

themselves

• The costs of purchasing the organisms did not accurately measure the actual ecosystem damages.


• Advantages

– May provide a rough indicator of economic value• subject to data constraints and the degree of similarity or

substitutability between related goods. – They are less data and resource-intensive

• It is easier to measure the costs of producing benefits than the benefits themselves, when goods, services, and benefits are non-marketed

– Data or resource limitations may rule out valuation methods that estimate willingness to pay

– Provide surrogate measures of value that are as consistent as possible with the economic concept of use value, for services which may be difficult to value by other means


• Issues and Limitations

– Do not provide a technically correct measure of economic value, which is properly measured by the maximum amount of money or other goods that a person is willing to give up to have a particular good, less the actual cost of the good

– Assume that expenditures to repair damages or to replace ecosystem services are valid measures of the benefits provided

– Do not consider social preferences for ecosystem services, or individuals’ behaviour in the absence of those services


• Issues and Limitations (cont’d)

– May be inconsistent because few environmental actions and regulations are based solely on benefit-cost comparisons, particularly at the national level

• the cost of a protective action may exceed the benefits to society

• the cost of actions already taken to protect an ecological resource will underestimate the benefits of a new action to improve or protect the resource


• Issues and Limitations (cont’d)– The replacement cost method requires information on

the degree of substitution between the market good and the natural resource

• Substitute goods are unlikely to provide the same types of benefits as the natural resource


• Issues and Limitations (cont’d)

– The goods/services being replaced probably represent only a portion of the full range of services provided by the natural resource

• the benefits of an action to protect or restore the ecological resource would be understated.

– Without evidence that the public would demand the least cost alternative for the affected ecosystem, this methodology is not an economically appropriate estimator of ecosystem service value

EEnd of nd of SSessions essions 33 & & 44

Thank YouThank You




Session 5THE TRAVEL COST METHOD



Session 5THE TRAVEL COST METHOD

Preferences

Revealedpreferences

StatedPreferences

MarketValues

Travel Cost Methods

HedonicMethod

Averting Behavior

ContingentValuation

ChoiceExperiments




TRAVEL COST METHODTRAVEL COST METHODOUTLINEOUTLINE

• Introduction• Theory• Forms of TCM• Illustration• Advantages• Issues and Limitations• Case-applications

– Environmental conservation– Improvements in water quality

• Case studies– Beach degradation in Morocco– The value of forestry in Britain

TRAVEL COST METHODTRAVEL COST METHODIntroductionIntroduction

• Used to estimate use values associated with ecosystems or sites that are used for recreation

• Assumes that the value of the site or its recreational services is reflected in how much people are willing to pay to get there

• Useful in planning for the provision and management of outdoor recreation, such as: – changes in access costs for a recreational site – elimination of an existing recreational site – addition of a new recreational site – changes in environmental quality at a recreational site

TRAVEL COST METHODTRAVEL COST METHODTheoryTheory

• Based on the premise that– The cost an individual incurs in visiting a site reflects his

valuation to the site – Individuals will react to an increase in entry fees the same

way as they would react to an increase in travel cost

• The most controversial aspects of the travel cost method include – accounting for the opportunity cost of travel time– handling multi-purpose and multi-destination trips– the fact that travel time might not be a cost to some

people, but might be part of the recreational experience


• A demand curve can be generated for the site in question– By collecting information

from people on• Where they had travelled

from• The costs they have

incurred– By deriving a trip generation

function– By deriving an aggregate

demand curve for visits to the site per year, and thus for the recreational or scenic services of the site

Number of visits per year

Trav

el c

ost p

er v

isit

(US

D)

Trip Generation Function

Consumer surplus

B

A

C

D

TRAVEL COST METHODTRAVEL COST METHODInterpreting TravelInterpreting Travel--Cost ModelsCost Models

Linear functional form: V = α + βC +γSWhere:

V = number of visits to a siteα = constantβ = coefficient of C, usually negativeC = cost of travel to gain access to siteγ = coefficient of S, probably negativeS = cost of travel to gain access to the respondent’s preferred substitute site

– The TCM is used to estimate α, β, and γ• Estimated consumer surplus (CS) for an individual

making q visits to the siteCS = -q2 / 2β

– This functional form implies finite visits at zero costs

– This functional form has a critical cost above which negative visits will be demanded

TRAVEL COST METHODTRAVEL COST METHODInterpreting TravelInterpreting Travel--Cost ModelsCost Models

Log-Linear functional form: lnV = α + βC +γSWhere:

V = number of visits to a siteα = constantβ = coefficient of C, usually negativeC = cost of travel to gain access to siteγ = coefficient of S, probably negativeS = cost of travel to gain access to the respondent’s preferred substitute site

– The TCM is used to estimate α, β, and γ• Estimated consumer surplus (CS) for an individual

making q visits to the siteCS = -q / β

– This functional form has been widely used in TCM models

• It implies a finite number of visits at zero cost• It never predicts negative visits even at very high costs


• This demand curve– shows how many visits people would

make at various travel cost prices– is used to estimate the willingness to

pay for people who visit the site– is downward sloping where travel cost

is inversely related to number of visits• people who live farther from a site will

visit it less often, because it costs more in terms of actual travel costs and time to reach the site.

• Other factors that may affect the number of visits to a site– Visitors’ income – The availability of alternative sites or

substitutes– Factors like personal interest in the

type of site, or level of recreational experience

Number of visits per year

Trav

el c

ost p

er v

isit

(US

D)

Demand curve for the travel cost method

Consumer

surplus

TRAVEL COST METHODTRAVEL COST METHODForms of the TCMForms of the TCM

Individual TCM

• Uses the number of visits per year made by an individual

• Requires more data collection and slightly more complicated analysis

• Gives more precise and statistically efficient results

• More flexible than ZTCM and applicable at a wider range of sites

Random utility approach

• The most complicated and expensive

• Allows for much more flexibility in calculating benefits

• Best suited to estimate benefits for specific characteristics of sites, rather than for the site as a whole

• Most appropriate when there are many substitute sites

Zonal TCM

• Concentric zones defined around each site

• Cost of travel in each zone is constant

• Site visitors grouped by zone of origin

• May rely on secondary data• The simplest and least

expensive approach• Suited when visitors origins

are relatively evenly distributed

• Unsuitable for linear recreational sites

TRAVEL COST METHODTRAVEL COST METHODZonalZonal TCMTCM

• Zonal TCM methodology1. Identify site and collect data from visitors on

• their points of origin• number of visits from each origin zone• round-trip mileage from each zone • travel costs per mile • demographic information about people from each zone

2. Define zones of origin and allocate visitors to the appropriate zone• Zones commonly defined based on

straight line distance from site• GIS techniques allow redefining zones

based on road distances or travel times3. Calculate zonal visits per household to the site

• Estimate number of households per zone• Divide number of household visits originating in zone h by the total number of

households in the zone4. Calculate average travel cost from each zone to the site5. Use census data to derive variables relating to zonal socio-economic

characteristics

32

1S

TRAVEL COST METHODTRAVEL COST METHODZonalZonal TCMTCM

• Zonal TCM methodology (cont’d)6. Use data collected above to estimate the trip generation

function

VVhh/N/Nhh = f(C= f(Chh,X,Xhh,S,Shh))

6. Derive demand curve7. Obtain zonal household consumer surplus estimates through

integrating under the demand curve8. Calculate aggregate zonal consumer surplus

• By multiplying consumer surplus per household by the number of households in each zone

10. Aggregate zonal consumer surplus estimates to obtain an estimate of total consumer surplus or the benefits of the site

where: Vh = # of visits from zone hNh = population of zone hCh = travel cost from zone hXh = a vector of socio-economic variables that

explain changes in VSh = a vector of substitute recreational site

characteristics for residents of zone h

TRAVEL COST METHODTRAVEL COST METHODIndividual TCMIndividual TCM

• Individual TCM methodology1. Identify site2. Use an on-site questionnaire survey to collect data from

visitors relating to• Cost of travel to the site• Number of visits to the sites• Recreational preferences• Socio-economic characteristics

3. Specify trip-generation function

Vij = f(Cij,Tij,Qi,Sj, Yi) where: Vi = # of visits made by individual i to site jCij = travel cost incurred by individual i when visiting site jQj = a vector of perceived qualities of the recreation site jSj = a vector of available substitute recreational site

characteristics Yi = household income of individual i


• More complicated, and thorough, applications may also collect information about: – exact distance that each individual travelled to the site – exact travel expenses – the length of the trip – the amount of time spent at the site – other locations visited during the same trip, and amount of time spent at

each – substitute sites that the person might visit instead of this site, and the

travel distance to each – other reasons for the trip (is the trip only to visit the site, or for several

purposes) – quality of the recreational experience at the site, and at other similar

sites (e.g., fishing success) – perceptions of environmental quality at the site – characteristics of the site and other, substitute, sites


• Individual TCM methodology (cont’d)4. Estimate travel cost model taking truncation

into account• Truncation accounts for non-visitors behavior

5. Derive demand curve and obtain household consumer surplus estimates through integrating under the demand curve


TRAVEL COST METHODTRAVEL COST METHODRandom utility approachRandom utility approach

• Focuses attention on the choice among alternative sites for any given recreational trip and assumes the visitor is comparing utilities for available destinations

• First models the individual’s decision on whether or not to participate in recreational activity

• Then models the decision on the number of visits

• Models used include– Probit, tobit, and logit

TRAVEL COST METHODTRAVEL COST METHODIllustrationIllustration

• Background– A site used mainly for recreational fishing is threatened by development

in the surrounding area• Pollution and other impacts from this development could destroy the fish

habitat at the site, resulting in a serious decline in, or total loss of, the site’s ability to provide recreational fishing services

• Resource agency staff want to determine the value of programs or actions to protect fish habitat at the site

• The Travel Cost Method was selected because– The site is primarily valuable to people as a recreational site– There are no endangered species or other highly unique qualities that

would make non-use values for the site significant– The expenditures for projects to protect the site are relatively low

• Alternative Approaches– Contingent valuation or contingent choice methods

• might produce more precise estimates of values for specific characteristics of the site

• could capture non-use values• are considerably more complicated and expensive to apply


Application of the Zonal Travel Cost Approach

Step 1• Define a set of zones surrounding the site by

– concentric circles around the site, or – geographic divisions

• metropolitan areas or counties surrounding the site at differentdistances

Step 2• Collect information on

– the number of visitors from each zone– the number of visits made in the last year

• For this example, assume that staff at the site keep records of the number of visitors and their zip code, which can be used to calculate total visits per zone over the last year



Step 3• Calculate the visitation rates per 1,000 population in each zone• This is the total visits per year from the zone, divided by the zone’s

population in thousands

Zone Total Visits/Year Zone Population

Visits/1000

0 400 1,000 400

1 400 2,000 200

2 400 4,000 100

3 400 8,000 50

Beyond 3 0

Total visits 1,600



Step 4• Calculate the average round-trip travel distance and travel time to

the site for each zone• Using average cost per mile and per hour of travel time, calculate

the travel cost per trip– Assume that this cost per mile is USD 0.30.

• The cost of time calculated using the simplest approach involving the average hourly wage– Assume that it is 9 USD/hour, $0.15 USD/minute for all zones,

although in practice it is likely to differ by zone

Zone Round Trip Travel Distance

(miles)

Round TripTravel Time

(mins)

Distance times Cost/Mile ($.30)

(cost a)

Travel Time times Cost/Minute ($.15)

(cost b)

Total Travel (Cost/Trip)

0 0 0 0 0 0

1 20 30 $6 $4.50 $10.50

2 40 60 $12 $9.00 $21.00

3 80 120 $24 $18.00 $42.00


Step 5• Estimate the trip generation function using regression analysis

– This allows the researcher to estimate the demand function for the average visitor

– The analysis might include demographic variables, such as age, income, gender, and education levels, using the average values for each zone

– To maintain the simplest possible model, calculating the equation with only the travel cost and visits/1000

Visits/1000 = 330 – 7.755*(Travel Cost)




Step 6• Construct the demand function for visits to the site

– The first point on the demand curve is the total visitors to the site at current access costs

• 1,600 visits per year– The other points are found by estimating the number of visitors

with different hypothetical entrance fees• Example: start by assuming a $10 entrance fee

Plugging this into the estimated regression equation, V = 330 – 7.755C, gives the following:

Zone Travel Cost plus $10 Visits/1000 Population Total Visits

0 $10 252 1000 252

1 $20.50 171 2000 342

2 $31.00 90 4000 360

3 $52.00 0 8000 0

Total Visits 954


Step 6 (cont’d)• This gives the second point on the

demand curve—954 visits at an entry fee of $10. In the same way, the number of visits for increasing entry fees can be calculated

• These points give the demand curve for tripsto the site.


0

10

20

30

40

50

60

0 400 800 1200 1600 2000

Total visits

Add

ed c

ost p

er tr

ip

Entry Fee Total Visits

$20 409

$30 129

$40 20

$50 0



Step 7• Estimate the total economic benefit of the site to visitors by

calculating the consumer surplus, or the area under the demand curve.

• The total estimate of economic benefits from recreational uses of the site is around $23,000 per year, or around $14.38 per visit

• If the actions to protect the site cost less than $23,000 per year, the cost will be less than the benefits provided by the site

• If the costs are greater than this, the staff will have to decide whether other factors make them worthwhile.


Application of the Individual Travel Cost Approach

• Conduct a survey of visitors on – location of the visitor’s home – distance travelled to the site – how many times they visited the site in the past year or season – the length of the trip – the amount of time spent at the site – travel expenses – the person’s income or other information on the value of their time – other socioeconomic characteristics of the visitor – other locations visited during the same trip, and amount of time spent at

each – other reasons for the trip (is the trip only to visit the site, or for several

purposes) – fishing success at the site (how many fish caught on each trip) – perceptions of environmental quality or quality of fishing at the site – substitute sites that the person might visit instead of this site


Application of the Individual Travel Cost Approach (cont’d)

• Estimate the relationship between number of visits and travel costs and other relevant variables using regression analysis– Use individual data rather than data for each zone– The regression equation gives the demand function for the

“average” visitor to the site– The area below this demand curve gives the average consumer

surplus

• Multiply the average consumer surplus by the total relevant population in the region where visitors come from

to estimate the total consumer surplus for the site


Application of the Individual Travel Cost Approach (cont’d)

• Value estimates can be improved by adding other factors to the statistical model– additional data about visitors, substitute sites, and quality

of the site has been collected– Including information about the quality of the site allows

the researcher to estimate the change in value of the site if its quality changes

• two different demand curves would be estimated—one for each level of quality

• the area between these two curves is the estimate of the change in consumer surplus when quality changes


Application of the Random Utility Approach

• The agency might want to value the economic losses from a decrease in fish populations, rather than from loss of the entire fish stock

• The random utility approach focuses on choices among alternativesites which have different quality characteristics.

• The random utility approach assumes that individuals will pick the site that they prefer, out of all possible fishing sites.

• This model requires information on– all possible sites that a visitor might select– their quality characteristics– the travel costs to each site


Application of the Random Utility Approach (cont’d)

• Conduct a telephone survey of randomly selected residents of the state– Ask residents if they go fishing or not– If they do, it would then ask a series of questions

• how many fishing trips they took over the last year (or season)• where they went• the distance to each site• and other information similar to the information collected in our individual travel cost survey

– Might also ask questions about fish species targeted on each trip, and how many fish were caught.

• Estimate a statistical model that can predict– the choice to go fishing or not,– the factors that determine which site is selected

• If quality characteristics of sites are included, the model can estimate values for changes in site quality, for example the economic losses caused by a decrease in catch rates at the site

TRAVEL COST METHODTRAVEL COST METHODAdvantagesAdvantages

• It is based on real data rather than stated willingness to pay and as such provides true values

• It is relatively inexpensive to apply

• On-site surveys provide opportunities for large sample sizes, as visitors tend to be interested in participating

• The results are relatively easy to interpret and explain

TRAVEL COST METHODTRAVEL COST METHODIssues and limitations (1)Issues and limitations (1)

• The method assumes that people perceive and respond to changes in travel costs the same way that they would respond to changes in admission price

• The most simple models assume that individuals take a trip for the single purpose of visiting a specific recreational site– if a trip has more than one purpose, the value of the site may be

overestimated

• Defining and measuring the opportunity cost of time, or the value of time spent travelling, can be problematic– There is no consensus on how to account for time – Travel time may be a benefit if people enjoy the travel itself leading

to an overestimation of the value of the site

• The availability of substitute sites will affect values– Ex: if two people travel the same distance, they are assumed to

have the same value. However, if one person has several substitutes available but travels to this site because it is preferred, this person’s value is actually higher. Some of the more complicated models account for the availability of substitutes.


• The assumption that travel costs reflect recreational value may not always be true– Those who value certain sites may choose to live nearby,

resulting in low travel costs, but high values for the site

• Visits to certain sites could be seasonal and thus survey results could be biased unless survey is conducted for a long period

• Interviewing visitors on site can introduce sampling biases to the analysis

• Measuring recreational quality and relating it to environmental quality can be difficult

• Standard travel cost approaches provides information about current conditions, but not about gains or losses from anticipated changes in resource conditions


• The demand function requires enough difference between distances travelled to affect travel costs and for differences in travel costs to affect the number of trips made– it is not well suited for sites near major population centers where many

visitations may be from "origin zones" that are close to one another

• The travel cost method is limited in its scope of application because it requires user participation– It cannot be used to assign values to on-site environmental features and

functions that users of the site do not find valuable– It cannot be used to value off-site values supported by the site– It cannot be used to measure non-use values– It excludes non-users who may have significant values for the site

• Certain statistical problems can affect the results, including – Choice of the functional form used to estimate the demand curve– Choice of the estimating method– Choice of variables included in the model

TRAVEL COST METHODTRAVEL COST METHODCase application 1Case application 1: Environmental Conservation : Environmental Conservation

Background*• Hell Canyon on the Snake River separating Oregon

and Idaho– offers spectacular vistas and outdoor amenities to visitors – supports important fish and wildlife habitat

• It also has economic potential as a site to develop hydropower.

– Generating hydropower there would require building a dam– Dam and the resulting lake would significantly and

permanently alter the ecological and aesthetic characteristics of Hell Canyon

Challenge• Controversies regarding the future of Hell Canyon

during the 1970’s• Environmental economists were asked to develop

an economic analysis justifying the preservation of Hell Canyon in its natural state

www.bigfootoutfitters.com

www.colonelkern.com

* www.ecosystemvaluation.org

TRAVEL COST METHODTRAVEL COST METHODCase application 1Case application 1: Environmental Conservation : Environmental Conservation

Methodology• the net economic value (cost savings) of producing

hydropower at Hell Canyon was $80,000 higher than at the "next best" site which was not environmentally sensitive

• The recreational value of Hell Canyon was estimated via a low-cost/low precision travel-cost survey at about $900,000

• Even if the "true value" of recreation at Hell Canyon was ten times less than their estimate, it would still be greater than the $80,000 economic payoff from generating power there as opposed to the other site

Results• Based largely on the results of this non-market valuation

study, Congress voted to prohibit further development of Hell Canyon

www.windingwatersrafting.com/hell_rates.php

TRAVEL COST METHODTRAVEL COST METHODCase application 2Case application 2: Improvements in Water : Improvements in Water

QualityQualityBackground*• The costs to farmers and taxpayers of implementing

on-farm best management practices to reducesediment and nutrient runoff to the ChesapeakeBay are well known

• Controversies arose over the benefits ofresulting improvements in water quality

Challenge • To assess the economic benefits of water quality improvements to beach

users in the Chesapeake Bay area• To establish linkages between differences in water quality and

differences in willingness to pay for beach use- reflected in the travel costs to visitors to particular beaches

• The hypothesis to be tested – average willingness to pay was positively correlated with water quality

• If the hypothesis was correct the results would allow the estimation of the increase in willingness to pay of improving water quality at all beaches

www.chesapeakebaysampler.com/

* www.ecosystemvaluation.org

TRAVEL COST METHODTRAVEL COST METHODCase application 2Case application 2: Improvements in Water : Improvements in Water

QualityQualityMethodology• The concentration of nitrogen and phosphorous in the water at the

monitoring station nearest to the beach was selected as an index of water quality at the beach

– reflect s the level of objectionable visual and other characteristics that affect the value of beach use.

• A cross-sectional analysis of travel cost data was used– collected from 484 people at 11 public beaches – a 20% increase in water quality was assumed to be associated with a 20% reduction in total

nitrogen and phosphorus

Results• The average annual benefits to all Maryland beach users of the

improvements in water quality were estimated to be $35 million in 1984 dollars

• These were thought to be conservative for several reasons, including: – The value of improvements in water quality was only shown to increase the value of current

beach use– Improved water quality can also be expected to increase overall beach use – Estimates ignore visitors from outside the Baltimore-Washington statistical metropolitan

sampling area– The population and incomes in origin zones near the Chesapeake Bay beach areas are

increasing, which is likely to increase visitor-days and thus total willingness to pay

The Travel Cost MethodThe Travel Cost MethodCaseCase--study 1: Beach degradation in Moroccostudy 1: Beach degradation in Morocco

The situation*• The coastline of Morocco

– 3,500 km long, 13 coastal zones, 174 beaches

Coastal zone/ Coastal zone/ Beach Beach

degradationdegradation

In 2002, “Monitoring Bathing Beach Waters in Morocco”campaign showed that 28% of beaches were unfit for swimming28% of beaches were unfit for swimming

•Domestic and industrial wastewater discharge, industrial accidents

•Offshore pollution from ships and boat harbors

•Haphazard construction along the coast, etc.

* World Bank, 2003


The situation• Annual cost of coastal degradation

– Willingness to Pay (WTP) of foreign tourists and Moroccan nationals living abroad to improve the coast

– Loss of local fishing (sardines)

– Lost recreational value for Moroccan residents •• Value of additional travel costs (including time) incurred by Value of additional travel costs (including time) incurred by

Moroccan residents Moroccan residents to find beaches of “better” environmental quality

The Travel Cost MethodThe Travel Cost MethodCaseCase--study 1: Beach degradation in Moroccostudy 1: Beach degradation in MoroccoThe methodology

• Additional travel cost incurred– Longer distances traveled by car, bus or taxi to visit

“less degraded” beaches• For citizens using cars, additional travel costs per beach visit

include:– Fuel costs– Vehicle operating costs, and– Journey time lost

– Estimation conducted for cities of• Rabat• Tangiers• Casablanca


The methodology• Step 1: Estimation of additional travel cost for beach visits by car

– Number of visits to beach by car per household = 10-20– Assumption that 50% of households with cars visit more distant beaches for environmental

reasons– Additional travel cost by car per visit

• vehicle functioning = Dh 4.6/km• additional distance of 30 km (round trip)• additional time of 2 hrs estimated at Dh 15/hr• = (Dh 4.6/km x 30 km) + (Dh 15/hr x 2 hrs) = 168 Dh

Parameter ValueUrban households in Rabat-Tangiers-Casablanca 937,143Households with cars (20%) 187,400Households with cars that visit more distant beaches 93,700Number of yearly visits to beach by car per household 10 - 20Total number of yearly visits to beach by car 937,000 - 1,874,000Average additional travel cost by car per visit (Dh/visit)* 168Total additional yearly travel cost for beach visits by car (Dh) 157,416,000 - 314,832,000


The methodology• Step 2: Estimation of additional travel cost for beach visits by bus/taxi

– 20% of population visit beach by bus or taxi– 50% of population without cars visit more distant beaches for environmental reasons– Number of visits to beach by bus or taxi per person = 10-20– Average additional travel cost by bus or taxi is 10 Dh

Parameter ValueUrban population of Rabat-Tangiers-Casablanca 5,248,000Persons from Rabat-Tangiers-Casablanca visiting beaches by bus/taxi 1,049,600Persons visiting, by bus or taxi, more distant beaches 524,800Number of yearly visits to beach per person 10 - 20Total number of yearly visits to beach by bus or taxi 5,248,000 - 10,496,000Average additional travel cost by bus or taxi per visit (Dh/visit) 10Total additional yearly travel cost for beach visits by bus or taxi (Dh) 50,248,000 - 100,496,000


The methodology• Step 3: Assessing the total additional travel cost due

to beach degradation

The average total additional travel cost due to The average total additional travel cost due to beach degradation in Moroccobeach degradation in Morocco

Dh 311.5 million, 0.09 % of the GDPDh 311.5 million, 0.09 % of the GDP

Lower bound

Higher bound

Additional travel cost for beach visits by cars (million Dh) 157.4 314.8Additional travel cost for beach visits by bus/taxi (million Dh) 50.2 100.5Total additional travel cost (million Dh) 207.6 415.3Percent of GDP (%) 0.06 0.12

TRAVEL COST METHODTRAVEL COST METHODCaseCase--study 2: The value of forestry in study 2: The value of forestry in

BritainBritainSituation*• Estimate the total recreational

value of Forestry Commission woodland in Great Britain

Methodology– Define representative forest types

• Cluster analysis used• 14 similar groups of forests identified• Sample forests selected from each group

– Conduct interviews with visitors of 15 forests– Allocate visitors into 20 concentric distance zones at 5-

mile intervals• Those further than this were allocated together in a single zone

www.ecastles.co.uk/thetford.html

* Hodge, I. 1995


BritainBritainMethodology (cont’d)

– Develop the Trip generation functionsby estimating the relationship betweenthe visit rate from each zone and thetransport cost• taking into account the socio-economic characteristics of the zones

– Use trip generation functions to estimate the consumer surplus or the total value of each visit

• Represented by the maximum willingness to pay minus the cost of each trip

– Estimate the total value for each site by summing up across all visitors to each site produced

www.avich-kilchrenan.co.uk/


BritainBritain

Consumer surplus estimates fornon-priced recreation for forest districts

Consumer surplus per recreational visitor (£)

Consumer surplus per

hectare of forest(£/ha)

Annual number of visitors to the cluster of forests

(1000s)

Total consumer surplus

(£million)

New Forest 1.43 425 8,000 11,440

Loch Awe 3.31 <1 34 0.114

Brecon 2.26 27 2,117 4.784

South Lakes 1.34 31 1,968 2.637

Thetford 2.66 14 4,742 10.718


BritainBritainResults• By summing up all estimates of

consumer surplus– The value of non-priced recreation

for Forestry Commission forests asa whole = £ 53 million

– This compares with £ 71 million income to Forestry Enterprise, the timber production arm of Forestry Commission from the sale of timber in 1988

– In this year, there was an annual net subsidy of £ 8.5 millionpaid on forestry recreation and amenity in terms of visitors centers, forest walks, wildlife, conservation, etc.

– Provision for non-priced recreation represents good value


Thank YouThank You




Session 6THE HEDONIC PRICING METHOD

THE AVERTING BEHAVIOR METHOD

WORKSHOP ONWORKSHOP ON

COST OF ENVIRONMENTAL COST OF ENVIRONMENTAL DEGRADATION METHODOLOGYDEGRADATION METHODOLOGY

Session 6HEDONIC PRICING METHOD

AVERTING BEHAVIOR METHOD

Preferences

Revealedpreferences

StatedPreferences

MarketValues

Travel Cost Methods

HedonicMethod

Averting Behavior

ContingentValuation

ChoiceExperiments



Environmental Valuation Environmental Valuation MethodsMethods

HEDONIC PRICING METHODHEDONIC PRICING METHODOUTLINEOUTLINE

• Theory• Methodology• Illustration• Advantages• Issues and limitations• Case-study 1

Values of environmental amenities in Southold, Long Island

• Case-study 2Values of Environmental Amenities in Marickville and Rockdale, Sydney

• Case-study 3Effect of landfill sites on housing value, Minnesota

• Case-study 4Impact of Solid Waste Dumping on Land Prices in Tunisia

• Case-study 5Quarries in Mount Lebanon

HEDONIC PRICING METHODHEDONIC PRICING METHODTheoryTheory

• Used to estimate the value or price of an environmental feature by looking at actual markets where the attributes are traded

• Most commonly applied in relation to the public’s willingness to pay for housing/ property

• Also applied in labour markets for health economic valuation


Hedonism is the philosophy that pleasure is of ultimate importance, the most important pursuit. The name derives from the Greek word for delight.

Based on the assumption that people value the characteristics of a good, or the services it provides, rather than the good itself. Thus, prices will reflect the value of a set of characteristics, including environmental characteristics, that people consider important when purchasing the good– Ex: the price of a car reflects the characteristics of that car—

transportation, comfort, style, luxury, fuel economy, etc.• We can value the individual characteristics of a car or other good by

looking at how the price people are willing to pay for it changes when the characteristics change


• Assumes that– The price of a product is a function of its characteristics– The range of product choices is continuous– The choice is based on perfect information and with no mobility

restrictions– The amount of a particular characteristic can be varied

independently

• Relatively straightforward and uncontroversial to apply, because it is based on actual market prices and fairly easily measured data– If data are readily available, it can be relatively inexpensive to

apply– If data must be gathered and compiled, cost can increase

substantially


• Usually applied ex post, to examine the effects of developments and policies after implementation

• Can be used to estimate economic benefits or costs associated with– Environmental risk

• Ex: effect of information of different levels of earthquake damage on property values

– Environmental quality • Water pollution

– Ex: Impact on waterfront property• Air pollution• Noise

– Ex: Impact of highway noise and aircraft noise• Soil quality and erosion

– Environmental amenities• Aesthetic views, proximity to recreational sites, hazardous sites, waste

management sites, etc.

If all characteristics of houses and neighbourhoods in a given area were the same, except for the level of air pollution, then houses with better air quality would cost more. This higher price reflects the value of cleaner air to people who purchase houses in the area.

HEDONIC PRICING METHODHEDONIC PRICING METHODMethodologyMethodology

• Applying the Hedonic Pricing Method using housing prices– The price of a house is related to

• Structural characteristics of the house– Plot size, number of rooms, garage space, structural integrity, etc.

• Local socio-economic and public sector characteristics– Unemployment rate, social conditions, quality of schools, etc.

• Local amenity– Environmental quality, access to services, communications, etc.

– Upon controlling for non-environmental factors, any remaining differences in price can be attributed to differences in environmental quality


• Applying the Hedonic Pricing Method using housing prices– Collect the needed information

• Data requirements fall into 2 broad categories– Specific: Cross-section and/or time-series data on property values and

property and household characteristics for a well-defined market area» Structural and locational information» Details of purchase or tenancy (price, date, personal and financial

particulars of the purchasers)– Local: area where transaction occurred

» Neighbourhood, amenity, environmental, and socio-economic factors

» A measure or index of the environmental amenity of interest

• Data sources depend on country/ state involved– Government agencies, estate agents and realtors, mortgage granting

institutions, etc.– GIS, postcode classification of neighbourhood types, etc.


• Applying the Hedonic Pricing Method using housing prices (cont’d)– Analyze the data using regression analysis, relating

the price of the property to its characteristics and the environmental characteristics of interest

Output is a function linking property value to characteristics:

property value=a0+a1size+a2rooms+a3environmental quality+…

where property value is the dependent variablesize, rooms, environmental quality are independent

variables


• Applying the Hedonic Pricing Method using housing prices (cont’d)

• The analysis will indicate how much property values will change for a small change in each characteristic, holding all other characteristics constant

property value=a0+a1size+a2rooms+a3environmental quality• The analysis may be complicated by a number of factors

– The relationship between price and characteristics of the property may not be linear – prices may increase at an increasing or decreasing rate when characteristics change

– Multicollinearity:» many of the variables are likely to be correlated, so that

their values change in similar ways» this can lead to understating the significance of some

variables in the analysis


• Applying the Hedonic Pricing Method using housing prices (cont’d)– Different functional forms and model

specifications for the analysis must be considered

property value=a0+a1size+a2rooms+a3environmental quality+a4size2

log(property value)=a0+a1size+a2rooms+a3environmental quality


• Applying Hedonic Wage Models– Hedonic technique applied to wage rates

• An individual choice of job may be influenced by the job location if it improves access to desirable services

Hourly earnings=b0+b1location+b2firm size+b3education+b4experience

– Problems with this technique• High unemployment

– Individuals cannot satisfy their demand for environmental improvement due to unavailability of suitable jobs in areas of higher environmental quality

HEDONIC PRICING METHODHEDONIC PRICING METHODIllustration*Illustration*

• Agency staff want to measure the benefits of an open space preservation program in a region where open land is rapidly being developed

• The Hedonic Pricing Method is used because– Housing prices in the area appear to be related to proximity to

open space– Data on real estate transactions and open space parcels are

readily available

• Alternative Approaches– The travel cost method, if the open space of concern is used

mainly for recreation– Survey-based methods, like contingent valuation or contingent

choice, but are more difficult and expensive to apply

Adapted from www.ecosystemvaluation.org

HEDONIC PRICING METHODHEDONIC PRICING METHODIllustrationIllustration

• Step 1– Collect and compile data on residential property sales

in the region for a specific time period includingdependent variable:• Selling prices and locations of residential propertiesindependent variables:• Structural characteristics (lot size, number and size of rooms,

number of bathrooms)• Local socio-economic characteristics• Local amenity including the environmental characteristic of

concern- the proximity to open space

– Collect data on the amount and type of open space within a given radius of each property, noting the direct proximity of a property to open space

• data may be obtained from computer-based GIS maps

HEDONIC PRICING METHODHEDONIC PRICING METHODIllustrationIllustration

• Step 2– Statistically estimate a function that relates property values to the

property characteristics, including the distance to open space– The resulting function measures the portion of the property price

that is attributable to each characteristic– Estimate the value of preserving open space by looking at how

the value of the average home changes when the amount of open space nearby changes

• Step 3– Evaluate agency investments in open space preservation– Determine the benefits of preserving each parcel, which can

then be compared to the cost

HEDONIC PRICING METHODHEDONIC PRICING METHODAdvantagesAdvantages

• Can be used to estimate values based on actual behaviour and choices

• Property markets are relatively efficient in responding to information, so can be good indications of value

• Property records are typically very reliable

• Data on property sales and characteristics are readily available through many sources, and can be related to other secondary data sources to obtain descriptive variables for the analysis

• Versatile method that can be adapted to consider several possible interactions between market goods and environmental quality

HEDONIC PRICING METHODHEDONIC PRICING METHODIssues and LimitationsIssues and Limitations

• The scope of environmental benefits that can be measured is limited to things that are related to property values

• It will only capture people’s willingness to pay for perceived differences in environmental attributes, and their direct consequences– if people aren’t aware of the linkages between the

environmental attribute and benefits to them or their property, the value will not be reflected in home prices

• It assumes that people are free to select the combination of characteristics satisfying their preferences, given their income– However, the housing market may be affected by outside

influences, like taxes, interest rates, or other factors

HEDONIC PRICING METHODHEDONIC PRICING METHODIssues and LimitationsIssues and Limitations

• It is relatively complex to implement and interpret, requiring statistical expertise

• The results depend heavily on model specifications

• It is susceptible to multicollinearity i.e. a high degree of correlation among the variables under study which makes it difficult to estimate their individual effect– Ex: Air pollution measures where the levels of one form of

pollution (PM) is closely related to levels of another (NO2)

• All relevant variables must be included for the results to be valid

• Large amounts of data must be gathered and manipulated

• The time and expense to carry out an application depends on the availability and accessibility of data

HEDONIC PRICING METHODHEDONIC PRICING METHODCase Study 1: Case Study 1: Values of Environmental Values of Environmental

Amenities in Southold, Long Island*Amenities in Southold, Long Island*• Background

– The town of Southold, Long Island, New York has coastlines on both the Peconic Bay and Long Island Sound

– Southold is a relatively rural area, with a large amount of farmland

– Population and housing density are rapidly increasing in the town, resulting in development pressures on farmland and other types of open space

• The Challenge– The Peconic Estuary Program is considering various

management actions for the Estuary and surrounding land areas– A hedonic valuation study was conducted to assess some of the

values that may result from these management actions, using 1996 housing transactions.

*Adapted from ecosystemvaluation.org


Amenities in Southold, Long IslandAmenities in Southold, Long Island

• Analysis– Variables relevant for local environmental management with

significant impact on property values in Southold

Location of properties % change in value as compared to similar properties elsewhere

Adjacent to open space 12.8% higher per-acre valueAdjacent to farmland 13.3% lower per-acre valueWithin 20 meters of a major road 16.2% lower per-acre valueWithin an area with two- or three-acre zoning

16.7% higher per-acre value

For every percentage point increase in the percent of a parcel classified as a wetland, the average per-acre value increased by 0.3%


Amenities in Southold, Long IslandAmenities in Southold, Long Island

• Results – Calculate the value of preserving a parcel of

open space, by calculating the effects on property values adjacent to the parcel

• Ex: the value of preserving a 10 acre parcel of open space, surrounded by 15 “average”properties, was calculated as 410,907 USD

HEDONIC PRICING METHODHEDONIC PRICING METHODCase Study 2: Case Study 2: Values of Environmental Amenities Values of Environmental Amenities

in Marickville and Rockdale, Sydney*in Marickville and Rockdale, Sydney*• Data collected from house sales in 1973-1974 (1,414

observations)

• Information collected on 20 characteristics of each property and the local environment– Size, age, and construction of the house– Size of the plot of land– Type and amount of traffic on the road outside the house– Access to public transport and shops– Aircraft noise– Zoning and plans for road widening

• Some variable were measured directly (number of rooms) and others on a subjective scale (road traffic levels-noisy/normal/quiet)

*Adapted from Hodge (1995)

HEDONIC PRICING METHODHEDONIC PRICING METHODCase Study 2: Case Study 2: Values of Environmental Amenities Values of Environmental Amenities

in Marickville and Rockdale, Sydneyin Marickville and Rockdale, Sydney• Results

– Various types of statistical relationships between house prices and characteristics were tested

– Major determinants of prices were house quality and size and plot size

– Aircraft noise significant determinant of price in Marickville• $1,250 - 3,250 difference depending on house price

– Price of noisy road was $1,400 (5.6% of the house price)– $440 difference between poor and average view and $440

difference between poor and good view

• Limitations– Difficulties in measuring several house attributes– Prices estimated may not have been based on buyers’ informed

judgement

HEDONIC PRICING METHODHEDONIC PRICING METHODCase Study 3: Case Study 3: Effect of landfill sites on Effect of landfill sites on

housing value, Minnesota*housing value, Minnesota*• A sample of 708 single-family homes located

within close proximity of a landfill site

• House values rose by about 0.2 percent per mile from the landfill (when a linear specification is used)

• The effect on house value varied with distance• 12% for houses located at landfill boundary• 6% for houses located one mile from landfill boundary• 0% for houses located more than 2 miles from landfill

boundary(when a non linear specification is used)

Adapted from Hussen (1999); study by Nelson et al. (1992)

HEDONIC PRICING METHODHEDONIC PRICING METHODCaseCase--study 4:study 4: Impact of Solid Waste Dumping on Impact of Solid Waste Dumping on

Land Prices in TunisiaLand Prices in Tunisia

The situation

•Lack of municipal waste collection

•Accumulation of waste

•Presence of unauthorized dumpsites

•Lack of hazardous waste treatment

• Public health risks• Deterioration of

quality of life• Risks on natural

resources through the contamination of soil and water resources



The methodology• Step 1

Identify and characterize dumpsites

• Step 2Determine the area of land affected by each dumpsite

Ex: Agricultural Region– 1st affected zone: 30m radius– 2nd affected zone : 100m radius

• Step 3Conduct a survey to collect information on

– the value of land not affected by the dumpsite in the neighborhood

– The rate of land devaluation due to proximity to dumpsite

• Step 4Estimate the annual value of neighborhood land

– 10% in this caseEstimate the annual devaluation of land prices

Ex: Agricultural Region– 1st affected zone : 15%– 2nd affected zone : 10%

• Step 5Calculate the total annual loss in land prices



Parameter Grombalia Korba Beni Khaled Other TotalDumpsites total area (ha) 5 8 10 48 71Total area of 1st affected zone (ha) 2.7 3.3 3.6 20 29.6Total area of 2nd affected zone (ha) 8.4 9.9 10.7 60.5 89.5Value of land not affected by devaluation (DT/ha) 7,500 11,000 10,000 9,500

Annual value land not affected by devaluation (DT/ha) --10% of value of land 750 1,100 1,000 950

Annual loss of land prices in the 1st affected zone

303=2.7×750×15%

543=3.3×1,100

×15%547 2,853 4,236

Annual loss of land prices in the 2nd affected zone

630=8.4×750×10%

1,086 1,070 5,746 8,538

Total annual loss (DT/year) 12,774

Application- Agricultural region



Parameter Peri-urban Urban

Dumpsites total area (ha) 143 118Total affected area (ha) 543 47.2Average value of land not affected by devaluation (000 DT/ha) 700 1,700

Annual value of land not affected by devaluation (DT/ha) 70,000 170,000Percent annual loss of land prices (DT/year) 30% 35%Total annual loss (DT/year) 11,403,000 2,808,000

Application- Peri-urban & Urban region

Total land price losses due to solid waste dumping isTotal land price losses due to solid waste dumping is

14,224,000 DT/Year14,224,000 DT/Year

HEDONIC PRICING METHODHEDONIC PRICING METHODCaseCase--study 5: study 5: Quarries in Mount LebanonQuarries in Mount Lebanon

The situation• More than 700 quarries in Lebanon, of which > 50%

are in Mount Lebanon• Constructed with little consideration for the

environment and surrounding human settlements• Many of them are abandoned with minimal or no

rehabilitation

Destruction of natural vegetation and habitatsDestruction of natural vegetation and habitatsAir pollution from dustAir pollution from dust

Reduction in aesthetic value in and around sitesReduction in aesthetic value in and around sites

HEDONIC PRICING METHODHEDONIC PRICING METHODCaseCase--study 5: study 5: Quarries in Mount LebanonQuarries in Mount Lebanon

The methodology• Annual cost of environmental degradation due to

quarries– Loss in residential land value around quarries

– Loss in apartment values around quarries

Land/apartments in areas visually affected by quarriesexperience a decline in value

Cost of degradation associated with loss of Cost of degradation associated with loss of aesthetic valueaesthetic value

HEDONIC PRICE METHODHEDONIC PRICE METHODCaseCase--study 5: study 5: Quarries in Mount LebanonQuarries in Mount Lebanon

The methodology• A survey of impacts on surrounding areas around 4

quarries in Mount Lebanon was conducted• Additional impacts recorded to occur during quarries

operation:– Structural damage to buildings and infrastructures from

explosives used– Dust pollution– Traffic congestion due to quarry transport activities

• Additional impacts during operation are fraction of the losses in land and apartment values and not included in this assessment


The methodology• Step 1: Estimation of loss in land value around surveyed

quarries– Area of land affected ranged between 100,000 and 600,000 m2

– Decline in land price ranged between 7.5 and 125 US$/m2

* Total losses in land value were annualized at a discount rate f 10% over 20 to 100 years (high and low estimates)

Quarry Areas affected Land area affected (m2)

Decline in land price (US $/m2)

Loss in land value (US $ million)

Shnanaayer Shnanaayer municipality 600,000 125 75.0=600,000×125

Abou-Mizan Shirine, Bteghrine, and other villages 175,000 7.5 1.3Antelias Raboueh and Qornet Chehouane municipality 100,000 50 5.0

Total 875,000 93 81.3Annualized loss (“low”)* 8.1Annualized loss (“High”)* 9.6

Average annual loss in land value:Average annual loss in land value:8.85 US $ million8.85 US $ million


The methodology• Step 2: Estimation of loss in apartment values around

surveyed quarries– Area of apartments affected ranged between 8,000 and 36,000 m2

– Decline in apartment price ranged between 100 and 225 US$/m2


QuarryAreas affected Apartments

affected (m2)Decline in apartment

price (US $/m2)

Loss in apartment value (US $ million)

Shnanaayer Shnanaayer municipality 36,000 225 8.1Antelias Raboueh and Qornet Chehouane municipality 8,000 100 0.8

Total 44,000 202 8.9Annualized loss (“low”)* 0.9Annualized loss (“High”)* 1.0

Average annual loss in apartment value:Average annual loss in apartment value:0.95 US $ million0.95 US $ million


The methodology• Step 3: Estimation of land value occupied by other quarries (not

surveyed)– The total cost of degradation for >700 quarries is estimated based on the

value of land occupied by quarries– Estimated land value ranges between 3 and 5 US$/m2

– Average size of quarry ranges between 15,000 and 20,000 m2


Average annual value of land occupied by quarries:Average annual value of land occupied by quarries:5.45 US $ million, 0.035 % of the GDP5.45 US $ million, 0.035 % of the GDP

Lower bound

Higher bound

Number of quarries 710Average area of quarry (m2) 15,000 20,000Average land value (US $/m2) 3 5Total land value (US $ million) 207.6 415.3Annualized loss (US $ million/year)* 5.0 5.9Percent of GDP (%) 0.03 0.04


The methodology• Step 3: Assessing the total cost of degradation

due to quarries

The average total annual degradation cost due to The average total annual degradation cost due to quarries in Lebanon:quarries in Lebanon:

15.25 US $ million, 0.10 % of the GDP15.25 US $ million, 0.10 % of the GDP

Lower bound Higher boundLoss in land value around surveyed quarries (US $ million) 8.1 9.6Loss in apartment value around surveyed quarries (US $ million) 0.9 1.0Annual land value occupied by all quarries (US $ million) 5.0 5.9Total loss (US $ million) 14.0 16.5Percent of GDP (%) 0.08 0.10

AVERTING BEHAVIORAVERTING BEHAVIOROUTLINEOUTLINE

• Theory• Methodology• Issues to consider• Case-study 1

– Cost of pesticide contamination of drinking water

• Case-study 2– Consumption of Bottled Water in Lebanon

AVERTING BEHAVIORAVERTING BEHAVIORTheoryTheory

• Actions are taken to reduce or avoid the consequences/ cost of environmental damage– Water pollution

• Drinking bottled water • Purchasing water filters

– Air pollution• Frequent painting of dwellings due to smoke emissions from

a nearby factory • Moving away from a polluted location• Installing air purifiers• Staying indoors

– Noise• Installing soundproof walling to reduce noise

AVERTING BEHAVIORAVERTING BEHAVIORTheoryTheory

• In many cases, several types of aversive expenditures are undertaken simultaneously– Ex: Possible action in response to a noisy road

• Install double glazing• Move to another area

– Total benefits estimated by summing up all expenditures

Cost of avertingactions undertaken

People’s value for environmentalimprovement

≈

AVERTING BEHAVIORAVERTING BEHAVIORMethodologyMethodology

Application– Step 1: Identification of the environmental hazard and the

affected population• Monitoring equipment used to measure variables indicative of the

environmental hazard• Common sense to be adopted in defining the population at risk

– Step 2: Observation of the responses of individuals• Survey design should avoid biased sample, strategic bias, and self-

selection• Identify public expenditures

– Step 3: Measurement of the cost of taking action• Understand why the individual is taking a certain action• Understand if the chosen course is enough to avoid the hazard

AVERTING BEHAVIORAVERTING BEHAVIORIssues to considerIssues to consider

• Some actions are difficult to monetize– Moving house and leaving a familiar neighborhood

The cost of the action is a minimum estimate

• Some impacts have consequences with no possible averting actions– Air pollution on reduced visibility– Air pollution on lake acidification

The cost of the action is not accurate or complete

• Some goods provide additional non-environmental benefits– Bottled water tastes better– Air conditioning ameliorates room temperature

The should be accounted for to avoid overestimation of benefits

• Some are only partial substitutes for the environment– Double glazing partially reduces noise– Discomfort may still occur

The should be included in the analysis

AVERTING BEHAVIORAVERTING BEHAVIORCase Study 1: Case Study 1: The cost of solvent contamination The cost of solvent contamination

of drinking water*of drinking water*

• Background– Water supplies in Perkasie, Pennsylvania

contaminated with Trichloroethylene (TCE)• Detected TCE levels = 35 ppb• EPA TCE limit = 5 ppb

– No means of reducing contaminationWater consumers were notified of the

contamination

Adapted from Hodge, 1995


of drinking waterof drinking water• Methodology

– Postal survey• A sample 1733 households• A response rate of 45%• Questionnaire inquired about actions to avoid

exposure to TCE– Increased purchase of bottled water– Installation of home water treatment systems– Bringing in water from other sources– Boiling water

– Estimation of the cost of actions


of drinking waterof drinking water• Results

– In the absence of a clear logic for choosing the value to attach to the time spent on averting behavior, two approaches were considered

Actions undertaken Low estimatea

(USD)High estimateb

(USD)Increased purchase of bottled water 11,135 11,135New purchases of bottled water 17,342 17,342Home water treatment systemsc 4,691 4,691Hauling water 12,513 34,013Boiling water 15,633 64,135Total cost 61,313 133,334

a Time valued at minimum wage rate (3.35 USD/hr)b Time valued at estimated hourly wagec Because such a system would last for longer than the contamination period,

a proportion of the cost was included


of drinking waterof drinking water• Results

– Represent a minimum estimate of the costs of chemical contamination

• Only 43% of respondents were aware of the contamination• Averting behavior did not remove all contamination• No allowance made for possible ecological impacts

– Analysis suggests that if contamination could be avoided for an expenditure of 60,000 USD, then it should be undertaken

Adapted from Hodge, 1995

AVERTING BEHAVIOR AVERTING BEHAVIOR CaseCase--study 2:study 2: Consumption of Bottled Water in Consumption of Bottled Water in

LebanonLebanonBackground Substandard quality and inadequate quantity of potable water

Inadequate sanitation facilities and sanitation practicesInadequate personal, food and domestic hygiene

Impact on human health and quality of life

Cost to societyCost to society

AVERTING BEHAVIOR AVERTING BEHAVIOR CCasease--study 2:study 2: Consumption of Bottled Water in Consumption of Bottled Water in

LebanonLebanonBackground

Individuals and communities at risk from waterborne illnesses and mortality resort to

Aversive expenditures such as purchase Aversive expenditures such as purchase of bottled waterof bottled water

Lebanon’s population consumes a large quantity of bottled water mostly due to the perception that municipal water is of a low quality. According to the State of the Environment Report (ECODIT, 2000), bottled water consumption is about 115 liters per capita per year

AVERTING BEHAVIOR AVERTING BEHAVIOR CCasease--study 2:study 2: Consumption of Bottled Water in Consumption of Bottled Water in

LebanonLebanonMethodology•Cost of municipal water of inferior quality could be estimated based on bottled water consumption •Cost is equivalent to the difference between

– actual bottled water consumption– estimated consumption associated with taste and lifestyle

preference

Cost of Environmental degradation Cost of Environmental degradation = =

Actual bottled water consumption Actual bottled water consumption –– Expected consumption Expected consumption associated with preferenceassociated with preference


LebanonLebanonResults• Actual bottled water consumption

– According to the State of the Environment Report (SOER):• Bottled water expenditure represent 0.60% of total per capita

expenditure• The average price of one liter of bottled water is 0.23 US$

Unit ValuePer capita expenditures in Lebanon US$/capita/yr 4,465Per capita bottled water expenditures in Lebanon % 0.60Bottled water expenditure in Lebanon US$/capita/yr 26.8Average price of bottled water in Lebanon US$/liter 0.23Actual bottled water consumption in Lebanon Liter/capita/yr 115


LebanonLebanonResultsExpected bottled water consumption associated with

preference• Bottled water consumption associated with preference

– taste, lifestyle, etc…

• Estimation based on bottled water consumption in Europe and the United States in the 1970’s – 1970 figures used because of the large increase in consumption in

1980’s and 1990’s due to perception of inferior municipal water quality

• Expected consumption adjusted for GDP per capita differentials and price differentials between several European countries (in the 1970) and Lebanon


LebanonLebanonResultsExpected bottled water consumption associated with preference

– Estimate of expected bottled water consumption in Lebanon if consumers perceived no health risk of potable municipal water

Unit ValueLow High

GDP per capita 2000 (Western Europe and USA) US$/capita 27,750

GDP per capita 2000 (Western Europe and USA) US$/capita 17,253

GDP per capita 2000 in Lebanon US$/capita 3,875

Bottled water consumption in several European countries in 1970’s Liter/capita/yr 30 30

Income elasticity of bottled water demand 0.25 0.4

Price elasticity of bottled water demand (“low”) -1.5 -1.5

Price elasticity of bottled water demand (“High”) -2 -2

Average price of bottled water in European countries US$/liter 0.3 0.3

Average price of bottled water in Lebanon US$/liter 0.23 0.23

Expected bottled water consumption in Lebanon “Low” Liter/capita/yr 30 24Expected bottled water consumption in Lebanon “High” Liter/capita/yr 34 27


LebanonLebanonResults• Estimation of bottled water consumption to protect against risk

– Equivalent to “actual” less “expected” (if perceived risk of illness from municipal water were zero)

Unit ValueLow High

Actual bottled water consumption in Lebanon Liter/capita/yr 115 115Expected bottled water consumption in Lebanon “Low” Liter/capita/yr 30 24

Expected bottled water consumption in Lebanon “High” Liter/capita/yr 34 27

Average expected bottled water consumption in Lebanon Liter/capita/yr 32 26

Bottled water consumption to protect against risk Liter/capita/yr 83 89Lebanese population in 2000 Million capita 4.2

Total bottled water consumption to protect against risk Million liter/yr 356 383

Total cost of bottled water consumption to protect against risk Million US$/yr 82=356×0.23

88=383×0.23

% GDP 0.49 0.53

Average annual cost of bottled water consumption = 85 million US$

= 0.51% of GDP


Thank YouThank You




Sessions 7 & 8The Revealed Preference

ApproachGROUP EXERCISES

Regional Training Workshop on the Cost of Environmental Degradation Methodology July 1‐5, 2008 Crowne Plaza Hotel, Beirut, Lebanon

1 of 11

SESSIONS 7-8

GROUP EXERCISE 1

Economic Valuation of the Environment and the Travel Cost Approach: The Case of Ayubia National Park

(Himayatullah, 2003)

Case description The Government of Pakistan is seeking to revitalize its nature-based tourism sector to an expanding system of national parks and reserves. The Government of Pakistan has, in recent years, felt a serious concern over the deforestation and has shown significant interest in the growth of a renowned national park system. Despite limited number of national parks and reserves their management is far from satisfactory. This may partly be because of insufficient governmental funds and open access of visitors to these places. There is a need for a thorough investigation of how these parks can be well managed and how these environmental resources can be valued. The present case aims at obtaining economic information about benefits that flow from recreational use of a national park, Ayubia National Park (ANP), Pakistan. Ayubia National Park is a small national park in the Murree hills, Pakistan. It is located North of Murree in the Himalayan Range Mountains. Ayubia consisting of four hill stations, namely, Khaira Gali, Changla Gali, Khanspur and Gora Dhaka is spread over an area of 26 kilometers. These hill stations have been developed into a hill resort known as Ayubia. The chairlifts provided at this place are a matter of great attraction. It is an important place from the viewpoint of wildlife, nature, ecotourism, and education. This park provides refuge to the elusive leopard and the black bear. Bird watching is excellent here. There are steep precipices and cliffs on one side and on the other are tall pine trees. The scenery is superb with huge pine forests covering the hills and providing shelter to the larger and smaller mammals. Wild animals are also found in the thick forests around. Mammals in the park include Asiatic leopard, Black bear, Yellow throated marten, Kashmir hill fox, Red Flying squirrel, Himalayan palm civet, Masked civet and Rhesus Macaque. Birds in the park are Golden eagle, Griffin vulture, Honey buzzard, Peregrine falcon, Kestrel, Indian sparrow hawk, Hill pigeon, Spotted dove and Collared dove. Note that the access to ANP is free of charge One of the objectives of this study is to estimate the consumer surplus and recreational value (benefits) of the ANP. 1. Why is the TCM selected in this case? Which values will it capture? ___________________________________________________________________________

___________________________________________________________________________

___________________________________________________________________________

2. What alternative method could have been used and why? ___________________________________________________________________________

___________________________________________________________________________

___________________________________________________________________________


2 of 11

3. What are the three forms of a TCM?

a. _________________________________

b. _________________________________

c. _________________________________

4. For this case-study, the Individual Travel Cost Method was used. 5. What are the three most controversial aspects of the travel cost method?

a. _________________________________

b. _________________________________

c. _________________________________

6. Steps of the TCM process

A. Identify site ___________________________________________________________________________

___________________________________________________________________________

___________________________________________________________________________

B. Use an on-site questionnaire survey to collect data from visitors relating to:

a. _________________________________

b. _________________________________

c. _________________________________

d. _________________________________

The data used in this study were collected from 300 visitors by following systematic random sampling. The results showed the following:

− the sample respondents visited nature-based recreation about 7 times per year − The respondents’ mean yearly spending on recreation was Rs 5300 − The respondents’ mean monthly income is Rs 12,500 − About 61 percent of the respondents are male − 60 percent of the respondents were married − The average age of the respondents was 43 years − The average household size was about 6 − More than 76 percent were literate − Half of the respondents (50 percent) considered quality of the park as good − 60 percent of the visitors were from urban areas − more than 62 percent of the respondents wanted improvement in the quality of services of the

park − On the question about how more resources should be allocated for the park management, 38

percent of the respondents preferred an increase in entrance fee, 40 percent chose reallocation of government budget, 22 percent advocated voluntary donations towards parks’


3 of 11

management funds. − The visitors visited the ANP for different reasons. More than two-third (80 percent) of visitors

came to Galliat for recreation purposes. Some 20 percent of visitors reported travelling as the reason for coming to Galliat.

− Regarding income distribution as many as 45 percent of sample households fall in income group of Rs 10,000–20,000 per month. More than one-fifth (22 percent) households have monthly income in the range of Rs 5,000-10,000. Some 19 percent households have income of Rs 20,000-50,000. Taken together 64 percent households fall in income range of Rs 5,000-20,000.

C. Specify trip-generation function: V = f( , , , ,…) Factors that influence the demand for visits include:

a. _________________________________

b. _________________________________

c. _________________________________

d. _________________________________

Would you expect the variables below to increase or decrease frequency of visitation? (Tick the appropriate box)

− Increased travel cost: Increases visitation Decreases visitation

− Increased travel time: Increases visitation Decreases visitation

− Advanced age: Increases visitation Decreases visitation

− Gender: Males more less likely to visit

− Increased education level: Increases visitation Decreases visitation

− Increased household income: Increases visitation Decreases visitation

− Improved quality of the site: Increases visitation Decreases visitation

D. Estimate travel cost model: The basic model used in this study depicts: vi = α0 + α1 TC + α2 Yi+ α3 STi+ α4 Ai+ α5 Ei+ α5+k ΣDk+ ei … …

where: vi = the number of visits by the ith individual to the Park per period of time TC = round trip total cost (Rs) to the site including travel time M = household income (Rs/month) ST = travel cost to and from a substitute site A = age of the visitor E = highest level of education gained by visitor FS = family size, D1 = 1 if male and 0 otherwise, D2 = 1 if urban dweller and 0 otherwise D3= 1 if the visitor’s perception about the site’s recreational facilities is good and 0 if bad.


4 of 11

Econometric model used: Specification of the functional form is crucial to the benefit estimates obtained. In practice the choice of the functional form needs to be determined empirically. There is some consensus that a semi-log gives the best results namely regressing the logarithm of visitation rates against travel cost, etc. However, we will also use double log functional form of the above model to estimate (own- and cross) price and income elasticity of demand for visitation of the Ayubia National Park. The table below reports the results of the travel cost regression models in a linear fashion. In these models, most coefficients have the expected algebraic signs. The coefficient on travel costs is negative and statistically significant. As expected high travel costs incurred by individuals are inversely related to park visitation rate.

Estimated Results of Linear Regression Equations Variable Coefficients (t-stats)

Dependent variable # of visits, v

Intercept 2.41 (2.32)

Travel costs -0.06 (-2.58)***

Household income 0.0057 (2.23)**

Price of substitute 0.00025 (1.75)

Age -0.024 (-1.69)

Education 0.0059 (1.17)

Family size 0.0029 (0.35)

Dummy 1 0.332 (1.54)

Dummy 2 0.018 (1.40)

Dummy 3 0.045 (2.33)**

R2 0.47

F-statistics 13.5

** indicate significance at 5 percent level *** indicate significance at 1 percent level

E. Derive demand curve and obtain household consumer surplus estimates through

integrating under the demand curve. Two linear demand curves for ANP visitation were estimated from the survey data. The figure below shows the actual user demand for the Park and a hypothetical demand for the Park in case of improvement in the quality of park services. It implies that improvement in the quality of the services at the park would shift the demand curve upward to the right. In addition, the log-linear (semi-log) demand curve was also estimated. The semi-log demand curve is curvilinear and convex to the origin, which is relatively flat at low prices and steep at higher prices.


5 of 11

Visits with improvement: vi = 35.14 – 0.013 tc (R2=0.6313) Visits without improvement: vi = 44.32 –0.018 tc (R2=0.5666)

ln vi = 4.56 e-0.0079tc (R2= 0.7878)

a. Linear demand curve b. Log-linear demand curve


The total recreational value equals the consumer surplus plus total cost of visit.

Recreational Value of the ANP

Consumer Surplus Recreational Value

Per Visitor (Rs) 240 1996

Total (Rs million) 24.0 200.6


6 of 11

SESSIONS 7-8

GROUP EXERCISE 2

Forest Recreation Areas in Malaysia (FRAs) (Garrod and Willis, 2001)

Case description There are 74 Forest Recreation Areas in spread throughout the States of Peninsular Malaysia. FRAs are relatively small areas of natural virgin forest containing a variety of attractive landscape, fauna and flora, rivers and unique geological features, making them attractive as sites for outdoor recreation. They provide open access to non-priced recreation. Activities pursued in FRAs range from hiking, camping, swimming, etc. to more passive pursuits such as picnics, walks, observing the ecology, and enjoying the scenic attractions of the forests. The state government incurs costs for visitors facilities in terms of maintaining footpaths and toilets, collecting litter, providing information, and patrolling the site with park rangers. The objective of this study is to assess the value of the benefits of open access, non-priced recreation at FRAs. 1. Why is the TCM selected in this case? Which values will it capture? ___________________________________________________________________________

___________________________________________________________________________

___________________________________________________________________________


___________________________________________________________________________

___________________________________________________________________________

3. What are the three forms of a TCM?

a. _________________________________

b. _________________________________

c. _________________________________

4. For this case-study, the __________________form was used. 5. What are the three most controversial aspects of the travel cost method?

a. _________________________________

b. _________________________________

c. _________________________________


7 of 11

6. Steps of the TCM process A. Identify site

___________________________________________________________________________

___________________________________________________________________________

___________________________________________________________________________

B. Use an on-site questionnaire survey to collect data from visitors relating to:

a. _________________________________

b. _________________________________

c. _________________________________

d. _________________________________

A random sample of visitors was interviewed on a next to sample basis. A sample size of 385 interviews was set at the Jeram Linang FRA.

C. Specify trip-generation function: V = f( , , , ,…) Factors that influence the demand for visits include:

a. _________________________________

b. _________________________________

c. _________________________________

d. _________________________________

Would you expect the variables below to increase or decrease frequency of visitation? (Tick the appropriate box)

− Increased travel cost: Increases visitation Decreases visitation

− Increased travel time: Increases visitation Decreases visitation

− Advanced age: Increases visitation Decreases visitation

− Gender: Males more less likely to visit

− Increased education level: Increases visitation Decreases visitation

− Increased household income: Increases visitation Decreases visitation

− Improved quality of the site: Increases visitation Decreases visitation

The dependent variable is: ________________________________

D. Derive demand curve and obtain household consumer surplus estimates through integrating under the demand curve


8 of 11

Vij = f(Cij, Eij, Si, Yi, Ai, Hi, Ni, Mi) where: Vij = number of visits made by individual i to site j Cij = individual i's total cost of visiting site j Eij = individual i's estimate of the proportion of the day’s enjoyment which is attributable to the FRA Yi = income of individual i's household Ai = age of individual i Hi = size of individual i's household Ni = size of individual i's party Mi = dummy variable: whether individual i is a member of an outdoor organization The functional form used was the linear truncated Maximum Likelihood model, which took account of the truncated nature of the data which excluded individuals who chose not to visit the site over that time period. The model reported variables significant at the 0.15 statistical significance level. The variables made sense intuitively, whereby visits to FRA were: − negatively related to time and travel cost − positively related to being single and living with parents, being educated only to

primary and secondary levels, engaging in fishing at the site

The ML model for Jeram Linang FRA

Variable Coefficient Std. deviation T-ratio Prob:t:>x Mean of x

ONE -4.00525 1.25289 -3.197 0.00139 1.00000

Time and travel cost -0.64479 0.24658 -2.615 0.00893 2.34940

First visit -11.07460 1.64099 -6.749 0.00000 0.26216

Single- living with parents 5.78721 1.09867 5.267 0.00000 0.18919

Educated to primary level 6.04067 2.31138 2.613 0.00896 0.04594

Educated to secondary level 3.20342 1.17677 2.722 0.00648 0.13784

Other FRAs visited in last 6 months 0.48544 0.25220 1.925 0.05426 0.97297

Fished at the site 8.02514 3.22890 2.485 0.01294 0.00540

Engaged in nature walking -1.97990 1.10616 -1.790 0.07347 0.25135

E. Calculate aggregate consumer surplus for the site Consumer surplus estimates for individual visits to the sites are calculated by substituting values from the linear equation and integrating over the demand function.

The average economic benefit of the marginal visit for each visitor to FRA

FRA Consumer surplus per visit per adult

Jeram Linang (Kelantan) 0.78

Telok Bahang (Penang) 2.38

Gunung Pulai (Johor) 3.74


9 of 11

SESSIONS 7-8

GROUP EXERCISE 3

Valuing Landscape and Amenity Attributes in Central England (Garrod and Willis, 2001)

Case description The study area consists of around 4,800 km2 in Central England offering a variety of landscape form and feature. The focus is on the County of Gloucestershire in addition to large areas of Hereford and Worcester and small areas of Gwent, Wiltshire, Oxfordshire, and Avon. The aim is to study the impact of individual landscape and amenity features on house prices. Questions 1. Why is the HPM selected in this case? Which values will it capture? ___________________________________________________________________________

___________________________________________________________________________

___________________________________________________________________________


___________________________________________________________________________

___________________________________________________________________________

3. What data need to be collected? Specific and local ___________________________________________________________________________

___________________________________________________________________________

___________________________________________________________________________

4. How can proximity to landscape and amenity features be defined? ___________________________________________________________________________

___________________________________________________________________________

___________________________________________________________________________

5. What are the possible data sources? ___________________________________________________________________________

___________________________________________________________________________


10 of 11

___________________________________________________________________________

6. What type of analysis may be used for the collected data? What is the dependent

variable? What are the independent variables? Dependent variable: __________________________________________________________

Independent variables: ________________________________________________________

___________________________________________________________________________

7. What important factor needs to be accounted for? ___________________________________________________________________________

8. What types of functional forms may be used? ___________________________________________________________________________ 9. For what should the variables measuring proximity to landscape and amenity attributes be

measured?

− Statistical significance − Freedom from the effects of multicollinearity − Freedom from the effects of omitted variable bias − Variables measuring structural attributes of the model and socio-economic characteristics of the

study were permitted to enter the model even if they displayed some degree of collinearity 10. Empirical Results The following variables were statistically significant and robust in terms of estimation

Variable Definition

FOR20 0-1 variable: over 20% woodland in same 1-km2 as property

RIVER 0-1 variable: river or canal in same 1-km2 as property

SETTLEMENT 0-1 variable: rural settlement in same 1-km2 as property

WETLAND 0-1 variable: area of wetland in same 1-km2 as property

WOODVIEW 0-1 variable: probable woodland view in same 1-km2 as property

URBANVIEW 0-1 variable: probable urban view in same 1-km2 as property

SLOPE Predominant gradient of slope in same 1-km2 as property

ROAD Kilometers of road in same 1-km2 as property

RAIL Kilometers of rail track in same 1-km2 as property


11 of 11

Coefficient values of the amenity variables and the marginal implicit price Variable Coefficient (t-value) Marginal Implicit Price

(Percent of sample average house value)

FOR20 0.0710 (2.53) 7.10

RIVER 0.0490 (2.74) 4.90

SETTLEMENT 0.0834 (5.34) 8.34

WETLAND -0.1800 (-1.75) 18.00

WOODVIEW -0.0735 (-3.10) 7.35

URBANVIEW -0.0580 (-3.55) 5.80

SLOPE -0.0030 (-2.50) 0.30

ROAD 0.0279 (3.66) 2.79

RAIL -0.0543 (-2.77) 5.43

Marginal implicit price was inferred directly from the coefficient values of the semi-log model. It was found that the factors inflating house prices were proximity to rivers and waterways, proximity to land with high woodland cover, and good reach of local amenities and communications. The factors that depressed house prices were proximity to wetlands, proximity to rail lines, and view on urban areas.



Sessions 7 & 8The Revealed Preference ApproachThe Revealed Preference Approach

GROUP EXERCISESGROUP EXERCISES

OUTLINEOUTLINE

Case study 1Case study 1Economic valuation of the Environment in AyubiaNational Park, PakistanCase study 2Case study 2Non-priced Forest Recreation Areas in MalaysiaCase study 3Case study 3Valuing Landscape and Amenity Attributes in Central England

Case study 1Case study 1Economic valuation of the Environment in Economic valuation of the Environment in

AyubiaAyubia National Park, PakistanNational Park, Pakistan

Case DescriptionCase DescriptionStudy areaStudy area

• Ayubia National Park (ANP), Pakistan

• 26-km2 park in the Murree Hills in the Himalayan Range Mountains

• Important place from the viewpoint of wildlife, nature, ecotourism, and education– Refuge for the elusive leopard & black

bear and other mammals such as Kashmir hill fox, Red flying squirrel, etc.

– Bird watching: Golden eagle, Griffin vulture, Hill pigeon, Indian sparrow hawk, etc.

– Superb scenery– Chairlifts

Case DescriptionCase DescriptionEcosystem threatsEcosystem threats

What are the factors affecting visitor’s WTP for recreational services of the park? What are the benefits of the park? Would the improvement in recreational benefits of the park lead to a

higher demand for park visitation?

• The park is threatened by various activities– Forest fires– Soil erosion– Human settlements inside park– Pollution caused by villagers or visitors – Encroachment by local villagers

• Unsatisfactory management– Insufficient governmental

funds– Open access⊕

Need for a thorough investigation of the value of the environmental resources and services provided by the park to prove the necessity of its revitalization & better management

Applied MethodApplied MethodTravel cost approach (1)Travel cost approach (1)

• The consumer maximizes utility subject to a budget constrained by his income, represented by the product of hours of work to wage rate. The algebraic form is:Max: U(x,v)Subject to: wL = pxx + p0vWhere U= utility, x= the market good, v= visits to the park, w= hourly wage rate, L= work hours, px= price of market good x, p0= out-of-pocket expenses for a visit to the park

• In addition to out-of-pocket expenses, the consumers take time to travel. Time has anopportunity cost. The time constraint has the following form:T = L + HvWhere T= total household time available, H= time associated with a single round trip to the park including time spent on site, L= hours of wage labour

• The maximization problem then takes the following form:Max: Max: U(x,v)Subject to: wT = pxx + [p0 + wH]v

• The Lagrange expression is:L = U(x,v) + λ(wT-pxx+pvv)Where pv is the price for visiting the park

• Solving these problems will yield the following demand function for visit to the park:V = f(pv, px, Y, Z)Where the consumer demand for a visit to the park depends on the price for a visit, the price for other goods (substitute sites), the income and other socio-economic variables

Applied MethodApplied MethodTravel cost approach (2)Travel cost approach (2)

• When the demand function is derived, next step is to derive a demand curve (number of trips to park at different travel costs)– by holding income & prices of substitutes constant– and varying the price of the travel cost of a single

round-trip to the park• Based on demand curve, total value of park can

be computed, and is referred to as consumer surplus

• Total recreational value of park = consumer surplus + travel cost

Applied MethodApplied MethodSurvey designSurvey design

• In-person questionnaire with 300 random visitors of the park

• Two parts: 1. General information: gender, education, marital

status, age, income, place of living, etc.2. Visitor’s recreational behaviour:

• If they wanted improvement? • What is the method of resource allocation they prefer

(increase in entrance fee, reallocation of government budget, voluntary donations, etc.)?

• Why they’re visiting the park?

ResultsResultsTravel cost econometric modelTravel cost econometric model

• Regression analysis yielded a model of the form:vi = α0 + α1TC + α2Mi + α3STi + α4Ai + α5Ei + α6FSi + α6+kDk + ei

Where vi= the # of visits of the ithindividual to the park per period of time, TC= round trip total cost, M= monthly household income, ST= travel cost to and from a substitute site, A= age of the visitor, E= level of education, FS= family size, D1= 1 if male & 0 otherwise, D2= 1 if urban dweller, D3= 1 if visitor’s perception about site’s recreational facilities is good

Variable Coefficients (t-stats)Dependent variable # of visits, v

Intercept 2.41 (2.32)

Travel costs -0.06 (-2.58)***

Household income 0.0057 (2.23)**

Price of substitute 0.00025 (1.75)

Age -0.024 (-1.69)

Education 0.0059 (1.17)

Family size 0.0029 (0.35)

Dummy 1 0.332 (1.54)

Dummy 2 0.018 (1.40)

Dummy 3 0.045 (2.33)**

R2 0.47

F-statistics 13.5

**, and *** indicate significance at 5% & 1% levels, respectively

ResultsResultsConclusions from econometric modelConclusions from econometric model

• High travel costs are inversely related to park visitation rate

• Household income has positive impact on recreational demand

• No significant relationship between cost of substitute & the demand for the park

• Education & age had positive & negative impact on demand, respectively; however their impact is insignificant

• Dummy variables had all a positive impact, with dummy 3 having a statistically significant influence → if quality of park services are improved, visitors would pay more visits to the park

ResultsResultsDemand curvesDemand curves

Improvement in the quality of the park services would shift the demand curve upward to the right

The semi-log demand curve is curvilinear & convex to the origin, which is relatively flat at low prices & steep at higher prices

ResultsResultsRecreational valueRecreational value

• Total recreational value of park = consumer surplus + travel cost

• The annual monetary value of the park is 200 million Indian rupee (Rs), projected to become 209 million in case of improvement

• ⇒ The park constitutes a valuable environmental resource (in terms use & non-use values) and can be a significant source of benefits with proper conservation & management

Consumer Surplus Recreational Value

Actual New scenario Actual New scenario

Per visitor (Rs) 240.0 320.0 1996.0 2082.4

Total (Rs million) 24.2 35.01 200.6 209.2

Total (USD million) 0.6 0.9 4.9 5.1

End of Case-Study

Thank You

Case study 2Case study 2NonNon--priced Forest Recreation Areas in priced Forest Recreation Areas in

MalaysiaMalaysia

www.cumbavac.org/Earth_Day.htm

Case DescriptionCase DescriptionForest Recreation Areas in MalaysiaForest Recreation Areas in Malaysia

• What are FRAs?– Small areas of natural virgin forest containing a variety of

• Attractive landscape• Fauna and flora• Rivers and unique geological features

– Activities pursued in FRAs• Hiking, camping, swimming, …• Picnics, walks, observing the ecology, enjoying the scenic

attractions, …– Government incurs costs for visitors facilities

• Maintaining footpaths and toilets• Collecting litter• Providing information• Patrolling the site with park rangers

• Throughout the State of Peninsular Malaysia – 74 FRAs– Provide open access, non-priced recreation

www.mymalaysiabooks.com

Applied MethodApplied MethodIndividual Travel Cost ModelIndividual Travel Cost Model

Individual TCMDefines the dependent variable as the number of visits to an FRA made by each visitor over a specified period

TCM can be used to estimate consumer surplus by observing the number of visits in relation to price and estimating the recreation demand curve


Integrating over this equation permits the calculation of consumer surplus


• Questionnaire survey collected info on– Visits to the FRA– Preferences and activities of visitors– Demographic details of visitors

• Questionnaire structured to allow visitors to think about– Their use of the FRAs– The availability and use of substitute sites and

activities• Random sample of visitors on a next-to-pass basis

– Sample size = 385 interviews– Respondents at the Jeram Linang FRA

wanzee.fotopages.com/?&page=20


Cost element Financial cost Economic costFuel cost 0.110 0.060

Lubricant oil 0.009 0.009

Tyre cost 0.009 0.008

Maintenance 0.053 0.045

Depreciation 0.062 0.051

TOTAL 0.243 0.173


Total annual number of trips to FRAsTotal annual number of trips to forests, countryside, mountains, and other sites

• The ML model for Jeram Linang FRA

– Model reported variables significant at the 0.15 statistical significance level

– Variables make sense intuitively whereby visits to FRA • negatively related to time and travel cost• positively related to being single and living with parents, being educated only to

primary and secondary levels, engaging in fishing at the site,

Variable Coefficient Std. deviation T-ratio Prob:t:>x Mean of x

ONE -4.00525 1.25289 -3.197 0.00139 1.00000

Time and travel cost -0.64479 0.24658 -2.615 0.00893 2.34940

First visit -11.07460 1.64099 -6.749 0.00000 0.26216

Single- living with parents 5.78721 1.09867 5.267 0.00000 0.18919

Educated to primary level 6.04067 2.31138 2.613 0.00896 0.04594

Educated to secondary level 3.20342 1.17677 2.722 0.00648 0.13784

Other FRAs visited in last 6 months 0.48544 0.25220 1.925 0.05426 0.97297

Fished at the site 8.02514 3.22890 2.485 0.01294 0.00540

Engaged in nature walking -1.97990 1.10616 -1.790 0.07347 0.25135


Applied MethodApplied MethodResultsResults

• Consumer surplus estimates for individual visits to the sites are calculated by substituting values fromthe linear equation and integrating overthe demand function

The average economic benefit of the marginal visitfor each visitor to FRA

FRA Consumer surplus per visit per adultJeram Linang (Kelantan) 0.78Telok Bahang (Penang) 2.38Gunung Pulai (Johor) 3.74

www.rrcap.unep.org/.../malaysia/results.html

End of Case-Study

Thank You

Case study 3Case study 3Valuing Landscape and Amenity Valuing Landscape and Amenity

Attributes in Central EnglandAttributes in Central England

www.cotswolds.info/gloucestershire/index.shtml

Case DescriptionCase DescriptionStudy AreaStudy Area

• Around 4,800 km2 in Central England offering a variety of landscape form and feature– Focus on County of Gloucestershire

• with large areas of Hereford and Worcester• Small areas of Gwent, Wiltshire,

Oxfordshire, and Avon

• Aim to study the impact of individual landscape and amenity features on house prices

Hedonic Price Model

http://www.gloucestershire-hotels.co.uk/tudorhousehotelgl205bh.html

Application MethodApplication MethodHedonic Price ModelHedonic Price Model

• Choice of study region• Identification of house sale

transactions– Data set of a large mortgage

lender– Post-code data to identify and

remove transactions within urban areas

– Local knowledgehttp://commons.wikimedia.org/wiki/Image:Middleyard_Gloucestershire_With_Branches.JPG

Nearly 2,000 mortgages processed by the lender between 1985 and 1989 to purchase properties within the non-urban part of the study region


• Socio-economic data obtained from 1981 census and other sources

• Data on landscape features– Ordnance survey (OS) 1:50,000 map sheets

• Over 50 variables influencing house prices were obtained through a tedious process

• GIS not feasible due to copyright restrictions– Concentration on simple 0-1 variables

• Whether the 1 km OS map square containing the property of interest contained a particular feature of interest

– Rivers, wetlands, overhead cables– Post offices, public houses, country parks

– Some variables were continuous• Approximate land cover from forestry, buildings, open water, in the 1-Km2

• Approximate distance in kms to the nearest urban center, settlement, school, etc.

• Length of roads and rail tracks• The predominant aspect of the 1-Km2 with its average height above sea

level and degree of slope


• Particular problems encountered– Defining the proximity of a property to forestry

• Wooded areas may change in size over a short period of time– The relationship between the selected variables and

incorporated houses was ill-defined• Data provided by OS maps give only an approximation of a given

property’s proximity to a landscape feature• Such an approximation was considered sufficient

• Accounting for market fluctuations and inflations over the 5-year period– The use of dummy variables reflecting the year and quarter

of sale– The use of socio-economic dummies and dummies reflecting

local authority areas

Application MethodApplication MethodEmpirical ResultsEmpirical Results

• A semi-log functional form to model the data– Variables measuring proximity to landscape and

amenity attributes subjected to rigorous examination with respect to

• Statistical significance• Freedom from the effects of multicollinearity• Freedom from the effects of omitted variable bias

– Variables measuring structural attributes of the model and socio-economic characteristics of the study were permitted to enter the model even if they displayed some degree of collinearity


• The following variables were statistically significant and robust in terms of estimationVariable Definition

FOR20 0-1 variable: over 20% woodland in same 1-km2 as property

RIVER 0-1 variable: river or canal in same 1-km2 as property

SETTLEMENT 0-1 variable: rural settlement in same 1-km2 as property

WETLAND 0-1 variable: area of wetland in same 1-km2 as property

WOODVIEW 0-1 variable: probable woodland view in same 1-km2 as property

URBANVIEW 0-1 variable: probable urban view in same 1-km2 as property

SLOPE Predominant gradient of slope in same 1-km2 as property

ROAD Kilometers of road in same 1-km2 as property

RAIL Kilometers of rail track in same 1-km2 as property


Variable Coefficient (t-value) Marginal Implicit Price(Percent of sample average house value)

FOR20 0.0710 (2.53) 7.10

RIVER 0.0490 (2.74) 4.90

SETTLEMENT 0.0834 (5.34) 8.34

WETLAND -0.1800 (-1.75) 18.00

WOODVIEW -0.0735 (-3.10) 7.35

URBANVIEW -0.0580 (-3.55) 5.80

SLOPE -0.0030 (-2.50) 0.30

ROAD 0.0279 (3.66) 2.79

RAIL -0.0543 (-2.77) 5.43


INFLATEDHOUSEPRICES

DEPRESSEDHOUSEPRICES

Application MethodApplication MethodDiscussionDiscussion

• Study failed to produce comprehensive estimate of the marginal value of landscape and other amenity attributes

• Limitations– Data used– Inaccuracies in linking properties to amenities

• Better linkage may be provided by using GIS• Lack of amenity value does not mean lack of non-

market value– Ex: Proximity to wetlands depressed market prices of

houses• Dampness, flooding, increased insurance and maintenance

costs• However wetlands offer a valuable habitat for a number of

important species

End of Sessions 7 & 8

Thank You




Session 9THE CONTINGENT VALUATION

METHOD



Session 9THE CONTINGENT VALUATION METHOD

Preferences

Revealedpreferences

StatedPreferences

MarketValues

Travel Cost Methods

HedonicMethod

Averting Behavior

ContingentValuation

ChoiceExperiments




Contingent Valuation MethodContingent Valuation MethodOUTLINEOUTLINE

• Overview• Steps in a CVM procedure• Application• Associated biases• Illustration• Advantages• Issues and limitations• Sample applications• Case-studies

– Exxon-Valdez oil spill– Wilderness designation in Colorado– Beach degradation in Lebanon– Beach degradation in Morocco

Contingent Valuation MethodContingent Valuation MethodOverview Overview

• The CVM is the most widely used method for estimating non-use values

• The CVM uses interview techniques to ask individuals to place values on environmental goods and services

• It is called “contingent” valuation, because it is contingent on simulating a hypothetical market for the good in question

• It involves directly asking individuals– how much they would be willing to pay (WTP) to preserve or

use a given good or service OR– the amount of compensation they would be willing to accept

(WTA) to forgo specific environmental services• It is used to

– estimate economic values for all kinds of ecosystem and environmental services

– estimate both use and non use values

Contingent Valuation MethodContingent Valuation MethodOverview Overview

• It is the most controversial of the non-market valuation methods– many economists, psychologists and sociologists, for many different

reasons, do not believe the dollar estimates that result from CV are valid– Many jurists and policy-makers will not accept the results of CV.

• A carefully composed and tested study, where the circumstances are not too distant from the experience of the respondent and the issue is not too emotive, can produce answers of value

• Applications included estimates of the value of– Landscape– Recreation– Beaches– Water quality– Nature conservation– Endangered species– Visibility and air quality, etc

Contingent Valuation MethodContingent Valuation MethodSteps in a CVM ProcedureSteps in a CVM Procedure


1. Setting up the hypothetical market• Devise a convincing CV scenario to demonstrate that

respondents are actually stating their values for these services when they answer the valuation questions

• Establish a reason for a good or service• Pictorial aids could be of use

2. Obtaining bids• Possible bid vehicles include income taxes, property

taxes, value added or sales tax, utility bills, entry fees, payments into a trust fund

• Bids obtained through a questionnaire survey


2. Obtaining bids (cont’d)• Not all bid vehicles are viable options in a

given situation• Chosen bid vehicle should

– Have a plausible connection with the valued amenity

– Be perceived as fair and equitable• People have different views on the

acceptability of different types of taxes


2. Obtaining bids (cont’d)• Focus groups

– Precede CV surveys– Provide insight on the respondents’ likely understanding of

and attitude towards the issue being investigated– Provide valuable information in framing and designing a

CV study and questionnaire survey– Drawn from a cross-section of the population, stratified by

social class– Meeting lasts between one to two hours– Around 8-10 participants in a focus group discuss

• Their understanding of the context of the good• The good itself, its value, who should provide it• How it should be paid for, whether they should contribute• How much they are willing to pay


2. Obtaining bids (cont’d)• Concerns with focus groups

– Responses may be influenced by the person conducting the focus group

– Focus group participants have a longer time to think about the issue than in a typical CV survey

– Focus group participants have more information to base their judgment on

– Individuals behave differently in group situations compared to situations when they are alone

• Eliciting WTP/ WTA bids– Bids obtained through questionnaire survey and elicitation format

where respondents are asked to state their• Maximum WTP to increase quantity/ prevent quantity decrease of an

environmental good• Minimum WTA compensation to forgo an increase in the quantity/

accept less of the good– Various elicitation methods may be used


2. Obtaining bids (cont’d) 2. Obtaining bids (cont’d)• Questionnaire surveys

– Three sets of information obtained from respondents

– Data on use, preferences and substitutes should be collected at the beginning of the questionnaire

– Respondents must be reminded of their budget constraints when eliciting their bids

• Attitudes to environmental goods in general and reference for the particular good under investigation

• Awareness of substitute goods

• Use of the good, in relation to other goods

• Perceived non-use benefits of the good

• WTP and/or WTA bids for the goods using one or more of the elicitation methods

• Questions exploring reasons for the bids which can be used to eliminate illegitimate responses

• Questions to gauge the respondent's ambivalence

• Socio-economic information on the respondent and his/her household

-Data gathered to assess:- Representativeness of the sample- The theoretical validity of the bids using regression models


2. Obtaining bids (cont’d)Questionnaire administered in a number of ways


2. Obtaining bids (cont’d)• Sample size

– Determines the precision of the sample statistics used(mean WTP/ WTA)

– The larger the sample the smaller the variation in mean WTP measured by

• Standard error• Confidence intervals

– Mitchell and Carson (1989) devised a system to determine sample size based on choice of acceptable deviation between the ‘true’ and estimated WTPs

• For a deviation of 5%, 95% of the time, a sample of 6,000 is needed• For a deviation of 20%, 90% of the time, a sample of 286 is needed

– Mitchell and Carson argue that a sample > 600 is needed for applications seeking to evaluate policy

• This ensures a deviation of 15%, 95% of the time


3. Estimating mean and median WTP/WTA• WTP means, medians, modes, trimmed and modified

estimators, standards of deviation can be found from individual bids– Mean WTP, or trimmed or modified estimators based on mean

WTP are the most appropriate• Represent cardinal measures of the utility individuals derive from the

good– Median WTP

• is recommended because unaffected by large bids• is lower than mean WTP and may underestimate the value

– Trimmed estimator involves trimming the top and bottom 5% or 10% of the distribution of WTP observations

• Some true estimates of WTP may be omitted– Modified estimator provides the truest value

• Identifies and excludes biased and illegitimate responses by a series of questions included in the questionnaire


3. Estimating mean and median WTP/WTA• Probit, logit and random utility models can be

used for close-ended referendum bids• Bid curves can also be estimated

– By regressing WTP against socio-economic variables

WTPi = f(Yi, Vi, Pi, Si, Ei)Y = income level; V = visitis; P = preferences = S = substitutes; E = socio-economic variables (age, education, etc.)

• Differentiating bid curves (dWTP/dV) provides demand curve for the good

• Area under the curve = consumer surplus


4. Aggregating WTP or WTA amounts• Mean WTP/ WTA from the sample survey

are aggregated across the total population– TOTAL VALUE of the good/ service =

(mean WTP) × (# of population units)

• While mean WTP/ WTA may be modest for non-use benefits, the populations over which they are aggregated can be large


5. Assessing the Validity of CV studies• Content validity

– The appropriate framing of the study and questions asked in relation to the good being valued

• Criterion validity– The comparison of CV estimates with actual market or

simulated market experience• Construct validity

– The convergence between a CV measure and other such as travel cost and hedonic price measures of the value of the same good

– The extent to which the findings of the CV study are consistent with theoretical expectations


Contingent Valuation MethodContingent Valuation MethodApplication Application

Contingent Valuation MethodContingent Valuation MethodApplicationApplication

Contingent Valuation MethodContingent Valuation MethodApplicationApplication

• A good CV study will consider the following– Thoroughly pre-test the valuation questionnaire

for potential biases – Include validation questions in the survey

• to verify comprehension and acceptance of the scenario

• to elicit socioeconomic and attitudinal characteristics of respondents

• Make sure that survey results are analyzed and interpreted by professionals before making any claims about the resulting dollar values

Contingent Valuation MethodContingent Valuation MethodAssociatedAssociated Biases (1)Biases (1)

Contingent Valuation MethodContingent Valuation MethodAssociatedAssociated Biases (2)Biases (2)

Contingent Valuation MethodContingent Valuation MethodExamples of MinimizingExamples of Minimizing Biases (1)Biases (1)

Question Bias

a. I cannot afford to pay more water charges at present

b. I have no interest in having different flow levels in rivers

c. I would not pay anymore in water charges but I would be prepared to pay by some other means of payment

Payment vehicle

d. Someone else should pay rather than me Strategic

e. The water company should pay not customers Bid vehicle

f. Low levels in rivers are not a problem

g. I require more information to answer this question

h. Other reasons. Please specify

i. Don’t know

j. Refused to answer

Show card to elicit legitimate and illegitimate reason for NOT being WTP towards low flow alleviation in rivers

Contingent Valuation MethodContingent Valuation MethodExamples of MinimizingExamples of Minimizing Biases (2)Biases (2)

Question Bias

a. It was the most I could afford to pay

b. Rivers and beaches are important for recreation and I am happy to pay to ensure that they are well looked after

c. I would pay this much each year to ensure that rivers and beaches are protected for future generations

d. Rivers and beaches are important for wildlife and ecology and I am happy to pay to ensure that they are well looked after.

e. I wanted to show my support for environmental improvement in general Strategic

f. It’s an important issue and by saying I’d pay such a large sum each year I hope to get something done about it

Strategic

g. I’m very concerned about this issue and although I’m not sure I could afford to pay this much each year I wish I could

Hypothetical

h. Rivers and beaches are important for a number of reasons and I am happy to pay to ensure that they are well looked after

i. Other reason. Please specify;

j. Don’t know

k. Refuse to answer

Show card to elicit legitimate and illegitimate reason for being WTP towards low water quality

Contingent Valuation MethodContingent Valuation MethodIllustration Illustration

Hypothetical Scenario*• A remote site on public land provides important habitat

for several species of wildlife• The management agency in charge of the area must

decide whether to issue a lease for mining at the site– they must weigh the value of the mining lease against the

wildlife habitat benefits that may be lost if the site is developed

– non-use values are the largest component of the value for preserving the site because few people actually visit it, or view the animals that rely on it for habitat

*www.ecosystemvaluation.org


Application of the Contingent Valuation MethodStep 1• Define the valuation problem

– Determine what services are being valued• the resource to be valued is a specific site and the services it provides are

primarily wildlife habitat– Determine who the relevant population is

• Because it is federally owned public land, the relevant population would be all citizens of the U.S.

Step 2• Make preliminary decisions about the survey itself

– whether it will be conducted by mail, phone or in person– how large the sample size will be– who will be surveyed, etc. – The answers will depend, among other things, on

• the importance of the valuation issue• the complexity of the question being asked• the size of the budget


Step 2 (cont’d)• The researchers decided to conduct a mail survey

– they want to survey a large sample, over a large geographical area– They are asking questions about a specific site and its benefits, which

should be relatively easy to describe in writing in a relatively short survey

Step 3 • The actual survey design may take six months or more to complete• It is accomplished in several steps

– Starts with initial interviews and/or focus groups with the types of people who will be receiving the final survey, in this case the general public

• the researchers would ask general questions– about peoples’ understanding of the issues related to the site– whether they are familiar with the site and its wildlife– whether and how they value this site and the habitat services it provides


Step 3 (cont’d)• In later focus groups, the questions would get more detailed and

specific– to help develop specific questions for the survey– to decide what kind of background information is needed and how to

present it• People might need information on the location and characteristics of the site, the

uniqueness of species that have important habitat there, and whether there are any substitute sites that provide similar habitat.

• The researchers would also want to learn about peoples’ knowledge of mining and its impacts, and whether mining is a controversial use of the site. If people are opposed to mining, they may answer the valuation questions with this in mind, rather than expressing their value for the services of the site

– At this stage, different approaches to the valuation question and different payment mechanisms would be tested

– Questions that can identify any “protest” bids or other answers that do not reveal peoples’ values for the services of interest would also be developed and tested at this stage.


Step 3 (cont’d)• After a number of focus groups, pretesting of the survey is

started– The survey should be pretested with as little interaction with the

researchers as possible– Pre-testing will continue until a survey is developed that people

seem to understand and answer in a way that makes sense and reveals their values for the services of the site

Step 4• Actual survey implementation.

– Select the survey sample• the sample should be a randomly selected sample of the relevant

population, using standard statistical sampling methods– obtain a mailing list of randomly sampled U.S. Citizens– use a standard repeat-mailing and reminder method to get the greatest possible

response rate for the survey


Step 5• Compile, analyze and report the results

– data must be entered and analyzed using statistical techniques appropriate for the type of question

– The researchers also attempt to identify any responses that may not express the respondent’s value for the services of the site

– they can deal with possible non-response bias in a number of ways• The most conservative way is to assume that those who did not respond have

zero value

Results• Estimate the average value for an individual or household in the

sample, and extrapolate this to the relevant population in order to calculate the total benefits from the site– If the mean willingness to pay is $.10 per capita, the total benefits to all

citizens would be $26 million.

Contingent Valuation MethodContingent Valuation MethodAdvantages Advantages

• The most widely accepted method for estimating total economic value including use values and all types of non-use values

• Straightforward and highly flexible– can used to estimate the economic value of virtually anything– best suited to estimate values for goods and services that are easily

identified and understood by users and that are consumed in discrete units

• Requires few theoretical assumptions• The nature and results of CV studies are easy to analyze and

describe– Dollar values can be presented in terms of a mean or median value per

capita or per household, or as an aggregate value for the affected population.

• A great deal of research is being conducted to– improve the methodology– make results more valid and reliable– better understand its strengths and limitations

Contingent Valuation MethodContingent Valuation MethodIssues and Limitations (1)Issues and Limitations (1)

• Considerable controversy over whether CVM adequately measures people's willingness to pay for environmental quality– CV assumes that people understand the good in question and

will reveal their preferences in the contingent market just as they would in a real market

– Most people are unfamiliar with placing dollar values on environmental goods and services and may not have an adequate basis for stating their true value

• Expressed answers to a willingness to pay question may be biased

• Respondents may make associations among environmental goods that the researcher had not intended– For example, if asked for willingness to pay for improved visibility

(through reduced pollution), the respondent may actually answer based on the health risks that he or she associates with dirty air

Contingent Valuation MethodContingent Valuation MethodIssues and Limitations (2)Issues and Limitations (2)

• WTA very significantly exceeds WTP– this result may invalidate the CVM approach, showing responses to be

expressions of what individuals would like to have happen rather than true valuations

• The “ordering problem”– In some cases, people’s expressed willingness to pay for something has

been found to depend on where it is placed on a list of things being valued

• Difficulty to validate externally the estimates of non-use values

• When conducted appropriately, contingent valuation methods can be very expensive and time-consuming, because of the extensive pre-testing and survey work

• Many people, including jurists policy-makers, economists, and others, do not believe the results of CV

Contingent Valuation MethodContingent Valuation MethodSample application 1Sample application 1-- Mono Lake* Mono Lake*

Background


Contingent Valuation MethodContingent Valuation MethodSample application 1Sample application 1-- Mono Lake Mono Lake

Initial Work • Initial mail survey where residents of California

– Were told that, according to biologists, the higher flows to thelake were needed to maintain food supplies for nesting and migratory birds

– Were asked whether they would pay more on their water bill for higher cost replacement water supplies, so that natural flows could once again go into Mono Lake

Results• Average WTP per household =

$13 per month = $156 per year• The total benefits exceeded the

$26 million cost of replacing thewater supply by a factor of 50.

http://www.ram.org/pictures/sights/cctrip/mono_lake.gif

Contingent Valuation MethodContingent Valuation MethodSample application 1Sample application 1-- Mono Lake Mono Lake

Follow-up Work • The State of California hired a consulting firm to perform a more

detailed CV survey• New survey

– Involved the use of photo-simulations showing what the lake would look like at alternative water levels

– Gave detailed information about effects of changing lake levels on different bird species

– Was conducted over the telephone, with people who had been mailed information booklets with maps and photo-simulations

– Survey respondents were asked how they would vote in a hypothetical referendum regarding Mono Lake.

• This study showed that the benefits of a moderately high (but not the highest) lake level were greater than the costs

• The California Water Resources Control Board reduced Los Angeles’ water rights by half, from 100,000 acre feet to about 50,000 acre feet, to allow more flows into Mono Lake

Contingent Valuation MethodContingent Valuation MethodSample application 2Sample application 2-- Water Over the FallsWater Over the Falls**

Background• The Federal Energy Regulatory Commission faced a licensing

decision regarding how much water the utility company should allow to flow over the falls at a recreation area

• Increasing the flow over the falls would result in less hydropower generated, but more water for recreation

• The previous license required only a minimum in stream flow of 50 cubic feet per second, which reduced the flow over the falls to a trickle

• A contingent valuation survey was developed to determine how much visitors to the falls would be willing to pay for increasedoverflow levels


Contingent Valuation MethodContingent Valuation MethodSample application 2Sample application 2-- Water Over the FallsWater Over the Falls

Application • The survey

– included pictures of the falls at four different flow levels– was mailed to a sample of previous visitors to the site

• The key survey questions asked– how much individuals would pay to visit the falls with each of the

four flow levels depicted in the photos, and how many times theywould visit each year at the four different flow levels

Results• A statistical analysis of the survey results used to estimate a total

recreation benefit function– the economic value of additional flows in each month was calculated,

and compared to the economic value of the foregone hydropower required to allow the additional flows

• The resulting optimum flow level during the summer months, when visitation was high, was calculated as 500 cubic feet per second, which was ten times larger then the existing minimum in stream flow

Contingent Valuation MethodContingent Valuation MethodSample application 3Sample application 3-- Glen Canyon Dam*Glen Canyon Dam*

Background• In the early 1980’s it became clear that continued operation of the

Glen Canyon dam to provide peak-load power – adversely affecting the downstream ecosystem in the Grand Canyon– significantly reducing the quality of recreational rafting

• The valuation question of concern was how much recreational rafting was worth compared to the market value of the peak-loadpower supply

Application • The study attempted to quantify

how the value of rafting in theGrand Canyon would change withmore even base flows, as compared to reduced flows during peak-power periods

http://www.hdprint.co.uk/ftp/CanyonLands/212%20-%20Glen%20Canyon%20Dam%20from%20plane.jpg


Contingent Valuation MethodContingent Valuation MethodSample application 3Sample application 3-- Glen Canyon DamGlen Canyon DamResults • Substantial economic values for rafting with increased water flows =

$2 million per year• CVM helped change perspectives about how economic tradeoffs should be

discussed– the challenge was now to find a release pattern that increased the economic value of

all uses of the river water• More even flows were put into place while the final environmental impact

studies were being prepared• The study represented one of the first federally-funded projects to estimate

non-use values

Additional Research • It became more obvious that citizens throughout the U.S., not just rafters,

cared about how dam operations affected the natural resources of the Grand Canyon

• The Bureau of Reclamation funded a major contingent valuation study of households throughout the U.S. to estimate their willingness to pay for flow regimes that would protect the natural resources in the Grand Canyon.

• Results showed strong support for a more natural flow regime

Contingent Valuation MethodContingent Valuation MethodSample application 4Sample application 4-- NonNon--commercial Fish* commercial Fish* Background• Rivers in the Four Corners Region provide 2,465 river

miles of critical habitat for nine species of fish that are listed as threatened or endangered

• Continued protection required habitat improvements to imitate natural water flows needed by fish– fish passageways– bypass releases of water from dams

• A CV survey was used to estimate the economic value for preserving the critical habitat


Contingent Valuation MethodContingent Valuation MethodSample application 4Sample application 4-- NonNon--commercial Fish commercial Fish

Application • Sample

– Random sample of 800 households in the Four Corners states of Arizona, Colorado, New Mexico, and Utah (with the proportions based on the states’ relative populations).

– An additional 800 households from the rest of the U.S.• Survey respondents

– provided detailed maps that highlighted the areas designated as critical habitat units for the fish

– told that some State and Federal officials thought the combined costs of the habitat improvements and the restrictions on hydropower were too costly and had put forward a proposal to eliminate the critical habitat unit designation

– asked if they would contribute to the Four Corners Region Threatened and Endangered Fish Trust Fund

– told that efforts to raise funds would involve contributions from all U.S. Taxpayers

• If a majority of households voted in favor of the fund, the fish species would be protected from extinction

– through water releases from Federal dams timed to benefit fish– through the purchase of water rights to maintain in stream flows

Contingent Valuation MethodContingent Valuation MethodSample application 4Sample application 4-- NonNon--commercial Fishcommercial Fish

Application (cont’d)• Respondents were told that• within the next 15 years, three fish species would increase in population to the point that

they would no longer be listed as threatened species• if a majority of households in the U.S. voted not to approve the fund, the critical habitats

shown on the map would be eliminated, causing the extinction of four of the nine fish species in 15 years

• The exact wording on the questionnaire was:

Suppose a proposal to establish a Four Corners Region Threatened and Endangered Fish Trust Fund was on the ballot in the next nationwide election. How would you vote on this proposal? Remember, by law, the funds could only be used to improve habitat for fish. If the Four Corners Region Threatened and Endangered Fish Trust Fund was the only issue on the next ballot and it would cost your household $______ every year, would you vote in favor of it?

(Please circle one.) YES / NO

The dollar amount, blank in the above illustration, was filled in with one of 14 amounts ranging from $1-$3 to $350, which were randomly assigned to survey respondents.

Contingent Valuation MethodContingent Valuation MethodSample application 4Sample application 4-- NonNon--commercial Fishcommercial Fish

Results • The average WTP = $195 per household• When extrapolated to the general

population, the value of preserving the habitat areas was determined to be far in excess of the costs.

Contingent Valuation MethodContingent Valuation MethodSample application 5Sample application 5-- Salmon RestorationSalmon Restoration**

Background• The removal of dams blocking salmon migration

routes has been proposed– the Elwha and Glines dams on the Elwha River on the

Olympic Peninsula in Washington• 200-foot dams • very old • have no fish ladders• block migration of fish to 70 miles of pristine

spawning grounds in Olympic National Park– Dam removal would more than triple

salmon populations on the Elwha River– Cost to remove the dams and the 50 years of

sediment build-up behind them was estimated = $100-$125 million

seattletimes.nwsource.com/html/localnews/2001


Contingent Valuation Method Contingent Valuation Method Sample application 5Sample application 5-- Salmon RestorationSalmon Restoration

Application • CVM survey developed to estimate the economic values associated

with the removal of the dams– Households in Washington and elsewhere were surveyed– Asked if they would vote in favour of removing the dams and restoring

the river, in order to triple salmon populations at an annual cost that varied across households.

Results • The estimated economic values per household ranged from $73 for

Washington households to $68 for the rest of the U.S. Households

• The economic value to Washington residents alone would nearly beenough to justify removing the dams and restoring the river– National willingness to pay was in excess of $1 billion

Contingent Valuation MethodContingent Valuation MethodCaseCase--study 1study 1-- Exxon Valdez Oil Spill* Exxon Valdez Oil Spill*

Background• In March 1989, the oil

tanker Exxon Valdezwent aground on Bligh Reef Prince William Sound, Alaska

• Around 11 million gallons of crude oil were spilled

• In 5 months,– Oil has moved across

nearly 10,000 square miles of water

– About 1,600 mile of the Sound’s shoreline was heavily oiled

http://menlocampus.wr.usgs.gov/50years/accomplishments/oil.html

http://menlocampus.wr.usgs.gov/50years/accomplishments/oil.html

* Hodge, I, 1995.

Contingent Valuation MethodContingent Valuation MethodCaseCase--study 1study 1-- Exxon Valdez Oil Spill Exxon Valdez Oil Spill

Spill Impact• On wildlife

– Over 20,000 dead birds recovered including 100 bald eagles

– Over 2,650 dead sea otters

– Seals and other species were also damaged or killed including plants and microorganisms

– None of the losses threatened species extinction

– Birds and mammal populations expected to recover within 3-5 years

http://www.alaska-in-pictures.com/data/media/4/exxon-victim_2901.jpg

www.channel6.dk/native/uk/page104.html


Spill Impact• On commercial and recreational fishing

and tourism– These impacts could be valued easily

• Non-use values: existence, options, and bequest values–– CVM study CVM study in connection with legal action by

State of Alaska against Exxon corporation


Methodology• Survey of residents across the US • Alaska excluded to focus on non-use values• In principle, survey should ask about the WTA compensation for the

damage arising from the spill• WTP approach adopted due to difficulty in survey design• Hypothetical market:

– Proposal for a scheme to prevent future oil spills of the sort that had been experienced

• Escort ships to accompany oil tankers through the Sound• Escort ships carry special booms to be used immediately in the case of an oil spill

to contain the damage• Spilled oil then skimmed off and taken for safe disposal

– Without the scheme, oil spill to occur within the next 10 years– Scheme financed from a special tax on oil company profits and from a

single tax on all households


Methodology• Survey

– Investigated respondents prior knowledge on the issue

– Provided respondents with info on the spill and its impacts

– Basic valuation question• Whether or not he/she would vote for a proposal to

implement the scheme given a specified level of a single one-time tax

– Tax values set at $10, $30, $60, and $120• If responded answered ‘yes’, the amount was raised and the

question asked just once more– Collected information on

• Interest in environmental issues, household composition, education and incomes


MethodologyExcerpts of the administered questionnaire:

“The only mammals killed by the spill were sea otters and harbour seals. This card shows information about what happened to Prince William Sound. According to scientific studies, about 580 otters and 100 seals in the Sound were killed by the spill. Scientists expect the population size of these two species to return to normal within a couple of years after the spill.

Many species of fish live in these waters. Because most of the oil floated on the surface of the water, the spill harmed few fish. Scientific studies indicate there will be no long-term harm to any of the fish populations.

#2. Of course, whether people would vote for or against the escort ship program depends on how much it will cost their householdAt present, government officials estimate the program will cost your household a total of $______. You would pay this in a special one-time charge in addition to your regular federal taxes. This money would only be used for the program to prevent damage from another oil spill in Prince William Sound.If the program cost your household a total of $______, would you vote for or against it?”


MethodologyExcerpts of the administered questionnaire:

“#3. What if the final cost estimate showed that the program would cost your household a total of $_______. Would you vote for or against the program?

#4. What is it about the program that made you willing to pay something for it?

#5. Before the survey, did you think the damage caused by the Valdez oil spill was more serious than was described to you, less serious, or about the same as described?

#6. Is anyone in your household an angler, birdwatcher, backpacker, or environmentalist?

#7. This card shows amounts of yearly incomes. Which category best describes the total income from all members of your family before.”


Results• 1,043 interviews completed successfully• Response rate = 75 percent• Proportion of respondents voting for the scheme

– Proportion decreases as cost of scheme increases– There is little difference between the $30 and $60

questions

Questionnaire version

Initial tax level per household (USD)

Percent of respondents willing to pay taxes

A 10 67

B 30 52

C 60 51

D 120 34

Positive response to alternative tax levels


Results• Median = 31 USD (adopted)• Mean = 94 USD (dismissed)

– Considered unreliable due to the nature of the questions asked

• One third were not willing to pay at either of the offered prices– They believed that oil companies should pay

• Total value for non-use values lost in the US:

31 USD 31 USD ×× 91 million households = 2.8 billion USD (CI: 2.491 million households = 2.8 billion USD (CI: 2.4--3.2)3.2)

Contingent Valuation MethodContingent Valuation MethodCaseCase--study 2study 2-- Wilderness designation in Colorado*Wilderness designation in Colorado*

Case• Evaluating increments of wilderness

designations in Colorado in Summer 1980

http://images.google.com/imgres?imgurl=http://lh3.google.com/_t45xPwpSIdI/RihTxMzwEeI/AAAAAAAAAhw/n7pXikvzgB0/s800/Pano%2B-%2BIMGP3464%2B-%2B3952x3754%2B-%2BPLIN%2B-%2BBlended%2BLayer.jpg&imgrefurl=http://picasaweb.google.com/lh/photo/S4znX1uzeyI6XzqK2vmmZQ&h=760&w=800&sz=109&hl=en&start=1&tbnid=gVAC_Q6WZm7qRM:&tbnh=136&tbnw=143&prev=/images%3Fq%3Dwilderness%2BColorado%26gbv%3D2%26hl%3Den

http://ridethegreatdivide.blogspot.com/2007/07/south-san-juan-wilderness.html

* Hussen, AM, 1999

Contingent Valuation MethodContingent Valuation MethodCaseCase--study 2study 2-- Wilderness designation in ColoradoWilderness designation in Colorado

Methodology• Mail survey• Sample size = 218 Colorado households• Participants shown 4 maps of the State of Colorado

– Map 1: • 1.2 million acres of land currently designated as wilderness• Represent 2 percent of the state land

– Map 2:• 2.6 million acres of land hypothetically designated as wilderness

– Map 3:• 5 million acres of land hypothetically designated as wilderness

– Map 4:• 10 million acres of land hypothetically designated as wilderness

• Participants provided with realistic and credible information about the hypothetical market


Methodology• Respondents asked to write their maximum annual

WTP four the preservation of the 4 maps• Respondents then asked to allocate their WTP

among four categories of value– Recreational use– Option demand– Existence demand– Bequest demand

• Data was gathered, processed, and a statistical demand analysis was employed to estimate preservation values

Preservation valueof wilderness


Results: Total annual consumer surplus (US$) from recreation use and preservation value

Value categories Map A(1.2 M acres)

Map B(2.6 M acres)

Map C(5 M acres)

Map D(10 M acres)

Recreation use valuePer visitor dayTotal, million

14.0013.2

14.0021.0

14.0033.1

14.0058.2

Option valuePer householdTotal, million

4.044.4

5.446.0

7.348.1

9.2310.2

Existence valuePer householdTotal, million

4.875.4

6.567.2

8.869.7

11.1412.3

Bequest valuePer householdTotal, million

5.015.5

6.757.4

9.1010.0

11.4612.5

Preservation value to Colorado residentsPer householdTotal, million

13.9215.3

18.7520.6

25.3027.8

31.8335.0

Total annual recreation use value and preservation value to Colorado households (million)28.5 41.6 60.9 93.2


Discussion:• Increasing the area of wilderness from 1.2 to 2.6

acres increase total value by 46 percent (28.5 to 41.6 million USD)

• For all wilderness designations, non-use values represented a significant portion of the total value

Contingent Valuation MethodContingent Valuation MethodCaseCase--study 3: Coastal Degradation of Jounieh study 3: Coastal Degradation of Jounieh

Beach, LebanonBeach, LebanonBackground •The coastal zones of Lebanon represent unique economic and recreational assets

• Coast line > 240 km• > 50% of population concentrated along the coast

•Untreated municipal wastewater disposal, seafront solid waste dumps, uncontrolled development of resorts and vacation homes, etc.

Coastal zone/Beach degradationCoastal zone/Beach degradation(Loss in ecological and non(Loss in ecological and non--use values of the beach)use values of the beach)

Contingent Valuation MethodContingent Valuation MethodCaseCase--study 3: Coastal Degradation of Jounieh study 3: Coastal Degradation of Jounieh

Beach, LebanonBeach, Lebanon

MethodologyA survey was conducted to estimate the WTP for

restoration of Jounieh beach as an ecological protected area

Cost of degradation associated with ecological Cost of degradation associated with ecological and nonand non--use value of coastal areas of Lebanonuse value of coastal areas of Lebanon

Results• WTP per household for restoration of Jounieh beach

– Survey respondents were asked about their WTP for Jounieh beach restoration with payments each year for 10 years

– The average WTP per year of survey respondents was adjusted to reflect average income of Lebanese households

– A regression analysis (with low and high coefficients) was undertaken to estimate WTP in relation to income

Contingent Valuation Method Contingent Valuation Method CaseCase--study 3: Coastal Degradation of Jounieh study 3: Coastal Degradation of Jounieh


Survey LebanonLow High

Number of survey responses 94

WTP in relation to income WTP per 500US$ monthly income 15.3 10Average WTP/yr for 10 years US$/household 80.05 57 66WTP/household over 10 years US$

10% discount rate 385.3 446.15% discount rate 462.1 535.1

Results• WTP per household for restoration of Jounieh beach

– Household WTP annualized over 30 years – Two discount rates applied:

• 5%, reflecting a combination of social rate of intertemporal substitution and opportunity cost of capital

• 10%, reflecting opportunity cost of capital



Survey LebanonLow High

WTP/household over 10 years US$10% discount rate 385.3 446.15% discount rate 462.1 535.1

Annualized WTP/household US$

10% discount rate over 30 years 37.2 43.0


Results• Total WTP for restoration of Lebanese coast

– Number of Lebanese households: 935,000



LebanonLow High

Annualized WTP/household US$



Lebanese households 935,000 935,000

Total annualized WTP US$ million (all Lebanon) 27 40% GDP 0.16 0.24

Average annual WTP = 33.5 million US$= 0.2% of GDP

Contingent Valuation MethodContingent Valuation MethodCaseCase--study 4: Beach degradation in Moroccostudy 4: Beach degradation in Morocco

Background• The coastline of Morocco: 3,500 km, 13 coastal

zones, and 174 beaches• Domestic and industrial wastewater discharge,

industrial accidents, offshore pollution from ships and boat harbors, haphazard construction along the coast, etc.

Coastal zone/Beach degradationCoastal zone/Beach degradationIn 2002, “Monitoring Bathing Beach Waters in

Morocco” campaign showed that 28% of beaches were unfit for swimming

Contingent Valuation Method Contingent Valuation Method CaseCase--study 4: Beach degradation in Moroccostudy 4: Beach degradation in Morocco

Methodology• Annual cost of coastal degradation

– Willingness to Pay (WTP) of foreign tourists and Moroccan nationals living abroad to improve the coast

Tourists are willing to pay an additional value for Tourists are willing to pay an additional value for ““unspoiled destinationsunspoiled destinations”” as opposed to as opposed to ““slightly slightly

spoiledspoiled”” or or ““very spoiledvery spoiled”” destinationsdestinations– Lost recreational value for Moroccan residents

– Loss of local fishing (sardines)


Moroccan touristic statistics(Department of Tourism, Statistics, 2000, 2001, 2002)

Total tourists (foreigners and Moroccans living abroad) 4,113,037Total foreign nationalities 2,462,894

North Americans 155,388North and West Europeans 1,127,211Total North American and European tourists 1,282,599

Moroccan nationals living abroad 1,650,143Number of nights occupied by foreign tourists in classed hotels 13,539,586Average length of stay of foreign tourists (days) 5.5Total tourism expenses (million Dh) 21,644Average daily tourism expenses (Dh/per/day) 957


Methodology• Step 1: Estimation of WTP by foreign tourists to conserve the coast

– Based on a study conducted by Huybers and Bernnett (2000)*, British touristsare willing to pay US$ 70 per day (or 35% of their daily tourist expenses) for “unspoiled destinations” as opposed to “slightly spoiled” or “very spoiled”destinations

– This same proportion was applied to European and North American tourists visiting the Moroccan coast

Low HighAverage stay of North American and European tourists on the coast, days(1/3 and 2/3 of average stay, 5.5 days)

2 4

Number of North American and European tourists 1,282,599Total number of North American and European tourist stay days, days 2,565,198 5,130,396WTP to improve the coast, Dh/per/day(35% of the average daily tourism expenses in Morocco, 957)

338

Total WTP by North American and European tourists to improve thecoast (Million Dh) 867 1,735

* Huybers, T and Bernnett, Impact of the Environment on Holiday Destination Choices for Tropical North Queensland, Tourism Economics, 6(1), pp. 21-46, 2000.


Methodology• Step 2: Estimation of WTP by Moroccan nationals living

abroad to conserve the coast– The same approach applied to foreign tourists was applied to Moroccan

nationals living abroad, except:• WTP was applied to the household, since Moroccans normally visit their

country with the entire family (rather than individually, like foreign tourists)• Stays of Moroccan nationals are longer, averaging between 7 and 14 days

Low HighMoroccan nationals living abroad visiting country 1,650,143Average persons per household in Morocco 5.6Total number of Moroccan households living abroad visiting country 294,668

of which, households having similar economic conditions as NorthAmerican and European tourists (20% of total) 58,934

Average stay of Moroccan nationals on the coast, days 7 14Total number of stay days of Moroccan households living abroad, days 412,538 825,076WTP to improve the coast, Dh/household/day(35% of the average daily tourism expenses in Morocco, 957)

338

Total WTP by Moroccan nationals living abroad to improve the coast (Million Dh) 139.4 278.9


Methodology• Step 3: Assessing the total WTP to improve the

coast, due to beach degradation

The average WTP to improve the coast, due to The average WTP to improve the coast, due to beach degradation in Morocco:beach degradation in Morocco:

Dh 1,510 million, 0.2 % of the GDPDh 1,510 million, 0.2 % of the GDP

Low HighWTP by foreign tourists, million Dh 867 1,735WTP by Moroccan nationals living abroad, million Dh 139.4 278.9Total WTP to improve the coast, million Dh 1,007 2,014Percent of GDP (%) 0.14 0.28


Thank YouThank You




Session 10aTHE DISCRETE CHOICE METHOD



Session 10aTHE DISCRETE CHOICE METHOD

Preferences

Revealedpreferences

StatedPreferences

MarketValues

Travel Cost Methods

HedonicMethod

Averting Behavior

ContingentValuation

ChoiceExperiments




Discrete Choice MethodDiscrete Choice MethodOUTLINEOUTLINE

• Overview• Types of formats• Choice modelling• Illustration• Summary• Advantages• Issues and limitations• Case-studies

Discrete Choice MethodDiscrete Choice MethodOverviewOverview

• Contingent choice, also referred to as conjoint analysis, was developed in the fields of marketing and psychology to measure preferences for different characteristics or attributes of a multi-attribute choice

• The contingent choice method is similar to contingent valuation– asks people to make choices based on a hypothetical scenario– can be used to estimate economic values for any ecosystem or environmental service– can be used to estimate non-use as well as use values

• Differs from contingent valuation because– it requires people to evaluate several alternatives separately– it does not directly ask people to state their values in dollars

• values are inferred from the hypothetical choices or tradeoffs that people make

• Asks the respondent to state a preference between one group of environmental services or characteristics, at a given price or cost to the individual, and another group of environmental characteristics at a different price or cost

Discrete Choice MethodDiscrete Choice MethodOverviewOverview

• Is especially suited to policy decisions where a set of possible actions might result in different impacts– For example, improved water quality in a lake will

improve the quality of several services provided by the lake, such as drinking water supply, fishing, swimming, and biodiversity

• While contingent choice can be used to estimate dollar values, the results may also be used to simply rank options, without focusing on dollar values

• There are a variety of formats for applying contingent choice methods

Discrete Choice MethodDiscrete Choice MethodTypes of FormatsTypes of Formats

Discrete Choice MethodDiscrete Choice MethodChoice ModelingChoice Modeling

• Choice experiments are used to examine the response of the individual to changes in the attributes of the scenario as well as the scenario as a whole– Allows breaking down the relevant attributes of the

situation and determining preferences over attributes– Allows for more flexibility than CVM

• Choice experiments attempt to identify the utility the individuals have for the attributes of the goods and services by examining the tradeoffs that they make between them when making choice decisions




• Initial screening of the attributes– Crucial stage in study design– Attributes should be familiar and relevant to

respondents– Attribute levels should be measureable using

quantitative or qualitative scales– Ways for portraying attributes

• Verbal, pictorial, etc.– It is important to define an appropriate number of

attributes• Too many attributes burden the respondents• Too few cause problems with estimation and reliability

– Pre-testing and focus groups helpful in defining attributes and determining their numbers


• Experimental design– The specification of a factorial or fractional

factorial experimental design to estimate the utility for the good in question

• Use an orthogonal main-effects plan sampled from the complete factorial design to select the profiles to be used in the choice experiment

– Procedures in computer packages such as SAS and SPSS may be used to create an orthogonal matrix based on the attribute levels specified by the researcher

Discrete Discrete Valuation MethodValuation MethodChoice ModellingChoice Modelling

Survey designQuestionnaire administered in a number of ways


• Analysis of the choices– Random utility theory used to model the choices

as a function of attribute levels• Based on the hypothesis that individuals make choices

based on the attributes of the alternatives along with some degree of randomness

– Based on repeated observations of choice, one can examine how the levels of various attributes affect the probability of choice

• An assumption of normality leads to the binary probit model

• An assumption of a Gumbel distribution means that the multinomial or Mother Logit can be employed

Discrete Choice MethodDiscrete Choice MethodIllustrationIllustration

Background• A remote site on public land that provides important habitat for

several species of wildlife• The management agency in charge must decide whether to issue a

lease for mining at the site• Suppose that there are several possible options for preserving

and/or using the site– allowing no mining and preserving the site as a wilderness habitat area– various levels and locations for the mining operation, each of which

would have different impacts on the site• The contingent choice method selected because

– the outcomes of several policy options needs to be valued– Non-use values are the largest component of the value for preserving

the site• The TCM will underestimate the benefits of preserving the site• The CVM might also be used but the survey questions might

become very complicated


Application• Contingent choice and contingent valuation have very similar

application – The main differences are in the design of the valuation

question(s), and the data analysis.

Step 1• Define the valuation problem

– Determine exactly what services are being valued, and who the relevant population is

• the resource to be valued is a specific site and the services it provides i.e. wildlife habitat

• because it is federally owned public land, the relevant population would be all citizens of the U.S.


Step 2• Make preliminary decisions about the survey

– whether it will be conducted by mail, phone or in person,

– how large the sample size will be, who will be surveyed, and other related questions

– In this case, the researchers decided to conduct a mail survey

• Administered to a large sample• Over a large geographical area• Questions about a specific site and its benefits should

be relatively easy to describe in writing


Step 3• Survey design is accomplished in several steps

– starts with initial interviews and/or focus groups with the types of people who will be receiving the final survey

– In the initial focus groups, the researchers would ask general questions• about peoples’ understanding of the issues related to the site• whether they are familiar with the site and its wildlife• whether and how they value this site and the habitat services it provides

– In later focus groups, the questions would get more detailed and specific• different approaches to the choice question are tested

– each choice might be described in terms of the site’s ability to support each of the important wildlife species.

– people will be making tradeoffs among the different species that might be affected in different ways by each possible choice of scenario

– Pre-testing the survey• People would be asked to assume that they’ve received the survey in the mail and to fill it out.• Then the researchers would ask respondents about how they filled it out, and let them ask

questions about anything they found confusing.– A mail pretest might be conducted.– This process is continued until a survey is developed that people seem to understand

and answer in a way that makes sense and reveals their values for the services of the site


Step 4• Survey implementation

– Select the survey sample• Randomly selected sample from a mailing list of randomly sampled U.S.

citizens– Use a standard repeat-mailing and reminder method, in order to

get the greatest possible response rate for the survey

Step 5• Compile, analyze and report the results

– The statistical analysis for contingent choice is often more complicated than that for contingent valuation

• requiring the use of discrete choice analysis methods to infer willingness to pay from the tradeoffs made by respondents.


Step 5 (cont’d)• Estimate the average value for each of the services of the site, for

an individual or household • Extrapolate to the relevant population in order to calculate the total

benefits from the site under different policy scenarios• The average value for a specific action and its outcomes can also be

estimated, or the different policy options can be ranked in terms of peoples’ preferences

• The results of the survey might show that– the economic benefits of preserving the site by not allowing mining are

greater than the benefits received from allowing mining• the mining lease might not be issued, unless other factors override these results

– the results might indicate that some mining scenarios are acceptable, in terms of economic costs and benefits

• rank different options and select the most preferred option

Discrete Choice MethodDiscrete Choice MethodSummarySummary

• Whatever format is selected,– choices that respondents make are

statistically analyzed using discrete choice statistical techniques, to determine the relative values for the different characteristics or attributes

– If one of the characteristics is a monetary price, then it is possible to compute the respondent’s willingness to pay for the other characteristics.


• A good contingent choice study will consider the following:

– Before designing the survey, learn as much as possible about how people think about the good or service in question

• Consider people’s familiarity with the good or service, as well as the importance of such factors as quality, quantity, accessibility, the availability of substitutes, and the reversibility of the change.

– Determine the extent of the affected populations or markets for the good or service in question, and choose the survey sample based on the appropriate population.

– The choice scenario must provide an accurate and clear description of the change in environmental services associated with the event, program, investment, or policy choice under consideration

• iI possible, convey this information using photographs, videos, or other multi-media techniques, as well as written and verbal descriptions


• A good contingent choice study will consider the following:

– The nature of the good and the changes to be valued must be specified in detail, and it is important to make sure that respondents do not inadvertently assume that one or more related improvements are included

– The respondent must believe that if the money was paid, whoever was collecting it could effect the specified environmental change

– Respondents should be reminded to consider their budget constraints – Specify whether comparable services are available from other sources,

when the good is going to be provided, and whether the losses or gains are temporary or permanent

– Respondents should understand • the frequency of payments required, for example monthly or annually,• whether or not the payments will be required over a long period of time in order to

maintain the quantity or quality change• who would have access to the good and who else will pay for it, if it is provided


• A good contingent choice study will consider the following– In the case of collectively held goods, respondents should

understand that they are currently paying for a given level of supply. The scenario should clearly indicate whether the levels being valued are improvements over the status quo, or potential declines in the absence of sufficient payments.

– If the household is the unit of analysis, the reference income should be the household’s, rather than the respondent’s, income

– Thoroughly pre-test the questionnaire for potential biases• test different ways of asking the same question• test whether the question is sensitive to changes in the description of

the good or resource being valued– Conduct post-survey interviews to determine whether

respondents are stating their values as expected.– Include validation questions in the survey

• to verify comprehension and acceptance of the scenario• to elicit socioeconomic and attitudinal characteristics of respondents

Discrete Choice MethodDiscrete Choice MethodAdvantages (1)Advantages (1)

• Can be used to value the outcomes of an action as a whole, as well as the various attributes or effects of the action

• it does not ask the respondent to make a tradeoff directly between environmental quality and money– the tradeoff process may encourage respondent introspection

and make it easier to check for consistency of responses.– respondents may be able to give more meaningful answers to

questions about their behavior (i.e. they prefer one alternativeover another), than to questions that ask them directly about the dollar value of a good or service or the value of changes in environmental quality

• Respondents are generally more comfortable providing qualitative rankings or ratings of attribute bundles that include prices, rather than dollar valuation of the same bundles without prices

Discrete Choice MethodDiscrete Choice MethodAdvantages (2)Advantages (2)

• Even if the absolute dollar values estimated are not precise, the relative values or priorities elicited by a contingent choice survey are likely to be valid and useful for policy decisions

• Minimizes many of the biases that can arise in open-ended contingent valuation studies where respondents are presented with the unfamiliar and often unrealistic task of putting prices on non-market amenities

• Has the potential to reduce problems such as expressions of symbolic values, protest bids, and some of the other sources of potential bias associated with contingent valuation

Discrete Choice MethodDiscrete Choice MethodIssues and LimitationsIssues and Limitations

• Respondents may find some tradeoffs difficult to evaluate, because they are unfamiliar

• The respondents’ behavior underlying the results of a contingent choice study is not well understood.

– Respondents may resort to simplified decision rules if the choices are too complicated, which can bias the results of the statistical analysis.

• If the number of attributes or levels of attributes is increased, the sample size and/or number of comparisons each respondent makes must be increased

• When presented with a large number of tradeoff questions, respondents may lose interest or become frustrated

• Contingent choice may extract preferences in the form of attitudes instead of behavior intentions

• By only providing a limited number of options, it may force respondents to make choices that they would not voluntarily make

• Contingent ranking requires more sophisticated statistical techniques to estimate willingness to pay.

• Translating the answers into dollar values, may lead to greater uncertainty in the actual value that is placed on the good or service of interest.

• Validity and reliability for valuing non-market commodities is largely untested.

Discrete Choice MethodDiscrete Choice MethodCaseCase--StudiesStudies

Discrete Choice MethodDiscrete Choice MethodCaseCase--study 1*study 1*-- Landfill Siting in Rhode IslandLandfill Siting in Rhode Island

Background• With its primary landfill nearing capacity, the State of Rhode Island

was faced with the need to choose locations for new landfills• Besides technical considerations, the State wanted to address the

social and economic tradeoffs and values related to the location of a landfill to avoid some of the controversy associated with landfill siting

Analysis• A contingent choice, paired comparison, survey was conducted

– The survey asked Rhode Island residents to choose between pairs of hypothetical sites and locations for a new landfill, described in terms of their characteristics

– The site comparisons described • the natural resources that would be lost on a hypothetical 500 acre landfill site• area surrounding the landfill

– Each comparison gave the cost per household for locating a landfill at each hypothetical site or location


Discrete Choice MethodDiscrete Choice MethodCaseCase--study 1study 1-- Landfill Siting in Rhode IslandLandfill Siting in Rhode Island

Results• Used by the State to predict how residents would

vote in a referendum on different possible landfill locations– First, 59 possible sites were selected, based on

geological and public health criteria.• sites were ranked using the contingent choice survey results,

in order to come up with a short list of potential sites

• The final decision, based on geological, public health, public preferences, and political considerations, was to expand the existing landfill site

Discrete Choice MethodDiscrete Choice MethodCaseCase--study 2*study 2*-- Management of the Peconic Estuary Management of the Peconic Estuary

SystemSystemBackground• The environmental and natural resources of the Peconic

Estuary System provide many services to the public– the bay waters, beaches, wetlands, ecosystems, habitats, and

parks and watershed lands

• The Peconic Estuary Program – established under the National Estuary Program– Responsible for creating a conservation and

management plan for the environment and naturalresources of the Estuary

• Information was needed about the value that the public holds for the ecosystem services of the Estuary.


Discrete Choice MethodDiscrete Choice MethodCaseCase--study 2study 2-- Management of the Peconic Estuary Management of the Peconic Estuary

SystemSystem

Analysis• Contingent choice survey to estimate the relative

preferences and economic values that residents and second homeowners have for preserving and restoring key natural and environmental resources– Open space– Farmland– Unpolluted shellfish grounds– Eelgrass beds– Intertidal salt marsh.

Discrete Choice MethodDiscrete Choice MethodCaseCase--study 2study 2-- Management of the Peconic Estuary Management of the Peconic Estuary

SystemSystemResults• The public has a strong attachment to environmental and amenity resources

of the Peconic Estuary, even if they do not use these resources directly• 97 percent of the respondents supported at least one hypothetical action to

protect resources, and indicated they would financially support such actions

• The survey results indicated that the resource priorities, or relative values of resources, are more reliable than are the dollar estimates of values,

– researchers recommended that relative values, rather than dollar values, be used in the process of selecting management actions.

Relative priorities for protecting natural resources

Per acre dollar values

Farmland $70 thousand Eelgrass $66 thousand Wetlands $54 thousand Shellfish $30 thousandUndeveloped land $13 thousand

Discrete Choice MethodDiscrete Choice MethodCaseCase--study 3*: Environmental Cost of Low River study 3*: Environmental Cost of Low River

Flows Flows

Background• South-west of England

encompasses– 176 beaches– 4,000 miles of rivers

• Abstraction for hydro-electric power stations

• Water imponded by reservoirs

• Abstraction by water companies from river or underlying aquifer

http://www.cornerwaysresidentialhome.co.uk/tavistock16.jpg

Low water flow

* Garrod and Willis, 2001

Discrete Choice MethodDiscrete Choice MethodCaseCase--study 3: Environmental Cost of Low River Flows study 3: Environmental Cost of Low River Flows

Background• Benefits of increased river flows comprise both use values

and non-use valuesExpressed preference approach most appropriate to elicit WTP

• Survey of general public will comprise a large proportion of non-users

Lack of familiarity makes it difficult to answer open-ended questions

• Choice experiment approach with the aim of estimating the marginal WTP of the general public for – unit improvements in low flow alleviation in rivers in south west

of England – unit improvements in the numbers of clean beaches in the area– unit improvements in the miles of unpolluted rivers in the area


Definitions• Non-use value arises from the knowledge

that the river remains healthy and viable and will persist

• Non-users were identified as those respondents who did not visit any of the flow rivers in the south west specified in the project


Questionnaire design• Series of focus groups undertaken

– Suggested that the public considered coastal and river pollutionto be most pressing problems in the south west

– When shown photographs, agreed that low flow was also important but not as pressing

• To obtain conservative welfare estimates that can be interpreted as lower-bound figures– A series of questions and statements reminded respondents of

other environmental issues that they might wish to support

• To introduce the notion of a multigood environment– Respondents asked about their donations to good causes and

their willingness to contribute more to these causes– Tested and refined over two separate pilot surveys


Questionnaire design• Respondents presented with a brochure

– Describing the Environmental Agency’s (EA) activities• Reducing river pollution• Monitoring marine pollution in coastal waters• Improving flows in low flow rivers

– Text kept to a minimum and illustrations used– Information limited to bullet points describing

• The problems being tackled• The causes and consequences• How EA was tackling the problem including the amount spent on a per

household• How much improvement has been achieved till now

• Survey sample limited to the south west where respondents relate better to familiar local issues


Questionnaire design• Survey used both

– Stated preference choice experiment– Dichotomous choice contingent valuation question

• Valuation scenario used– Whether or not respondents were willing to pay a specified amount to

increase the overall levels of environmental quality along rivers and beaches in the south west

• Environmental quality expressed as the level of three attributes

• In choice experiments, cards chosen randomly from an orthogonal set of 64 choice cards– Respondents given a card and asked to choose one of three choices– Then given three cards and asked to choose their preferred choice from

each

• Low flow embedded within a more holistic set of EA water qualityobjectives


Questionnaire design• Example of a choice experiment card:

CARD 06Please choose one column

CHOICE 1(current situation)

CHOICE 2 CHOICE 3

Increase in water charges needed to achieve targets

No increase £5 increase £10 increase

Beaches in the South West NOT MEETING European standards on cleanliness

9 beaches 5 beaches 3 beaches

Rivers in the South West WITHOUTgood quality water


Rivers in the South West WITHOUTacceptable flow levels



Empirical results• River usage by the general public

– Distance from rivers• Half the interviewed households live one mile or less from a

river• More than two-thirds lived within 2 miles of a river

– Recreational activities• 77 % of household regularly undertook recreational activities

along rivers• 88 % of the households had visited more than one river over the

past 12 months• Frequency of visits to beaches had the same frequency as visits

to rivers during the summer season


Empirical results• Public perception

– ¾ of respondents thought that rivers were an important source of water

– ½ thought too much water is being abstracted from rivers

• Public WTP for good causes– 80% prefer to see additional public expenditure

on the nature environment– 40% were willing to contribute more towards

‘good causes’


Empirical results• Responses from choice experiments used to

estimate discrete choice model of the probability Pr(i) of choosing a given alternative I

Pr(i) = exp (sVi)/Σexp(sVj)• Models were estimated using

– linear functional form• Used mainly for benefit estimation

– quadratic functional form• Some attribute coefficients were not statistically significant

– Coefficient values and t-statistics for various variables were estimated


Reduction Basic specification Extended specification

1 polluted beach £1.307 £1.431

1 mile of polluted river £0.017 £0.019

1 mile of low flow river £0.052 £0.058

Results


• Median welfare estimates for CVM scenarios

• Estimated using same models• Estimation involves

– an examination of how utility levels change as a result of a specified improvement

– calculating the magnitude of the associated increase in water charges that would be required to make the utility the same before and after the improvements

Model Annual welfare measure

Basic specifications £12.80

Extended specifications £15.36

Results


User and non-user populations for low-flow rivers from the south west

• The population of users:– all households who had visited a given low flow river in the two-year

period immediately preceding the survey• Nearly 45% of households in the south west are users

– Rivers are linear features– All rivers located in areas with scenic attractions

River User Households Non-user households

Miles affected by low flows

Allen 85,897 1,562,533 20Upper Avon 230,717 1,417,713 35Meavy 166,760 1,481,670 7Otter 157,982 1,490,448 5Piddle 157,854 1,490,576 16Tavvy 240,155 1,408,275 16Wylye 162,481 1,485,949 30

Results


Approximate aggregate annual benefits for improving low flows across the entire length of all low flow rivers in the south west

River Aggregate benefits for user Households

Aggregate benefits for visitor households

Aggregate benefits for non-user households

Allen £1,115,925 £130,563 £795,414

Upper Avon £1,952,868 £613,707 £1,391,975

Meavy £390,574 £88,716 £278,395

Otter £278,981 £60,033 £198,854

Piddle £892,740 £191,950 £636,331

Tavvy £892,740 £292,028 £636,331

Wylye £1,673,887 £370,457 £1,193,121

734,161 low-flow river users × length of river affected by low flow × £0.076 per mile

Results


• The costs of the cheapest available options for low flow alleviation estimated by ERM for each river

• Present value of the benefits of low flow alleviation estimated for the user sub-sample– Calculated by assuming a constant flow of benefits for the period

1997 to 2017 and discounting at 6%• Benefits exceeded costs by a wide margin

– Avon, Meavy, Wylye• Costs prohibitive on Otter• Benefits and costs similar for Allen

• Benefits based only on user samples and ignore benefits to non-users but include non-use benefits for low flow river visitors who do not visit the river in question

Results


Net present value of aggregate benefits for improving low flows across the entire length of all low flow rivers in the south west

River Present value of costs

Present value of aggregate user benefits

Net present value Benefit-cost ratio

Allen 11,867,000 13,915,000 2,048,000 1.17

Upper Avon 763,000 24,252,000 23,589,000 31.92

Meavy 80,000 4,870,000 4,790,000 60.88

Otter 34,430,000 3,480,000 -30,950,000 0.10

Piddle 5,471,000 11,132,000 5,661,000 2.03

Tavvy Unknown 11,132,000 - -

Wylye 224,000 20,873,000 20,649,000 93.18

Annual stream of costs and benefits discounted at 6% between 1997 and 2017

Results

Discrete Choice MethodDiscrete Choice MethodCaseCase--study 4*: Recreational Choice and Water study 4*: Recreational Choice and Water

QualityQuality

• Case study by Alberta Environment

• A combination of– Stated preference choice

experiment– Revealed preference

approach

• Choice experiment– Three options related to choice of

recreational activities• Recreation at standing water site• Recreation at running water site• Recreation at non-water site

* Garrod and Willis, 2001

Discrete Choice MethodDiscrete Choice MethodCaseCase--study 4: Recreational Choice and Water Qualitystudy 4: Recreational Choice and Water Quality

• Water-based recreation described using various attributes– Travel distance to reach the site– Water quality

• Presented as either good or bad– Terrain– Camping facilities– Presence of beach– Various attributes related to fishing

and other recreational activities

• Attributes of standing and running water alternatives– Treated as a collective factorial– An orthogonal main effects design

chosen which would vary all attribute levels simultaneously

http://www.hickerphoto.com/data/media/11/travel_alberta_T3258.jpg

http://www3.nationalgeographic.com/places/images/photos/photo_lg_alberta.jpg


Recreational attributes offered in stated preference choice cardsAttribute Description Attribute Description

Fish size LargeSmall

Fish species

Still: Pike & PerchPickerel, Pike & Perch

Running: Mountain whitefishRainbow trout & Mountain whitefishRainbow trout, Mountain whitefish & Brown troutCutthroat trout, Mountain whitefish & Bull trout

Fish catch rate

1 fish per 4 hrs1 fish per 80 mins1 fish per 45 mins1 fish per 35 mins

Terrain Flat prairieRolling PrairieFoothillsMountains

Distance to site

25 km50 km100 km150 km

Facilities NoneDay-use onlyLimited facilities campsiteFully serviced campsite

Boating Still: NoneSmall craftsPower boats (limited)Unrestricted

Running: None

Swimming YesNo

Water feature

Still: Natural lakeReservoir

Running: RiverStream


• Data collected using a telephone survey conducted by Alberta Environment– Respondents asked to participate in choice

experiment– Respondents asked to provide information to

be used in revealed preference model• 730 separate recreational trips in August 1991• Destinations characterized by the attributes under

study


• Selected attribute coefficient values from choice experiment model

Attribute Standing water Running water

Distance(km)

-0.007(0.0004)

0.007(0.0004)

Catch rate(fish per unit time)

0.062(0.028)

0.105(0.026)

Fish size(large=1 vs. small=-1)

0.058(0.028)

0.090(0.0250

Water quality(good=1 vs. bad=-1)

0.394(0.027)

0.321(0.025)

Swimming(yes=1 vs. no=-1)

0.274(0.026)

0.158(0.025)

Beach(yes=1 vs. no=-1)

0.198(0.026)

0.123(0.024)

Boating N/A N/A


• Empirical results (Choice Experiment)– Multinomial logit discrete choice model

• All parameters except distance estimated as interaction terms associated with standing or running water

– Model fit the observed data– Model has parameters with coefficients consistent with authors’

expectations– Factors with positive influence on utility

• Larger fish• Good water quality• Increased catch rates• Availability of swimming• Presence of beaches

– Respondents preferred upland topography to prairie– Respondents preferred higher diversity to fewer species of fish

• Most popular package: rainbow trout, mountain whitefish, and brown trout– Respondents preferred fully-serviced campsites– Differences noted between estimated coefficient values for parameters

between running water and standing water sites• Increased fish sites and catch rates preferred more in running water


• Selected attribute coefficient values from revealed preference modelAttribute Standing water Running water

Distance(km)

-0.0282(0.001)

-0.0282(0.001)

Catch rate(fish per unit time)

2.0338(0.237)

2.0338(0.237)

Fish size N/A N/A

Water quality(poor=1 vs. good=0)

-0.8197(0.494)

-3.129(0.3749)

Swimming(yes=1 vs. no=0)

2.7477(0.290)

0.9148(0.251)

Beach(yes=1 vs. no=0)

0.9918(0.302)

-1.955(0.369)

Boating(unrestricted=1 vs. none=0

6.6620(1.024)

1.7335(0.289)


• Empirical results (Revealed preference model)– Multinomial discrete choice model estimated for the site

choice decision• Travel cost and site attributes used to explain site choices

– Significant factors in explaining site choice include• Distance to be travelled• Water quality• Catch rates• Availability of swimming, fishing, and boating

• A joint version of the choice experiment and revealed preference was conducted– Data used from both models– Allows to improve quality of the estimates based on

revealed preference• Reduces collinearity

EEnd of nd of SSession ession 10a10a

Thank YouThank You




Session 10bTHE BENEFIT TRANSFER

METHOD



Session 10bTHE BENEFIT TRANSFER METHOD

Benefit Transfer MethodBenefit Transfer MethodOUTLINEOUTLINE

• Overview• Methodology• Application• Illustration• Advantages• Issues and limitations• Case-applications• Case-study

Benefit TransferBenefit TransferOverviewOverview

• Represents another alternative for obtaining non-market values• Involves transferring values that have been estimated for a similar

good or service from another location/context to the current location/context– Estimating benefits for one context by adapting an estimate of benefits

from some other context• Represents a useful method under

– Budget constraints– Time constraints

• Has been applied to value the impact of improved water quality on– Recreation values– Public health

• Has been the normal procedure adopted in regulatory command and control mechanisms in which common standards are applied– EU assumes that benefits of environmental improvement are of equal

value in different areas of the EU• Benefit transfers can only be as accurate as the initial study

Benefit TransferBenefit TransferOverviewOverview

• The simplest type of benefit transfer is the unit day approach– existing values for activity days are used to value the same activity at

other sites– estimates are based on expert judgment in combining and averaging

benefit estimates from a number of existing studies– “unit day values” may be adjusted for characteristics of the study site

when they are applied.

• A more rigorous approach involves transferring a benefit function from another study– The benefit function statistically relates peoples’ willingness to pay to

characteristics of the ecosystem and the people whose values were elicited

– adjustments can be made for differences in these characteristics, thus allowing for more precision in transferring benefit estimates between contexts

Benefit TransferBenefit TransferMethodologyMethodology

• Approaches for applying benefit transfer and assessing the validity of the attempts– The ‘unit day value’ applied by the US Forest Service in the 70s and

80s• Federal guidelines in 1982 recommended

– $6.10-$17.90 per day for specialized recreation» Wilderness use, trout fishing, big-game hunting, white water boating

– $1.50-$4.50 per day for general recreation» Picnicking, swimming, small game hunting, camping, boating

• When applied to a new site, unit day values are adjusted on the basis of the demand functions of site-visitors

• Demand depends on site attributes such as– Congestion– Accessibility and parking conditions– Environmental quality; scenery, pests, water, air, climate– Socio-economic characteristics of recreationalists– Preferences– Price– Availability of substitute sites

• None of these factors will be identical across different sites• Expert judgment is required to assess what the benefits of a new site might be from a

range of possible values• Unit day values can be updated to account for

– Inflation– Observed changes in price and income elasticities for recreation over time


• Approaches for applying benefit transfer and assessing the validity of the attempts– Real-estate agents’ judgments

• To estimate utility loss associated with noise from the proposed airport site

• Some research suggested close correlation between estate agents’ estimates of total house price and estimates derived from an hedonic price model

• Other research revealed discrepancies


• Approaches for applying benefit transfer and assessing the validity of the attempts– Via TCM

• By transferring demand functions from existing facilities, resembling closely the prospective facility in the type of recreation provided

• If the catchment areas of the two sites are mutually exclusive– Existing site coefficients × values for independent variables of the

new site = estimates of number of visits and benefits attributable to the new site

– This approach expected to yield more accurate results than simply applying an average value of benefit per visitor day to the site

• If proposed facility situated within the catchment area of an existing facility

– Apply existing demand function to the new site as if unique– If new consumer surplus exceeds the existing one, the net gain

from having the new facility is the difference between the two


• Approaches for applying benefit transfer and assessing the validity of the attempts– Via TCM (cont’d)

• The lack of homogeneity in product mix may be remedied by valuing the different recreational activities separately and then aggregating, rather than developing a demand curve for the site as a whole

• Errors in BT via TCM– Choosing the wrong functional form– Selecting an incomplete or inappropriate set of arguments– Measuring arguments incorrectly

» Value of time, income, cost of access– Measuring the dependent variable with error– The presence of substitute sites

» Could be cancelled out if sites are randomly distributed via a simulation models


• Approaches for applying benefit transfer and assessing the validity of the attempts– Via CVM

• Application can be affected by– Ex ante- Ex post valuation perspective

» Some estimates elicited after the uncertainty about the good is removed are employed in an ex ante project appraisal

– Scale or quantity value» If the new good or policy is identical to the old and lies within the

same market area, then it represents an additional quantity of the good and should be valued less than the existing good at the site

– Sequential position of the supply of the good» Where goods are complements or substitutes, the sequence in

which a particular good is provided in relation to others determines its value

– Differences in attributes– Compositional effects

» When respondents have difficulty in disentangling the structure of the substitution and complementary interrelationships among attributes within the same holistic set


• Approaches for applying benefit transfer and assessing the validity of the attempts– Meta-analysis

• Uses data-based aids to explain variations in estimated benefits across different studies with the aim of applying past results to future resource policy decisions

• Attempts to assess environmental values by investigating the relationship between

– benefit estimates (WTP) – the features of the goods– the assumptions of the models

• Entails the systematic application of statistical methods to assess common features and variations across a wide range of prior studies

• Undertaken using a variety of techniques encompassing qualitative and quantitative econometric methods

• Relatively underdeveloped in the field of benefit transfer• Important as a means of investigating the factors and issues

involved in the derivation and construction of values


• Meta-analysis (illustration)– Study by Walsh et al. (1989) to explain variations in

net economic benefits per activity day in terms of site, location, and methodological variables

– 287 benefit estimates compared• 156 based on TCM• 129 based on CVMs• 2 based on HPMs

– Some main findings• Omitting travel time in TCM studies reduced benefit

estimates by 34%• ITCM estimates were about 46% greater than ZTCM

estimates using the same functional form• If TCM accepted as the standard for benefit estimation, then

CV estimates needed to be adjusted upwards by 20-30%


• Different standards for benefit transfer may be applied in different contexts– a higher standard of accuracy may be required when the costs of

making a poor decision are higher– a lower standard of accuracy may be acceptable when costs are lower

• when the information from the benefit transfer is only one of a number of sources of information, or when it is used as a screening tool for the early stages of a policy analysis.

• The benefit transfer method is most reliable– when the original site and the study site are very similar in terms of

• quality, location, and population characteristics– when the goods/services in both sites have similar characteristics– when the original valuation study has been carefully conducted and

used sound valuation techniques– when values in original study have not been valuated a long time ago

since preferences change over time


• Three tests have been suggested to determine the accuracy of benefit transfer

– Comparing benefit transfer values with primary data values obtained from the policy site

– Determining whether different populations have the same preferences for the same non-market good, after controlling for differences in socio-economic characteristics

– Determining whether transfers are stable over time

Benefit TransferBenefit TransferApplicationApplication

1. Identify existing studies or values that can be used for the transfer– There are a number of valuation databases that can be

useful2. Evaluate the existing values to determine whether they

are appropriately transferable– Consider whether:

• the service being valued is comparable to the service valued in the existing studies

– Site features– Site qualities– Availability of substitutes.

• the characteristics of the relevant population are comparable– Demographics– Peoples’ preferences

Benefit TransferBenefit TransferApplicationApplication

3. Evaluate the quality of studies to be transferred– The better the quality of the initial study, the more accurate and

useful the transferred value will be– Requires professional judgment of the researcher

4. Adjust the existing values to better reflect the values for the site under consideration, using available and relevant information

– Supplemental data may need to be collected • survey key informants• talk to the investigators of the original studies• get the original data sets• collect some primary data at the study site to use to make

adjustments5. Estimate the total value by multiplying the transferred values

by the number of affected people

Benefit TransferBenefit TransferIllustration*Illustration*

Background• A park being upgraded to provide

additional recreational opportunities– A proposal is to add a swimming beach

to the lake– The benefits of the new beach needs to

be explored – Limited budget for valuation study

• Benefit Transfer Method preferred because – No large budget available for site-

specific benefits studies– Values for recreational uses are

relatively easy to transfer


http://www.inetours.com/England/London/images/Parks/Hyde/Hyde_Park_9466.jpg

Benefit TransferBenefit TransferIllustrationIllustration

Methodology• Step 1

– Identify existing studies or values that can be used for the transfer– Look for studies that value beach use, specifically for lake beaches if

possible• Assume that the researcher has found two travel cost studies that estimated

values for swimming at lake beaches

• Step 2– Decide whether the existing values are transferable by examining

various criteria– The existing values or studies would be evaluated based on several

criteria, including: • Is the service being valued comparable to the service valued in the existing

studies?– similar types of sites (e.g., lake beaches in a park)– similar quality of sites (e.g., water quality and facilities)– similar availability of substitutes (e.g., the number of other lake beaches nearby)

• Are characteristics of the relevant population comparable?– are demographics similar – if not, are data available to make adjustments


Methodology (cont’d)• Step 2 (cont’d)

– In the example, the first study is for a similar lake beach• The beach is also in a park, has comparable water quality and facilities,

and a similar number of substitute sites in the area• It is located in an urban area, while the beach being valued is in a rural

area– The characteristics of visitors can be expected to be different for the two sites

– The second study is in a rural area with similar types of visitors, but the lake has many more available substitutes.

• Step 3: – Evaluate the quality of studies to be transferred

• In this example, the researcher has decided that both studies are acceptable in terms of quality


Methodology (cont’d)• Step 4• Adjust the existing values to better reflect the values for the site

under consideration– In this case, the sites valued in each of the existing studies differ from

the site of interest• The researcher might adjust the values from the first study by applying

demographic data to adjust for the differences in users• If the second study has a benefit function that includes the number of substitute

sites, the function could be adjusted to reflect the different number of substitutes available at the site of interest

• Because the beach will be new, the researcher will need to estimate how many people will use the beach– Survey of park visitors, asking whether they would use a beach on the

lake, and how many times they would use it– Then multiply these visitation estimates by the value per day for beach

use (adjusted for differences in population and site characteristics), to get an estimate of the economic benefits for the new beach

Benefit TransferBenefit TransferAdvantagesAdvantages

• Less costly than conducting an original valuation study

• Economic benefits estimated faster than when undertaking an original valuation study

• Can be used as a screening technique to determine if a more detailed, original valuation study should be conducted

• The method can easily and quickly be applied for making gross estimates of recreational values– The more similar the sites and the recreational

experiences, the fewer biases will result.

Benefit TransferBenefit TransferIssues and LimitationsIssues and Limitations

• Lack of accuracy, except for making gross estimates of recreational values, unless the sites share all of the site, location, and user specific characteristics

• Unavailability of good studies for the policy or issue in question• Difficulty in finding appropriate studies, since many are not

published• Reporting of existing studies may be inadequate to make the

needed adjustments• Difficulty in assessing the adequacy of existing studies• Extrapolation beyond the range of characteristics of the initial study

is not recommended• Benefit transfers can only be as accurate as the initial value

estimate• Unit value estimates can quickly become dated

http://www.duckboats.net/images/poison13.jpg

Benefit TransferBenefit TransferCaseCase--application 1*: application 1*: Wetlands Wetlands

Restoration in Saginaw Bay, Michigan Restoration in Saginaw Bay, Michigan

Background• The State of Michigan is considering plans

to protect and restore coastal wetlands along the southern shore of Saginaw Bay

• The State must estimate the potential benefits from protecting and restoring the wetlands

• A survey asked people about their support for restoring wetlands, but did not include a valuation question

• The researchers used benefit transfer methods to estimate the value of protecting and restoring wetlands around the Bay

www.fws.gov/midwest/alpena/images/sagbay.jpg


Benefit TransferBenefit TransferCaseCase--application 1: application 1: Wetlands Restoration Wetlands Restoration

in Saginaw Bay, Michigan in Saginaw Bay, Michigan Methodology• A valuation study for proposed wetlands protection and restoration

of Ohio’s Lake Erie coastal wetlands was used for the benefit transfer

• Researchers assumed that the values estimated for Ohio were similar enough to be transferable to Michigan– The study valued similar programs and quantities of wetlands to those

proposed in Michigan– However, coastal residents were not surveyed

• the transfer of values from the Ohio study to coastal residents in Michigan requires the assumption that coastal residents have values similar to those of residents of other areas of the drainage basin

Results• Estimates of wetland values for Michigan, based on the Ohio study

– $500 - $9,000 per acre for residents of the drainage basin– $7,200 - $61,000 per acre for residents of the State of Michigan

Benefit TransferBenefit TransferCaseCase--application 2*: Benefits of Water application 2*: Benefits of Water

Pollution Controls on Pulp and Paper MillsPollution Controls on Pulp and Paper Mills

Background• The Clean Water Act provides standards for water quality that

affect the pulp and paper industry

• The industry must implement technological improvements to bring water quality up to standards

• Researchers attempted to assess the benefits of water quality improvements in a particular set of stream segments where pulp and paper mills discharge effluent

• This determines downstream water quality, which in turn affects benefits to recreational users of the streams, as well as non-use benefits from improved water quality.


Benefit TransferBenefit TransferCaseCase--application 2: Benefits of Water application 2: Benefits of Water

Pollution Controls on Pulp and Paper MillsPollution Controls on Pulp and Paper MillsMethodology• Researchers used benefit transfer to estimate the economic benefits of

improved water quality– streams affected by 68 mills were selected for the study– data on existing water quality and pollution control costs for the streams was collected – feasible uses assigned for each stream, based on existing water quality

• The benefits transfer was based on three studies of other rivers that valued changes in water

– the Charles River in Boston– the Monongahela River in western Pennsylvania quality– Two were contingent valuation studies, and one was a travel cost study.

• Both recreational and non-use benefits were considered.

Results • Even using the upper bound estimate of benefits ($66 million), total

benefits for the 68 mills were only two-thirds of the costs to these mills ($95.5 million)

• The total costs to the entire pulp and paper industry were estimated at $310 million

BENEFIT TRANSFER METHODBENEFIT TRANSFER METHOD

Case studyCase studyTransferability of WTP estimatesTransferability of WTP estimates

Valuation of water quality improvements Valuation of water quality improvements in Jaco and Puntarenas along the Pacific in Jaco and Puntarenas along the Pacific

Coast of Costa Rica Coast of Costa Rica


• The study tests the transferability of WTP estimates of improvements in coastal water quality:– Between two urban areas: Jaco and

Puntarenas– Within Puntarenas between 3 city

districts: Centre, Chacarita and Baranca

• Jaco:– Small town in rural area (3000

inhabitants, 840 households)– Dedicated to sun and sea tourism

• Puntarenas:– Second largest airport (65500

inhabitants, 14770 households)– National tourist attraction

Case DescriptionCase DescriptionEcosystem deteriorationEcosystem deterioration

• Puntarenas:– Deteriorating water quality drove visitors to other beaches such as Jaco– In Baranca, a wastewater treatment plant was constructed in 1992

(6000 households) → saturation of available treatment capacity– Rest of City: septic tanks & direct outfalls

• Jaco:– Septic tanks & direct outfalls are used

• General problems:– Surface and groundwater pollution – High water tables– Frequent flooding of septic tanks in rainy season– Deterioration of local sanitary conditions

What is people’s WTP for improvements in coatsal water quality???

Are the estimates transferable between different areas?

Applied MethodApplied MethodContingent ValuationContingent Valuation

• In-person survey: 380 households in Jaco & 1049 in Puntarenas

• Presentation of maps showing– Current quality of sea water, river and estuarine water, and well and

groundwater– Water quality deterioration at the ‘without treatment’ scenario after 5

years– Water quality deterioration at the ‘with treatment’ scenario after 5 years

(new wastewater treatment plant + connection of all households to network)

• WTP question– Respondents were asked whether they would vote ‘for’ or ‘against’ a

wastewater treatment plan– If voting ‘for’, were asked if they would be WTP a monthly sewage fee to

the local water authority– If voting again ‘for’, a double-bounded WTP question was asked

Applied MethodApplied MethodDescription of water quality situation Description of water quality situation

& scenarios& scenarios

Puntarenas sample was divided to 2 subsamples:

1.Full improvement scenario: to highest improvement level2.Partial improvement scenario: river & estuarine water level 2, the rest to highest

levels (inefficiency in wastewater treatment in the proposed plant

Classification level used in showcards

Coastal water resources

Seawater (A-C)

River & estuarine water (1-3)

Well & groundwater (I-III)

Class A/1/I Fit for swimming all year

Fit for human contact all year

Potable well water: no faecal pollution in groundwater

Class B/2/II Fit for swimming dry season

Fit for human contact dry season

Potable well water: contamination risk from faecal pollution in surrounding groundwater

Class C/3/III Not fit for swimming all year

Not fit for human contact all year

Well water not potable: faecal pollution in groundwater

Current classification X: Jaco X: Puntarenas

X

X

X

X X X

Applied MethodApplied MethodBenefit transfer reliabilityBenefit transfer reliability

• Benefit transfer = the application of primary non-market valuation estimates to a secondary setting for which the original study was not expressly designed

• Estimates from the original ‘study site’ are applied to a target ‘policy site’ at a different time and/or place

• Four hypotheses were tested

H1.1 Unadjusted transfer: Benefits transferred are robust to differences in site characteristics

H1.2 Simple adjusted transfer: Values generated at the study site are identical to those at the policy site after adjustment for changes in consumer prices & average differences in income

H2 Benefit function adjusted transfer: The values generated with the coefficients of the WTP regression function estimated at the study site, & the policy site characteristics, are identical to the values that would be obtained from a primary study at the policy site

H3 Slope coefficients of benefit function: Estimated benefit functions at the policy site and study site are drawn from the same population

ResultsResultsWTP responsesWTP responses

The least conservative WTP estimates are 65% (Jaco) & 193% (Puntarenas) higher

than the most conservative estimates

Jaco Puntarenas

Population (hh) sample frame

840 14770

Sample size 380 1049

Sample non-response

83 (21%) 273 (26%)

Water quality scenario

Full improveme-nt

Full improveme-nt

Partial improveme-nt

Freq % Freq % Freq %

= sample response 297 100 398 100 378 100

- protest bids 13 4 6 2 3 1

- Incomplete/ do not know 3 1 16 4 13 3

= valid WTP reponses 281 95 376 94 362 96

- Zero WTP 18 6 25 6 22 6

= item responses (WTP>0)

263 89 351 88 340 90

Valuation approach Estimated WTP

Bid format*

Distribution assumption

Data treatment

Jaco sample Puntarenas pooled sample

Mean Median Mean Median

DC-DB Truncated normal

All zeros included

3085 2598 2347 1966


Only true zeros included

3080 2764 2382 2096


No zeros3089 2963 2404 2268

DC-DB Lognormal No zeros 3168 2557 2467 1918

DC-SB Lognormal No zeros 4789 3247 6617 3093

Sensitivity of WTP to responses treatmentSample sizes and response rates

*DC: dichotomous DB: double bounded SB: single bounded

ResultsResultsBenefit transfer Benefit transfer –– RuralRural--urban testsurban tests

• There are significant differences between the two sites• Full models indicated that socio-demographics mainly

explained WTP in Puntarenas, while sanitation, resource use & environmental attitudes were important in Jaco

Model Transfer error factor at policy site

Jaco Greater Puntarenas Absolute average transfer error %

No covariates -22.5%H1.1 rejected

29.0%H1.1 rejected 25.8

Income adjusted -10.4%H1.2 not rejected

11.7%H1.2 rejected 11.1

Socio-demographic covariates

-20.7%H2 rejectedH3 not rejected

28.1%H2 rejectedH3 not rejected

24.4

Full model covariates

-20.3%H2 rejectedH3 rejected

-1.6%H2 not rejectedH3 rejected

11.0

Note

Transfer error % = 100 (wp/s –ws/s) / Ws/s

Where W is the WTP estimate, p is the policy site, & s is the source site

Transfers are rejected when errors are in the range 11-26%

ResultsResultsBenefit transfer Benefit transfer –– IntraIntra--urban testsurban tests

• Unexpectedly! Absolute transfer errors have not been reduced by geographical proximity relative to the urban-rural transfer

• The simple income-adjusted method outperformed the more sophisticated methods

Model Transfer error factor at policy site

Jaco Greater Puntarenas Absolute average transfer error %

No covariates -22.8%H1.1 rejected

29.5%H1.1 rejected 26.2

Income adjusted -16.3%H1.2 rejected

19.5%H1.2 rejected 17.9

Socio-demographic covariates

-20.9%H2 rejectedH3 not rejected

29.1%H2 rejectedH3 not rejected

25.0

Full model covariates

-22.6%H2 rejectedH3 rejected

28.4%H2 not rejectedH3 rejected

25.5

Note

Transfer error % = 100 (wp/s –ws/s) / Ws/s

Where W is the WTP estimate, p is the policy site, & s is the source site

Transfers are rejected when errors are in the range 11-26%

ResultsResultsBenefit transfer Benefit transfer –– General remarksGeneral remarks

• Populations who have similar socio-demographic characteristics may be different while sanitation & recreation practices, environmental & institutional attitudes are considered

• Having information on socio-economic differences across sites e.g. census data may be a necessary but not sufficient condition for successful benefit transfer

• There are significant & differences in variables explaining WTP between different districts of the same area → very localised phenomena can play large role in determining WTP


Thank YouThank You




Session 11Stated Preference Approach

GROUP EXERCISES


1 of 13

SESSION 11

GROUP EXERCISE 1

Air Quality Improvement Estimation and Assessment using Contingent Valuation Method: A Case Study in Beijing

(Wang et al., 2006)

Case description Beijing has been experiencing a rapid economic development, with a GDP growth rate of more than 9% per year since 1995, and a maximum of 10.2% in 1999. However, as a negative result of the rapid economic growth, Beijing’s environmental quality has deteriorated significantly, especially for air quality in the urban area. Because air pollution may impact many aspects of society, including human health, agriculture yield and industrial production, it is a difficult task to measure the benefit of air quality improvement. According to the Beijing Statistical Bureau, there are around 2,351,000 households in Beijing. The aim of the study is to estimate and assess residents’ willingness to pay to improve air quality in the urban area of Beijing using the Contingent Valuation Method (CVM). 1. Why is the CVM selected in this case? ___________________________________________________________________________

___________________________________________________________________________

2. What alternative method (s) could have been used and why? ___________________________________________________________________________

___________________________________________________________________________

3. Work through the steps of the CVM process to estimate the WTP to improve air quality in

Beijing.

A. Set up the hypothetical market (a convincing scenario, aids, …) ________________________________________________________________________

________________________________________________________________________

________________________________________________________________________

________________________________________________________________________

B. Obtain the bids

a. Select a method for obtaining a bid (income taxes, property taxes, value added or sales tax, utility bills, entry fees, payments into a trust fund)

________________________________________________________________________

________________________________________________________________________

________________________________________________________________________


2 of 13

b. Select the type of questionnaire survey to be adopted (In-person survey, mail

survey, phone survey, etc.): ________________________________________________________________________

c. Types of information obtained

i. What are the various types of information to be obtained via the questionnaire? − _____________________________________________________________

− _____________________________________________________________

− _____________________________________________________________

ii. How can you help the interviewee understand the question, and reduce the tension

during the interview − _____________________________________________________________

iii. Would you use an open or close-ended question to elicit the WTP, and why?

− _____________________________________________________________

− _____________________________________________________________

iv. What wording would you use to elicit the WTP?

− What would your household be willing to pay annually during the next 5 years in order to fulfill the goal of air quality improvement in Beijing (reducing the concentration of air pollutants by 50% in urban areas)?

v. How would you convince the interviewee that his answer will influence the decision-making process − _____________________________________________________________

d. Sample size: Total sample size was 1,500 in 8 sampling districts. The number of households targeted in each of the eight districts was proportional to the household density of that district (the total number of households divided by the total area). Each district was divided into a number of communities according to the number of targeted households in it.


3 of 13

Figure 1. Eight sampling districts in Beijing City

e. What is an important step in questionnaire development and administration?

_____________________________________________________________

f. How long in your opinion should the duration of the questionnaire be?

_____________________________________________________________________

C. Estimate the mean WTP/WTA

Willingness to pay for improving air quality in Beijing

WTP in RMB (in USD) /year # of interviewees Percentage (%)

0 460 33.6

≤ 10 (1.4) 65 4.7

11-50 (1.5-7.0) 161 11.7

51-100 (7.2-14.1) 302 22.0

101-500 (14.2-70.3) 331 24.1

501-1000 (70.4-140.5) 42 3.1

≥ 1001 (140.7) 10 0.7

Total 1371 100


4 of 13

Statistical description of the willingness to pay WTP/year

Sample Mean in RMB

(in USD) SD

(RMB) Median (RMB)

Maximum (RMB)

N=1371 (whole) 143 (20.1) 346 50 7000

N=911 (positive) 215 (30.2) 406 100 7000

D. Aggregate WTP/WTA amount

____________________________________________________________

___________________________________________________________________________

___________________________________________________________________________

_________________________________________________________


5 of 13

SESSION 11 GROUP EXERCISE 2

Applying contingent valuation in China to measure the total economic value of

restoring ecosystem services in Ejina region (Zhongmin et al., 2003)

Case description Ejina lies in the lower reaches of Hei River, one of the two largest inland river basins in China, and is situated south of Monogolia and western Inner Mongolia. The Ejina oasis covers an overall area of 3115.88 km2, which is a detached island oasis encompassed by peripheral desert. With a current population of near 16 thousand, Ejina is one of the world’s most sparsely populated district in the world’s most populated country. The Ejina region also has an extreme and harsh natural environment. The climate of the area is characterized by frequent and severe droughts and large differences in temperature. Mean annual temperature at Ejina is 8.2 °C, with a maximum of 41 °C (July) and a minimum of -36.4 °C (January). Mean annual precipitation is only 36.6 mm. The Hei River’s water resources are the basis of the Ejina environment, economic development and people lives. Water use has grown rapidly over the past 40 years due to economic growth and population increasing in the middle of the Hei River. The flow of the Hei River into the lower reaches in the Zhengyi Xia has decreased by 44.4%, from 11.90 × 108 m3 year-1 in the 1950s to 6.9 × 108 m3 year-1 or so in 1995. The drying up of runoff directly threatens the existence of Ejina ecosystem. About 3.07 × 104 ha of cultivated land in 1960 has now been reduced to only 0.3 × 104 ha and the rest of the cultivated oasis has turned into desert. The area of degraded forest and harsh desert grassland has increased by 35.09 × 104 ha since 1960. The shape of the Ejina oasis has been reduced to three riverine areas: West River, East River and Gurinai. Due to the desert area increasing and oasis area decreasing in Ejina, sandstorms have increased recently in the middle of the Hei River. This deterioration of the Ejina ecosystem has a huge influence on much of northern China. In the spring of 2000, an unprecedented heavy sandstorm event took place in Beijing, Tianjin and their neighboring areas. This storm had adverse effects on the environment, as well as other aspects of people’s daily life and work. The Ejina oasis is the first barrier to sandstorms in the middle of the Hei River valley and north-western China. As a result, the government and the Hei River management bureau decided to adopt conservation measures to restore Ejina’s ecosystem. These measures include restoring the natural vegetation to establish an effective ecological protective shield in Ejina and to reduce the magnitude of this problem. Restoring Ejina ecosystem could allow for controlling soil erosion and reducing sandstorms, provide habitat for wildlife, natural purification of water, dilution of wastewater, and curbing land salinization. It is estimated that this restoration effort will cost approximately 600 million RMB in total over 5 years. The five key ecosystem services that restoring Ejina ecosystem could provide, which are (1) control soil erosion and reducing sandstorms, (2) provide habitat for wildlife, (3) natural purification of water, (4) dilution of wastewater, and (5) curb land salinization.


6 of 13

Fig. 2. Sketch of restoring Ejina ecosystem The aim of the study is to assess whether these costs are worth the benefits to Chinese people living in this area. 1. Why is the CVM selected in this case? ___________________________________________________________________________

___________________________________________________________________________


___________________________________________________________________________

3. Work through the steps of the CVM process to estimate the WTP to improve ecosystem

services in Ejina.


________________________________________________________________________


7 of 13

________________________________________________________________________

________________________________________________________________________

________________________________________________________________________

B. Obtain the bids

a. Select a method for obtaining a bid (income taxes, property taxes, value added or sales tax, utility bills, entry fees, payments into a trust fund)

________________________________________________________________________

________________________________________________________________________

________________________________________________________________________

________________________________________________________________________

b. Select the type of questionnaire survey to be adopted (in-person, mail, phone, etc.):

_____________________________________________________________________

c. Types of information obtained

vi. What are the various types of information to be obtained via the questionnaire? − _____________________________________________________________

− _____________________________________________________________

− _____________________________________________________________

vii. The people living in the upland rural western area of China are still not familiar with

the market prices. Accordingly, how would you elicit their WTP? − _____________________________________________________________

viii. What wording would you use to elicit the WTP?

− If the majority of households vote in favor of restoring Ejina ecosystem, the Ejina’s ecosystem will be restored to the level of the early age of 1980s.

− If a majority vote against, the Ejina ecosystem will remain the conditions and deteriorated as is the current tendency, at last, it has the likelihood to disappear in the world like the historic country ’LouLan’.

− If the project of restoring Ejina ecosystem is at the stage of raising capital, if you vote in favor of it, please draw a circle around the maximum amount your household would vote for and draw a line under the lowest amount your household will switch (i.e. to a no) each year in the following 20 years.

− 0 2 5 10 20 35 50 75 100 200 300 − If current raising capital is a lump-sum payment, would your household be in favor

of cost _____ (yuan) to restoring the Ejina’s ecosystem. (Please fill in the blank).


8 of 13

ix. How would you convince the interviewee that his answer will influence the decision-making process − _____________________________________________________________

_____________________________________________________________

d. Sample size:

Total sample size was 700 households in Hei valley. To save travel time, randomized cluster sampling was adopted. 24 villages and towns were chosen randomly by region as the sampling site. The relative sample amount in each region is determined by population density. Response rates were beyond 99% among main valley and surrounding district.

e. How can you detect and account for protests against the suggested bid vehicle? _____________________________________________________________________

_____________________________________________________________________

A series of follow-up check questions were asked after the WTP question to determine if those refusing to pay represent a valid representation of their value or reflect a protest about some feature of the simulated referendum.

Table 1. Distribution of survey willingness to pay responses (vote for)

Percent of respondents Response

Main Valley % (n)

Surrounding district % (n)

Willing to pay some amount 92.37 (448) 92.09 (198)

‘Restoring ecosystem service is not worth this money to me’ 0.00 (0) 0.00 (0)

‘I can’t afford to pay this amount’ 1.03 (5) 0.93 (2)

‘It is unfair to expect me to pay for increasing ecosystem services’ 2.06 (10) 3.26 (7)

‘Restoring Ejina ecosystem services cannot get expected effect’ 1.65 (8) 0.00 (0)

‘I am opposed to paying for this government program’ 2.27 (11) 2.79 (6)

Other reasons (protest response) 0.62 (3) 0.93 (2)

Total 100.00 (485) 100.00 (215)

Deleted as protest 6.60 6.98

f. What is an important step in questionnaire development and administration? _____________________________________________________________

g. How long in your opinion should the duration of the questionnaire be?

_____________________________________________________________________


9 of 13

C. Estimate the mean WTP/WTA. Time discounting was used by researchers in this case.

Frequency distribution of respondents by bid amount they would vote in favor WTP amount (RMB) 0 2 5 10 20 35 50 75 100 200

Frequency distribution (%) 7.3 8.5 10.4 22.4 17.2 8.2 11.8 2.3 8.5 3.4 Median WTP per household is 19.37 RMB/year (2.72 USD/year) in Hei Valley; ranging between 20.78 RMB (2.92 USD) in Main Valley and 16.41 RMB (2.31 USD) in the Surrounding District. Accordingly, it can be inferred that people living in different areas view differently the services provided by an ecosystem. D. Aggregate WTP/WTA amount by completing the table below (fill in the highlighted

cells)

Total benefits of households in Hei valley Regions Household

annual median WTP

Number of households

Number of households which have WTP

Annual aggregate WTP (millions)

Discount rate (%)

Time scale (year)

Present value aggregate benefits (million RMB)a

Main valley 20.78 223,895 222,187 15 20 28.90

Surrounding

district

16.41 259,328 257,277 15 20 26.43

Total 55.33

Note that the WTP was aggregated on time scale by adopting the mean environment discount rate (15%) based on compounding interest. The aggregate present value of benefits (55.33 millions over a 5-year period) is less than the present value of restoration cost (400 millions), calculated from 600 millions, at the 15% discount rate, over 5 years.

E. Assess the validity of the CV study Three full statistical models including all survey demographic and attitude variables were estimated by maximum likelihood regression. − Respondent’s education & income level were positively correlated with WTP and were significant − Suburban and urban residents have higher willingness to pay than rural/farm residents − On the average, WTP amounts of 20 and 100 represent 1 and 5% of per capita yearly income,

respectively.


10 of 13

GROUP EXERCISE 3

Economic valuation of environmental services sustained by water flows in the

Yaqui River Delta (Ojeda et al., 2007)

Case description The Yaqui River is located in a trans-boundary 72,540 km2 basin, largely situated in the Mexican State of Sonora and a small part in Chihuahua, as well as small portions of Arizona and New Mexico in the United States. The Yaqui River Basin is within one of the driest hydrologic regions in Mexico. The predominant climate is arid and semi-arid throughout the Basin, except in the eastern portion where the high mountains are located. The average annual rainfall in the area is 527 mm. The majority of the precipitation falls in the months of July to September and is dominated by the North American Monsoon. The runoff from precipitation is captured by several reservoirs on the Yaqui River and its tributaries, and is used mainly for irrigation purposes. The Yaqui River Delta occurs where the River meets the Gulf of California, also called the Sea of Cortez. The Delta is the location of two of the more important ecosystems in the lower part of the Yaqui River Basin: the riparian ecosystem, and the coastal wetlands and estuaries. The Yaqui Valley farming region, which is encompassed within the Delta, is the most important agricultural area (more than 250,000 ha) in Sonora State. Agriculture is the largest user of water, representing more than 96% of the total water withdrawal in the Delta. Water demand from cities and towns of more than 800,000 inhabitants in the entire Basin is increasing due to the accelerated migration of rural inhabitants to nearby cities. Other significant economic activities that exert a water demand in the Delta include manufacturing, animal husbandry, aquaculture and fisheries. The environmental concerns associated with water management in the Yaqui River Delta are clearly linked to a decrease in water flows and deterioration of water quality. These concerns can be summarized by five major problems: salinity intrusion, agrochemical pollution, deterioration of wetlands and estuaries, habitat destruction, and loss of biodiversity. After completion of Oviachic Dam in 1952, the majority of the flow in Yaqui River has been used for irrigation. As a result, the Yaqui River has not reached the Gulf of California for several decades. This situation has deteriorated the quality of the environmental services provided by the ecosystems that depend on the water flows in the Yaqui River Delta. Until the Oviachic Dam began its operations, the Delta consisted of lush, riparian forests of mezquite, alamo, willows and coastal scrubs. This vegetation, however, has effectively disappeared over the last few decades. The loss of riparian vegetation, coupled with the loss of wetlands and estuaries because of desiccation and the expansion of aquaculture farms, has reduced the habitat for resident and migratory birds and other animals, including several protected species. The lack of water in the rivers has also greatly reduced the deposition of silt that formerly replenished the wetlands and estuaries with nutrients. The reduction in freshwater flow in the Yaqui River Delta has also reduced the influx of nutrients to the Gulf of California, one of the world's most productive marine ecosystems, and has reduced critical nursery habitat for fisheries that thrive in the upper portion of the Gulf. The lack of flow in the River downstream of the Oviachic Dam has also reduced recharge of the aquifers in the Delta. The reduction of recharge, combined with the groundwater extraction for irrigation could generate the saline intrusion problems that have occurred in several neighboring aquifers. Water rights have been allocated in the Rio Yaqui basin to municipal, industrial and agricultural users, with the majority of the water rights being allocated to agriculture (95%). In Mexico, water rights law has been historically based on the principle that water resources are the property of the state and thus should be a free, constitutional right for every citizen. Recent reforms, however, have been designed to promote private water rights and to allow for water rights to be traded and leased by users.


11 of 13

The objective of this study is to estimate non-market values for water in the Yaqui River Delta, Sonora, Mexico, based on residents' willingness-to-pay for existing or potential environmental services sustained by water flows in the Yaqui River. 1. Why is the CVM selected in this case? ___________________________________________________________________________

___________________________________________________________________________


___________________________________________________________________________

3. Work through the steps of the CVM process to estimate the WTP to improve air quality in

Beijing.


________________________________________________________________________

________________________________________________________________________

________________________________________________________________________

B. Obtain the bids

a. Select a method for obtaining a bid (income taxes, property taxes, value added or sales tax, utility bills, entry fees, payments into a trust fund). Given that all farmers in the Yaqui Valley are organized into irrigation districts which hold water rights for almost 3,000 million m3/year. All water diverted for irrigation purposes is controlled by the irrigation districts.

________________________________________________________________________

________________________________________________________________________

________________________________________________________________________

________________________________________________________________________

b. Select the type of questionnaire survey to be adopted (In-person survey, mail survey,

phone survey, etc.): ________________________________________________________________________


12 of 13

c. Types of information obtained x. What are the various types of information to be obtained via the questionnaire?

− _____________________________________________________________

− _____________________________________________________________

− _____________________________________________________________

xi. Would you use an open or close-ended question to elicit the WTP, and why? − The WTP elicitation format consisted of a single-bound dichotomous choice (DC)

bid followed by an open-ended question eliciting maximum WTP. − The bid amount X was assigned randomly to the respondents and came from a set

of 15 possible values in the range of 10 to 150 pesos per month, in increments of 10 pesos. The maximum bid amount was estimated based on data on the distribution of typical household expenditures, the purchase of agriculture water rights in Sonora, and average water bills in Ciudad Obregon.

xii. What wording would you use to elicit the WTP? − The respondents faced a single DC question of the form “Are you willing to pay X

monthly for the next five years?

xiii. What is the main type of bias that could be associated with this form of WTP elicitation? − _____________________________________________________________

xiv. How would you convince the interviewee that his answer will influence the decision-

making process − _____________________________________________________________

− _____________________________________________________________

d. Sample size: 197 households

g. What is an important step in questionnaire development and administration?

_____________________________________________________________

e. How long in your opinion should the duration of the questionnaire be?

_____________________________________________________________________

C. Estimate the mean WTP/WTA

Bid level in pesos

# of respondents per given bid level

# of respondents agreeing to bid level

# of respondents not agreeing to bid level

10 8 8 0

20 8 7 1


13 of 13

30 10 9 1

40 9 8 1

50 9 5 4

60 10 5 5

70 11 10 1

80 10 6 4

90 7 7 0

100 6 5 1

110 7 5 2

120 8 4 4

130 8 6 2

140 7 4 3

150 7 3 4

Total 125 92 33

− 23 responses were excluded: 3 because of lack of confidence by the interviewer, 2

unrealistically large WTP, 18 ‘protest zeros’ − 125 respondents: 5% non protest zero WTP, 95% non-zero

DC question: ∑=

=N

i ii yXN

MeanWTP1

1 = ___________________________________

Where N is the total number of responses, Xi the bid level, and yi the number of yes responses to that bid level Open ended question mean WTP = 73 pesos/month (6.8 USD/month) D. Assess validity of the CV exercise Multivariate statistical analyses were performed to understand households’ determinants of WTP responses: Significant determinants were:

− initial bid amount (- correlation) − # of years of formal education (+ correlation) − # of children in household younger than 15 yrs (+ correlation) − Household monthly income (+ correlation)

Results from linear and logit models were relatively similar ⇒ robust WTP-determinant relationships



Session 11Stated Preference ApproachStated Preference Approach


CASECASE--STUDIESSTUDIES

1. Coastal Ecosystems in Phang Nga Bay, Thailand

2. Air quality in Beijing

3. Ecosystem services in Ejina, China

4. Environmental Services in the Yaqui River Delta, Mexico

5. Sustainable development in Swedish coastal zone

Case study 1Case study 1Economic Valuation of Coastal Ecosystems Economic Valuation of Coastal Ecosystems

in Phang Nga Bay, Thailandin Phang Nga Bay, Thailand


• Phang Nga Bay, Thailand: large bay in Andaman Sea, covering costs of Phuket, Phang Nga and Krabi provinces

• Habitat for– 60228 ha mangrove area– Coral reefs– Sea-grass beds

• Intensive aquaculture activities: shrimps, cockles, oysters, bivalves

• Distinctive & attractive tourism assets: beaches, islands, parks, etc.


• Reduction of mangrove forests in all Thailand:– From 367000 ha to less than 168676

ha in the period 1961-1993– Converted to other uses:

aquaculture, mining, settlement sites, ports & roads, salt ponds, marine shrimp aquaculture

• Damages to coral reefs:– Natural forces– Increased tourism activities such as

snorkeling• Dangers on fauna: mangroves and

coral reefs provide habitat for fish & migratory birds as well as rare & endangered plant & animal species Major mangroves and reefs

in Thailand

Case DescriptionCase DescriptionEcosystem deterioration & threatsEcosystem deterioration & threats

What is the economic value of changes to the quality of the mangroves & coral reefs

ecosystems in the Bay???

• The Bay was designated as an area for the Southern Seaboard Development Project which– Is intended to capitalize on natural resources to

attract foreign development– Consists of: 2 deep-sea ports, industrial estates,

urban centres & highways• Some people are arguing that net benefits of

planned developments exceeded those of mangrove forests

Applied MethodApplied MethodConjoint analysisConjoint analysis

• Face-to-face interviews with randomly selected 300 thais from 2 areas: Phang Nga Bay & other major provinces

• Surveyors received a training• Four ecosystem attributes were valuated: living coral

cover, income from fishery, flood occurrence, area protected

• Payment method: increase in income tax• The survey (pre-tested with 60 individuals) included 4

parts:– General attitudes towards the environment (ranking exercise)– Background information on current use of mangroves & reefs

(illustrated with maps & graphics)– The WTP question in the form of a choice experiment– Socio-economic characteristics

Applied MethodApplied MethodThe WTP experimentThe WTP experiment

• Intended to estimate the value resource users place on ecosystem quality changes from its present average level (status quo) to good(plan A) & excellent (plan B) for the 4 selected attributes

• Cost values – Were determined based on pre-tests: a payment card

where respondents were asked to tick the amount they were sure they would pay & to cross the amount they were sure they would not pay

– Varied between 200 & 1500 Baht (5 & 37.5 USD)

Applied MethodApplied MethodThe WTP experiment The WTP experiment –– Alternative choicesAlternative choices

Set 1 2 3 4 Cost (Baht)

1 Good Good Good Ave 200

2 Good Good Exc Ave 200

3 Good Good Ave Good 1000

4 Good Good Ave Exc 1000

5 Good Exc Good Exc 1500

6 Good Exc Good Exc 700

7 Good Exc Exc Good 1500

8 Good Exc Exc Ave 700

9 Good Ave Good Good 1000

10 Good Ave Exc Good 1500

11 Good Ave Ave Exc 700

12 Good Ave Ave Ave 200

13 Exc Good Good Good 700

14 Exc Good Good Ave 1500

15 Exc Good Exc Exc 1500

16 Exc Good Exc Exc 1000

17 Exc Good Ave Ave 1500

18 Exc Exc Good Good 200

19 Exc Exc Good Ave 1000

20 Exc Exc Exc Good 700

Set 1 2 3 4 Cost (Baht)

21 Exc Exc Exc Ave 200

22 Exc Exc Ave Good 1000

23 Exc Ave Good Exc 200

24 Exc Ave Exc Exc 200

25 Exc Ave Ave Good 700

26 Exc Ave Ave Ave 1500

27 Ave Good Good Good 1500

28 Ave Good Good Good 200

29 Ave Good Exc Good 200

30 Ave Good Exc Ave 700

31 Ave Good Ave Exc 700

32 Ave Exc Good Exc 1500

33 Ave Exc Exc Ave 1000

34 Ave Exc Ave Good 200

35 Ave Exc Ave Exc 200

36 Ave Exc Ave Ave 1500

37 Ave Ave Good Ave 1000

38 Ave Ave Good Ave 700

39 Ave Ave Exc Good 1500

40 Ave Ave Exc Exc 1000

Attributes Level

1- Increased living coral cover

Average (no change), Good (25%), Excellent (65%)

2- Increased income from fishery

Average (no change), Good (35 %), Excellent ( 60%)

3- Flood occurrence

Average (every year), Good (every 2 years), Excellent (every 4 years)

4- Increased area protected

Average (no change), Good (20 %), Excellent (50 %)

Increased income tax in 2002 (Baht)

0, 200, 700, 1000, 1500

Respondents were asked to chose among the following alternative choices

ResultsResultsAttitudes on use of the BayAttitudes on use of the Bay

Problems Rank

Protecting natural habitats and wildlife 35% (1)

Reducing water pollution 23% (2)

Improving quality of education 16% (3)

Increasing agricultural productivity 12% (4)

Inflation 9% (5)

Other social and environmental problems 6% (6)

Problems Rank

Degraded mangroves and coral reefs 44% (1)

Deforestation 22% (2)

Floods 11% (3)

Other environmental problems 11% (3)

Water pollution 10% (5)

Air pollution 2% (6)

4 8 %

2 6 %

2 0 %3 % 3 % 1

2

3

4

5

5 0 %3 0 %

1 4 % 4 % 2 % 1

2

3

4

5

Ranking of social & environmental problems

Ranking of environmental problems

We have a duty to protect the environment

from development regardless of the cost

We should minimize environmental damage

for the benefit of our grandchildren

1= Strongly agree 2=Agree 3=No opinion 4=Disagree

5=Strongly disagree

ResultsResultsAttitudes on use of the BayAttitudes on use of the Bay1 4 %

1 5 %

1 4 %1 3 %

4 4 %

1

2

3

4

5

Thailand needs to develop her forests, sea, and land to increase jobs and incomes, regardless of

the environmental damage

7 0 %

1 8 %

8 % 2 % 1

2

3

4

5

2 %

2 4 %

3 2 %

3 0 %

7 % 7 % 1

2

3

4

5

Mangroves and coral reefs should be protected because rare birds and marine lives depend on

them

I should pay for the protection of parks and nature reserves even if I do not visit them

2 7 %

3 6 %

2 7 %

6 % 4 % 1

2

3

4

5

Even if I do not use the mangroves and coral reefs now, I am prepared to pay now to protect them in case I want

to use them in the future

3 9 %

3 4 %

1 9 %5 % 3 % 1

2

3

4

5

It is worth spending money to protect mangroves because they help to protect

agricultural productivity in the area

9 %1 4 %

1 8 %2 0 %

3 9 %

1

2

3

4

5

We have more important things to think about than the loss of the mangroves and coral reefs

1= Strongly agree 2=Agree 3=No opinion 4=Disagree 5=Strongly disagree

ResultsResultsWTP estimationWTP estimation

• Diversity of flora & fauna is the most important attribute

• Aggregate WTP is 5784 million Baht (144.6 million USD)per year (computed by multiplying by number of beneficiary people

Attributes (Baht/person/year)

Average Good Excellent

Flora and fauna -699 265 434

Local livelihood -257 257 -

Ecological function -252 - 252

Rare and endangered species 46 -204 158

Attributes WTP Baht (USD)/person/yr

Percent(%)

Flora and fauna(from average to excellent)

434 – (-699)= 1,133 (28)

50

Local livelihood(from average to good)

257 – (-257)= 514 (13)

22

Ecological function(from average to excellent)

252 – (-252)= 504 (13)

22

Rare and endangered species(from average to excellent)

158 – 46 = 112 (3)

6

Total 2,263 (57) 100

End of Case Study Case study 2Case study 2Economic valuation of air quality Economic valuation of air quality

improvement in Beijing, Chinaimprovement in Beijing, China


• 8 districts of Beijing, China:– 4 urban (Dongcheng, Xicheng, Chongwen and

Xuanwu)– 4 suburban (Chaoyang, Fengtai, Shijingshan and

Haidian)


• Beijing has been experiencing a rapid economic development:– GDP growth rate of more than 9% per year since 1995– Maximum of 10.2% in 1999

• Significant deterioration of environmental quality, especially air quality

• Impacts on:1. Human health 2. Agricultural yield 3. Industrial production

What is Beijing’s residents WTP to improve air quality in its urban areas???


• In-person 15-20 min interviews on WTP to improve air quality, with a sample of 1500 households distributed among districts proportionally to their household density

• Steps taken to reduce common biases in the contingent valuation method:

– Part-whole bias: interviewee was asked about the payment his household would do annually, to eliminate biases related to individual/household payments and time horizon

– To get a conservative estimate: open-ended question about WTP estimate was used (generally leads to lower mean WTP than other types of questions)

• Interviewees were told that their WTP answers would contribute to air quality improvement where they live and work and influence the decision-making process

• A follow-up question was used at the end of the interview regarding degree of understanding of the survey: 70.3% fully understood, 15.9% partially understood, 13.8% didn’t understand

Applied MethodApplied MethodQuestionnaire designQuestionnaire design

• Three parts (revised through pre-tests):1. The general attitudes of interviewees toward the

environmental quality in Beijing (willing or unwilling to pay, and why?)

2. The WTP of residents to improve air quality in Beijing: designed question‘What would your household be willing to pay annually during the next 5 years in order to fulfil the goal of air quality improvement in Beijing (50% reduction in air pollution levels)?’Note: before responding, the interviewee was shown some pictures of ‘deteriorated air’ and some of ‘improved air’

3. The social and economic features of interviewees: age, gender, employment, income, household population, etc.

ResultsResultsWTP responses (1)WTP responses (1)

• 1371 out of 1500 questionnaires were recovered

Proposed area for expenditure Ratio (%)

Scientific research on air quality improvement

28.9

Air quality improvement project (e.g. improve energy consumption)

53.2

Compensation to the unemployed workers due to the closure of enterprises which cause air pollution

16.8

Pay for the environmental management expenditure (e.g. air quality monitoring system development)

15.9

Improvement of gas exhaust equipment of vehicles

18.3

Assisting the air pollution enterprises renovation/relocation

24.0

WTP in RMB (in USD) /year

# of interviewees

Percentage (%)

0 460 33.6

≤ 10 (1.4) 65 4.7

11-50 (1.5-7.0) 161 11.7

51-100 (7.2-14.1) 302 22.0

101-500 (14.2-70.3) 331 24.1

501-1000 (70.4-140.5) 42 3.1

≥ 1001 (140.7) 10 0.7

Total 1371 100

WTP/yearSample

Mean in RMB (in USD)

SD (RMB)

Median (RMB)

Maximum (RMB)

N=1371 (whole) 143 (20.1) 346 50 7000

N=911 (positive) 215 (30.2) 406 100 7000

Includes true zero and protest zero WTP ⇒ conservative estimationLess than 1% of household income


• Aggregate annual WTP (by multiplying the mean WTP per household per year by total # of households in survey area):– Conservative: 3.36 billion RMB (0.5 billion USD) per

year– Upper limit: 4.98 billion RMB (0.7 billion USD) per

year

• Beijing residents prefer better air quality and have a clear idea about the trade-off between economic growth and environmental protection

ResultsResultsStatistical analysisStatistical analysis

• 4 variables were found to have significant influence on WTP:– Household income

(+ correlation)– Education

(+ correlation)– Household population

(- correlation)– Age

(- correlation)

End of Case Study

Case study 3Case study 3Restoration of ecosystem services Restoration of ecosystem services

in Ejina region, Chinain Ejina region, China


• Ejina region, China • 3116 km2 oasis, surrounded by

desert in Hei River Basin• Sparsely populated (16000)• Temperature extremes: average

8.3°C (range: -36.4 to 41°C)• Hei River’s water resources

are the basis of the environment, economic development and people lives

• Ejina oasis is the first barrier to sandstorms originating in the middle of Hei River Valley andn orth Western China


• Decrease in water flow over past 40 years as a result of:

– Low mean annual precipitation (36.6 mm)

– Increase in water use as a result of economic growth and increase in population

– Stop in water flow from May to July as a result of agriculture production peak

– Drying up of runoff after November

Cultivated land has been reduced from 3.07x104 ha in 1960 to 0.3x104 ha in 2000, the rest turned down into desertArea of degraded forest and harsh desert grassland has increased by 35.09x104

ha since 1960Ejina oasis has been reduced to 3 riverine areas: West River, East River and Gurina (figure)

Case DescriptionCase DescriptionImpacts of ecosystem Impacts of ecosystem

deteriorationdeterioration• Increase of desert area and decrease of oasis area in Ejian

resulted in the increase of sandstorms in the middle of the Hei River, which reached sometimes Northern China (Beijing, Tianjin and neighboring areas). These storms:– Generated thick dust and created traffic problems – Resulted in economic losses: reduced sunlight to cultivated land

caused a decrease in production– Reduced the visibility: increased traffic hazard– Adversely affected mental health of the population

• Cost of restoration of Ejina’s ecosystem to prevent sandstorms was estimated at 600 million RMB (Renminbi, ~85 million USD) over 5 years

• Is this cost worth the benefits to people living in the area???


• In-person 30-min interviews on WTP estimation for ecosystem restoration, with a sample of randomly chosen 700 households in Hei Valley

• Steps taken to reduce common biases in the contingent valuation method:– Providing a 2 RMB (~0.3 USD) currency bill as appreciation – Printing questionnaire on good quality paper bound in booklet form– Providing background information on Ejina ecosystem– Deciding the bid range (range of WTP amounts) by pretest and adopting

the ‘payment card’ (PC) method– Keeping responses anonymous

• Respondents were told that feedback from survey will be used as input by the government in its study of restoring Ejina ecosystem

Applied MethodApplied MethodQuestionnaire designQuestionnaire design

• Four major parts:1. Background on Ejina region: maps showing location and condition of

the ecosystem; history of the area2. Card listing the 5 services that restoring Ejina ecosystem could provide:

• Control soil erosion & reducing sandstorms• Provide habitat for wildlife • Natural purification of water• Dilution of wastewater• Curb land salinization

3. Valuation portion: • Portrayal of ecosystem to be valued• Choice of payment vehicle: donation, ecological protection tax, water bill, a

fourth option left as blank to fill by a method preferred by respondent• Payment Card: used to elicit respondent’s RMB amount of WTP

4. Series of questions on respondent’s personal & socio-economic information

Applied MethodApplied MethodExact wording of Payment CardExact wording of Payment Card

If the majority of households vote in favor of restoring Ejina ecosystem, the Ejina’s ecosystem will be restored to the level of the early age of 1980s.If a majority vote against, the Ejina ecosystem will remain the conditions

and deteriorated as current tendency, at last, it has the likelihood to disappear in the world like the historic country ‘LouLan’.

If the project of restoring Ejina ecosystem is at the stage of raising capital, if you vote in favor of it, please draw a circle around the maximum amount your household would vote for and draw a line under the lowest amount

your household will switch (i.e. to a no) each year in the following 20 years.

0 2 5 10 20 35 50 75 100 200 300

If current raising capital is a lump-sum payment, would your household be in favor of cost …………. (RMB) to restoring the Ejina’s ecosystem. (Please

fill in the blank)

1% of per capita yearly income


Response Percent of respondents (%)

Main Valley Surrounding district

Willing to pay some amount 92.37 (448) 92.09 (198)

‘Restoring ecosystem service is not worth this money to me’ 0.00 (0) 0.00 (0)

‘I can’t afford to pay this amount’ 1.03 (5) 0.93 (2)

‘It is unfair to expect me to pay for increasing ecosystem services’ 2.06 (10) 3.26 (7)

‘Restoring Ejina ecosystem services cannot get expected effect’ 1.65 (8) 0.00 (0)

‘I am opposed to paying for this government program’ 2.27 (11) 2.79 (6)

Other reasons (protest response) 0.62 (3) 0.93 (2)

Total 100.00 (485) 100.00 (215)

Deleted as protest 6.60 6.98

WTP amount (RMB) 0 2 5 10 20 35 50 75 100 200

Frequency distribution (%) 7.3 8.5 10.4 22.4 17.2 8.2 11.8 2.3 8.5 3.4

5% of per capita yearly income


• Median WTP per household is 19.37 RMB/year (2.72 USD/year) in Hei Valley; ranging between 20.78 RMB(2.92 USD) in Main Valley and 16.41 RMB (2.31 USD) in the Surrounding District ⇒ People living in # areas view differently the services provided by an ecosystem

• Respondent’s education & income level were positively correlated with WTP and were significant

• Suburban and urban residents have higher willingness to pay than rural/farm residents

ResultsResultsTime Discount for WTPTime Discount for WTP

• Economic theory requires that the utility of a lump sum be equal to that of a series of annual payments as a result of a discount rate:A(PVIFAei,n) = FVn1(PVIFr,n1)Where A is the annual payment (32.18 RMB as per survey results); PVIFAei,n is the present value interest factor for ei and n (n=20); ei is the interest rate of environment goods (?); FVn1 is the future value lump sum investment at the beginning or end of n1 years; PVIFr,n1 is the present value interest factor for r and n, r is the risk-free interest rate (2.25%) and n1 is the time lump sum investment provided (20 years).

• If lump sum is provided at the end of the 20th year:32.18x(PVIFAei,20) = FV20(PVIF2.25%,20)⇒ ei = 19.8%

• If lump sum is provided at the beginning of 1st year ⇒ ei = 11.5%• ei = 11.5-19.8% ⇒ High discount rate for environmental goods!!

– Encourages public to underestimate the importance of future benefit– Demonstrates that humans should take action on environmental restoration & protection

ResultsResultsExpansion from sample to populationExpansion from sample to population

• The following steps were taken:– Providing a conservative estimate of WTP: non-respondents and protests

have a zero WTP– Using the median annual WTP per household of Main Valley and

surrounding district– Multiplying the median by the number of households in the respective

regions– Aggregating WTP on time scale by adopting the mean environment

discount rate (15%)• Aggregate present value of benefits: 55.33 millions over a 5-year

period • Less than present value of restoration cost: 400 millions

(calculated from 600 millions, at the 15% discount rate, over 5 years)

• Implications: – Limitations of using WTP approach in less developed countries– Need to determine if there are additional benefits in other regions

End of Case Study

Case study 4Case study 4Economic valuation of environmental Economic valuation of environmental

services of water flows in the services of water flows in the Yaqui River Delta, MexicoYaqui River Delta, Mexico


• Yaqui River basin, largely in Sonora, Mexico, small part in Chihuahua, Mexico, and small portions in Arizona and New Mexico in US

• 72540 km2 basin, in driest hydrologic regions of Mexico

• 800000 inhabitants• The delta occurs where River meets

the gulf of California: location of 2 important ecosystems, the riparian ecosystem, and the coastal wetlands and estuaries

• Home of most important agricultural area in Sonora State: > 250000 ha of wheat, soybeans, cotton, maize, sorghum, and alfalfa; withdraws 96% of delta water


• Decrease in water flow and deterioration of water quality:– Low annual precipitation (527

mm)– Increase in water demand in

urban centres of the basin: migration of rural inhabitants to nearby cities

– Economic activities exerting water demand: manufacturing, animal husbandry, aquaculture and fisheries

Disappearance of riparian vegetation, loss of wetlands and estuaries – habitat for resident and migratory birds and several protected species –Reduction of the influx of nutrients to the Gulf of CaliforniaReduction of critical nursery habitat for fisheriesReduction of the recharge of the aquifers in the Delta, coupled to groundwater extraction for agriculture ⇒ saline intrusion problems

What is the WTP for restoring in-stream flows in the Yaqui River delta???


• Face-to-face interviews on WTP estimation for water flows restoration, with a sample of 197 households

• Steps taken to reduce common biases in the contingent valuation method:– Sample selection bias: next available house was approached

immediately after end of each interview– Sampling frame bias: a full range of survey sites were used– Starting point bias: was assessed by determining the dependence of the

WTP on bid starting point (yes-saying bias)– Potential sponsor bias: respondents were told the research was

sponsored by a university – a neutral body –– Interviewer and misspecification biases: interviewers were trained on

CVM surveys

Applied MethodApplied MethodSurvey designSurvey design

• Major parts (revised through pre-tests):– Presentation of background on the Yaqui River Delta: geographic

information, agricultural activity, environmentally sensitive issues (+ visual aids)

– Presentation of information on environmental services sustained by Delta if in-stream flows are restored:

• Preservation of habitat for birds and other fauna• Maintenance of local fisheries• Dilution of pollutants• Recreation• Non-use values: existence & cultural values, use value for future

generations– Scenario description: explaining how ecosystem functions under

current conditions versus with increased ecosystem services– Payment vehicle description: water bill increase to purchase water

from farmers– WTP elicitation– Collection of demographic information (gender, age, income, etc)

Applied MethodApplied MethodWTP elicitation formatWTP elicitation format

1. Single dichotomous choice (DC) format question (yes/no):‘Are you willing to pay X monthly for the next five years?’X randomly lying between 10 and 150 pesos (~1 and 14 USD) in increments of 10 pesos

2. Open-ended question eliciting maximum WTP(N.B: a main rule was the avoidance of repeated questioning and iteration)

ResultsResultsDC question responsesDC question responses

• 148 out of 197 households (75%) responded:

– 23 responses were excluded: 3 because of lack of confidence by the interviewer, 2 unrealistically large WTP, 18 ‘protest zeros’

– 125 respondents: 5% non protest zero WTP, 95% non-zero

•

= 52 pesos/month (4.9 USD/month)Where N is the total number of responses, Xi the bid level, and yi the number of yes responses to that bid level

Bid level in pesos

# of respondents per given bid level

# of respondents agreeing to bid level

# of respondents not agreeing to bid level

10 8 8 0

20 8 7 1

30 10 9 1

40 9 8 1

50 9 5 4

60 10 5 5

70 11 10 1

80 10 6 4

90 7 7 0

100 6 5 1

110 7 5 2

120 8 4 4

130 8 6 2

140 7 4 3

150 7 3 4

Total 125 92 33

∑==

N

1iiiyX

N1MeanWTP

ResultsResultsOpenOpen--ended question responsesended question responses

• Mean WTP = 73 pesos/month (6.8 USD/month)

• Higher than that obtained with DC question because:– respondents accepting the initial

bid added a maximum WTP estimate significantly higher than the initial bid

– respondents rejecting the initial bid chose a slightly lower WTP

WTP (pesos/month)

Freq

uenc

y

Distribution of WTP responses

Normal distribution

ResultsResultsStatistical analysisStatistical analysis

• Multivariate statistical analyses were performed to understand households’ determinants of WTP responses:– Significant determinants were:

• initial bid amount (- correlation)• # of years of formal education (+ correlation)• # of children in household younger than 15 yrs (+ correlation)• Household monthly income (+ correlation)

– Results from linear and logit models were relatively similar ⇒ robust WTP-determinant relationships

End of Case Study

Workshop on Workshop on the Cost of Environmental Degradationthe Cost of Environmental Degradation

Case study 5Case study 5Economic valuation for sustainable Economic valuation for sustainable

development in the Swedish Coastal Zonedevelopment in the Swedish Coastal Zone


• The Swedish coastal zone• Open-access conditions &

public nature of provided services ⇒ conflicting interests– Between people of remote coastal

areas interested in coastal ecosystem goods & services, & the urban population interested in high quality recreation facilities

– Between those who demand coastal ecosystem goods & services (e.g. swimmers) and those influencing their supply (e.g. emitters of nutrients)


What is people’s WTP for environmental improvements or for avoiding environmental damage???

• Emission of nutrients– Agriculture and forestry in catchment

areas– Municipal wastewater treatment plants– Atmospheric deposition (e.g. from

traffic)

• Open access to fishing grounds– Increased harvest– High fishing pressure in commercial

catches in the North Sea– By-catches from bottom trawling

• Deterioration of marine water quality

Decreased fish stocks, e.g. cods & blue skates

Increase in nitrogen load to seawater → decrease in water transparency →Eutrophication

Frequent violation of water quality standard; Medium biological diversity; Low average cod catch per trawling hour

Case Study 1Case Study 1Reduced eutrophication of the Reduced eutrophication of the

Stockholm ArchipelagoStockholm Archipelago• Nutrient emissions in catchment area → Increased N

concentration in coastal water → eutrophication• Do the benefits of reduced eutrophication outweigh the costs

of reduced eutrophication effects?• Cost of mitigation:

– 1-m increase in the average water transparency → 40% reduction in N load (annual reduction of 2725 tonnes of nutrients) → increased sewage water treatment & reduced fertilizer use ⇒ 57 million SEK (~9.5 million USD) per year

• Benefits: – Recreational benefits were estimated by travel cost method– Other benefits were captured by contingent valuation method (5500

random mail questionanires in Stockholm & Uppsala)– ⇒ 60 million SEK per year (travel cost)

+ 500 million SEK per year (CV)= 560 million SEK (~93 million USD) per year

Case Study 2Case Study 2Improved coastal fisheries in SwedenImproved coastal fisheries in Sweden

• Open access to fishing grounds → decreased fish stocks• Do the benefits of increased fish stock outweigh the cost of

underlying support to fish reproduction?• Study is on-going• Benefits:

– Travel cost method: 2500 random questionnaires in Stockholm & Uppsala counties to collect info on sites visited by respondents, the distance travelled, travel time, travel costs, catch rates, etc.

– Average response rate was 55% – Preliminary results:

• Positive relationship between probability that a fishing is chosen & the catch of fish

• Negative relationship between probability that a site is chosen & travel cost• Based on these results, economic benefits of improved fishing in the

archipelago; they will be compared to mitigation costs to assess actual profitability

Case Study 3Case Study 3Improved water quality at the Swedish Improved water quality at the Swedish

WestcoastWestcoast• Deterioration of sea water

quality → decrease in biodiversity and fish stock and violation of water quality standards

• What are people’s preferences for improved water quality?

• A choice experiment framework was used:– Water quality was represented

by 3 attributes– People were asked about their

WTP for a change in the attributes from current level to the highest level

Attribute Description Levels*

Bathing water quality (%)

Frequency of west-coast sites violating the quality standard

12, 10, 5

Biodiversity Biological diversity or ecosystem balance

Low, Medium, High

Cod stock (kg)

Catch per trawling hour with a research vessel

2, 25, 100

Cost (SEK) The total cost for an individual for each alternative

0, 120, 240, 600, 960, 1800

WTP (SEK) for a change from current level to highest level

MWTP*

Attribute

Water Cod High biod

600 1200 600

*marginal willingness to pay

*present level is in bold format

End of Session 11

Thank YouThank You




Session 12Miscellaneous Case-studies



Session 12aPolitical Economy of the Coastal Zone

in Northern Lebanon

OutlineOutline

• Background• Study Objective• Study Scope and Methodology• Main Drivers and Pressures• Municipal Capacity• Economic Activity• Willingness to Preserve & CZ Intention• Conclusion and Discussion

BackgroundBackground

• The Mediterranean coast experienced drastic changes over the last decades:

• based on the increased pressures from drivers such as population and economic growth, globalization and trade, urbanization, industrialization, tourism, fisheries, extraction and agriculture

• with impacts on environmental health, sustainable development (air, water, soil and biota), land-use, encroachment on natural habitat, ecosystems, agricultural areas, watersheds and pristine areas; climate change


Source: Plan Bleu (2006).

Unplanned development disrupted the coastal “striped”slopes and altered the land-use hence putting more pressures on most coastal zones


• due to poor CZ regulatory framework/instruments across most East/South Med countries since the 1960s (Plan Bleu)

• With a regional response led by the EC (Plan Bleu, MAP, PAP/RAC, etc.) in conjunction with UNEP, World Bank-METAP and other Donors to reduce the land-based pollution of the coast, improve the integrated management of the coastal zones, etc. (1995 Barcelona Convention and its Protocols including Land-based and ICZM)

Objective Objective

Gauge the political economy of improving the management of the northern coastal zone by:

In a First Phase (EC-SMAP III and University of Balamand):- Determining drivers and coastal pressures in conjunction with fiscal and human resource municipal capacity, and economic activity- Assessing the willingness to preserve the coast (DD for change)

Objective Objective

In a Second Phase (EC-SMAP III and METAP):- Assessing the legal/institutional framework - Valuing the coastal environmental degradation and remedial actions- Suggesting policy (ways to achieve change) and institutional (instruments to achieve change) reforms

ObjectiveObjectiveThe objective of the First Phase will be

achieved through:

1. Analyzing the human resource and financial capacity of the coastal municipalities

2. Determining the coastline (including the municipalities) Gross Domestic Product (GDP)

3. Calculating the direct & indirect value of the northern coast and dweller ICZM intentions (Survey)

ScopeScope

Each coastal zone definition is meant to answer a specific purpose

The cadastre boundaries of the coastal municipalities were retained for this exercise

MethodologyMethodology

• Coastal municipalities: fiscal analysis based on surveys and budgets

• Coastal GDP: VAT, rapid surveys, etc. used to generate the value added based on National Account Input/Ouput ratios

• Coastal direct/indirect (UN Millennium Ecosystem Assessment) and dweller ICZM intention based on surveys (CVM and Behavior Theory of Collective Action)

Drivers and PressuresDrivers and Pressures

Coastal Population: 413,000 to 567,000Density: 390-1,000 population/km2

Northern Population 1997-2030: +41%GDP net growth: +6% (04) +2% (05-11)Urbanization: 74%; air, solid/liquid wasteIndustrialization: cement and fertilizersTrade: Tripoli port serving the hinterlandTourism: 42 beaches, resorts and hotelsFisheries: unsustainable practicesExtraction: salt marshes: a dying activityAgriculture: land erosion, water qualityWatersheds: municipal effluents; runoffMountain: deforestation; terrace collapse

MunicipalitiesMunicipalitiesThree typologies of municipalities are available:•• Four Cities/Towns (population: 312,000):Four Cities/Towns (population: 312,000):

Minieh, Tripoli, El Mina and Batroun

•• 17 Villages (population: 98,000):17 Villages (population: 98,000):Qlayaat, Qobet Shamra, Bebnin, El Mhamra, Bhanine, Deir Ammar, Bedawi, Ras Maska, Kalamoun, Kelhat, Anfeh, Chekka, El Herri, Hamat, Selaata, Koubba and Kfar Abida

•• Two Federations of Municipalities (population: 338,000)Two Federations of Municipalities (population: 338,000)Minieh Federation (Minieh, Deir Ammar, Bhanine) Fayha’a Federation (Tripoli, El Mina and Beddawi)

Villages w/o municipalities (Mohafaza Jurisdiction)•• Three orphaned Villages (population: 3,000):Three orphaned Villages (population: 3,000):

Arida, Cheikh Zennad and Rmoul

MunicipalitiesMunicipalitiesMohaf. /Casa

City/Village (north to south)

Number of Municipal Council members

Actual Number of Staff

Number of Staff Authorized by Law

Arida Under Mohafaza Jurisdiction Cheikh Zennad Under Mohafaza Jurisdiction Rmoul Under Mohafaza Jurisdiction Qleiat 12 3 6 Qoubbet Chamra 9 0 3 Bebnin 18 2 13

Akkar

Al Mehamra 10 2 5 Minieh Federation 3 NA NA Bhanine 15 2 6 Minieh 21 11 12 Deir Ammar 15 3 7

Minieh- Dennieh

Beddawi 15 30 42 Fayha’a (Tripoli) Federation 3 35 240 Tripoli 24 640 1,000 Tripoli El Mina 21 63 132

Koura Ras Maska 9 4 5 Tripoli Qalamoun 15 8 9

Kelhat 9 2 2 Koura Anfeh 15 5 5 Chekka 15 8 8 El Herry 9 2 7 Hamat 12 2 8 Selaata 8 4 7 Koubba 9 2 3 Batroun 15 15 20

Batroun

Kfar Abida 9 1 3 Total 291 844 1,543 Source: Municipal Budget Statements (2004-06); and study’s compilation.

Aggregate Municipal Budgets Aggregate Municipal Budgets

Municipal Revenues

Direct RevenuesCity: 39%

Village: 25%

Indirect RevenuesCity: 42%

Village: 52%

Taxes e.g. wastewater network

and sidewalkmaintenance tax ($2/capita/year)

Government transfers are decreasing/

backloggedCity: 36%

Village: 43%

Utility transfers are decreasing/ backlogged

City: 6%Village: 9%

Aggregate Municipal Revenues Aggregate Municipal Revenues

Municipal Expenses

Administrative ExpensesCity: 70%

Village: 19%

Maintenanceand Investments

City: 17%Village: 67%

MaintenanceCity: 8%

Village: 45%

InvestmentsCity: 9%

Village: 22%

Extra Budgetary Investments (CDR)

??

Aggregate Municipal Expenses Aggregate Municipal Expenses

Municipal ServicesMunicipal Services

Solid Waste:Solid Waste:Solid waste is being mishandled across the board except in:

1.Tripoli has a sanitary landfill under CDR implementation with possible carbon funding

2.Minieh (sorting and composting) under CDR implementation

3.Beddawi (sorting and composting) under CDR implementation

4.Hamat (sorting but unsanitary landfill)

Municipal ServicesMunicipal ServicesWaste Water Treatment:Waste Water Treatment:

With few exceptions, effluents are unloaded without any treatment into the sea or river beds.

1. A small wastewater treatment plant is operational in Batroun (old city)

2. Waiting to be connected to the sewer network in Chekka

3. Being built in Greater Tripoli4. Being planned in Bebnin and Batroun

Air Pollution:Air Pollution:No serious efforts at curbing air pollution from point and non-point sources, except in Chekka

Municipal Qualitative AssessmentMunicipal Qualitative Assessment•• Promising & GrowingPromising & Growing in Bebnin, Minieh & Ras

Maska - try to improve their urban environmental health with their meager means

•• Well establishedWell established in Tripoli & El Mina – TEDO funded under SMAP II to collect environmental indicators on air pollution to help decision-makers make informed choices

•• OpportunisticOpportunistic in Batroun- first coastal treatment plant funded by SMAP I and World Bank

•• ControversialControversial in Chekka & Selaata an industrial cluster –may greatly impact the environment &other stakeholders of the coastal zone

•• FrustratedFrustrated Anfeh cannot capitalize on its cultural heritage (salt marshes), because of bureaucratic complexity

Municipal Assessment SummaryMunicipal Assessment SummaryAlthough there is a municipal will to preserve the

commons, create healthy environments and ensure livelihoods:

• Decentralization and Land use Strategy are on the back burner; poor governance; financial dependability (unclear policy to borrow, backlog, no powers to change taxes and fees); uncoordinated investments; etc.

• Narrow financial base: Limited revenues/funds (especially small municipalities)

• Limited municipal environmental services• Limited environment-related human resource

capacity except for Al Fayha’a (TEDO)

Coastline Economic ActivityCoastline Economic Activity

The partial GDP of the northern coast was estimated at $292.5 million in 2005:

- Industrial sector 57.5 % - Energy & Water Supply 14.5 %- Market Services mainly tourism 12.1 %- Agriculture (fishing & extraction) 7.5 %- Government mainly municipalities 6.8 %-Transportation & Communications 1.7 %-Construction N.A. except for Tripoli Port-Trade N.A.

Source: MoET (2007); and Author.

Selected Coastline Economic Selected Coastline Economic Activity Activity

ActivityActivity CharacteristicsCharacteristics Output Output Impact(2005)Impact(2005)

Household Household Income(2005)Income(2005)

Employment Employment (2005)(2005)

Beaches/MarinasBeaches/Marinas/Resorts/Resorts

On of the main contributors to the coastal economic activity

US$ 39.9 million/year Not obtained Not obtained

Fishing ActivityFishing Activity4 major fishing ports: Abdeh

(Bebnin), El Mina, Qualamoun, & Batroun

US$ 27.7 million/year US$ 14.4 million/year 3,347 fishermen

Traditional Salt Traditional Salt ExtractionExtraction

7,000 tons/yearbought at about US$ 20/ ton &

sold at US$ 50/ ton

US$ 0.14 million/year US$ 0.12 million/year 60 workers

BoatingBoating

seasonal activity in El Mina mainly between May and Nov.

Trip costs US$ 100/day, 30 boats, passenger capacity 30,

3 crew members

US$ 0.12 million/year US$ 0.09 million/year 90 sailors

Boat Boat ConstructionConstruction

performed on the El Mina quayside and consists of

wooden and fiberglass boats

US$ 0.33 million/year US$ 0.13 million/year 30 craftsmen

Selected Coastline Economic Selected Coastline Economic ActivityActivity

Industry in 2005:Industry in 2005:A.A. Cement industry in Chekka:Cement industry in Chekka: 2 major players

supplying Lebanon, Cyprus, Syria, Iraq, etc. • $100 million in value-added• 4,500 million tons• 918 employees

B.B. Fertilizer Industry in Selaata:Fertilizer Industry in Selaata: The Lebanon Chemicals Company (LCC) operates in a free zone and exports exclusively to Europe.

• $68 million in value-added• 664,000 ton/year of sulfuric acid,

180,000 ton/year of phosphoric acid &85,000 tons/year of phosphatic fertilizers

• N.A. employees

Economic Activity SummaryEconomic Activity Summary• Coastal Area hosts a variety of economic activities:

industries, tourism (boating, marinas), agriculture, resource extraction, etc.

• National GDP main contributor is tourism; coastline GDP main contributor is industries, which give them some leverage (free loader)

• The fishing sector is the single largest employer on the coast second to the tourism sector (unavailable labor statistics)

• Trade-offs between stakeholders are distorted: Industry vs. Tourism vs. Extraction; Community vs. Municipality; Community vs. Industry/Tourism, etc.

DirectDirect--Indirect Resource UseIndirect Resource UseThe survey targeted several issues:

1. Relative importance of the CZ

2. CZ risk perception

3. The entity most suited to managing the CZ

4. Trust among community members for collaboration on a hypothetical ICZM program

5. Willingness to pay to preserve the CZ yearly value on the CZ in general and the marine resource in particular

And performed regressions

DirectDirect--Indirect Resource UseIndirect Resource UseBut first let us define the total economic value of a

resource:

1. Direct-use consumptive (goods and services consumed by users in terms of resource extraction such fish, oil, gas, sand, salt, pearls, etc.)

2. Direct-use non-consumptive (services such as recreational, educational, etc.)

3. Indirect use (services provided by ecological systems)

4. Passive use (option and bequest to endow the resource to future generations)

5. Intrinsic value (all organism valuable regardless of the monetary value placed by society)

DirectDirect--Indirect Resource UseIndirect Resource Use

• Sample: 382 Dwellings: 95% confidence level; ±5% confidence interval

• Male (67%); Female (33%); HHH (73%)

• Education > secondary (60%); < secondary (40%)

• Income > relative poverty (42% -close to IMF 2005); < relative poverty (58% -close to 2002 USJ)

• Village Dweller (26%); City Dweller (74%)


• ICZM knowledge: 23% knew about it

• Relative CZ importance: DCU (74%); DNCU (80%); IU (83%); EH (93%)

• Combined risk perception of the CZ is very high (96%) with minimal variation across determinants

• Liquid/solid waste fee = $7 per capita from the survey whereas it is only $2 per capita from actual municipal budgets


4 ICZM Choices were suggested

Non-Gov Mgt

PS: 37%

NGO: 6%

Other: 3%

Distrust in Community: 39%

DirectDirect--Indirect Resource UseIndirect Resource UseWTP Acceptability rate is 64%; WTP represents 0.5% of

income and 0.4% (truncate WTP at 88%)WTP Mean: $41/year/HH; $12/year/per capitaWTP Median: $12.9/year/HH; $2.4/year/per capitaIncome ε of mean WTP: 0.62 with 0.51 (Non-Gov)


Regression results:• 9 predictors determined ¼ of the WTP results• Risk perception significantly predicted non-Gov WTP• Risk perception suggestively predicted other 3 choices• In a distrust intention, odds are the highest for WTP

choices against non-WTP, i.e., the more distrust, the highest the odds of having a choice with WTP:– Odds of 4.4 selecting choice GovWTP over GovnonWTP– Odds of 2 selecting choice nonGovWTP over GovnonWTP– Odds of 2.4 selecting choice WTP over nonWTP

Direct & Indirect Res. Use Direct & Indirect Res. Use SummarySummary

• Although people are aware of the CZ risks:Median WTP per capita ≈ Waste Fee per capita

• This warrants an awareness campaign to sensitize the population on CZ risks and management responsibilities

• There is a distrust across the board: municipality in themselves to manage the commons (financial/HR constraints); the community in the government to manage the CZ (governance); the community in other members and stakeholders among themselves (vested interest).

ConclusionConclusion• There is a clear demand for change from the

population, however, there is a distrust: in the government to deliver (devolve resp.); and other members to participate in an ICZM program

• There is land-use competition between urbanization, tourism, industry & agriculture along the coast

• The trade-off between economic development/growth, social equity and the protection of the commons is poorly considered

• The government (at the central and municipal level) is constrained financially and does not have a federative approach: CDR is an executing agency

ConclusionConclusion• How do we resolve the policy (ways to achieve

change) and institutional (instruments to achieve change) and market failures that are leading to the CZ problems in this context?

• Is the introducing a tax for CZ conservation appropriate given the circumstances?

• Does building trust in an informal forum and starting discussing trade-offs among various stakeholders a solution?

• What are the appropriate ways and instruments of reform?

Discussion!Discussion!

Paper available on:Paper available on:www.balamand.edu.lb/english/IMAC.asp?ID=www.balamand.edu.lb/english/IMAC.asp?ID=

87638763


Thank YouThank You



Session 12bClimate Change Adaptation in the Water Sector in the Middle East & North Africa

Region: A Review of Main Issues

ContentContent

• What is Climate Change (CC)?

• What is the MENA Water Sector State (WS)?

• How Will CC Affect the MENA WS State?

• What are the Suggested Responses?

What is Climate Change?What is Climate Change?• CC is any long-term significant change in

the average temperature of the Earth's near-surface air and oceans that a given region experiences

• CC is human-made: science established a causal effect between the acceleration of Green House Gas (GHG) emissions and CC effects in the IPCC 4

• GHG (CO2, CH4 and N2O) emissions shot past a safe level of 350 ppm by the end of the 80s and stand at 385 ppm per volume in the Earth’s atmosphere

What is Climate Change?What is Climate Change?• Six scenarios with different assumptions

were developed to simulate GHG projections and their effects on CC until 2100

Source: IPCC 4 (2007).

What is Climate Change?What is Climate Change?• Main GHG emission effects on CC are:

-Average global surface temperature will likely rise between 0.6° to 4° Celsius by 2100


What is Climate Change?What is Climate Change?• Increases in the amount of precipitation are

very likely in high-latitudes, while decreases are likely in most subtropical land and semi-arid regions (by as much as about 20% in the A1B scenario in 2100

• In semi-arid areas, droughts will increase and runoffs will decrease

• The ice cap will shrink and sea level will rise by a likely range between 0.18 and 0.59 meter by 2100

What is the MENA Water Sector What is the MENA Water Sector State?State?

• 3 aspects of the water sector are covered:-Renewable Water (RW) Availability in 2004-Water Use in 2004-Water Services in 2004

• But first, a definition of RW availability:-Water security: ≥ 1,700 m3 pc pa of RW-Water stress: ≥ 1,000 and < 1,700 m3 pc pa of RW-Water scarcity: ≥ 500 and < 1,000 m3 pc pa of RW-Water absolute scarcity: < 500 m3 pc pa of RW


RW Availability. MENA region: • Most water stressed region in the world (1,100 m3)• 3 water groups: arid, hyper-arid and transboundary

RW

PC

PA

RW PC PA

Source: FAO-AQUASTAT (2002) compiled in World Bank (2007a).


RW Availability. MENA region is characterized by• Aridity, desertification and coastal density; and

by



• low precipitation, high evaporation, and increased droughts, flooding and weather extreme

People Affected by Floods, Droughts and Extreme WeatherMENA Region 1988-07 (000')

0

10

1,000

1988

1989

1990

1991

1992

1993

1994

1995

1996

1997

1998

1999

2000

2001

2002

2003

2004

2005

2006

2007

2008

Year

Peop

le A

ffect

ed

(log

sca

le)

FloodDroughtExtreme Weather

Source: FAO-AQUASTAT (2002) compiled in World Bank (2007a); and Author.


Water Use. MENA region is characterized by:• Highest RW withdrawal region (75%); and by

Water Use Share Total Water Withdrawal to Total Renewable



• An important share allocated to the agriculture sector (±85%) with low value-added GDP per km3 (US$ 701) and low yield

• An increased reliance on desalination to augment water supply

Sector Water Use



Water Services are characterized by:• Inadequate governance (accountability, planning,

financing, organizational capacity, etc.) affecting both access (87%) and water-related diseases (22 death per 100,000 from diarrhea mainly in rural)

• Poor utility performance (water losses between 30 and 60% and operating cost coverage ratio less than 1; and

• Low agricultural water requirement ratio that measures the agricultural efficiency and ranges between 0.3 and 0.5.

How Will CC Affect the MENA WS How Will CC Affect the MENA WS State?State?


Impact on 5categories but we willfocus onwater

and health


• The CC effects in the MENA region by 2050 are (figures should be used with care):– Higher temperatures by +2.5 degree C– Lower precipitation by >-10.5%– Lower runoffs between -20 and -30%– Sea level rise by 0.39 meter.– Accelerating drought cycle especially in NAfrica– Burden of disease marginal increase (water-

related, cardio-respiratory and vector-borne diseases, malnutrition and injuries)


• Demographic growth (+2% in 2000s) will put more pressure on RW with an urban population increasing by 93% between 1995-2050

• RW pc pa will decrease by more than half to less than 550 m3 putting the region in water absolute scarcity state

• Water Use: Domestic share will exceed 20% putting additional stress on the agriculture sector

• Water Services: all governance, access, efficiency and water-related disease indicators will deteriorate


2050 MENA RW: M3 Per Capita

-

500

1,000

1,500

2,000

2,500

3,000

Iran

Leba

non

Morocc

o

Tunisi

a

Algeria

Djibouti

Oman

WB & Gaz

aYem

enJo

rdan

Bahrai

nLib

ya

Saudi

Arabia

Qatar

UAEKuw

aitIra

qSyri

aEgy

pt

Arid Hyper-Arid Transboundary

M3 p

er c

apita

2050 RW PC

Without CC With CC

Source: FAO-AQUASTAT (2002); United Nations (2007); and Author

2050 MENA RFW with -20% CC Effects: M3 Per Capita

-

500

1,000

1,500

2,000

2,500

3,000

Iran

Leba

non

Morocc

o

Tunisi

a

Algeria

Djibouti

Oman

WB & Gaz

aYem

enJo

rdan

Bahrai

nLib

ya

Saudi

Arabia

Qatar

UAEKuw

aitIra

qSyri

aEgy

pt


M3 p

er c

apita

2050 RW PC

2004 MENA RFW: M3 Per Capita

-

500

1,000

1,500

2,000

2,500

3,000

Iran

Leba

non

Morocc

o

Tunisi

a

Algeria

Djibouti

Oman

WB & Gaz

aYem

enJo

rdan

Bahrai

nLib

ya

Saudi

Arabia

Qatar

UAEKuw

aitIra

qSyri

aEgy

pt


M3 p

er c

apita

2004 RW PC


Runoff Reduction by 2100


Drought Severity by 2100

What are the Suggested What are the Suggested Responses?Responses?

• The MENA region public and private human, social, capital, natural and cultural assets at stake from future CC effects

• Three responses are suggested:– Knowledge response– Mitigation response– Adaptation response


• Better knowledge response-Transparent awareness campaign (proactive

media and universities) could help ensure an inclusive and participatory CC mitigation/ adaptation planning and implementation process

-Mainstream CC in school and university curriculum-Adapt/set up knowledge-based CC infrastructure

(GIS, meteorological indicators, hydrological cycle, etc.)


• Better adaptation response– MENA region contributes between 3.5 and

5% to the global GHG emissions but the emissions growth has outpaced all the other regions (1995-2004)

– Opportunity to improve energy efficiency (electricity and energy) by tapping carbon funding mechanism and switching to abundant renewable energies (solar and wind in some regions)


• Better mitigation response requires climate-proof sector-wide water reforms by:– Balancing water demand (water allocated to

its highest use value) and supply (e.g., drip irrigation, water reuse, desalination) and build-in system responsive to variations


-Improving governance (e.g., integrated planning, organization, decision-making, management and resource mobilization), equity, justice and preservation of the commons

-Increasing efficiency (agriculture, domestic)

-Enhancing natural disaster and health service preparedness


• Water sector reforms could already help contain, delay and mitigate CC effects (Morocco has embarked on a long term programmatic reform with the World Bank)

• Looking at reducing the distortions of drivers could also help increase the effectiveness of water sector reforms (coherent growth strategy that encompasses population, poverty, urbanization, tourism and agriculture drivers) and increase the climate-proofing efforts


Thank YouThank You

Introduction of the Carbon Funding of Introduction of the Carbon Funding of Waste Emission PresentationWaste Emission Presentation


Global GHG emissions (in Giga tons or billion of tons) are illustrated by compound and sector over the 1970-2004 period.



Session 12cCarbon Finance Instrument to Improve Coastal Zone Solid Waste Management

Waste and ICZMWaste and ICZM

• Situation for solid waste? – Collection of municipal solid waste; – Waste separation/recycling;– Controlled sanitary landfills & composting

Collection of MSWCollection of MSW Waste separation / recyclingWaste separation / recycling

Controlled sanitary landfillsControlled sanitary landfills The real world is differentThe real world is different

Challenges Challenges • Need to design and enact adequate financial

and economic incentives to encourage behavioral changes in human activities in the coastal areas

• Self-standing ICZM capacity building interventions often do not accomplish much

• Support for institutional strengthening, restructuring, and policy reform works best in the context of a holistic, longer-term programmatic operation that links policy interventions with tangible benefits on the ground

• Catalyzing and sustaining ownership at the national, local, and community levels

How can CF solve some of How can CF solve some of these problems?these problems?

• Promoting environmental protection and sustainable economic activities

• By providing revenues for:– Improvements in solid waste management;– Improvements in sewage systems.

• By channeling revenues at the local level (Municipalities)

The Kyoto ProtocolThe Kyoto Protocol• Kyoto commitments

– In 1997, 38 Industrialized Countries committed to reduce GHG emissions by 5% below 1990’levels (entered into force in 2005)

• Kyoto targets are basically achieved by – Domestic reduction of GHG emissions– Trading emission permits (“allowances”) among

companies (EU-ETS) and Assigned Amounts Units (“AAUs”) among governments

– Purchase GHG emission reductions from projects

» In developing countries (Clean Development Mechanism – CDM)

» In economies in transition (Joint Implementation)

Carbon Finance is NOT about Carbon Finance is NOT about ……• CF is NOT about carbon only, but 6 Greenhouse Gases (GHGs):

– Carbon dioxide (CO2), – Methane (CH4) = 21x more potent than CO2, – Nitrous oxide (N2O) = 310x,– Sulphurhexafluoride (SF6) = 23,900x,– Hydrofluorocarbons (HFCs) HFC23 = 1,300x,– Perfluorocarbons (PFCs) CF = 46,500x, C2F = 69,200x– To promote understanding and facilitate calculations, all GHGs are

measured in tons of CO2 equivalents (CO2-e): 1 ton CO2-e = 1 “carbon credit”

• CF is NOT about Finance:– No loan, no grant, no line of credit, but PURCHASE

• CF is NOT about Financing (i.e., promoting) Carbon, but:– Purchase of GHGs REDUCTION, mainly through long-term agreements

(ERPAs)• GHGs can be avoided (e.g., CH4 avoidance in composting projects),• GHGs can be mitigated (e.g., CO2 mitigation in RE/EE projects) or• GHGs can be sequestered (e.g., CO2 sequestration in LULUCF activities)

Industrialized country with an emissions cap

Baseline em

issions

Baseline Scenario

Developing country/economy in

transition with no emissions cap

EmissionReductions (ERs)

Project em

issions

Project Scenario

Emissions target

Developing country/economy in transition benefits from technology

and financial flows

$$

ER

Purchase of ERs

Domestic action

Carbon CreditsCarbon Credits How does Carbon Finance work How does Carbon Finance work for a landfill?for a landfill?

• Baseline Scenario = generation of CH4• Capture of CH4:

– Avoided emissions = emission reductions (ER)• ER will be generated during the lifetime of the

landfill• ER can be sold: additional revenues to

improve IRR and cash flow • Works also with composting, wastewater

treatment, etc.• Incentive to collect and operate the landfill

adequately, otherwise no ER will be generated

BORG EL ARAB & EL HAMMAM BORG EL ARAB & EL HAMMAM LANDFILLS, ALEXANDRIA EGYPTLANDFILLS, ALEXANDRIA EGYPT

© Veolia Propreté/Onyx

LandfillsLandfills

• Before

• After

Description of the ProjectDescription of the Project• 18 districts within the city covering a 7,200km2 area• The services provided under the contract include:

– Street Cleaning Program: daily manual and mechanical sweeping covering over 12,000 km of city streets and roadways

– Household waste collection: collection of 1 million tons of waste per year

– Waste Transfer: 3 transfer stations were put in service to limit the vehicle number transporting waste from the city to treatment centers

– Landfills : 2 modern landfills were constructed– Composting : 3 composting centers are operated

and produce over 120,000 tons of compost pa

Environmental Benefits of Environmental Benefits of controlled landfillscontrolled landfills

• Flaring of the collected LFG does not only destroy methane, but also destroys compounds in the LFG, such as volatile organic compounds and ammonia.

• Prevention of risks associated with landfill gas at uncontrolled landfills:– Risk of explosion– Risk of fire– Unpleasant odor nuisances– Potential atmospheric pollution– Damage to vegetation by asphyxia

Benefits of the ProjectBenefits of the Project• Environmental Benefits:

– Preservation of water resources – Uncontrolled dumps have been replaced by engineered modern landfills with fully lined disposal areas for leachate (wastewater produced by the landfill) containment

– Fight against desertification and depleted soil – The local production of compost provides much needed organic soil amendments

• Social benefits:– Improvement of environmental health– Employee training

• Economical benefits:– Creation of 4,500 job opportunities– Retrocession of a percentage of the value of the

generated credits to the Governorate of Alexandria

CF: What is next ?CF: What is next ?• Project by

project: higher transaction costs, lower predictability for project owners, and non-transformational impact on emissions

• Programmatic:larger scale, better planning environment for project owners, and transformational impact on emissions

Future of Carbon FinanceFuture of Carbon Finance

• New methodologies, including small scale• Post 2012 regime?• Carbon Partnership Facility: sustaining the

market under transitional phase and increase investments by ensuring long-term C-revenues. Open to consider future assets and regimes

EEnd of nd of SSession ession 12c12c

Thank YouThank You




Session 13COST BENEFIT ANALYSIS



Session 13COST BENEFIT ANALYSIS

COST BENEFIT ANALYSISCOST BENEFIT ANALYSISIntroductionIntroduction

• One of the most widely used techniques for project appraisal in the public sector

• Represents a framework for policy decision-making

“Measure a hundred times but cut only once” Proverb

• A technique to evaluate the worth of an idea or project

• A comparison of alternatives• An aid to decision making• A means of looking back to

evaluate choices that have been made

COST BENEFIT ANALYSISCOST BENEFIT ANALYSISIntroductionIntroduction

• Origins of Cost Benefit Analysis: US Flood Control Act of 1936– The Federal government should improve or

participate in the improvement of navigable waters and their tributaries, including watersheds thereof, for flood control purposes if benefits to whomsoever they may accrue are in excess of the estimated costs, and if the lives and social security of people are otherwise adversely affected.

COST BENEFIT ANALYSISCOST BENEFIT ANALYSISTheoryTheory

• Measures of benefit– Demand curve also

referred to as marginal benefit curve

• Indicates the benefit of consuming one extra unit of a good

• Provides an idea of changes in ‘utility’ or level of satisfaction

B0

A

Demand Curve

Quantity

Pric

e pe

r un

it C

D

The price one is willing to pay for a good depends on the satisfaction one derives from consuming it, which is taken

as a MEASURE OF BENEFITS

E


• Measures of benefit– For environmental goods

the benefit or WTP exceeds the market price (if it exists)

– Valuation methods discussed earlier are used to obtain estimates of WTP

B0

A

Demand curve

Quantity

Pric

e pe

r un

it C

D

E

Total benefits = Total Revenue + Consumer Surplus= Area of 0CED + Area of ΔACE


The concept of costs– Opportunity cost (OC) to carrying out the

investment

– Under perfect competition, the OC of a good is the same as the market price of that good

– For environmental goods, there is no market-price

• Alternative methods to be used to measure OC


The concept of Net Social Benefits– It is important to distinguish between a social CBA and a

private CBA• Social CBA

– conducted from a society’s perspective– Referred to as economic analysis

• Private CBA– Carried out from an individual investor’s view point– Referred to as financial analysis

• A project may be financially viable but socially undesirable

– The objectives of a social CBA is to determine whether a project is socially beneficial

• Net Social Benefit (NSB) = WTP – OC > 0

– If NSB > 0, then the state can use the surplus to compensate the losers

COST BENEFIT ANALYSISCOST BENEFIT ANALYSISSteps in Conducting an SCBASteps in Conducting an SCBA

Case Case descriptiondescription

COST BENEFIT ANALYSISCOST BENEFIT ANALYSISSteps in Conducting an SCBASteps in Conducting an SCBA

CEA Vs. CBA

CaseCase--StudyStudyBintuli Wastewater Treatment ProjectBintuli Wastewater Treatment Project

• City of Bintuli in the Republic of Kabastan– Center of commerce and industry– Main industries include

• Metal manufacturing• Coal extraction• Chemical manufacturing• Construction• Paper making• Food processing

– Value of industrial output estimated at 200 million USD in 1990 as compared to 16 million USD for agriculture

CaseCase--StudyStudyBintuli Wastewater Treatment (WWT) ProjectBintuli Wastewater Treatment (WWT) Project

• Quantities of domestic and industrial effluents in water bodies increased– Total industrial effluent = 163,700 m3/day– Total effluent including domestic WW = 271,700 m3/day– 30% of industrial effluent treated– 0% of domestic effluent treated

• River courses in the city turned black and emit unpleasant odors

• Proposal to build a WWT facility with pumping stations and drainage networks– To treat 28% of industrial waste and the remaining domestic

waste– Treated effluent discharged in river to be used by industries and

agriculture

COST BENEFIT ANALYSISCOST BENEFIT ANALYSIS1. Defining objectives and project scope1. Defining objectives and project scope

• Objective often specified by decision-makers in the bureaucracy

• Objective should be clear and unambiguous

• Bintuli WWT Project– Objectives:

• To improve the health of the community• To increase economic activity by improving

wastewater treatment facilities in the city

COST BENEFIT ANALYSISCOST BENEFIT ANALYSIS2. Identifying and screening alternatives2. Identifying and screening alternatives

• List all possible options for reaching objectives• The ‘do nothing’ option should be considered• Preliminary screening of alternatives

• Bintuli Wastewater Treatment Project– Alternatives:

• Maintaining the status quo• Expanding the existing WWT facilities

– Ruled out because it uses outdated technology and would be difficult to maintain

• Building a new WWT facility• Various locations and site options

– Only one potential site considered in this application

COST BENEFIT ANALYSISCOST BENEFIT ANALYSIS3a. Identifying benefits and costs3a. Identifying benefits and costs

• Costs and benefits differ for an SCBA as compared to private investors– Benefit in an SCBA

• An outcome resulting in an increase in an individual’s utility– Cost in an SCBA

• An outcome resulting in a decrease in an individual’s utility

• Important notes– An incremental approach adopted in assessing costs and

benefits• Identify and value costs and benefits of the project• Compare with the situation to prevail without the project• The difference is the net incremental benefit arising from the project• Only additional changes in costs and benefits are considered, and not

total costs and benefits


• Important notes (cont’d)– Sunk costs and benefits incurred before project commencement

must be excluded• Previous costs are not an opportunity cost as they do not represent a

loss of future income from an alternative use of resources

– Transfer payments must be excluded• Taxes, subsidies, loans, and debt services do not result in an increase

in net benefits• Taxes by foreign investors should be included

– Depreciation and interest are excluded from the cost in a SCBA• SCBA involves discounting values of capital items at their opportunity

costs– Including depreciation as a cost will result in double counting

• The discount rate in an SCBA already takes into account the interest– Including interest as a cost will result in double counting


• Costs and benefits are normally classified into– Primary costs and benefits

• Related directly to the project– Secondary costs and benefits

• Arise from events and activities triggered by the project• Should be handled with care as they could exaggerate

estimates– Opportunity cost must be used as a guideline– Resources are sometimes merely transferred from one part of the

economy to another

• Costs and benefits may also be classified into – Market costs and benefits– Non-market costs and benefits


Costs• Primary

– Investment• Construction of a pumping

station, office building, WWT facilities

• Purchase of equipment– Operation and

maintenance• Wages and salaries• Fuel and chemical costs• Other costs (project

management, preparation, training and commissioning

Benefits• Primary

– Economic • User charges

– Reduction in health costs and mortality rates

– Reduction in costs of treating increasingly polluted water supplies

– Increase in labor productivity due to reduction in absence from work due to illness

• Secondary– Benefits to industry and

agriculture from using recycled water

– Additional revenues from re-afforestation

– Increase in reed harvesting for the paper mill industry

Bintuli WWT Project

COST BENEFIT ANALYSISCOST BENEFIT ANALYSIS3b. Valuing benefits and costs3b. Valuing benefits and costs

• Allows comparison between alternatives• Valuation should be done according to the opportunity cost

principle– Prices of inputs that do not reflect their true value to the society

are adjusted- shadow pricing• Comparison of costs and benefits should focus on the with vs

without the project, rather than before vs. after the project

Valuing the costs• Find market prices for the inputs and outputs• All costs may be in present day or constant prices

– Costs incurred over the project lifetime must be valued at prices prevailing at the time of the project appraisal

– Assumes that annual costs increase at the inflation rate


Valuing the costs• Residual values

– For assets with an economic life that exceeds the planning horizon or project life

• Economic life is the estimate of the duration of the operation of an asset before it requires refurbishment

– The residual or salvage value of the asset must be included as a cash inflow at the end of the planning horizon

– Calculated using• Linear method• Diminishing value method


Valuing the costs• The linear method

– Assumes that the asset value declines linearly over time

Residual value at time t is (1-t/n)PWhere t = time; n = economic life; P = initial price

Ex: an asset purchased at $100,000 and has an economic life of 20 years, at the end of the planning period of 15 years, its residual value is (1-15/20)*100,000 = $25,000

• The diminishing value method– Assumes that the asset value declines by a fixed

proportion of the beginning of year value per annum Residual value at time t is (1-1/n)t P


Valuing the costs• Land and pre-existing building and plant

– Property already owned by operating authority must be valued at opportunity costs– Opportunity costs should be current variations based on the most profitable alternative

uses• Staged construction

– When a project is to be constructed in stages• Only the portion of investment and operating costs to satisfy demand in the current planning

horizon must be attributed to the project• Working capital

– Often constitutes 2% of the total capital outlays– Must be considered as cash outflow at the time when capital expenditures are made

and cash inflow at the end of the project• Operating costs

– Include labor, utilities, supplies, repairs and maintenance, equipment hiring and leasing, insurance and administrative overheads

– To be estimated on an annual basis• Implicit costs

– Opportunity costs and social costs• Use of land, buildings, plants, already purchased by the local authority• Time spent on project by agency staff


Bintuli WWT Project

Item Cost (million USD)

Investment costsBuildings and structures 3.42

Equipment and supplies 13.15

Total investment cost 16.57

O&M costsElectricity 0.68

Salaries 0.09

Chemicals 0.06

Maintenance 0.58

Other 0.21

Total O&M costs 1.62


Valuing the benefits• Benefits of the Bintuli WWT project include

– Revenues from user charges– Economic benefits derived from WWT

• Reduced mortality• Productivity gained from reduced morbidity• Water treatment cost savings• Sale of recycled waster• Afforestation benefits• Reed harvesting


Valuing the benefits• User charges

– New charges• Based on the principle of full cost recovery

– Estimated at 6.9 cents/m3

– 54.75 million m3/year of effluent treated– Annual revenue = 3.78 million USD per year

• Existing charges– 11.4 million m3/year already being treated– User charged set at 6.9 cents/m3

– Annual revenue = 0.61 million USD per year

Net incremental sales revenue 3.17 million USD by year 6 when the new plant is at full capacity


Valuing the benefits• Recycled water benefits

– 60% of treated wastewater will be reused for irrigation and industrial purposes

– Opportunity cost estimated at 10 cents/m3

– Economic benefits = 66,000 USD at year 4– Economic benefits = 3.29 million USD per year by year 8

• Afforestation benefits– Pine and hard wood species planted on 142.8 ha– Net return for experimental plots =689 USD– Net benefit = 10,000 USD in year 8– Net benefit = 100,000 USD by year 17

• Reed harvesting– Reed harvesting for the paper mill industry over an area of 95.25 ha– Net returns = 258.4 USD per ha– Net benefit = 20,000 USD starting Year 6


Valuing the benefits• Reduced mortality benefits

• Total income benefits– 10,000 USD in Year 4– 110,000 USD by the end of the project– Note that only half of the morbidities and mortalities were considered since the

project is planned to serve 50% of the Bintuli population



YearRecycled

water AfforestationReed

harvestingReduced mortality

Reduced morbidity

Water treatment

cost savings


1234 0.66 0.01 0.18 0.11 0.965 0.99 0.02 0.38 0.11 1.506 1.64 0.02 0.03 0.57 0.11 2.377 2.63 0.02 0.05 0.76 0.11 3.578 3.29 0.01 0.02 0.06 0.95 0.11 4.449 3.29 0.02 0.02 0.06 1.00 0.11 4.5010 3.29 0.03 0.02 0.06 1.06 0.11 4.5711 3.29 0.04 0.02 0.07 1.11 0.11 4.6412 3.29 0.05 0.02 0.07 1.17 0.11 4.7113 3.29 0.06 0.02 0.07 1.23 0.11 4.7814 3.29 0.07 0.02 0.08 1.30 0.11 4.8715 3.29 0.08 0.02 0.08 1.37 0.11 4.9516 3.29 0.09 0.02 0.09 1.44 0.11 5.0417 3.29 0.10 0.02 0.09 1.51 0.11 5.1218 3.29 0.10 0.02 0.10 1.59 0.11 5.2119 3.29 0.10 0.02 0.10 1.68 0.11 5.3020 3.29 0.10 0.02 0.11 1.76 0.11 5.39

Incremental economic benefits of Bintuli WWT project (million USIncremental economic benefits of Bintuli WWT project (million USD)D)

COST BENEFIT ANALYSISCOST BENEFIT ANALYSIS4. Calculating discounted cash flows and 4. Calculating discounted cash flows and

project performance criteriaproject performance criteria• Involves reducing future streams of benefits

and costs to their present values to enable comparisons to be made between alternatives

• Given a stream of benefits (B0, B1…Bn) and a stream of costs (C0, C1…Cn),

Net Present Value

where r = discount rate

∑= +

−=

+

−+

+

−+

+

−+−=

n

t nrnCnB

nrnCnB

r

CBrCB

CBNPV0 )1()1(

...2)1(22

111

00


project performance criteriaproject performance criteriaChoice of discount rate• Discount rate in SCBA reflects society’s preferences between

present and future consumption– High discount rate

• implies that society has a stronger preference for present consumption over future consumption

– Low discount rate• implies that society has a stronger preference for future consumption

over present consumption– Choice of discount rate controversial– Environmentalists argue against high discount rates– Economists tended to use long-term interest rates on

government bonds as a measure of opportunity cost of capital• Rate of 10 percent in US• Rate of 8 percent in Australia

• Discount rate must be real rate– Interest rate minus inflation rate


project performance criteriaproject performance criteriaPeriod of analysis• Planning period varies with nature of project

– Should be determined by a period within which estimates are made with a certain degree of confidence

– Should correspond to the economic life of the projectChoice of project performance criteria• These include

– Net present value (NPV)– Benefit-cost ratio (BCR)– Internal rate of return (IRR)– Payback period


project performance criteriaproject performance criteriaChoice of project performance criteria• BCR

– the ratio of the present value of project benefits to the present value of the project costs

• Payback period– The number of years required for a project to recover its costs

• Discriminates against projects with high capital expenditures and long-term benefits

• Not recommended as a measure of project worth

∑=

+

∑=

+=

++

++

++

++

++

++

= n

tnrnC

n

tnrnB

nrC

r

Cr

CC

nrB

r

Br

BBBCR

n

n

0)1(

0)1(

)1(...2)1(10

)1(...2)1(10

21

21


project performance criteriaproject performance criteriaChoice of project performance criteria• IRR is the discount rate at which the present value of

project benefits equals the present value of project costs– It represents the maximum interest rate at which a project

could recover the investment and operating cost and still break even

– May not exist or may not be unique– Difficult to calculate– Trial and error method must be used

0)1(

...2)1(22

111

00 =+

−+

+

−+

+

−+−

ninCnB

i

CB

i

CBCB


project performance criteriaproject performance criteriaChoice of project performance criteria• The rule is to accept a project when

– NPV ≥ 0– BCR ≥ 1– IRR > the social OC of capital

• NPV most preferred criterion because it provides an estimate of the size of the Pareto improvement

• If two or more projects have NPVs > 0, then IRR can be used to rank them


project performance criteriaproject performance criteriaBintuli WWT project• Discount rate

– Real rate of 12%

• Planning period– 20 years

• NPV at the 12% discount rate = 12.08 million USD

• IRR = 21% which is above the opportunity cost of capital of 12 %

• The project is economically viable

YearIncremental

economic costsIncremental

sales revenue


Incremental net benefits

1 2.01 -2.012 8.45 -8.453 6.11 -6.114 2.42 1.91 0.96 0.455 1.62 1.91 1.5 1.796 1.62 3.17 2.37 3.927 1.62 3.17 3.57 5.128 1.62 3.17 4.44 5.999 1.62 3.17 4.50 6.05

10 1.62 3.17 4.57 6.1211 1.62 3.17 4.64 6.1912 1.62 3.17 4.71 6.2613 1.62 3.17 4.78 6.3314 1.62 3.17 4.87 6.4215 1.62 3.17 4.95 6.5016 1.62 4.43 5.04 7.8517 1.62 4.43 5.12 7.9318 1.62 4.43 5.21 8.0219 1.62 4.43 5.30 8.1120 -0.02 4.43 5.39 9.84

Incremental economic benefits of Bintuli WWT project (million USIncremental economic benefits of Bintuli WWT project (million USDD

COST BENEFIT ANALYSISCOST BENEFIT ANALYSIS5. 5. Conduct a sensitivity analysis and/or Conduct a sensitivity analysis and/or

risk analysisrisk analysis• Risk

– Potential outcome whose magnitude and probability of occurrence are known or can be determined

• Uncertainty– Situation where the magnitude of the outcome may or may

not be known and the probability of occurrence is unknown• Distinction between the two may not be clear-cut• Common methods for accounting for risk and

uncertainty– Sensitivity analysis– Break-even analysis– Cross-over values– Risk analysis


risk analysisrisk analysisSensitivity analysis• Used to assess the possible impact of uncertainty by posing

‘what if’ questions• Highlights the critical factors affecting the project’s viability• Parameters subjected to sensitivity analysis include

– Discount rate– Length of project planning horizon– Different timing of the project’s operation– Changes in the capital outlays– Changes in the price of non-market goods– Changes in social and environmental benefits and costs

• Carried out by recalculating project performance criteria using a range of values for the uncertain parameter

• Project performance criteria commonly used are NPV and IRR


risk analysisrisk analysisSensitivity analysis• Methodology

– Determine a realistic range of values for the variables that are subject to uncertainty.

• Example– Capital cost ± 30 percent– O&M costs ± 30 percent– Product prices ± 30 percent

– Calculate the effect of possible changes on the project selection criteria, while varying one variable and holding the others constant

– Reconsider the economic viability of the project in light of the performed calculations


risk analysisrisk analysisBreak-even analysis• Break even value is the value of the discount rate at which the NPV

is zero or the value at which the entire costs will be recovered• On the benefit side

– If a variable appears to be higher than the break-even level, that increases confidence in the project’s viability

• On the cost side– An estimate lower than the break-even level means that the project is

likely to be economically viable

Switching (cross-over) values• Is the discount rate at which the ranking of two projects changes• Recommended when considering only one uncertain variable

COST BENEFIT ANALYSISCOST BENEFIT ANALYSIS5. Conduct a sensitivity analysis and/or 5. Conduct a sensitivity analysis and/or

risk analysisrisk analysisBintuli WWT project• Conducting sensitivity analysis

– Critical uncertain variables chosen for analysis• Changes in capital and O&M costs• Changes in the net incremental economic benefits

– Results indicate that• IRR is robust

– 30% decline in economic benefits reduce IRR to 17% assuming no change in capital and O&M costs

– 30% increase in capital costs assuming no change in economic benefits reduces the IRR to 17%

– 30% increase in operating costs reduces IRR to 19%– The estimate is insensitive to large changes in the

projected economic costs and benefits

COST BENEFIT ANALYSISCOST BENEFIT ANALYSIS5. Conduct a sensitivity analysis and/or 5. Conduct a sensitivity analysis and/or

risk analysisrisk analysisBintuli WWT project

Change in net economic benefit

Changes in Capital costs

-30% -15% 0% +15% +30%-30% 23 25 27 29 31-15% 20 22 24 25 270% 17 19 21 23 24

+15% 15 17 19 21 22+30% 14 16 17 19 20

Changes in O&M costs

-30% 19 21 23 25 26-15% 18 20 22 24 250% 17 19 21 23 24

+15% 16 18 20 22 23+30% 15 17 19 21 22


risk analysisrisk analysisRisk analysis• Suitable in the cases where the values of several parameters are

uncertain• Involves the use of the probabilities of occurrence of the key

variables as weights to recompute the project performance criteria• Carried out using special purpose computer packages (@RISK)

– Generates probability distributions for NPV and IRR• A major difficulty is obtaining probability estimates• Common probability distributions include

– Uniform• Requires minimum and maximum estimates

– Triangular• Requires most pessimistic (minimum), most likely (mode), and most optimistic

(maximum)– Beta

COST BENEFIT ANALYSISCOST BENEFIT ANALYSIS6. Recommendations6. Recommendations

• Water pollution in Bintuli is a serious problem– 70% of untreated effluent dumped in rivers– Project implementation urgently required to

• Protect health of the community• Reduce environmental degradation

• Project would yield substantial economic benefits– IRR estimated at 21%– Sensitivity and risk analysis indicate that estimate

insensitive to costs and benefits

• Recommended to implement the project– With the institution of a good monitoring program

COST EFFECTIVENESS COST EFFECTIVENESS ANALYSIS (CEA)ANALYSIS (CEA)

• CEA may be used when– It is impossible to value a project's major benefits in

dollar terms– two projects have similar economic benefits,– For example, if the decision problem is to choose

between building two hospitals, a CEA would be appropriate since the social benefits in either case would be similar .

• Both CBA and CEA are based on the principle of economic efficiency and therefore do not consider equity or distributional issues.


• CEA looks only at financial costs– A CEA takes the objective as given, and then works out the

costs of the alternative ways of achieving that objective

• The decision on whether to use CEA instead of CBA will depend on a number of factors including the following: – The size and complexity of the project; – The extent to which there are quantifiable benefits; and – The extent to which the benefits can be valued in monetary

terms.

• Unlike CBA, CEA does not have absolute criteria by which to judge the economic viability of projects– CEA not recommended when a decision about the level of

output or service to be provided is at issue


• Examples of situations in which CEA could be used:– Given a desirable pollution abatement standard, what will

be the least cost, out of various alternatives, of achieving the standard?

– Can buying up all the property rights in a flood plain and moving people out by constructing dykes save the same number of lives more cheaply?

– Given two parks with similar recreation benefits, which should be developed? Park A requires extensive filling and flood control and Park B involves buying warehouse sites.

– Choosing between alternative ways of constructing a town's water supply system.


The steps involved in a CEA similar to CBA:

COST BENEFIT ANALYSISCOST BENEFIT ANALYSISCASECASE--STUDYSTUDY

The Effect of Food Waste Disposers on The Effect of Food Waste Disposers on Municipal Waste and Wastewater Municipal Waste and Wastewater

ManagementManagement

OUTLINEOUTLINE

Introduction

GBA ExistingConditions

Study Objectives & Methodology

Results

IntroductionIntroduction

Rapid urbanization ++ associated industry & services growth key feature of economic & demographic development in many developing countries

Cities are absorbing 2/3 of total population amount of solid waste generated surpasses the capacity of municipalities to handle it

Limited areas for landfilling ++ social acceptance ++political conflict need to consider other waste minimization alternative, Food Waste DisposersFood Waste Disposers

Food Waste Disposers / Garbage Food Waste Disposers / Garbage Grinders Grinders

MSW stream

Sewage stream

Grinding by mechanical means + tap water

Food Food wastewaste

Other Other wastewaste

Food Waste Disposers (FWD)Food Waste Disposers (FWD)

Literature ReviewLiterature Review-- FWD FWD

Extra water use amount to 4.3 L/c/d ~ 2.2% of total household water use

21-month pilot project in NY:FWD are used 2-3 times/d for a total of 0.6 min.Using industry upper limit of 2 min/d, the most common ½ HP FWD consumes <75 Watt light bulb uses in 10 min

Increased loadings by: 50% for BOD and SS (Sweden)12% for TN and neg. for P (100% MP)

Literature ReviewLiterature Review-- FWDFWD

Optimal usage for 15 yrs did not exhibit operational problems within plumbing system

Theoretical calculations showed a 57% increase in sludge at WWTP, affecting bio-stage & sludge treatment

Galil & Yaacov (2001): increase sludge by 60-62% Vs.Vs. 18.1% for WMR (1994)

AdvantagesAdvantages of FWDof FWD

Dispose of almost all types of biodegradable food waste

Eliminate nuisance from waste handling & storage, smell and risk of exposure to disease-spreading vectors

Advantages of FWD (contAdvantages of FWD (cont’’d)d)

Leave mostly better quality recyclables stored nuisance and risk-free

Reduce volume of garbage to be disposed of in Landfills

Reduce acidic leachate produced in the landfill & greenhouse gases emitted (CH4)

Ground food waste is quickly and cleanly transported through the sewer system to sewage treatment plants

Gases produced at the plants can be collected and used as an energy source

FWDs are among the safest household appliances

Advantages of FWD (contAdvantages of FWD (cont’’d)d) Questions raisedQuestions raised??????

Installation cost of grinders??

Amount of additional water required for the transfer of the particles??

Change of raw sewage quality in terms of Suspended solids & Organic substances??

Influence of additional loads on the treatment plants??

Greater Beirut Area (GBA) Greater Beirut Area (GBA) under studyunder study

Solid Waste Generation

Collection and Transport

Recyclables Organic Bulky Rejects

Warehousestorage

Composting(Coral)

Landfill (1)(Bsalim)

Industries Farmers

Landfill (2)(Naameh)

Emergency Plan for the SWM in Emergency Plan for the SWM in the GBAthe GBA

Implemented since 1997

Transfer/Processing(1)(Amroussieh)

Transfer/Processing(2)(Karantina)

System Performance EvaluationSystem Performance Evaluation

ISMW targets

(tonnes/d)

Actual Achievements LACECO(1999)

(tonnes/d)

LACECO(2002)

(tonnes/d)Total wastes received at transfer/processing plants

1,700 1,922 2,085

Waste handling meansComposting 850 216 300Recycling 160 99 143Landfilling 690 1,607 1,640

Majority of the waste generated is disposed of at the landfill !!!

Costs of the SWM systemCosts of the SWM system

Activity Cost (USD/tonne)

Collection/Transport 59Sorting 18Baling 12Wrapping 9Hauling to Coral composting plant 4Composting 18-40Landfill disposal 25-35

> 119 USD/tonne

Low market demand oncompost & recyclables+Difficulty in locating new landfillsfor municipal waste disposal along

Lebanese coast (High Cost + Limited Suitable Land +High Social & Political Oppositions) ⇓the need for OTHER ALTERNATIVES

…… GBA MSW CharacteristicsGBA MSW Characteristics

15

106 3 2 1

63

Food waste Paper and cardboardPlastic Glass/chinaMetal FabricOther

70% 70% water water

2,000 tonnes/d2,000 tonnes/d

GBA WW CharacteristicsGBA WW CharacteristicsParameter 2005Population projections (×1,000 persons) 1,782Wastewater generation rates (m3/d) 257,609BOD (kg/d) (1) 114,378COD (Kg/d) (2) 162,294SS (Kg/d) (3) 146,837(1) Based on an average of 444 mg/L(2) Based on an average of 630 mg/L(3) Based on an average of 570 mg/L

No wastewater treatment plants exist. Many are planned to comply with the GoL signed protocol for the protection of the Mediterranean.

Objectives of this StudyObjectives of this Study

Examine the feasibility of introducing food disposers as a waste management option in GBA

Assess FWD operational impacts on

Solid waste stream

Wastewater stream quality and treatment options

Assess FWD economic impacts

MethodologyMethodology

OPERATIONAL IMPACTSOPERATIONAL IMPACTS

Solid waste composition & distribution

Domestic water consumption

Wastewater loadings & flow

Methodology (contMethodology (cont’’d)d)

S1S125% marketpenetration



75% 95% 75% 95% 75% 95%

% of food ground% of food ground

Laboratory InvestigationLaboratory Investigation

Collected from 3 households over weekend

Mixed thoroughly

3 batches of blended food

Methodology (contMethodology (cont’’d)d)

ECONOMIC IMPACTS

Costs (conventional + environmental)

Savings (conventional + environmental)

Benefits

Breakeven Points

Economy of Scale

Capital & Operating cost of

FWD

Wastewater secondary

treatment cost of added wastewater volume (min-max)

Sludge treatment cost of added

wastewater volume for most common used technologies

(min-max)

Cost of increased domestic waste consumption

Cost of sludge treatment(min-max)

Cost forgone of management of food wastes diverted from solid waste stream

Cost foregone of leachate remediation

Cost forgone of abating pollutant discharge from management of food wastes

++

++

++

++Cost of electricity

negligible

Foregone earnings from potential energy recovery from food waste insignificant particularly in anaerobic digestion

Conventional Costs

Environmental Costs

Conventional Savings

Environmental Savings

Capital & Operating cost of FWDAnnual cost of 43 US$/unit/yr assuming average cost/unit of

US$ 400 with expected life span of 12 yrs & 5 % opportunity cost/household

Wastewater secondary treatment cost of added wastewater volumeMin Max

0.03 US$/kg of added BOD 0.25 US$/kg of added SS

0.22 US$/kg of added BOD 0.48 US$/kg of added SS

Sludge treatment cost of added wastewater volume for most common used technologies *

Min Max39 US$/dry tonne 292 US$/dry tonne

* Examined technologies: centrifuge thickening & dewatering; belt filter press; composting; recessed-plate filter; aerobic digestion; anaerobic digestion; alkaline stabilization; thermal aerobic pre-treatment & anaerobic digestion; pre-pasteurization & anaerobic digestion; reactor composting; anaerobic digestion & thermal drying; & incineration.

Cost of increased domestic water consumptionDomestic water charging rate of 150 US$/m3-d/yr

CONVENTIONAL COSTSCONVENTIONAL COSTS

ENVIRONMENTAL COSTSENVIRONMENTAL COSTS

Cost of sludge treatmentMin Max

45 US$/dry tonne 336 US$/dry tonne

Equivalent to 15% of conventional cost Equivalent to 15% of conventional cost

MinMin--Max CONVENTIONAL COSTSMax CONVENTIONAL COSTS++ MinMin--Max ENVIRONMENTAL COSTSMax ENVIRONMENTAL COSTS

= Min= Min--Max TOTAL COSTSMax TOTAL COSTS

CONVENTIONAL CONVENTIONAL SAVINGSSAVINGSCost forgone of management of food wastes diverted from solid

waste stream119 $/tonne of municipal waste

ENVIRONMENTAL SAVINGSENVIRONMENTAL SAVINGSCost foregone of leachate remediation

Equivalent to 46.46 $/tonne

Cost foregone of abating pollutant discharge from management of food waste

Equivalent to 7.5% of conventional management cost of solid waste

CONVENTIONAL SAVINGS + ENVIRONMENTAL CONVENTIONAL SAVINGS + ENVIRONMENTAL SAVINGS SAVINGS = TOTAL SAVINGS= TOTAL SAVINGS

865

455

1,215

228 288

577 683

33

6358 56

4752

43

0

250

500

750

1,000

1,250

Food wastegeneration

in 2005

S1a S1b S2a S2b S3a S3b0

25

50

75

100Food groundPercent composition of food waste

% w

eigh

t

tonn

es/d

ay

25% MP 50% MP 75% MP

Food Waste Composition after Installation of FWDs Food Waste Composition after Installation of FWDs (2005 values)(2005 values)

ResultsResults

Scenario %Reduction in solid waste

to be managed

% Increase in domestic

waterconsumption

% Increase in wastewater

flow

% Increase in BOD loading

% Increase

in SS loading

25% MP + 75% food ground 11.8 0.7 1.1 16.9 1.9

25% MP + 95% food ground 14.7 0.8 1.4 21.3 2.4

50% MP + 75% food ground 23.0 1.3 2.3 33.4 3.8

50% MP + 95% food ground 29.1 1.6 2.9 42.0 4.8

75% MP + 75% food ground 34.3 1.9 3.4 49.5 5.6

75% MP + 95%food ground 43.4 2.4 4.4 62.2 7.1

Impacts of introducing food disposers on SWM and Impacts of introducing food disposers on SWM and WWM schemes WWM schemes (2005 values)(2005 values)

Scenario %Reduction in solid waste

to be managed

% Increase in domestic

waterconsumption

% Increase in wastewater

flow


% Increase

in SS loading

25% MP + 75% food ground 11.8 0.7 1.1 16.9 1.9

25% MP + 95% food ground 14.7 0.8 1.4 21.3 2.4

50% MP + 75% food ground 23.0 1.3 2.3 33.4 3.8

50% MP + 95% food ground 29.1 1.6 2.9 42.0 4.8

75% MP + 75% food ground 34.3 1.9 3.4 49.5 5.6

75% MP + 95% food ground 43.4 2.4 4.4 62.2 7.1

Impacts of introducing food disposers on SWM and Impacts of introducing food disposers on SWM and WWM schemes WWM schemes (2005 values)(2005 values)

4.6

0.4

0.6

0.1

4.6

2.4

1

0.15

1.6

6

9.9

0.02

4.2

8.7

0 2 4 6 8 10 12

Cost of food disposer units

Cost of added volume of domestic water

Cost of wastewater treatment

Conv. cost of sludge mangt

Env. cost of sludge mangt

Conv. solid waste savings

Env. solid waste savings

Net Benefts based on Conv.Costs/Savings

Net Benefits based on Conv. & Env.Costs/Savings

Million US$/yr

Costs, Savings & Benefits achieved underCosts, Savings & Benefits achieved underS1aS1a (25% MP+ 75% food ground)(25% MP+ 75% food ground)

CostsCosts

SavingsSavings

BeneBenefitsfits

13.7

1.5

2.3

0.5

37.6

17.4

9.0

3.8

0.60.1

9.6

26.436.8

19.5

0 10 20 30 40

Cost of food disposer units

Cost of added volume of domestic water

Cost of wastewater treatment

Conv. cost of sludge mangt

Env. cost of sludge mangt

Conv. solid waste savings

Env. solid waste savings

Net Benefts based on Conv.Costs/Savings

Net Benefits based on Conv. & Env.Costs/Savings

Million US$/yr

Costs, Savings & Benefits achieved underCosts, Savings & Benefits achieved underS3bS3b (75% MP+ 95% food ground)(75% MP+ 95% food ground)

CostsCosts

SavingsSavings

BeneBenefitsfits

Benefits achieved under Benefits achieved under S1aS1a (25% MP+ 75% food (25% MP+ 75% food ground) & ground) & S3bS3b (75% MP + 95% food ground) (75% MP + 95% food ground)

as as % of existing SWM cost% of existing SWM cost

1.9 3.7

12.29.0

28.8

49.4

5.0

10.4 7.2

23.3

11.5

44.0

31.6

6.8

16.9

37.0

0

10

20

30

40

50

1

%

based on min conventional costbased on max conventional costbased on min conventional and environmental costbased on max conventional and environmental cost

20052005 20202020

S1a S3b S1a S3b

68.757.1

-10

0

10

20

30

0 40 80 120

US$/tonne

Ben

efit

s

S1a S3b

Breakeven points for Breakeven points for S1a S1a (25% MP + 75% food (25% MP + 75% food ground) & ground) & S3bS3b (75% MP + 95% food ground) (75% MP + 95% food ground)

taking into consideration the lower & upper range of taking into consideration the lower & upper range of costs costs (2005 values) (2005 values)

(a) Based on MIN. CONVENTIONAL costs(a) Based on MIN. CONVENTIONAL costs

(MU

S$/y

r)

Breakeven points for Breakeven points for S1aS1a (25% MP + 75% food (25% MP + 75% food ground) &ground) & S3bS3b (75% MP + 95% food ground) (75% MP + 95% food ground)


(b) Based on MIN. CONV. + ENV. costs(b) Based on MIN. CONV. + ENV. costs

13.92.4

-10

0

10

20

30

0 40 80 120

US$/tonne

Ben

efit

s

S1a S3b

(MU

S$/y

r)

50

100.188.5

-10

0

10

20

30

0 40 80 120

US$/tonne

Ben

efit

s

S1a S3b

Breakeven points for Breakeven points for S1aS1a (25% MP + 75% food (25% MP + 75% food ground) & ground) & S3bS3b (75% MP + 95% food ground) (75% MP + 95% food ground)


(c) Based on MAX. CONV. costs(c) Based on MAX. CONV. costs

(MU

S$/y

r)

46.935.3

-10

0

10

20

30

0 40 80 120

US$/tonne

Ben

efit

s

S1a S3b

Breakeven points for Breakeven points for S1aS1a (25% MP + 75% food (25% MP + 75% food ground) &ground) & S3bS3b (75% MP + 95% food ground) (75% MP + 95% food ground)


(d) Based on MAX. CONV. + ENV. costs(d) Based on MAX. CONV. + ENV. costs

(MU

S$/y

r)

Dynamics of Economy of ScaleDynamics of Economy of Scale

To define the cost of managing the remaining solid waste generated ‘X’ (US$/tonne) after integration

of FWD at which the proposed system would breakeven Savings = Costs OR Benefits = 0

worst case scenario [100% MP + 100% food ground]was assessed

ARGUMENTARGUMENTCost of managing the remaining solid waste wouldif total quantity waste to be managed

Total savings achieved from the SWM scheme as a result of integrating FWD

Excluding environmental externalities, FWD remain profitable until cost/tonne of managing remaining solid waste reaches

US$ 223/tonne = 1.8 US$ 223/tonne = 1.8 ×× current charging ratecurrent charging rate

Including environmental externalities, FWD will still be profitable up to a management cost of

US$ 315/tonne = 2.6 US$ 315/tonne = 2.6 ×× current charging ratecurrent charging rate

Dynamics of Economy of ScaleDynamics of Economy of Scale

Certainly, it is not expected that the cost of Certainly, it is not expected that the cost of SWM would reach such levels in study area in SWM would reach such levels in study area in near future, which justifies the adoption of near future, which justifies the adoption of FWDFWD


Thank YouThank You

Waste Manage Res 2005: 23: 20–31Printed in UK – all right reserved

Copyright © ISWA 2005Waste Management & Research

ISSN 0734–242X

20 Waste Management & Research

The effect of food waste disposers on municipal waste and wastewater management

This paper examines the feasibility of introducing food wastedisposers as a waste minimization option within urban wastemanagement schemes, taking the Greater Beirut Area(GBA) as a case study. For this purpose, the operational andeconomic impacts of food disposers on the solid waste andwastewater streams are assessed. The integration of foodwaste disposers can reduce the total solid waste to be man-aged by 12 to 43% under market penetration rangingbetween 25 and 75%, respectively. While the increase indomestic water consumption (for food grinding) and corre-sponding increase in wastewater flow rates are relativelyinsignificant, wastewater loadings increased by 17 to 62%(BOD) and 1.9 to 7.1% (SS). The net economic benefit ofintroducing food disposers into the waste and wastewatermanagement systems constitutes 7.2 to 44.0% of the existingsolid waste management cost under the various scenariosexamined. Concerns about increased sludge generation per-sist and its potential environmental and economic implica-tions may differ with location and therefore area-specificcharacteristics must be taken into consideration when con-templating the adoption of a strategy to integrate food wastedisposers in the waste–wastewater management system.

Natasha Marashlian Mutasem El-FadelDepartment of Civil and Environmental Engineering, AmericanUniversity of Beirut, Lebanon.

Keywords: Food waste disposers, solid waste/wastewater management: wmr 708-1

Corresponding author: M. El-Fadel, American University ofBeirut, Faculty of Engineering and Architecture, Bliss Street, POBox 11-0236, Beirut, Lebanon

Tel: 961 3 228 338; fax: +961 1 744 462; e-mail: [email protected]

DOI: 10.1177/0734242X05050078

Received 12 September 2003; accepted in revised form 14 October 2004

Introduction

Rapid urbanization coupled with the associated growth ofindustry and services constitute a key feature of economicand demographic development in many developing coun-tries. Cities are currently absorbing two-thirds of the totalpopulation increase throughout the developing world(UNCSD 1999). An important environmental concern ofurbanization is the amount of solid waste that is generated ata rate that surpasses the capacity of municipal authorities tomanage it, resulting in potential adverse impacts on theenvironment, human health, and the quality of urban life.With limited land areas around many urban centrers, thesearch for environmentally safe as well as socially and politi-cally acceptable sites for landfills has become a perennial

problem, and for several cities, seemingly unsolvable, thuscreating the need to consider other waste minimization alter-natives at the source. In this context, the use of food wastedisposers enables the separation of a considerable fraction offood-waste ingredients out of the entire municipal solidwaste (MSW) stream by grinding the waste using mechani-cal means with the addition of tap water, and allowing themixture into the sewage system. This paper evaluates therole of food waste disposers within the waste managementsystem of urban areas, taking the Greater Beirut Area (GBA)as a case study. Background information on food waste dis-posers is first presented followed by an examination of theirimpact on the solid waste and wastewater management

Effect of food waste disposers on municipal waste and wastewater management

Waste Management & Research 21

schemes with emphasis on operational and economic feasi-bilities taking area-specific characteristics into considera-tion.

Background

A garbage grinder or a food waste disposer unit is a kitchenappliance that is mounted directly under the kitchen sinkand connected to the sewer pipe. These units are designed togrind biodegradable organics such as meat scraps, vegetables,fruit pits, citrus fruit peelings, coffee grounds and small bones(Nilsson et al. 1990).

Food waste disposers were first introduced back in the1930s in the US where their usage evolved to reach morethan 94% of all cities and they are included as a standarditem in more than 80% of new home construction and arefound in almost half of all US households (Macnair 2000).This use, however, was surrounded by scepticism in certainlarge cities. New York City for instance, had banned wastegrinders for a long time because of concerns that the city’sold sewer infrastructure could not handle the additionalload. It was not until 1995 when the City started to worrymore about where it was going to dispose of its garbage afterthe closure of its major landfill, that it commissioned themost comprehensive pilot project ever conducted to investi-gate the impact of food grinders on the sewer system. Basedon the positive outcome of the study, the City lifted the banand legalized the installation of food waste grinders in resi-dential buildings in 1997 (Dunham 2001). Today, food wastedisposers are sold to households under limited or no restric-tions in approximately 50 countries including England, Ire-land, Italy, Spain, Japan, Canada, Mexico and Australia.Although a ban was in effect in France, it was lifted in 1986after another in-depth investigation by the French authori-ties (Nilsson et al. 1990).

Although food waste disposers allow diversion of organicwaste from the solid waste stream and hence save on theirassociated management costs, their use raises numerous ques-tions regarding the additional energy required to run theseunits, the amount of additional tap water required for thetransfer of particles into sewage, the alteration of sewagequality in relation to the addition of suspended solids andorganic substances, and the additional loads on the sewagesystem and wastewater treatment plants. Various studieshave been conducted to investigate these issues.

The extra water use due to disposers is reported to be neg-ligible, amounting to 4.3 L/capita per day on average andrepresenting 2.2% of the total household water use (Nilssonet al. 1990, Waste Management Research Unit 1994, Ketzen-berger 1995, New York City DEP 1997, Wainberg, et al.2000,The Plumbing Foundation City of NY 2001).

Further, it has been widely reported that the cost of elec-tricity to run food disposers and its associated pollution is rel-atively insignificant. The most comprehensive study in thiscontext is the 21-month pilot project in New York where itwas estimated that food waste disposers are used two to threetimes a day for a total of 0.6 min. If, to be conservative, usingthe industry upper limit of 2 min/day, the most common0.5 hp motor of a food waste disposer consumes less than a75 W light bulb uses in 10 min (The Plumbing FoundationCity of NY 2001).

On the other hand, it is expected that the quantity oforganic material and suspended solids will exhibit radicalchange with the use of food disposers as the idea with thedisposer is after all to move as much organic material fromthe solid waste to wastewater stream. In a pilot study, Nilssonet al. (1990) reported that the biochemical oxygen demand(BOD) and suspended solids (SS) increased by almost 50%after integrating disposers in the city of Staffanstorp in Swe-den, whereas the increase in total nitrogen was 12% andnegligible for phosphorus. In contrast, a study examining theimpacts of food waste disposal units and compost bins used inthe Ashmore suburb of the Goald City Coast in Queenslandshowed an increase of 16.5% for BOD, 3.0% for total nitro-gen and 4.6% for total phosphorus (based on a 100% marketpenetration) (Waste Management Research Unit 1994).Similarly, a study conducted in the town of Shurammar inSweden showed no increase in the amount of incomingBOD, nitrogen or phosphorus. Only an increase in the BOD/nitrogen was detected (Sudo 2000).

In long-term testing for clogs, Nilsson et al. (1990)showed that a stimulated optimal usage of disposers for aperiod of 15 years did not exhibit operational problemswithin the plumbing system. Regular inspection and yearlyvideotaping of the piping system revealed that the outersewage pipes function well during the use of food disposersand the buildup of sewage was reported at the water levelwith a width of 2–3 cm along the envelope surface and athickness of 0.5–1.5 cm. Similar results were reported bySinclair Knight (1990), Waste Management and ResearchUnit (1994), Koning & Van Der Graaf (1996), NYCDepartment of Environmental Protection (1997) and Strutz(1998).

At the wastewater treatment plants, theoretical calcu-lations showed that the amount of sludge increases by57% due to usage of food disposers, affecting primarily thebio-stage and sludge treatment (Nilsson et al. 1990). Simi-lar theoretical estimations by Galil & Yaacov (2001)reported that the contribution of garbage disposers isexpected to increase the weight of raw sludge by about60% in the case where biological treatment is applied and62% in the case of primary sedimentation followed by bio-

N. Marashlian, M. El-Fadel


logical treatment. In contrast, a study by Waste Manage-ment Research Unit (1994) indicated an increase of only18.1% in sludge production due to the introduction offood disposers. Concerns about increased sludge genera-tion persist and its potential environmental and economicimplications may differ with location and therefore area-specific characteristics must be taken into considerationwhen contemplating the adoption of a strategy to inte-grate food waste disposers in the waste-wastewater man-agement system.

Existing conditions

At present, MSW in the study area is managed through anintegrated solid waste management system (ISWM) involv-ing: (1) collection and transport of raw municipal waste;(2) sorting and processing of raw municipal waste at twotransfer facilities; (3) recycling of the waste fraction com-posed of glass, metals, papers and plastics; (4) compostingof the waste that is rich in highly biodegradable organic

content; and (5) transport and disposal of sorted/baled aswell as bulky waste at two old nearby quarries converted tolandfills (Fig. 1).

At nearly US$ 120 per tonne (UNEP 2000, MoE &LEDO 2001), the ISWM has suffered from many deficienciessince its implementation, the major ones being at the treat-ment and disposal levels. Indeed, the system has failed toachieve its targets, with more than 80% of the total wastesgenerated routed to the landfills, raising into question thepurpose of the sorting-processing–composting facilities aswell as the recycling programme (Table 1).

Apparently, the market demand for compost and recycla-ble materials may be either less than the generation rate ornot economically competitive. Thus, whether viewed as ahierarchy or as complementary components, the currentwaste management activities, particularly source separationand recycling have not measured up favorably with the stepsoutlined in an ISWM system (El-Fadel & Chahine 1999).On the other hand, the difficulty associated with locatingand approving a suitable site for landfilling has been increas-

Fig. 1: Current integrated solid waste management scheme of the Greater Beirut Area (El Fadel & Khoury 2001).

Table 1: Performance of adopted ISWM.

Actual achievements

ISWM targets (1997) LACECO (1999) LACECO (2002)

tonnes/day tonnes/day tonnes/day

Total wastes received at transfer/processing plants 1700 1922 2085

Waste handling means

Composting 850 216 300

Recycling 160 99 143

Landfilling 690 1607 1640



ing continuously, which dictates the adoption of policies thatwill minimize the amount of waste to be disposed of in alandfill. Therefore, solid waste minimization alternativessuch as food waste disposers are examined in this study.

The waste stream in the study area is characterized by a highproportion of food waste (63%) (Table 2), and has a projectedsolid generation rate exceeding 2000 tonnes/day. With morethan 70% moisture content in typical food waste (Tchoba-noglous et al. 1993), processing such waste at a wastewater treat-ment plant appears to be a suitable approach, if technically andeconomically feasible. Currently, while no wastewater treat-ment plants exist in the study area, many are planned to complywith the Government’s signed protocol for the protection of theMediterranean (MoE & LEDO 2001). The implementation ofplanned investments in wastewater treatment is still in its earlyphase, so that an increase in capacity is feasible at the designlevel. The projected wastewater generation in the study area issummarized in Table 3 with corresponding loadings.

Methodology

The operational impacts associated with the integration offood waste disposers include primarily: (1) solid waste com-position and distribution; (2) domestic water consumption;and (3) wastewater loading. Six scenarios (Fig. 2) wereexamined using a variable market penetration rate (25 to75%). The latter covers all the possible market penetrationscenarios reviewed in the literature with 25 and 50% beingthe most realistic ones since after 60 years of marketing gar-bage disposers in the US (which is considered the oldestmarket worldwide), their distribution reached a maximum of

50% (Galil & Yaacov 2001). A variable amount of foodground at the household level was adopted (75 to 95%). Thelowest value (75%) was reported by Wainberg et al. (2000).The upper range (95%) was used since only a limited numberof food wastes could not be ground including highly fibrouswastes and shells of certain seafood.

The current and anticipated future loadings to wastewatertreatment plants from the use of food waste disposers wasestimated based on a laboratory investigation that was con-ducted to assess the BOD and SS contents of ground foodwaste from the study area. The investigation was performedat the Environmental Engineering Research Center at theAmerican University of Beirut (AUB). Kitchen food wasteconsisting of vegetable, fruit, meat and other food waste con-stituents was collected from several households. The wastewas mixed thoroughly and divided into three batches ofequal size. The three batches were blended with tap water. Avolume of 11.7 L of water was used to grind 1 kg of foodwaste (Hartmann 2000, Wainberg et al. 2000). The resultingmixtures were then analysed for BOD and SS using StandardMethods for the Examination of Water and Wastewater (APHA1998). Each of the experiment was duplicated to assure repli-cability and consistency of the results.

Economic impacts entailed the evaluation of the conven-tional (tangible or direct) and environmental (non-tangibleor indirect) costs/savings for all scenarios. The conventionalcosts included the capital and operating cost of food disposerunits, the cost of wastewater and sludge treatment of theadded wastewater volume (loadings and flow), and the costof increased domestic water. As indicated in the backgroundsection, the cost of electricity needed to run food waste dis-

Table 2: Average solid waste composition in the study area.

Waste category Waste composition (%)

Food waste 63

Paper and cardboard 15

Plastic 10

Glass/china 6

Metal 3

Fabric 2

Other 1

Source: Ayoub et al. (1996); Baldwin et al. (1999); El-Fadel & Khoury (2001); MoE & LEDO (2001).

Fig. 2: Scenarios examined.

Table 3: Projected municipal wastewater generation rates in the study area.

Parameter 2005

Population projections (× 1000 persons) 1782

Wastewater generation rates (m3/day) 257 609

BOD (kg/day) 1 114 378

COD (kg/day) 2 162 294

SS (kg/day) 3 146 8371Based on an average of 444 mg/L (MoE & LEDO 2001); 2Based on an average of 630 mg/L (Khatib & Alami 1994); 3Based on an aver-age of 570 mg/L (El Fadel & Abou Ibrahim 2002).



posers was considered as negligible. Similarly, foregone earn-ings from potential energy recovery from food waste wereassumed to be insignificant particularly in cases where thewastewater treatment process involves anaerobic digestion.The conventional savings included the costs forgone due toreduced management requirements of food wastes divertedfrom the solid waste stream. Environmental costs/savings areassociated with potential impacts that are usually notdirectly perceived by the community. Due to the complexityand the inter-connection of the environmental media (air,water, soil and humans), the valuation of these environmen-tal impacts is difficult. Nonetheless, they can be estimatedusing the abatement cost method in which costs required toabate pollution resulting from solid waste management

(SWM) alternatives are used to estimate the value of poten-tial damages (Fig. 3). Note that all values used in the presentanalysis are at constant year zero therefore inflation was nottaken into consideration.

Results

The variation of the food waste composition in the studyarea as a result of introducing food disposers is summarized inFig. 4. Naturally, the integration of food disposers is expectedto reduce the total food waste generated and collected. Thepercentage of food waste within the MSW stream decreasesfrom 63 to 58% under S1a (25% market penetration with75% of the food ground), and down to 33% under S3b (75%

Fig. 3: Methodology for costs/savings estimation. 1Adopted from reported values in Carawan (1996), Union Sanitary District-California (2003) and WSC (2003). 2Examined technologies include centrifuge thickening and dewatering; belt filter press; composting; recessed-plate filter; aero-bic digestion; anaerobic digestion; alkaline stabilization; thermal aerobic pre-treatment and anaerobic digestion; pre-pasteurization and anaero-bic digestion; reactor composting; anaerobic digestion and thermal drying; and incineration. Adopted from Hallvard et al. 2001; EC 2002; Minett & Fenwick 2001; US EPA 2000a, b, c. 3Current domestic water charging rate as assigned by the Beirut Water Authority. 4Adopted from the European Commission (2002). 5Adopted from Massoud (2000) and MoE & LEDO (2001). 6Adopted from CIWMB (1990). 7Equal to: Min conven-tional costs + min environmental costs. 8Equal to: Max conventional costs + max environmental costs.



market penetration with 95% of the food ground), whichtranslates into 12 to 43% reduction in wastes to be landfilled(Table 4). Consistent with values reported in the literature(USEPA 1980, Nilsson et al. 1990, Waste ManagementResearch Unit 1994, Ketzenberger 1995, New York CityDEP 1997, Wainberg, et al. 2000, The Plumbing Foundationof New York 2001), the increase in water consumption is rel-atively insignificant, ranging from 0.72 to 2.35% under S1a(25% market penetration with 75% of food ground) to S3b(75% market penetration with 95% of food ground). Thecorresponding increase in the wastewater flow is equallyinsignificant accounting for 1.1 to 4.4% under the same sce-narios. The anticipated increase in sewage loadings from theuse of food waste disposers was based on the laboratory inves-tigation that indicated BOD and SS concentrations of 7042and 1537 mg/L, respectively. The anticipated increaseranged from 17 to 62% in terms of BOD loading under S1a(25% market penetration with 75% of food ground) to S3b(75% market penetration with 95% of food ground) andfrom 1.9 to 7.1% for SS loading under the same scenarios.

Measurements of the output from food waste disposersshow that about 98% of the input is reduced in size to

< 0.20 cm (CIWEM 2003). Kitchen sewer connection pipesin the study area are of standard size of 3.18 and 5.08 cminner diameter (ID). On the other hand, the ID of the sew-age connection system ranges from 0.4 to 2 m. Thus, no clog-ging is expected to occur in the plumbing connectionswithin households or in sewers, which is consistent with theabsence of clogging problems in cities where food waste dis-posers were installed (Nilsson et al. 1990, Strutz 1998, Sudo1998, Galil & Yaacov 2001).

Table 5 presents the details of the cost-saving analysisassociated with the integration of food waste disposers intothe solid waste/wastewater management schemes. The bene-fits achieved constitute 1.9 to 5.0% of the existing solidwaste management cost under S1a (25% market penetrationwith 75% of food ground) and 11.5 to 23.3% under S3b(75% market penetration with 95% of food ground), respec-tively. The benefits increased with the inclusion of environ-mental externalities to reach 7.2 to 10.4% of the existingsolid waste management cost under S1a (25% market pene-tration with 75% of food ground) and 31.6 to 44.0% underS3b (75% market penetration with 95% of food ground),respectively. Figure 5 illustrates the savings achieved with

Fig. 4: Food waste composition after installation of food disposers (2005 values).

Fig. 5: Benefits achieved under S1a (25% market penetration + 75% of food grinded) and S3b (75% market penetration + 95% of food grinded) as percentage of existing SWM cost.



the integration of food waste disposers as percentages of theexisting SWM cost, based on the minimum and maximumtotal costs (conventional and environmental) for 2005 and2020. It indicates that using current costs of solid waste andwastewater management, the benefits of integrating foodwaste disposers would increase with time as the quantities ofsolid waste to be managed increase.

The above analysis assumes that the cost of managing1 tonne of MSW will remain constant at the current charg-ing rate of US$ 119 per tonne. Locally, this cost is perceived

as relatively high and efforts are directed towards reducing itthrough competitive tendering by the private sector. Hence,the savings and net benefits that are achieved from the inte-gration of food disposers are expected to decrease. Abreakeven analysis would allow decision makers to define thepercentage reduction required under which the integrationof food waste disposers within the management schemewould become non-profitable. Assuming that all other val-ues are constant, the breakeven points for S1a (25% marketpenetration with 75% of the food ground), and for S3b (75%

Table 4: Impacts of introducing food disposers on SWM and WWM schemes (2005 values).

Scenario % Reduction in solid waste to be managed

% Increase in domestic water consumption

% Increase in wastewater flow


% Increase in SS loading

(1) (2) (3) (4) (5)

S1a 11.8 0.7 1.1 16.9 1.9

S1b 14.7 0.8 1.4 21.3 2.4

S2a 23.0 1.3 2.3 33.4 3.8

S2b 29.1 1.6 2.9 42.0 4.8

S3a 34.3 1.9 3.4 49.5 5.6

S3b 43.4 2.4 4.4 62.2 7.1

(1)

(2)

(3)

(4)

(5)

YTXaYFoodb

YFoodGroundmgS

RLWT

cWFoodGroundw

IWVWWeVFoodGround

MCIWWXNewBODXBODWW

XBODFood

VwwVFoodMFood MExpFood VExpFood IBODXNewSSXSSWWXSSFood

ISS

=====

====

==

===

====

====

=

=========

=

Total solid waste generated in the GBA in 2005 (tonnes/day)Estimated population in 2005 (capita)Average daily solid waste generation rate (kg/capita per day)Food waste generated (tonnes/day)% Composition of food waste in waste stream (assuming con-stant waste composition)Food waste ground (tonnes/day)% Market penetration of food waste disposers (25, 50 or 75%)Estimated food ground (75 or 95%)Solid waste expected to be landfilled if the current ISWM scheme remains operational (without integration of food waste disposers) (tonnes/day)% Reduction in solid waste to be landfilledEstimated total water demand for GBA in 2005 (without integra-tion of food waste disposers) (m3/day)Average daily water demand rate (L/capita per day)Amount of water needed to grind food waste (m3/day)Amount of water needed to grind 1 Kg of organic food (11.7 L/kg)% Increase in domestic water consumptionEstimated wastewater flow for the GBA in 2005 (m3/day)Average daily wastewater generation rate (L/capita per day)Volume of food waste ground expected to be disposed down the drain (m3/day)Moisture content (%)% Increase in wastewater flowNew BOD loading after installation of food disposersBOD loading of the current wastewater sewage of the GBA (mg/L)Average BOD of food waste based on experimental results (7042 mg/L)Total volume of the wastewater generation in the GBA (m3/d)Volume of food ground (in m3/day)Mass of food disposed in the sink (tonnes/day)Mass of food sample used in the experiment (500 g)Volume of the food sample used in the experiment (6.5 L)% increase in BOD loading after integration of food disposersNew SS loading after installation of food disposersSS loading of the current wastewater sewage of the GBA (mg/L)Average BOD of food waste based on experimental results (1537 mg/L)% increase in SS loading after integration of food disposers

YT X a×=

YFood YT b×=

YFoodGround YFood m g××=

RL 100 YFoodGround S÷( ) 100×[ ]–=

WT X c×=

WFoodGround w YFoodGround×=

IW WFoodGround WT÷[ ] 100×=

VWW X e×=

VFoodGround YFoodGround MC WFoodGround+×=

IWW VFoodGround VWW÷( ) 100×=

XNewBOD

XBODWW VWW×( ) XBODFood VFood×( )+

VWW VFood+( )------------------------------------------------------------------------------------------------------=

VFood

MFood VExpFood×( )MExpFood

----------------------------------------------=

IBOD

XNewBOD XBODWW–( )XBODWW

--------------------------------------------------------- 100×=

XNewSS

XSSWW VWW×( ) XSSFood VFood×( )+

VWW VFood+( )---------------------------------------------------------------------------------------------=

ISS

XNewSS XSSWW–( )XSSWW

------------------------------------------------ 100×=



market penetration with 95% of the food ground), takinginto consideration minimum and maximum conventional

costs, are depicted in Fig. 6a and c, respectively. Thebreakeven points for the same scenarios, taking into consid-

Table 5: Economic benefit of food waste integration (2005 values).

Scenario S1a S1b S2a S2b S3a S3b

% of market penetration 25 25 50 50 75 75

% of food ground 75 95 75 95 75 95

Costs (MUS$/year) Cost of food disposer units (1) 4.6 4.6 9.1 9.1 13.7 13.7

Cost of added volume of domestic water (2) 0.4 0.5 0.8 1.0 1.2 1.5

Cost of wastewater treatment (3) 0.6–2.4 0.8–3.0 1.2–4.7 1.6–6.0 1.8–7.1 2.3–9.0

Conventional cost of sludge management (4) 0.1–1.0 0.2–1.3 0.3–2.0 0.3–2.5 0.4–3.0 0.5–3.8

Environmental cost of sludge management (5) 0.02–0.15 0.03–0.19 0.04–0.30 0.05–0.38 0.06–0.45 0.08–0.57

Conventional costs (6) 5.7–8.3 6.0–9.3 11.4–16.6 12.0–18.6 17.1–25.0 18.0–28.0

Environmental and conventional (7) 5.7–8.5 6.0–9.5 11.5–16.9 12.1–19.0 17.2–25.4 18.1–28.5

Savings (MUS$/year) Conventional solid waste savings (8) 9.9 12.5 19.8 25.1 29.7 37.6

Environmental solid waste savings (9) 4.6 5.8 9.1 11.6 13.7 17.4

Net benefits (MUS$/year) Based on conventional costs/savings (10) 1.6–4.2 3.2–6.5 3.1–8.4 6.4–13.0 4.7–12.5 9.6–19.5

% of existing SWM cost (12) 1.9–5.0 3.8–7.8 3.8–10.0 7.7–15.6 5.6–15.0 11.5–23.3

Based on conventional and environmental costs/savings (11) 6.0–8.7 8.8–12.3 12.0–17.5 17.6–24.6 18.0–26.2 26.4–36.8

% of existing SWM cost (12) 7.2–10.4 10.5–14.7 14.3–20.9 21.0–29.3 21.5–31.3 31.6–44.0

(1)

(2)

(3)

(4)

(5)

(6)

(7)

(8)

(9)

(10)

(11)

(12)

H

fAPinCFWD

CW

hBODdMinCWW

MinCBOD

MaxCWW

MaxCBOD

SSMinCSS

MaxCSS

Sll

MinCSl

MinCST

MaxCSl

MaxCST

MinECSl

MaxECSl

MinCMaxCMinECMaxECCSp

NEBmin

NEBmax

Pmin

Pmax

EPmin

EPmax

=

============

==

==

=

==

=====

=

======

=

=

=

=

=

=

Expected number of food waste disposers to be installed under studied scenarios (disposer unit)Number of capita/householdAnnual cost of a food waste disposer unit (US$) Initial cost of the food waste disposer unit (assumed to be US$ 400)Opportunity cost per household (5% per household) (US$)Expected economic life of a disposer (12 years – Wainberg et al. 2000)Total cost of food disposer units (MUS$/year)Cost of added domestic water consumption (MUS$/year)Annual domestic water charge rate (refer to Fig. 3).Added BOD loading (kg/day)Water density (1000 L/tonne)Total minimum cost of wastewater treatment (MUS$/year)Minimum cost of secondary wastewater treatment per kg of BOD (refer to Fig. 3) (US$/kg)Total maximum cost of wastewater treatment (MUS$/year)Maximum cost of secondary wastewater treatment per kg of BOD (refer to Fig.3) (US$/kg)Added SS loading (kg/day) Minimum cost of secondary wastewater treatment per kg of SS (refer to Fig. 3) (US$/kg)Maximum cost of secondary wastewater treatment per kg of SS (refer to Fig. 3) (US$/kg)Generated sludge due to integration of food disposers (tonnes/day):Average settable solids of the food waste adopted from the literature (3327 mg/L: Wainberg et al. 2000) Minimum conventional cost of sludge treatment (MUS$/year)Minimum cost of sludge treatment per dry tonne (refer to Fig. 3) (US$/dry tonne)Maximum conventional cost of sludge treatment (MUS$/year)Maximum cost of sludge treatment per dry tonne (refer to Fig. 3) (US$/dry tonne)Minimum environmental cost associated with sludge treatment (refer to Fig. 3) (MUS$/year)Maximum environmental cost associated with sludge treatment (refer to Fig. 3) (MUS$/year)Minimum conventional costs (MUS$/year)Maximum conventional costs (MUS$/year)Minimum environmental & conventional costs (MUS$/year)Maximum environmental & conventional costs (MUS$/year)Conventional savings (MUS$/year)Cost of managing 1 tonne of food waste under current ISWM scheme(collection, transportation, sorting, baling, and landfill disposal) (US$/tonne)Net benefits based on minimum environmental and conventional savings/costs (MUS$/year)Net benefits based on maximum environmental and conventional savings/costs (MUS$/year)% of existing SWM cost based on minimum conventional savings/costs (MUS$/year)% of existing SWM cost based on maximum conventional savings/costs (MUS$/year)% of existing SWM cost based on minimum environmental and conventional sav-ings/costs (MUS$/year)% of existing SWM cost based on maximum environmental & conventional sav-ings/costs (MUS$/year)

H X f+( ) m×=

A P A/P i n, ,( ) A→ Pi 1 i+( )n

1 i+( )n 1–---------------------------= =

CFWD A H×=

CW VW h×=

BOD VFoodGround XBODFood d××=

SS VFoodGround XSSFood d××=

MinCWW BOD MinCBOD SS MinCSS×+×[ ] 365×=

MaxCWW BOD MaxCBOD SS MaxCSS×+×[ ] 365×=

Sl VFoodGround l d××=

MinCSl Sl MinCST× 365×=

MaxCSl Sl MaxCST× 365×=

MinECSl MinCSl 0.15×=

MaxECSl MaxCSl 0.15×=

MinC CFWD CW MinCWW MinCSl+ + +=

MaxC CFWD CW MaxCWW MaxCSl+ + +=

MinEC CFWD CW MinCWW MinCSl MinECSl+ + + +=

MaxEC CFWD CW MaxCWW MaxCSl MaxECSl+ + + +=

CS YFoodGround p 365××=

ES YFoodGround q 365××=

NBmin CS MinC–=

NBmax CS MaxC–=

NEBmin CS ES+( ) MinEC–=

NEBmax CS ES+( ) MaxEC–=

Pmin NBmin YFood r 365××( )÷[ ] 100×=

Pmax NBmax YFood r 365××( )÷[ ] 100×=

EPmin NEBmin YFood r 365××( )÷[ ] 100×=

EPmax NEBmax YFood r 365××( )÷[ ] 100×=



eration the maximum conventional and total (conventionaland environmental) costs, are depicted in Fig. 6b and d.Clearly, the introduction of food waste disposers under thelowest market penetration S1a, excluding environmentalexternalities, becomes non-profitable if the current cost ofSWM decreases by 52% (to reach US$ 57.1 per tonne)(Fig. 6a), taking into consideration the minimum conven-tional costs defined in Fig. 3 and Table 5. However, if themaximum conventional costs for wastewater and sludgemanagement were considered, the proposed system becomesnon-profitable if the current cost of SWM decreases by 26%only or down to US$ 88.5 per tonne (Fig. 6c). With theinclusion of externalities, the introduction of food disposersunder the same scenario (S1a) remains largely profitableeven if the cost of SWM decreases by 70% to reach US$ 35.3per tonne (if minimum costs were assumed) (Fig. 6d) or 98%to reach US$ 2.4 per tonne (if maximum costs wereassumed) (Fig. 6b). On the other hand, using the highestmarket penetration S3b, the proposed system becomes non-profitable under higher SWM cost reductions ranging from42% (reaching US$ 68.7 per tonne) (Fig. 6a) to 16% (reach-ing US$ 100.1 per tonne) (Fig. 6c), taking into considera-tion minimum and maximum conventional costs, respec-tively. With the inclusion of externalities, the margin ofsafety is higher and the proposed system remains profitableeven if the current SWM charging rate decreases by 60%(reaching US$ 46.9 per tonne) (Fig. 6b) to 88% (reachingUS$ 13.9 per tonne) (Fig. 6d).

Conversely, decision makers should consider the dynam-ics of economy of scale whereby it can be argued that the

cost of managing the remaining solid waste will increase ifthe total quantity of the waste to be managed has decreased.This will ultimately affect the total savings achieved fromthe SWM scheme as a result of integrating food waste dispos-ers. In this context and in order to define the cost of manag-ing the remaining solid waste generated ‘X’ (US$ per tonne)after the integration of food waste disposers at which the pro-posed system would breakeven, the worst case scenario interms of economy of scale (100% market penetration with100% food ground) was assessed. Excluding environmentalexternalities, the integration of food waste disposers willremain profitable until the cost/tonne of managing theremaining solid waste reaches 1.8 times the current chargingrate (US$ 223 per tonne). If environmental externalities areincluded, the integration of food waste disposers will still beprofitable up to a management cost of US$ 315 per tonne(2.6 times the current charging rate). Certainly, it is notexpected that the cost of SWM would reach such levels inthe study area in the near future, which justifies the adoptionof food waste disposers.

Limitations

Accomplishing the penetration levels assessed in this studyconstitutes the main constraint in the analysis presentedabove. The study assumes that food waste disposers are useddaily for the adopted market penetrations. This may not bethe case when residents are travelling or dine outsidealthough it is reasonable and relatively easier to integratefood waste disposers in restaurants. Furthermore, the labora-

Fig. 6: Breakeven points for S1a (25% market penetration + 75% of food grinded) and S3b (75% market penetration + 95% of food grinded) tak-ing into consideration the lower and upper range of costs (2005 values) (a) Based on minimum conventional costs; (b) Based on minimum total (conventional + environmental) costs; (c) Based on maximum conventional costs; (d) Based on maximum total (conventional + environmental) costs.



tory investigation was conducted using a relatively smallnumber of food waste samples, which may not be statisticallyrepresentative of the waste stream in the study area. Thefood waste was blended instead of ground, resulting in SSconcentrations that are lower than those reported in the lit-erature (Table 6). However, if the highest values for BODand SS loadings reported in Table 6 (Koning & Van derGraaf 1996) were considered in the economic analysis, theintegration of food waste disposers in the study area wouldremain beneficial. More explicitly, the net benefits achieved(2005 values) would constitute 7.5 to 32.7% of the existingSWM cost, taking into consideration the lower and upperrange of market penetration based on the minimum (con-ventional and environmental) total costs. The net benefitswould range from 0.5 to 6.3% of the existing SWM cost,based on the maximum total costs.

Concluding remarks

This study revealed that a reduction of 12 to 43% in the totalsolid waste stream could be achieved by integrating foodwaste disposers under market penetration ranging between25 and 75%, respectively. While no significant increase indomestic water consumption (for food grinding) and waste-water flow rates were expected, wastewater loadingsincreased by 17 to 62% (BOD) and 1.9 to 7.1% (SS). Con-cerns about increased sludge generation persist and its poten-tial environmental and economic implications may differwith location and therefore area-specific characteristics mustbe taken into consideration when contemplating the adop-tion of a strategy to integrate food waste disposers in thewaste-wastewater management system. In this study, theintroduction of food disposers into the waste and wastewater

Table 6: Effluent quality from food waste disposers.

Source Suspended solids (mg/L)4 BOD5 (mg/L)4

Wainberg et al. (2000)1 5834 11150

Koning and Van der Graaf (1996)2 10667 11648

Waste Management Research Unit (1994)1 10369 7524

NYC DEP (1990)1 5634 8078

Sinclair Knight (1990)3 6356 4000

Current study 1537 70421 Based on experimental works. 2Based on generic literature values. 3Based on theoretical calculations. 4Average of two experiments with triplicate analysis per experiment. BOD5, Biochemical Oxygen demand – 5 day.

Fig. 7: Proposed action plan for the integration of food disposers within an urban area.



management systems led to net economic benefits thatranged between 7.2 and 44.0% of the current solid wastemanagement cost. Food waste disposers can constitute a via-ble option (economically and environmentally) that couldreduce the load on the solid waste stream and minimize theamount of end waste requiring landfilling. The main techni-cal constraint lies in the ability to increase the loadingcapacity of wastewater treatment plants. Administratively, aproper action plan is needed to integrate food waste disposerswithin the SWM scheme of an urban area. Inadequate legis-lative and administrative frameworks and limited institu-tional capacity, as is the case in many developing countries,coupled with overlapping responsibilities of line Ministries,necessitate the role definition of the various parties involvedwhere such a policy is to be adopted. In this context, themain components that are required include: (1) Legislation,which entails the promulgation of a law for integrating food

waste disposers within new homes in form of building coderequirements; (2) Implementation, which comprises activi-ties or processes associated with law implementation withcorresponding responsibilities of line Ministries; and (3)Monitoring which consists of supervising the proper imple-mentation of the law. Equally important is a public awarenesscampaign with the first two phases of the plan. Figure 7depicts a typical action plan for the integration of food dis-posers within an urban setting taking into consideration thecurrent institutional framework in the study area.

Acknowledgements

Special thanks are extended to the United States Agency forInternational Development for its support to the Environ-mental Program and the Water Resources Center at theAmerican University of Beirut.

References

APHA (American Public Health Association), AWWA (American WaterWorks Association) and (WEF) Water Environment Federation.(1998): Standard Methods for the Examination of Water and Wastewater,20th edition. APHA, Washington, DC.

Ayoub G., Acra A., Abdallah R., & Merhebi F. (1996): Fundamental Aspectsof Municipal Refuse Generated in Beirut and Tripoli. Department of Civiland Environmental Engineering, American University of Beirut, Leb-anon.

Baldwin, D., Cui, S., & Dussek, C. (1999): Technical aspects of the disposalof plastic-wrapped baled wastes at Naameh landfill. Beirut, Lebanon.In: Proceedings Sardinia 99, Seventh International Waste Management andLandfill Symposium. CISA, Environmental Sanitary EngineeringCenter, Cagliari, Italy.

Carawan, R. (1996): Cut Waste to Reduce Wastewater Surcharges. WaterQuality and Waste Management. Available from http://www.bae.ncsu.edu/programs/extension/publicat/wqwm/food.html; Internet; accessed Sep-tember 2004.

CIWMB (California Integrated Waste Management Board). (1990): Chap-ter 3: Is Solid Waste Reduction Economically Beneficial? Available from:http://www.ciwmb.ca.gov/bizwaste/Baxter/Feasibility.htm; Internet;accessed September 2004.

Dunham, A.B. (2001): Garbage grinders. Gotham Gazette. Available from:http://www.gothamgazette.com/environment/mar.01.shtml; Internetaccessed September 2004.

EC (European Commission) (2002): Disposal and Recycling Routes for SewageSludge: Economic Report. Available from: http://europa.eu.int/comm/environment/waste/sludge/sludge_disposal4.pdf; Internet; accessedSeptember 2004.

El Fadel, M. & Abou Ibrahim, A. (2002): Coastal Zone in Lebanon: Manage-ment and Policy Options. United Stated Agency for InternationalDevelopment. Water Resources Center. American University of Bei-rut, Lebanon.

El Fadel, M. & Chahine, W. (1999): An integrated solid waste managementsystem for the Greater Beirut Area. In: ICSW-99-000 1st InternationalConference on Solid Waste, Pontificia Universitas Urbaniana, Rome, 7–9 April.

El-Fadel, M. & Khoury, R. (2001): Municipal Solid Waste Management inLebanon: Impact Assessment, Mitigation, and the Need for IntegratedApproach. Submitted to United States Agency for InternationalDevelopment (USAID).

Galil, N.I. & Yaacov, L. (2001): Integrated solid waste systems includingdomestic garbage disposers. Water Science and Technology. 44, 27–34.

Hartmann, T. (2000): Email of 31/1/00 from Thomas Hartmann InSink Era-tor to Paul Droop, InSink Erator.

Hallvard, Ø., Paulsrud, B., & Karlsson, I. (2001): Sludge Disposal Strategiesand Corresponding Treatment Technologies aimed at Sustainable Handlingof Wastewater Sludge. Available from www.euro-case.org/Activities/wastewater_sludge_010425/degaard.pdf; Internet accessed September2004.

In-Sink Erator. (2000): Disposer installation steps; available from http://www. insinkerator.com/install.html; Internet; accessed September2004.

Ketzenberger, B.A. (1995): Water Use by Kitchen Food Waste Disposers iInHouseholds and Effects of Ground Food Wastes on Septic Tank/SoilAdsorption Fields-Critical Review of Literature. Department of Civil andEnvironmental Engineering. University of Wisconsin-Madison, Madi-son, WI.

Khatib and Alami. (1994): Lebanon staged Wastewater Program. Beirut, Leba-non.

Koning, J. & Van der Graaf, J.H. (1996): Kitchen Food Waste Disposers’Effects on Sewer Systems and Wastewater Treatment. Delft University ofTechnology, Delft, The Netherlands.

LACECO. (2002): Supervision of Greater Beirut Sanitary Landfills: FifthAnnual Report. Council for Development and Reconstruction, Beirut,Lebanon.

LACECO. (1999): Supervision of Greater Beirut Sanitary Landfills: FirstAnnual Report (19 January 1998 to 18 January 1999). Council forDevelopment and Reconstruction, Beirut, Lebanon.

Macnair, D. (2000): A new look at food disposers. Reeves Journal; availablefrom http://www.reevesjournal.com/CDA/ArticleInformation/features/Features_Index/1,3816,27-619,00.html; Internet; accessed September2004.

Massoud, M. (2000): Privatization of Waste Management Services in Lebanon:Municipal Solid Waste and Wastewater. MS Thesis, Department of Civiland Environmental Engineering, American University of Beirut, Leb-anon.

Minett, S. & Fenwick, K. (2001): Advanced System to Recycle, Dispose ofSludge in Sweden. Available from http://twri.tamu.edu/watertalk/archive/2001-Jun/Jun-12.1.html; Internet; accessed September 2004.

MoE & LEDO. (2001): Lebanon State of the Environment Report. Ministry ofEnvironment (MoE), Lebanon.

New York City Department of Environment Protection (DEP). (1997): TheImpact of Food Waste Disposers in Combined Sewer Areas of New YorkCity. New York DEP, New York.



Nilsson, P., Hallin, P., Johansson, J., Lennart, K. Lilja, G. Petersson, B., &Petersson, J. (1990): Waste Management at the Source Utilizing FoodWaste Disposers in the Home: a Case Study in the Town of Staffanstorp.Department of Environmental Engineering, Lund Institute of Tech-nology, Lund, Sweden.

Sinclair Knight. (1990): Review of Residential Waste Disposal Units: Prelimi-nary Analysis and Interim Report. Report for Sydney Water. SinclairKnight, Sydney, Australia.

Steiner, H. (1996): Engineering Economic Principles. McGraw Hill, NewYork.

Strutz, W.F. (1998): Life Cycle Comparison of Five Engineered Systems forManaging Food Waste. Department of Civil and Environmental Engi-neering. University of Wisconsin, Madison, WI.

Sudo, K. (1998): Japan: Food Disposers. US Department of Commerce- NationalTrade Data Bank; available from http://www.csjapan.doc.gov; Internet;accessed September 2004.

Sudo, K. (2000): US/European Studies. Available from http://www.csjapan.doc.gov/horizon/disposer/struz/sld001.htm; Internet; accessed Septem-ber 2004.

Tchobanoglous, G., Theisen, H., & Vigil, S. (1993): Integrated Solid WasteManagement. Mc Graw Hill, Inc: New York.

The Catered Institution of Water and Environmental Management(CIWEM). (2003): Food Waste Disposers. Available from http://www.ciwem.com/policy/policies/food/index.asp; Internet; accessedSeptember 2004.

The Plumbing Foundation of New York. (2001): Benefits of Food Waste Dis-posers to New York City. Available from http://www.kverna.no/no/artik-ler/benefitsnyc.htm; Internet; accessed September 2004.

UNCSD (United Nations Commission on Sustainable Development).(1999): Solid Waste Management and Sewage Related Issues. Availablefrom http://www.un.org/documents/ecosoc/cn17/1994/background/ecn171994-bpch21.htm; Internet; accessed September 2004.

UNEP, Blue Plan Regional Activity, and CEDARE. (2000): Policy and Insti-tutional Assessment of Solid Waste Management in Five Countries:

Cyprus, Egypt, Lebanon, Syria, Tunisia. Policies and Institutional Assess-ment of Solid Waste Management in Lebanon. Available from http://www.planbleu.org/pdf/wasteLBN.pdf; Internet; accessed September2004.

Union Sanitary District-California (2003): Wastewater Discharge PermitApplication Available from http://www.unionsanitary.com/; Internet;accessed September 2004.

US EPA. (1980): Design Manual: Oonsite Wastewater Treatment and DisposalSystems. Environmental Protection Agency, EPA 625/1-80-012. USEPA, Washington DC.

US EPA. (2000a): Biosolids Technology Fact Sheet: Alkaline Stabilization ofBiosolids. Available from http://www.epa.gov/owmitnet/mtb/alkaline_stabilization.pdf; Internet; accessed September 2004.

US EPA. (2000b): Biosolids Technology Fact Sheet: In-vessel Composting ofBiosolids. Available from http://www.epa.gov/owmitnet/mtb/inves-sel.pdf; Internet; accessed September 2004.

US EPA. (2000c): Biosolids Technology Fact Sheet: Land Application of Biosol-ids. Available from http://www.epa.gov/owmitnet/mtb/land_application.pdf; Internet; accessed September 2004.

Wainberg, R., Nielsen, J., Lundie, S., Peters, G., Asholt, N., Russel, D., &Janckelson, C. (2000): Assessment of Food Disposal Options in Multi-unit Dwellings in Sydney. CRC for Waste Management and PollutionControl Limited; available from: http://www.crcwmpc.com.au/Publi-cations/FoodWasteDisposalReport/FoodWasteDisposalReport.pdf;Internet; accessed September 2004.

Waste Management Research Unit. (1994): Economic and EnvironmentalImpacts of Disposal of Kitchen Organic Wastes using Traditional Landfill –Food Waste Disposer – Home Composting. Griffith University, Queens-land, Australia.

WSC (Wastewater Service Charges). (2003): Public and Private Sewers.Available from http://www.sleepy-hollow.il.us/codebook/cb-s6-ch4a.html#6-4-1; Internet; accessed September 2004.




Session 14THE VALUE OF LIFE AND

HEALTH



Session 14aThe Value of Life and HealthThe Value of Life and Health

Burden of DiseaseBurden of Disease

Useful referencesUseful references

• Quantifying Environmental Health Impacts & Environmental Burden of Disease Series published by World Health Organization (WHO) @ www.who.int/quantifying_ehimpacts/en/

• Global Burden of Disease Methodology and Documentation @ www.who.int/whosis/en/

• WHO 2002. World Health Report 2002- reducing risks, promoting healthy life. Geneva, World Health organization. @ www.who.int/whr/2002/en/

(Figure 2.1 from EBD Series no. 1)

Environmental hazards/risk factorsEnvironmental hazards/risk factorsEnvironmental hazards/risk factorsEnvironmental hazards/risk factors

• Examples of environment-related health problems– Noise Hearing loss; Cardiovascular (?)– Biologically-contaminated water Diarrhea– Air pollution Exacerbation of asthma– Pesticide Acute poisoning; Neurotoxicity– Asbestos Lung cancer– Lead Neurocognitive deficits– Lack of safety Fatal and non-fatal injuries

Environmental hazards/risk factorsEnvironmental hazards/risk factors

Health is not the absence of disease and disability but the attainment of physical, mental, and social well-

being.

The role of environment becomes more significant if the holistic definition of health is adopted

• How does the environment affect health?– Direct effect

• Agent causes health problem– Indirect effect

• Reduction of immunity• Exacerbation of an existing health problem

Environmental hazards/risk factorsEnvironmental hazards/risk factors

CAUSAL WEBDistal social & economic causes

A causal web for lead exposureA causal web for lead exposure


Health burden of environmentHealth burden of environment

We ask:We ask:

1. What fraction of the national 1. What fraction of the national burden of disease is attributable to burden of disease is attributable to environmental risk factors? environmental risk factors?

2. What is the burden of disease 2. What is the burden of disease attributable to a specific attributable to a specific environmental risk factor?environmental risk factor?

Measuring Health Impacts Measuring Health Impacts

• Two links to be established in estimating monetary values of changes in human health associated with environmental changes

1.Link between environmental change and change in health status• Establishing DRRs• Establishing DALYs

2.Link between change in health status and its monetary equivalent • Establishing a WTP

Link 1

Link 2

Measuring Health ImpactsMeasuring Health Impacts

• Health impacts of pollution may be well recognized– Air pollution

• From itchy eyes and chest discomfort• To chronic bronchitis, asthma attacks, and premature death

– Inadequate water supply and sanitation• Diarrhea, intestinal nematodes and other diseases

• Health impacts measured via– Various types of studies including

• Epidemiology and field studies• Human clinical studies• Laboratory and toxicology studies

• Epidemiologic studies allow the establishments of– Dose-response relations (DRR) linking environmental variables

with observable health effects• DRRs established for air pollutants

– Burden of disease in terms of DALYs


• Types of studies that provide evidence of the impacts of exposure to pollutants– Epidemiology and field studies

• Involve estimating a statistical relationship between the frequency of specific health effects observed in a study population and measured levels of pollutants

• Types of studies– Cohort studies

» Analyze the incidence of health effects in a sample of identified individuals usually selected specifically for the study

» Allows better control of risk factors since characteristics of individuals are well known

– Population studies» Rely on the data available for the population as a whole rather

than tracking the effects on specific individuals» Readily available and cost-effective


• Types of studies that provide evidence of the impacts of exposure to pollutants– Epidemiology and field studies

• Advantages– Provide sufficient information to infer a concentration-

response function used to predict a change in the number of cases of a given health effect and pollutant concentration

– Define health effects in terms of factors that can be directly related to perceived welfare

» Risks of premature death» Days with noticeable symptoms

• Limitations– Uncertainty about whether the causal factors for the

observed association with health effects has been fully and accurately specified


• Types of studies that provide evidence of the impacts of exposure to pollutants– Human clinical studies

• Examine response of human subjects to pollutant exposure in a controlled laboratory setting

• Can provide evidence of causation because confounding variables are well controlled

• Advantages– Provides more accurate dose-response information

• Limitations– Limited to considerations of short-term reversible health

effects that can be induced on purpose in human subjects– Requires assumptions to link human exposure in real life to

health effect


• Types of studies that provide evidence of the impacts of exposure to pollutants– Laboratory and toxicology studies

• Use animal subjects and human tissue or cells to study biological responses to pollutants in a controlled laboratory setting

• Provide important information about specific biological pathwaysand mechanisms by which pollutants cause harm to living organisms

• Advantages– Pollutant exposures well-controlled and variations in confounding

factors reduced– Can consider both long term and short term exposures

• Limitations– Requires analysis and assumptions to link human exposure in real-life

to laboratory exposure – Uncertainty in extrapolating data from animal subjects to human

populations– Sometime focus on health effects that are difficult to interpret in terms

of specific symptoms


• DRRs– Correlate mortality and morbidity outcomes for susceptible

population groups with ambient concentration of a given air pollutant

– Epidemiological studies associated with air pollution• Time series• Cross-sectional

– Most studies have focused on mortality effects

Measuring Health ImpactsMeasuring Health Impacts• DRRs- Example

Estimated increments in annual health effects associated with unit change in pollutantsOutcome PM10

(10μg/m3)SO2

(10μg/m3)Ozone(pphm)

Lead(1.0 mg/m3)

NO2(pphm)

Premature mortality (% change) 0.96 0.48

Premature mortality/ 100,000 6.72

Respiratory hospital admissions/100,000 12 7.7

Emergency room visits/100.000 235.4

Restricted activity days/person 0.575

Lower respiratory illness/child 0.016

Asthma symptoms/asthmatic 0.326 0.68

Respiratory symptoms/person 1.83 0.55

Chronic bronchitis/100,000 61.2

Minor restricted activity days/person 0.34

Respiratory symptoms/1,000 children 0.18

Respiratory symptoms per adults 0.1 0.1

Eye irritations/person 0.266

Burden of Disease (Burden of Disease (BoDBoD))

• BoD study– aims to quantify the burden of premature mortality and disability

for major diseases or disease groups– Uses a summary measure of population health (DALY) to

combine estimates of the years of life lost and years lived withdisability

– Data are broken down by age, sex, and region

• Global Burden of Disease (GBD)– Constituted the most comprehensive set of estimates of mortality

and morbidity yet produced (Murray and Lopez, 1996)– WHO now regularly develops BoD estimates at regional and

global level for a set of more than 135 causes of disease and injury

– National BoD studies involve obtaining country-specific estimates for input to national policy

Measuring the Measuring the Burden of Disease Burden of Disease

is a TOOL for is a TOOL for setting health prioritiessetting health priorities

Why do we need robust processes Why do we need robust processes for setting priorities?for setting priorities?

• To ensure that health care resources are used in the most appropriate manner

• To achieve maximum health benefits using available resources

Prioritize actions

Other uses of measures of Other uses of measures of Burden of DiseaseBurden of Disease

• Evaluating health interventions• Predicting health gains from interventions• Evaluating policy programs• Estimating performance indicators and

assessing trends over time • Making comparison between regions• Identifying high risk groups

Criteria/ information used in Criteria/ information used in setting health prioritiessetting health priorities

• Goals for population health• Current and projected estimates of

burden of disease• Human, financial, and logistic resources• Historical trends in policy focus• Interest & pressure groups (local and

international)

Indicators that can be used to Indicators that can be used to assess theassess the Burden of DiseaseBurden of Disease

• Incidence• Prevalence

• Mortality• Morbidity• Disability

Cost:– Medical care– Lost productivity– Social burden

PrevalencePrevalence• Prevalence is the number of existing

cases in a given population over a specified time period

– Number of diabetics per 100,000 in 2007– Number of asthmatics per 1000 children less

than 5 years old in 2007

Old + New cases in a given population

IncidenceIncidence• Incidence is the number of new cases in a

given population over a specified time period

– Number of new cases of diabetes in 2007 among 100,000 who had no diabetes on January 1, 2007

– Number of new cases of asthma in 2007 among 1,000 children less than 5 years old with no asthma on January 1, 2007

Challenges to interpretationChallenges to interpretation

• How does a death at the age of 20 years compare with a death at the age of 70 years?

• How do 200 acute respiratory infections compare to 400 cases of infectious diarrhea?

(EBD Series no. 1; page 4)

None of the individual None of the individual indicators is sufficient. indicators is sufficient. There is a need for a There is a need for a

composite indexcomposite index to assess BODto assess BOD

A summary measure of life lost due to disease plus disability associated with

living with the disease relevant to a measure of expected population health


Disability Adjusted Life Years (DALY)– Measures health gaps as opposed to health

expectancies, using time measures– Measures the difference between

• a current situation and an ideal situation

An ideal situation is where “everyone lives up to the standard life

expectancy and in perfect health”

Burden of Disease (Burden of Disease (BoDBoD)) Burden of Disease (Burden of Disease (BoDBoD))

Disability Adjusted Life Years (DALY)– YLL

• Corresponds to the number of deaths multiplied by the standard life expectancy at the age at which death occursYLL = N × L

Where:N = number of deathsL = standard life expectancy at age of death in years


Disability Adjusted Life Years (DALY)– YLD

• Estimated by measuring the incidence of disability and the average duration of each disability

• Number of disabilities is multiplied by the average weight factor that reflects the severity of the disease on a scale from 0 (perfect health) to 1(dead)

• Years of Life with Disability (without applying social preferences):

YLD = I × DW × LWhere:

I = number of incident casesDW = disability weightL = average duration of disability (years)

Burden of Disease (Burden of Disease (BoDBoD))Disability Adjusted Life Years (DALY)

– Disability weights• Quantify societal preferences for different health states• DO NOT represent the lived experience of any disability or

health state• DO NOT imply any societal value for the person in the disability• Example

– A weight for paraplegia of 0.57– Does NOT mean

» Person in this health state is half dead» Person experience life as half way between life and death» Society values them less as a person compared to healthy people

– It means» Society judges a year with blindness (0.43) to be preferable than a

year with paraplegia» Society would prefer living for 3 years followed by death (1.7 lost

healthy years) than have one year of good health followed by death (2 lost healthy years)


Disability Adjusted Life Years (DALY)– Disability weights example (Murray and Lopez, 1996)



AIDS 0.5 Asthma, cases 0.10

Infertility 0.18 Deafness 0.22

Diarrhea disease, episode 0.11 Brain injury, long term 0.41

Measles episode 0.15 Spinal cord injury 0.73

Tuberculosis 0.27 Sprains 0.06

Malaria episode 0.20 Burns (> 60%) long term 0.25

Cancer, terminal stages 0.81 Congestive heart failure 0.32

Parkinson disease cases 0.39 Benign prostatic hypertrophy 0.04

Alzheimer disease cases 0.64

Birth 80 years60 years

2. Cerebrovascular death at age 60expected lifetime - age at death

Birth 60 years50 years

1. Stroke at age 50 with paraplegia (death 60)duration of disability x severity weight

20 YLL

10 x 0.7 YLD

TOTAL DALYs for this person=YLD + YLL= 7 + 20 = 27


1 DALY = one lost year of healthy life


• Challenges/ Ethical issues– What is the ideal condition? What is perfect health?– How sensitive are measures to gender and regional

differences?– Are all years of life equivalent?

• Age weighting: young adulthood more valuable than infancy and old age

• Discount rate: a year of perfect health today is more valuable than a year of perfect health 10 years later

– Is there an “objective” measure of disability?


Disability Adjusted Life Years (DALY)– Other social values

• Age weightsA year of healthy life lived at younger and older ages was weighted lower than for other ages

– Various studies have shown a broad social preference to value a year lived by a young adult more than a year lived by a young child or lived at older ages

– Age weights in DALYs are controversial• Time discounting

The net present value of lives lost was estimated using a 3% discount rate

– Studies have shown that people prefer a healthy year of life immediately, rather than in the future

– BoD studies may or may not include time discounting and age weights depending on local preference


Disability Adjusted Life Years (DALY)– Calculating DALYs with a 3% discount rate

• YLL:

• YLD:

)1( e rL

rNYLL −−=

Where:N = number of deathsL = standard life expectancy at age of deathr = discount rate (0.03)

)1( e rL

rLDWIYLD −−

××=

Where:I = number of incident casesDW = disability weightL = duration of disability in yearsr = discount rate (0.03)


Disability Adjusted Life Years (DALY)– Calculating DALYs with age weight and a 3%

discount rate• YLL:

)1(1]]1)([]1))(([[)(

)())((2 e rLaraLr

ra

rKareaLre

rKCeYLL −+−++− −

−+−+−−−++−

+= ββ

βββ

Where:a = age of death (years)r = discount rate (0.03)β = age weight ing constant (Ex: β = 0.04)K = age weighting modulation constant (Ex: K =1)C= adjustment constant for age-weights (Ex: C = 0.1658)L = standard life expectancy at age of death (years)


Disability Adjusted Life Years (DALY)– Calculating DALYs with age weight and a 3%

discount rate• YLD:

)}1(1]]1)([]1))(([[)(

{ )())((2 e rLaraLr

ra

rKareaLre

rKCeDWYLD −+−++− −

−+−+−−−++−

+= ββ

βββ

Where:a = age of death (years)r = discount rate (0.03)β = age weighting constant (Ex: β = 0.04)K = age weighting modulation constant (Ex: K =1)C= adjustment constant for age-weights (Ex: C = 0.1658)L = duration of disability (years)DW = Disability weight


• Disability Adjusted Life Years (DALY)– Example

YLL for diarrhea


• Disability Adjusted Life Years (DALY)– Example

YLD for Alzheimer


• Other issues to remember– Priorities are not decided on the basis of

numbers only– DALY measure is health focused– other gains

are not considered

Same principles for..Same principles for..

• Global Burden of Disease (GBD) studies• National Burden of Disease (NBD) studies

• Nationally:– Sub-regional data– Modify disability weights, adjustment rates,

life expectancy, etc. for national purposes– Comparison to other countries requires

adherence to universal method

Illustration from the Illustration from the National Burden of Disease National Burden of Disease

Study in LebanonStudy in Lebanon

The case of Coronary The case of Coronary Heart Disease (CHD)Heart Disease (CHD)

Presented at the “Fighting Together Cardiovascular Diseases” ConferenceLebanese Order of Physicians, BeirutFebruary 19-21, 2004

CHD: Flow chartCHD: Flow chart

Lebanese population

CHD: Prevalence/ incidence

Angina Myocardial infarction

HospitalizationComplications

Mortality/ Survival rate

Social burden

Surgical intervention

What do we need?What do we need?• Prevalence data by age, sex, and region.• Proportion hospitalized• Long-term survival Cohort• Complications studies

Need for gender and social analysis

Prevalence of CHD (%) Prevalence of CHD (%)

< 40 40-49

50-59

60-69

>=70 All ages

Beirut 1994(Nuwayhid et al., 1997)

M 0.0-0.9 4.6 11.7 15.0 24.1 4.3F 0.2-0.9 3.2 8.2 16.4 24.4 3.9

NHHEUS 1999 M + + + + + 3.8F + + + + + 3.3

Prostate screening 1997-98

M + + + 15

Several hospital-based studies:Male hospitalized patients toFemale hospitalized patients = 3-4


Lebanese population


Angina MI



Social burden


Hospital admissionsHospital admissions

• 4500 out of 60000 (7.5%) hospital discharge diagnoses in Beirut were due to CHD.

• 13% of all 3981 hospital admissions were due to vascular/circulatory diseases (NHHEUS 1999)


Lebanese population


Angina MI



Social burden


Hospital survivalHospital survival• 201 MI patients in 1993 (Jazra et al., 1995):

– 29 (18 M) died within 1 week (14.4%)– Death rate increased with age

• 443 MI admissions in 1996 (Sawaya et al., 1999): – 99 Females– Mortality rate (males): 7%, 5%, 11%, and 13% for

<=50, 51-60, 61-70, and >70– Mortality rate (females): 8%, 12%, 16%, and 25% for

<=50, 51-60, 61-70, and >70

Proportionate CVD mortality Proportionate CVD mortality (1966(1966--1996)1996)

Study (year) % CVD mortality

Beirut (1966-67):Death certificates 48.4Beirut (1983-84):Household survey 17.2Beirut (1992-93): F/U HH survey 42.4Beirut (1998): Death certificates 32.0Beirut (PHS 1996): Household survey 39.4Lebanon (PHS 1996): HH survey 34.9

PopulationPopulation--based mortality based mortality rate rate (Sibai et al., 2001) (Sibai et al., 2001)

• 1567 (50+ years) men and women (Beirut- 1983)

• Follow-up in 1993• Total deaths in 10 years = 416

– 40% ischemic heart disease– 13% Cerebrovascular disease– 7% Other CVD

PopulationPopulation--based mortality based mortality rate (continued) rate (continued) (Sibai et al., 2001)(Sibai et al., 2001)

• Total deaths in 10 years = 416• Mortality rate due to IHD:

– 16.2 per 1000 person-years (95% CI 3.5-19.5) for males

– 7.6 per 1000 person-years (95% CI 5.8- 9.9) for females

CHD: Can we calculate its CHD: Can we calculate its DALY?DALY?

Lebanese population


Angina MI



Social burden


Global Burden of Disease Global Burden of Disease (WHO)(WHO)

Selected results

Results Results -- global leading causes of global leading causes of deaths deaths …… and and …………DALYsDALYs

Cardiovascular 29%Neoplasms 13%Injuries 9% Respiratory 7%HIV/AIDS 5% Perinatal 4%Diarrhoea 4%TB 3%Malaria 2%Traffic accidents 2%Depression <1%

Cardiovascular 10%Neoplasms 5% Injuries 12%Respiratory 6% HIV/AIDS 6% Perinatal 7%Diarrhoea 4%TB 3%Malaria 3%Traffic accidents3%Depression 5%

2001 data, World Health Report 2002

Projected Change in Rank Order:2020 vs. 1990

Burden of Disease and Injury Attributable to Selected Risk Factors in the World in 1990

Environmental Burden of DiseaseEnvironmental Burden of Disease

We ask:We ask:

1. What fraction of the national 1. What fraction of the national burden of disease is attributable to burden of disease is attributable to environmental risk factors? environmental risk factors?

2. What is the burden of disease 2. What is the burden of disease attributable to a specific attributable to a specific environmental risk factor?environmental risk factor?

What is attributable risk?What is attributable risk?


Attributable risk is the excess risk that can be attributed to a specific exposure

(risk among exposed - background risk i.e., risk among non-exposed)

More on attributable riskMore on attributable risk

• If the specific exposure is controlled, then the excess risk attributed to it will be reduced

• However, the excess risk can be attributed to more than one risk factor– For example, lung cancer is attributed to cigarette

smoking and ambient air pollution

• Hence, if one exposure is controlled, the fraction of the remaining excess risk attributed to the other exposure will change

Two types of attributionsTwo types of attributions

• Categorical attribution– Event attributed to a single cause even if it is

associated with multiple causes (e.g., death can result from a combination of malnutrition and measles, but the case of death is attributed to either malnutrition or measles)

– Used in GBD and NBD studies

• Counterfactual attribution– The scenario with (e.g., 50% of population smoke) is

compared to the scenario without (e.g., none smoke)– Used in Environmental Burden of Disease studies

and the like

Why measure the Environmental Why measure the Environmental Burden of Disease (EBD)?Burden of Disease (EBD)?

• To set action priorities in health and the environment• To plan for preventive action• To assess performance• To compare/ contrast benefits from different actions• To identify high-risk populations• To plan for future needs• To predict impact of future environmental changes• To set research priorities in health and the environment• To guide policy making

Environmental Burden of Disease Series, No. 1Introduction and methods: Assessing the environmentalBurden of disease at national and local levels. WHO, 2003

Conceptual frameworkConceptual framework

*

Health Outcome

Risk factor A

Risk factor B

Risk factor C

Environmental risk factor

* Proportion attributed to environmental risk factor

ExamplesExamples

• What fraction of asthmatic cases can be attributed to air pollution?

• What proportion of cancer can be attributed to smoking? Asbestos?

• What proportion of diarrhea can be attributed to lack of hygiene and sanitation?

• What proportion of new cases of malaria can be attributed to climate change?

Sources of evidence on association Sources of evidence on association between environmental risk factor between environmental risk factor

and health outcome is neededand health outcome is needed

• Epidemiological (population-based) studies • “Natural” experimental studies

– Intentional and non-intentional incidents– Disasters (human-made or natural)

• Designed experimental studies– Human– Animal

Key information neededKey information needed

1. Indicators for selected environmental problems/areas

Examples of Examples of environmental indicatorsenvironmental indicators

• Water, sanitation and hygiene

• Ambient air pollution

• Lead

• Climate change (coastal floods)

• Water supply coverage• Sanitation coverage

• PM10 (annual mean)

• Blood lead

• Sea level rise• Frequency of coastal

floods



2. Data on these indicators (regularly)



2. Data on these indicators (regularly)3. Health indicators

Examples of health indicatorsExamples of health indicators

• Water supply coverage• Sanitation coverage

• PM10 (annual mean)

• Blood lead

• Sea level rise• Frequency of coastal

floods

• Diarrhea

• Mortality from cardiopulmonary disease

• Mental retardation• Loss of IQ points• Anemia

• Deaths and injuries



2. Data on these indicators (regularly)3. Health indicators4. Data on health indicators



2. Data on these indicators (regularly)3. Health indicators4. Data on health indicators5. Fraction of health outcomes attributed to

the environmental hazard

Causal web for fecalCausal web for fecal--oral transmission: oral transmission: which indicator to use?which indicator to use?


Practical approachPractical approach

• Complicated causal pathways are simplified

• Easier parameters are used more

Main global findings Main global findings (Pruss(Pruss--Ustun et al. 2008)Ustun et al. 2008)

• 13-37% of countries’ disease burden (about 13 million deaths per year) can be prevented by environmental improvements– 4 million deaths could be prevented by

improving water, sanitation, and hygiene• Environmental DALYs = 14 – 316 per

1000 capita per year

The Burden of The Burden of Outdoor Air PollutionOutdoor Air Pollution

Ostro B. Environmental Burden of Disease Series, No. 5.

Outdoor air pollution: Assessing the environmentalburden of disease at national and local levels.

WHO 2004.

Global burden of Global burden of Outdoor Air PollutionOutdoor Air Pollution

• 1.4% of total mortality• 0.4% of all DALYs• 2% of all cardiopulmonary

disease

Indicators (markers)Indicators (markers)

• Indicator for air pollution (environmental)– Particulate matter (PM10 or PM2.5)

• Health indicator– Mortality

• Mainly older people (with pre-existing cardiovascular and respiratory disease) and infants

– Morbidity• Hospitalization and emergency room visits• Asthma attacks, bronchitis, respiratory symptoms• Lost work and school days

Sources of evidenceSources of evidence

• Smog incidents in cities in Europe and the USA

• Epidemiological studies– Fixed monitor sites (hot spots might be

missed)– Limited to large cities (> 100,000)– Most limited to PM (many excluded ozone

and other air pollutants)– Most limited to PM10 not PM2.5

Increase of daily total mortality Increase of daily total mortality per 10 ug/mper 10 ug/m33 PM10PM10

• 0.8% (0.5-1.1%) (Schwartz et al. 1996)

• 0.7% (0.2-1.2%) (Burnett et al. 2000)

• 0.6% (Katsouyanmi et al. 2001)

• 0.27% (Dominici et al. 2002)

0.8% (0.5-1.6%) for the USA

• 0.6% for Europe (WHO 2004)

• 0.4-0.6% for Asia (HEI 2004)

Increase of daily total mortality Increase of daily total mortality per 10 ug/mper 10 ug/m33 PM10 PM10

in nonin non--Western countries and citiesWestern countries and cities

• Bangkok: 1.7% (1.1-2.3%) • Mexico City : 1.83% (0.9-2.7%)• Santiago: 1.1% (0.9-1.4%)• Inchon (S. Korea): 0.8% (0.2-1.6%)

In some cities, the air pollution is very high with PM10 exceeding 200 ug/m3. Recommended to cap the range for the assumption of linearity

PM2.5= 0.5 PM2.5= 0.5 –– 0.65 PM100.65 PM10

Model inputs for determining cityModel inputs for determining city--specific PM10 concentrationsspecific PM10 concentrations

• Energy consumption• Atmospheric and geographical factors • City and national population density• Local urban population density• Local intensity of economic activity• National income per capita• Time trends• Binary variable for each country

Relative risk functions for Relative risk functions for mortality and lung cancermortality and lung cancer Relative Risk (RR)= exp [Relative Risk (RR)= exp [ββ (X(X--XX00)] )]

Child and infant mortality Child and infant mortality related to PM10related to PM10

Steps 1Steps 1--55• Assess ambient exposure of the population to

PM (PM10 or PM2.5)– Fixed monitoring vs. Modeling– Need for background concentration (comparison

region)• Determine size of population groups exposed to

PM10 and PM2.5• Determine type of health effect of interest

(Cardiopulmonary diseases and Lung cancer)• Estimate the incidence of health effect• Use concentration-response functions

(epidemiological studies) that relate ambient concentrations of PM to selected health effects

Step 6Step 6• Estimate the following:

– Number of cases of premature mortality and DALYs (cardiopulmonary and lung cancer) attributed to long-term exposure to PM2.5 for people > 30 years old

– Number of cases of premature mortality and DALYs from respiratory diseases attributed to short-term exposure to PM10 for children < 5 years old

– Number of cases of premature mortality from all causes from short-term exposure to PM10


Thank YouThank You



Session 14bThe Value of Life and HealthThe Value of Life and Health

Value of Life and HealthValue of Life and HealthIntroductionIntroduction

• Two links to be established in estimating monetary values of changes in human health associated with environmental changes

1.Link between environmental change and change in health status• Establishing DRRs• Establishing DALYs

2.Link between change in health status and its monetary equivalent • Establishing a WTP

Link 1

Link 2

Value of Life and Health Value of Life and Health Valuating Health ImpactsValuating Health Impacts

Methods for valuing health impacts

Value of Life and Health Value of Life and Health Human Capital Approach (HCA)Human Capital Approach (HCA)

• Considers individuals as units of human capital that produce goods and services for society

• Measures loss of productivity resulting from an individual’s – death (Work Loss Days-WLD)– injury (Restricted Activity Days-RAD)

• Human life and time spent ill or recovering are valued using forgone earnings

• WLD and/or RAD estimated for– specific individuals in a detailed study– average individuals, which is most commonly applied

• HCA usually provides a lower-bound estimate

HCA =(# of Life Years Lost due to premature death or due to illness)

×(Average Wage Rate)

Value of Life and Health Value of Life and Health Human Capital ApproachHuman Capital Approach

• Values calculated are dependent on income, skill level, and country of residence

• Considered as the most difficult and controversial aspect of valuing health effects associated with environmental changes

Age group (yrs) Life years lost Mortality cost (1992 US$)

< 5 75 502,4215-14 68 671,88915-24 57 873,09625-44 42 785,58045-64 25 278,35065+ 10 22,977

Human capital and mortality cost by age in the US

Cost estimates are based on life-expectancy at the time of death and includelabor-force participation rates, average earnings, the value of home-makingservices, and a 6% discount rate


• Applying HCA1. Specify the type of economy for the population of

interest2. Specify the characteristics of the economy for the

population of interest3. Specify the family and community structure4. Specify the unit of analysis5. Specify the desired measure of productivity changes6. Estimate the maximum loss in productive time as a

result of the health outcome• Requires information as to the groups of patients that are

working• Requires decisions about value of time of children and retired

people


• Problems with HCA– Faces difficulty in accurately estimating forgone earnings

• Employee’s compensation includes more than wages– Pension plans, health insurance, flexible hours

– Does not provide information about the individual’s WTP to reduce probability of loss of life

– Does not measure net contribution to society• Assumes full employment and no substitutability of labor• Assumes a dominant cash economy where there exists market prices

which is not the case in developing countries– Ignores non-market activities important to individuals– Undervalues retired people, children, and home-makers– Estimated value highly depends on discount rate used

• The higher the discount rate, the lower the economic value of children– Does not value pain and suffering, the individual’s own well-

being and preferences, and the sentiments of the society


Value of Life and Health Value of Life and Health Cost of Illness (COI) ApproachCost of Illness (COI) Approach

COI provides a lower-bound estimate


Direct costs• Useful economic tool as it indicates the direction and

magnitude of the economic flows resulting from health shocks to the economy

• Easily understood and often readily available being based on available market and expenditure data

• COI provides an estimate of an individual welfare loss– Direct expenditures do not correspond to a drop in income

or consumption for the economy as a whole, but constitute a redirection of economic activity, with some sectors benefiting from increased activity


Direct costs• COI does not provide a direct measure of disease severity

– Direct medical expenditures are influenced by income distribution

• Increased income is accompanied with increased consumption of health care

– Direct medical expenditures reflect the ability of current medical techniques to treat the disease under consideration

• Example treatment of malaria is expected to generate fewer expenditures than treatment of cold because the former has few remedies as compared to the latter

• COI not only measures disease severity but also the population’s education, skill level, income, insurance coverage, types of medical interventions currently available, etc.


Direct costs• Issues pertaining to its application

– Difficulty to disaggregate hospital payments• Drugs administered on the premise• Salaries paid to health professionals and staff…

– Inaccuracies in hospital diagnostic data and the fact that expenses might not be attributed to the correct illness

– A number of illnesses may be grouped under one diagnostic code making it hard to decipher individual expenses

– Large data sets assume the same charge for all types of physician services

• A visit for a routine checkup does not cause the same as a visitfor cancer

– Treatment of multiple conditions where all expenses are allocated to the patient’s primary condition

Value of Life and Health Value of Life and Health Cost of Illness ApproachCost of Illness Approach

• Estimation of direct cost of medical care (WASH, 1991)1. Estimate proportion of those affected at each level of severity of the

disease2. Estimate the proportion of those desiring treatment who have access

to treatment3. Specify the process of treatment for each level of severity of the

disease• Resource use• Number of inpatient days• Outpatient visits

4. Estimate the unit costs of resources used for treatment and the side effects for each level of severity of the disease taking into account that many fixed costs are not affected by reductions in the use of the health service

5. Estimate total treatment costs for each level of severity of the disease without intervention

6. Determine the proportion of the costs that can be avoided in theshort- and long-run

7. Determine the direct costs that would have been avoided

Value of Life and Health Value of Life and Health Hedonic PricingHedonic Pricing

• Involves the valuation of incremental morbidity or mortality by identifying wage differentials due to risk differences

• Based on the assumption that there is a fixed supply of jobs and a freely functioning job market where individuals choose jobs based on perfect information and with no mobility restrictions

• Based on the theory that workers have to be paid a premium to undertake jobs that are inherently risky, which can be used to estimate the implicit value individuals place on sickness or premature death

• Estimates of VOSL in the US– 1.9-10.7 million USD (1990 dollars)


Where: P = payment rate for a given jobS = vector of skills required to do the jobJ = vector of other job-related attributes (working hours, holiday, sickness benefits)R = risk of death


• Issues and limitations– Difficulty in assessing an objective measure of

the risk of death– Contains a high degree of uncertainty– Requires considerable data sets for

regression analysis, containing data on all relevant and confounding variables

– Results are not transferable between countries due to differences in attitudes to risk and incomes

Value of Life and Health Value of Life and Health Contingent Valuation MethodContingent Valuation Method



• Benefit transfer• Values may be adopted from the other countries by

adjusting for per capita income as follows

Per capita income of country i = Xi

⇒ Income ratio Xj/Xi

Per capita income of country j = Xj

⇓

Value of mortality or morbidity outcome in country i = Yi

⇒ Multiply Yi by Xj/Xi ⇒

Value of mortality or morbidity outcome in country j = Yj


• Advantages of a CVM– Can take into account non-use values– Can be designed to include only the variables or

characteristics of the market relevant to the objective of the study

– Allows individuals to consider the true costs to themselves of a particular injury or illness

– CVM results are repeatable• In terms of similarity in results across different settings• Using a test-retest methodology


• Problems associated with CVM– Does not require cash transactions– Biases: Strategic, design, hypothetical, etc.– Survey responses cannot be verified except

through comparison with actual behavior following survey

– WTP vs. WTA– Short time given to respondents to think about the

answer– In developing countries, questionnaires need to

be adapted carefully and trained researchers are required to administer the surveys


• Issues to consider– WTP questions should be clear and unambiguous– Respondents must be familiar with the valued

commodity• Health risk studies involving common, mild illnesses

have a greater chance of being understandable, meaningful, plausible, than severe, rare diseases

– Respondents should have prior valuation/ choice experience with respect to consumption levels of the commodity in order to give it well-formed values

Value of Life and Health Value of Life and Health Disability Adjusted Life Years (DALY)Disability Adjusted Life Years (DALY)

• The VOSL obtained from wage differential and contingent valuation studies may be linked with the corresponding number of DALYs lost in a specific study and so estimate the implicit value per DALY

• The cost of a DALY lost valued by two approaches– DALY (yrs) × GDP/capita (USD/year)

• Based on the rationale that the economic value of a year lost to illness or early death is the productive value of that year, which is approximated by GDP per capita

• Usually represents the lower bound estimate• Has nothing to do with the non-economic value of life in general

– DALY (yrs) × WTP for mortality reduction• Based on the willingness-to-pay (WTP) by an individual to reduce the

risk of death.• Valuations arrived at, in studies in the United States and Europe that

apply WTP, are substantially higher than the GDP per capita approach (at least for adults)

Value of Life and Health Value of Life and Health Valuating Health ImpactsValuating Health Impacts

Types of benefits Market value(COI, HCA)

Avertive expenditure

Hedonic pricing


Improved health-related quality of life

Improved life expectancy

Medical cost avoided ( )Reduced time spent in care ( )Reduced travel expenses to care ( )Reduced avertive expenditure ( ) ( )Increased productivity ( )Reduced sick leave ( )

= preferred method; ( ) = second best method

Recommended methods of valuation for health-related benefits of environmental health interventions


Thank YouThank You




Sessions 14 & 15Valuation of Life and Health

CASE-STUDIES



Sessions 14 & 15Valuation of Life and HealthValuation of Life and Health


CaseCase--StudiesStudies

• Drinking Water Quality: A health-based socio-economic assessment

• Socio-Economic Benefits of Leaded Gasoline Phase-out, Lebanon

• PM in urban areas: health-based economic assessment, Lebanon

• Economic Benefits of Reducing Particulate and Sulfate Emissions from the Cement Industry, Lebanon

Drinking Water Quality: A Drinking Water Quality: A healthhealth--based sociobased socio--

economic assessmenteconomic assessment

CASE DESCRIPTIONCASE DESCRIPTIONOutlineOutline

SOCIO-ECONOMICBURDENS

UNCLEANwater

Water-related DISEASES

INADEQUATEwater supply

&

CASE DESCRIPTION CASE DESCRIPTION Public Water Supply in BeirutPublic Water Supply in Beirut

• Falls under the responsibility of 2 water authorities under the jurisdiction of the MEW– Beirut Water Authority

• Serves Beirut Municipal District and Northern Suburbs

– Ain El Delbe Water Authority• Serves the Southern Suburbs of Beirut and

Damour

• Water supply situation in Beirut is inadequate

Intermittent supply in most areas&

Lack of piped water for a large number of the poor

CASE DESCRIPTION CASE DESCRIPTION Existing conditions of water supplyExisting conditions of water supply

Weak water utilities+

Water scarcity

22 % of the population are NOT connectedto the public water supply system -mainly in the suburbs(CDR, 1998; CAS, 1997)

CASE DESCRIPTION CASE DESCRIPTION Other Sources of Water SupplyOther Sources of Water Supply

• Two main supplementary water sources are– private wells and/or on water vending

• Water Vending– Consumers, in the direct southern suburbs of

Beirut, resort to small-scale private water providers – Water Vendors

– Around 800 water shops are distributed randomly all over Lebanon

– The majority of shops operate without any regulatory authorization

CASE DESCRIPTIONCASE DESCRIPTIONForms of water vendingForms of water vending

Mobile water trucks

Water tanks within shops Water tanks providedby political organizations

CASE DESCRIPTIONCASE DESCRIPTIONWater Quality in LebanonWater Quality in Lebanon

• Various studies were conducted to examine the quality of water in Lebanon– Analyses by Central Laboratory– Study on vended vs. bottled water– Study on water quality in pilot area

• The studies revealed that– Water quality in the Lebanese public supply

system is variable– Quality of vended water is unacceptable

CASE DESCRIPTION CASE DESCRIPTION Analyses by Central LaboratoryAnalyses by Central Laboratory

• 2001 statistics reported by the Central Laboratory, revealed pollution of water supply

Water source TotalBottled water Network water Groundwater

Samples exhibiting microbiological pollution

97 (24%) 345 (24%) 450 (37%) 892 (36%)

Total number of samples analyzed

403 863 1215 2481

CASE DESCRIPTION CASE DESCRIPTION Study on Vended Water Vs. Bottled WaterStudy on Vended Water Vs. Bottled Water

• Samples included– 65 samples from water vending shops– 23 samples of bottled water

• Analysis conducted at AUB– pH, Conductivity, Salinity– Total Hardness, Nitrates– Total and Fecal Coliform

• Results– Poor quality of some samples of Vended Water with

respect to microbiological indicators– Acceptable quality of Bottled Water

Percent of samples exceeding MOI bottling Percent of samples exceeding MOI bottling water guidelineswater guidelines

• 15 samples were collected from the public supply– Main station along the coast, the two reservoirs

uphill and dead ends in the network– Household taps, after being stored in roof tanks

to be used for domestic purposes– Public locations

• Results– Acceptable quality in terms of physico-chemical

and microbiological characteristicsNo microbiological contamination

CASE DESCRIPTION CASE DESCRIPTION Study on the water quality in a pilot Study on the water quality in a pilot

areaarea

CASE DESCRIPTION CASE DESCRIPTION Study on the water quality in a pilot area Study on the water quality in a pilot area

(cont(cont’’d)d)• 20 private wells were randomly sampled• Results of analysis

Percent of samples exceeding USEPA drinking water guidelines

CASE DESCRIPTION CASE DESCRIPTION Health Impacts of Water PollutionHealth Impacts of Water Pollution

Polluted water + Water shortage + Unsanitary living conditions

• 2.3 billion people in the world suffer from water-borne diseases

with over 12 million death per year

• Water-related health impacts in Lebanon

Short & Long-term HEALTH RISKS

Hence, providing clean water & ensuring proper sanitation facilities reduces prevalence of water-related diseases

Disease Geographic extent Number of cases a Deaths per year WATER-BORNE DISEASES

Diarrheal disease Worldwide 500 million/yr 3-4 million

Cholera South America, Asia, Africa 384,000/yr 20,000

Hepatitis A Worldwide 600,000 – 3 million/yr 2,400-12,000

Paratyphoid & typhoid

80% in Asia, 20% in Latin America, Africa 16 million currently 600,000

Polio 66% in India, 34% in Near East, Asia, Africa 82,000 currently 9,000

WATER-BASED DISEASES Ascariasis Africa, Asia, Latin America 250 million currently 60,000

Clonorchiasis Southeast Asia 28 million currently None reported

Dracunculiasis (guinea worm)

78% Sudan, 22% in other Sub-Saharan Africa, and few cases in India and Yemen

153,000/yr None reported

Paragonimiasis Far East, Latin America 5 million currently None reported

Schistosomiasis (bilharzias)

Africa, Near East, rain forest belt in Central Africa, Western Pacific, Cambodia, Laos

200 million currently 20,000

WATER-RELATED VECTOR DISEASES

Dengue Tropical environments, concentrated in Asia, Central and South America

50-100 million/yr 24,000

Filariasis Africa, Eastern Mediterranean, Asia, South America

120 million currently None reported

Malaria Africa, Southeast Asia, India, South America 300-500 million/yr (clinical)

2 million

Onchocerciasis Sub-Saharan Africa, Latin America 18 million currently None reported b

Rift Valley Fever (RVF)

Sub-Saharan Africa NA c 1% of cases

Major WaterMajor Water--Related DiseasesRelated Diseases(Muller & (Muller & MoreraMorera, 1994; WHO, 1996, 1998), 1994; WHO, 1996, 1998)

CASE DESCRIPTION CASE DESCRIPTION Health impacts of water pollution in Health impacts of water pollution in

LebanonLebanon• Lebanon suffers from adverse health

impacts as a result of water pollution• Prevalent water related-diseases include

– Diarrhea– Typhoid & paratyphoid– Hepatitis A

• Data are limited due to absence of proper disease reporting mechanism– Mortality data– Morbidity data

CASE DESCRIPTION CASE DESCRIPTION Mortality dataMortality data

“Each child under 5 is exposed, on average, to 3.5 incidents of diarrhea each year, causing the death of 750 children per year”

(UNDP, 1990)Note: This value may be an overestimation due to improvement in

water supply & sanitation

CASE DESCRIPTION CASE DESCRIPTION Morbidity dataMorbidity data

Average annual number of reported incidents for the years 1995-2000(Ministry of Health, Directorate of Preventive Medicine)

CASE DESCRIPTION CASE DESCRIPTION Mitigating adverse waterMitigating adverse water--related related

health impactshealth impacts

• Several studies examined the impact of improved water supply and sanitation– Expected reduction in morbidity rates

– Variation of effect of intervention with type of disease

Expected reductions in morbidity from water & Expected reductions in morbidity from water & sanitation improvements (sanitation improvements (EsreyEsrey et al., 1991; et al., 1991;

Dougherty & Hall, 1995)Dougherty & Hall, 1995)

Disease Percent reduction in morbidity Diarrheal diseases 26-50 Typhoid 80 Paratyphoid 40 Infective hepatitis 10 Ascariasis 29-40 Cholera 90 Dracunculiasis 78 Onchoserciasis 20? Hookworm infection 4 Schistosomiasis 60-77 Trachoma 27-60 Guinea worm 100

Potential relations between water and sanitation Potential relations between water and sanitation interventions and morbidity from selected diseases interventions and morbidity from selected diseases

((EsreyEsrey et al., 1991)et al., 1991)

Disease Intervention

Improved drinking water

Water for domestic hygiene

Water for personal hygiene

Sanitation

Ascariasis +1 ++2 -3 ++

Diarrhea + ++ ++ ++

Dracunculiasis ++ - - -

Hookworm infection

- - - ++

Schistisomiasis - ++ ++ ++

Trachoma - + ++ - 1 + = strong impact; 2++ = stronger impact; 3- = little or no impact

ECONOMIC VALUATION OF ECONOMIC VALUATION OF HEALTH IMPACTSHEALTH IMPACTS

• Associated with few constraints– Actual identification and measurement of health

impacts– Estimation of monetary values for associated

mortality and morbidity– Establishing dose response functions (DRFs)

“it is not ambient water quality per se that affects health but access to clean drinking water and adequate sanitation along with household level of income and education”

ECONOMIC VALUATION OF ECONOMIC VALUATION OF HEALTH IMPACTSHEALTH IMPACTS

• Methods of economic valuation– Economic valuation of mortality effects

• Human Capital Approach (HCA)• Willingness to Pay (WTP)/Willingness to Accept

(WTA) Approach– Economic valuation of morbidity effects

• Cost of Illness Approach (COI)• WTP/WTA Approach

• Economic valuation of health impacts in Lebanon

APPLIED METHODAPPLIED METHODHuman Capital Approach (HCA)Human Capital Approach (HCA)

• Measures loss of productivity resulting from an individual’s death or injury

HCA provides a lower-bound estimate.Yet, it is the best alternative in the absence of WTP data

HCA =(# of Life Years Lost due to premature death)

×(Average Wage Rate)

Advantages Disadvantages• Easily measured • Biased against groups with low wages

• Assigns no value to lives of old and retired

• Assigns no value to leisure time

APPLIED METHOD APPLIED METHOD WTP/WTA ApproachWTP/WTA Approach

• Assesses from market behavior– WTP for reduced risks of increased

mortality/morbidity– Or WTA increased risk of increased

mortality/morbidityTotal Benefit/Cost= ∑WTP of all concerned members of the society

= Value of Statistical Life (VOSL)

Advantages Disadvantages

• Correct valuation method• Captures the value of less

tangible changes in productivity

• Difficult to collect data

APPLIED METHOD APPLIED METHOD Cost of Illness (COI) ApproachCost of Illness (COI) Approach

• Measures the direct cost of morbidity in terms of– Medical expenditure for treatment– Lost wages during days spent in bed– Days missed from work, etc.

• Compared to the WTP approach,

COI provides a lower-bound estimate.Yet, it is the best alternative in the absence of WTP data

Advantages Disadvantages

• Required data generally obtained with more accuracy

• Obtained estimates can be better communicated

• Incomplete due to insufficient information

• Does not include non-tangible entities

RESULTSRESULTSEconomic Assessment of water quality Economic Assessment of water quality

in Lebanonin Lebanon

• Mortality estimation using HCA

• Mortality estimation using WTP

• Morbidity estimation using COI

• Estimation of mortality and morbidity reduction

• Uncertainty

RESULTS RESULTS Mortality estimation using the HCA Mortality estimation using the HCA

approachapproach• Assumptions– Average Lebanese monthly salary = 400 USD– Productivity age range = 20-65 years

• Reported mortality– 750 deaths/year– Children less than five years of age

• Calculation– (400 USD/month) × (12 months/year) × (45 years of lost

productivity) = 216,000 USD per case

According to the HCA, the total economic cost of premature death incurred in Lebanon, and caused by water pollution is 162 million USD, assuming the same number of premature deaths as reported by the UNDP in 1995

RESULTS RESULTS Mortality estimation using the WTP Mortality estimation using the WTP

approachapproach

• WTP data are lacking in Lebanon• Values are adopted from the US as

followsPer capita income of country i = Xi



⇓


⇒ Multiply Yi by Xj/Xi ⇒

Value of mortality or morbidity outcome in country j = Yj

RESULTS RESULTS Mortality estimation using the WTP Mortality estimation using the WTP

approach (contapproach (cont’’d)d)

• WTP estimates range between 0.6-13.5 MUS$

• Ratio of GNP per capita in Lebanon to US = 0.1 (World Bank, 1999)

• Value of statistical life in Lebanon ranges between 0.06-1.35 MUS$

The total cost of mortality ranges from 45 to 1,012.5 MUS$

RESULTS RESULTS Morbidity estimation using the COI Morbidity estimation using the COI

approachapproach• COI incurred by society from water-related

diseases consists of several elements– Cost of hospitalization– Cost of medication– Lost productivity– Cost of transportation

• The diseases are restricted to– Dysentery– Hepatitis A– Typhoid & Paratyphoid

• Total of COI



• The water-related illnesses are gastroenteritic in nature– They have the same hospitalization cost

per day– Cost varies with class of admission

Admission class Social Security Third class Second class First classPhysician visits (USD per day)

13.3 24 40 64

Hospital room (USD per day)

22.5 33 56 90

Laboratory (USD per stay)

127 218 240 285

Variation of hospitalization cost with class of admission

• Length of hospital stay, recovery at home, and medication cost vary with type and severity of disease

Hospital stay Recovery at home Length of stay

(days) Medication

cost (USD per day)

Length of stay (days)

Medication cost

(USD per day) Range Average Range Average Dysentery 2-4 3 15 0-1 1 0 Hepatitis A 3-7 5 10 7-14 10 10 Typhoid 3-7 5 40 5-10 7 30

Average length of stay and medication cost per disease



• Productivity loss– Productivity loss per day of leave from work = 18.2 USD

(per capita GDP = 400 USD; working days per month = 22)

– 52.2 % of cases are in the productive age(Based on age distribution of Lebanese population)

– Productivity loss of care-providers was disregarded• Transportation costs

– Roundtrip visit costs 1-3 USD– 3 roundtrips conducted per day of illness



Total # of days of restricted activity

Lost productivity (USD/case)

Transportation cost (USD/case)

Dysentery 2-5 36.4-91 6-45 Hepatitis A 10-21 182-382.2 30-189 Typhoid 8-17 145.6-309.4 24-153

Lost productivity & transportation costs per disease


approach (contapproach (cont’’d)d)• Upon summing up all input data, the total

COI incurred by society from the water-related diseases under study ranges between 613,295 - 2,664,502 USD

Number of reported cases

Cost of illness per case (USD/case)

Total cost of illness (USD)

Dysentery 529 254-1,054 134,366-557,566 Hepatitis A 287 389-1,822 111,643-522,914 Typhoid 809 454-1,958 367,286-1,584,022 Total 613,295-2,664,502

RESULTSRESULTSEstimation of Mortality & Morbidity Estimation of Mortality & Morbidity

ReductionsReductions• Estimates the benefits associated with the reduction in

prevalence of diseases/mortality upon provision of clean water supply and ensuring proper sanitation

Estimated benefit =Reported # of cases × Expected % reduction × Cost per case

Parameter Diarrhoeal diseases

Typhoid paratyphoid

Infective hepatitis

Mortality

Number of cases 529 809 287 750 Percent reduction (%) 26-50 60 10 55 Number of cases avoided 138-264 485 29 412

Cost per case (USD) 254-1,054 454-1,958 389-1,822 (0.06-1.35) × 106 Economic benefit (million USD/year)

0.035-0.278

0.220-0.949

0.011-0.053 25-557

RESULTS RESULTS Summary of socioSummary of socio--economic benefits due to economic benefits due to

provision of clean water & proper provision of clean water & proper sanitationsanitation

Endpoint Number of cases avoided

Total economic benefits (MUS$/yr)

Mortality 412 25-557 Morbidity 652-778 0.27-1.28 Total 1,064-1,190 25.27-558.28

RESULTS RESULTS UncertaintyUncertainty

• Absence of population-based vital and disease registries

• Assumptions adopted related to transportation costs, age distribution of ill, cost of lost productivity

• The reliance on epidemiological studies reviewed in the literature (WTP)

• The assumption that all diarrhea, typhoid, paratyphoid, and hepatitis A are due to water pollution

• The disregard of effects of other water contaminants

EEnd of nd of CCasease SStudytudy

Thank YouThank You

SocioSocio--Economic Benefits of Economic Benefits of Leaded Gasoline PhaseLeaded Gasoline Phase--out, out,

LebanonLebanon

Case DescriptionCase DescriptionStudy area Study area -- LebanonLebanon

• Relatively old vehicle fleet → high pollutant emission rates

• Prior to August 2002, leaded gasoline (max 0.66 g/l lead (Pb)) predominantly used

• Physiological effects of Pb:– Biochemical effects (anemia, interference with enzyme

synthesis)– Neurobehavioral effects (IQ deficiency, mental

retardation, hyperactivity) in children– Increased probability of hypertension & cardiovascular

diseases

Correlation Between Lead used and BLL in the US

Correlation betweengasoline lead contentand airborne leadlevels in the US

0

30

60

90

120

150

180

1974 1976 1978 1980 1982 1984

Lead

con

sum

ed in

gas

olin

e (1

000

tons

)

0

0.25

0.5

0.75

1

1.25

1.5

Com

posi

te m

axim

um q

uarte

rly a

vera

ge le

ad

leve

ls (m

g/m3

)

lead in gasoline Lead in air

9

10

11

12

13

14

15

16

17

1976 1977 1978 1979 1980 1981

Aver

age

BLL

(mg/

dl)

40

50

60

70

80

90

100

110

Tota

l lea

d us

ed p

er

6 no

nths

(100

0 to

ns)

BLL Lead used in Gasoline

Correlation betweengasoline lead contentand airborne leadlevels in the US

Case DescriptionCase DescriptionPb levels measurement in LebanonPb levels measurement in Lebanon

Air (μg/m3) Blood (μg/dl)

WHO 0.5-1 20

US 1.5 10

EU 2 NA

Australia 1.5 10

Canada 5 10

Regulations for Pb in air & blood (Hashisho and El-Fadel, 2001)

12.6

21.4

37

17.513.4

20.3 19.5

44

14.3 13.5

2215.8 15.8

0

20

40

60

Ret

ail S

hop

Mec

hani

cs

Rad

iato

r

Aut

obod

y

Car

ele

ctric

Grin

ding

/Iron

Car

lice

nce

plat

es

Smel

ting

Gas

Sta

tions

Car

pent

ry/W

ood

Fact

ory

Veh

icle

/Stre

et

Ave

rage

Workplace

Blo

od le

ad le

vels

(g/

dl)

Average Standard Deviation

BLL for working adults: Mean BLL = 15.8 μg/dl

BLL in 10-17 year old Mean BLL = 9.75 μg/dl for students; 11.36 for

working students; 13.54 for workers

0

20

40

60

<5 5-9 10-14 15-19 >=20BLL (μg/dL)

Perc

ent w

ithin

BLL

rang

e

Students Students/workers Workers

0

20

40

60

80

<5 5-9 10-14 15-16BLL (μg /dL)

Perc

ent

with

in B

LL ra

nge

BLL in 1-3 year old healthy children: Mean BLL = 9.75 μg/dl

Manual job of the father

Living in an area with traffic jams

Low income

Applied methodApplied methodImpact assessment & CBAImpact assessment & CBA

• Impact of use of leaded gasoline was examined: using dose-response functions for children & adults derived by the USEPA

• Cost benefit analysis of Pb phase out was conducted: using income-adjusted mortality and morbidity figures from US

Per capita income of country i = Xi



⇓


⇒ Multiply Yi by Xj/Xi ⇒ Value of mortality or morbidity outcome in country j = Yj

Applied methodApplied methodValuation techniquesValuation techniques

• Costs– Cost of switching consists of:

cost of lubricant additives, additional transportation costs to access new suppliers

– 0.01-0.02 USD/liter(worldwide experience)

– Not adjusted for income as driven by international market

– Distribution system adjustment costs are minor & can be neglected (same system is used but once-and-for-all cleaned, different pump nozzle size)

• Benefits– Mortality: WTP approach– Morbidity: COI approach– Morbidity & mortality values

estimated for the US in the year 1990 were used

– Figures were adjusted for• Inflation to the year 1998 (by

multiplication with an inflation factor of 1.2547)

• Income (using an income ratio of 0.12, as GNP in 1999 was 29340 USD for USA, & 3560 USD for Lebanon)

Applied MethodApplied MethodAssumptions Assumptions

• Population of major urban areas only are considered (about 1,639,000)

• Lebanese population age distribution• Population within each age group evenly distributed by age• BLLs independent of age in children and adults• Average BLL in women 64 percent of that in men• BLL in pregnant women the same as that in non-pregnant ones

80 60 40 20 0 20 40 60 80

0-4

10-14

20-24

30-34

40-44

50-54

60-64

70-74

>80

Age distribution per thousand personsMale Female

ResultsResultsHealth impact assessmentHealth impact assessment

• Tables show respectively health impact assessment in– Children: number of

avoided cases for a 77% reduction in BLL

– Adults: number of avoided cases for a 78% reduction in BLL

EffectNumber of cases

Average Range

Total IQ point loss (points) 42,689 35,176 - 48,495

Mental retardation (cases) 167 NA

Child Mortality (cases) 31 NA

Effect Age group

Number of cases

Mean Minimum Maximum

Hypertension men 20 - 74 1,688.3 NA NA

CHD men 40 - 59 54.6 15.3 133.0

60 - 64 24.1 10.0 482.6

65 - 74 30.7 3.3 152.4

CHD women 45 - 74 46.8 12.2 122.2

CA men 45- 74 42.7 7.7 162.5

CA women 45 - 74 22.9 4.6 77.3

BI men 45 - 74 26.5 3.8 127.4

BI women 45 - 74 16.8 3.0 63.3

Mortality men 40 - 54 50.9 21.1 0.01

55 - 64 36.1 6.8 121.2

65 - 74 12.9 5.4 78.0

Mortality women 45 - 74 25.0 5.3 75.8

ResultsResultsHealth benefitsHealth benefits

Health outcomeCost per case

(1998 USD)Cost (1998 USD) x 1000

Average RangeTotal IQ Loss 450 19,217.5 13,414.7 – 25,175.5

Mental retardation 8,023 1,069.3 NA

Child Mortality 730,754 22,427.0 NA

Total - 42,713,8 36,911.0 – 48,671.8

Effect Cost per case (1998 USD)

Average cost(1998 USD) x 1000

Cost range (1998 USD) x 1000

Hypertension men 104 175.0 NA

CHD men 7,916 867.5 226.3 – 6,079.7

CHD women 7,916 370.8 96.4 – 967.7

CA men 30,448 1,301.2 233.6 – 4,947.1

CA women 22,836 521.9 104.2 – 1,766.4

BI men 30,448 808.2 114.6 – 3,880.6

BI women 22,836 384.2 69.4 – 1,444.4

Mortality men 730,754 66,799.7 21,796.8 – 145,578.9

Mortality women 730,754 18,286.6 3,836.7 – 55,379.5

Total - 89,514.2 26,653.1 – 220,219.3

Children

Adults

ResultsResultsCar related benefitsCar related benefits

• In the US, car maintenance savings from the use of unleaded gasoline are about 0.003 to 0.024 USD/Liter

• Cost savings from improved fuel efficiencyare about 0.0024 USD/Liter

Car-related maintenance Leaded UnleadedSpark plug changes Every year Every other year

Oil changes and filter Twice per year One per year

Muffler replacements Twice per 5 years One per 5 years

Exhaust pipe replacements One per 5 years None

ResultsResultsCBA CBA –– Summary tableSummary table

Economic parameters

Costs(1998 million USD)

Benefits*(1998 million USD)

Capital investments 11.3 7.6 - 15.1 NA NAHealth

Not applicable

132.2 63.6 - 268.9Car maintenance 1.4 0.3 - 2.4 Energy efficiency 0.3 NATotal per capita benefit (USD/capita) 35.9 17.2 - 72. 9

LEAD PHASE OUT IS ECONOMICALLY HIGHLY FEASIBLE

*Previous values were adjusted for income


Thank YouThank You

PM in urban areas: healthPM in urban areas: health--based economic based economic

assessment, Lebanonassessment, Lebanon

Case DescriptionBackground

• PM is recognized as most important air pollutant in terms of health impacts, reportedly associated with:

– Increase in cardiac and respiratory mortality;– Decrease in levels of pulmonary lung function in children and adults;– Increase in daily prevalence of respiratory symptoms in children and adults;– Increase in functional limitations (school absenteeism, restricted activity days);– Increase in physician and emergency department visits for asthma and other

respiratory conditions.

International Standard Long-term (μg/m3) Short-term (μg/m3)1

PM10 BS TSP PM10 BS TSP

EU limit values NA2 80 150 NA 250 300

EU guide values NA 40-60 NA NA 100-150 NA

USEPA 50 NA 260 150 NA 75

WHO guidelines NA 40-60 60-90 NA 100-150 150-230

WHO guidelines for Europe NA 50 NA 70 125 120

1 24 hours2 NA = not available

Case DescriptionCase DescriptionStudy area - Lebanon

• Air quality measurements showed that:– TSP concentrations range

from 102 to 291 μg/m3 (with an average of 166 μg/m3)

– High levels– Contributed by:

• Vehicle-induced emissions• Vehicle movement on dusty

roads• On-going construction

activities• Anthropogenic sources• Dry climate (→ high dust

levels in the atmosphere)

0 1 2 3Km

SAMPLING LOCATION

NORTH

NORTH

9

87

26

131 4

1 2

3

1110

212019

17

18

1

22

5

4

16

MED

ITE

RR

AN

EA

N S

EA

0

100

200

300

1 2 3 4 6 8 9 10 11 12 14 16 17 18 19 20 21 22

TSP,

g/

m3

Average166 μg/m3

USEPA 24-hr standard

Applied methodApplied methodHealth assessment & valuationHealth assessment & valuation

• Assessment of health impacts– Average levels of ambient concentrations are related

to health effects through dose-response functions(DRFs)

– DRFs are related to the population at risk• Assessment of associated monetary value:

unit economic values are applied to the cases avoided by a 10 μg/m3 reduction in PM10 values using the following approaches– Human capital approach– COI approach: does not include the less tabgible

impacts to the individual’s well-being– WTP approach

Applied methodApplied methodWorldwide health impacts DRFsWorldwide health impacts DRFs

Particulate Percent increasein mortality

Percent increasein morbidity

Morbidity Type

Increase of 10 μg/m3 in PM10 0.1-4.6 0.2-2.90.8-11.50.2-6.40.6-1.20.4-6.00.3-0.41.1-24.90.4-13.01.6-17.6

Pneumonia hospital admissionsCOPD1 hospital admissionsRespiratory hospital admissionsCardiac hospital admissionsEmergency cases of asthmaBronchitis hospital admissionsLRI2 symptomsURI3 symptomsCough symptoms

Increase of 10 μg/m3 in PM2.5 0.4-3.7 0.41-24.63.7-20.9

Respiratory hospital admissionsAsthma hospital admissions

Increase of 10 μg/m3 in BS NR4 0.07-18.20.3-5.31.2-16.5

Respiratory hospital admissionsAsthma hospital admissionsCOPD hospital admissions

Increase of 100 μg/m3 in TSP 3.3-8.3 NR NR1 COPD = chronic obstructive pulmonary disease2 LRI = lower respiratory illness3 URI = upper respiratory illness4 NR = not reported

Applied methodApplied methodData for mortality assessment in Data for mortality assessment in

Lebanon Lebanon –– HC approachHC approach

Sex1 Age group2

0-9 10-19 20-39 40-59 60-69 >70 Unknown TotalMale (51) 5.54 2.67 9.60 18.31 21.35 37.48 5.07 100.00

Female (49) 6.50 2.91 4.00 13.30 17.18 51.13 5.00 100.00

Total (100) 5.93 2.77 7.31 16.27 19.66 43.03 5.03 100.00

Sex Age group0-9 10-19 20-39 40-59 60-69 >70 Unknown Total

Male (51) 0-17 0-8 1-30 1-58 1-67 3-118 0-16 6-314

Female (49) 0-20 0-9 0-12 1-40 1-52 3-155 0-15 5-303

Total (100) 0-37 0-17 1-42 2-98 2-119 6-273 0-31 11-617

Percent distribution of death in Lebanese households

Distribution of predicted lives saved per year for a 10 μg/m3 reduction in PM10

The annual death rate of 8.2 deaths/1000 persons was multiplied by the urban Lebanese population of 1.64 million → 13440 deaths multiplied by the percent distribution of Lebanese population by age

group and sex

Applied methodApplied methodData for mortality assessment in Data for mortality assessment in

Lebanon Lebanon –– WTP approachWTP approach

Study 1 Valuation/ case(MUS$/yr)2

Study Valuation/ case(MUS$/yr)

R. S. Smith 1974 7.2 Herzog & Schlottman 1987 9.1

R. S. Smith 1976 4.6 Leigh 1987 10.4

V.K. Smith 1976 4.7 Gerking et al 1988. 3.6

Viscusi 1978 4.1 Moore and Viscusi 1988 2.5

Olson 1981 5.2 Moore and Viscusi 1988 7.3

Viscusi 1981 6.5 Gaten 1988 13.5

Marin et al. 1982 2.8 Cousineau et al. 1988 3.6

Butler 1983 1.1 Jones-Lee3 1989 3.8

Leigh and Folson 1984 9.7 Kneisner and Leeth 1991 0.6-7.6

Smith and Gilbert 1984 0.7 Miller and Guria3 1991 1.2

Dillingham 1985 0.9-3.9 Viscusi et al. 1991 2.7

Gegax et al.3 1985 3.31 Labor market estimate; 2 1990 dollar value; 3 Contingent valuation estimate

Values from US studies; to be adjusted for country GNP

Applied methodApplied methodData for morbidity assessment in Data for morbidity assessment in

Lebanon Lebanon –– COI approachCOI approachTotal hospital admissions per year

LebanonBeirutOther urban areas

400,000133,000

53,200

Type of hospital admission per year

Emergency visits in BeirutEmergency visits in other urban areasa. Respiratory and cardiac hospital admissions (%)b Respiratory admissions (% of a)c. COPD admissions (% of b)d. Pneumonia admissions (% of b)

145,00058,000

15373763

Percent decrease in morbidity due to 10 μg/m3 reduction in PM10

Pneumonia hospital admissionsCOPD hospital admissionsEmergency visits

0.2-2.90.8-11.50.3-12.6

Endpoint Change Total Cases avoided

Respiratory hospital admissions/100,000 6.6-17.3 108-284

Emergency department visits/100,000 116.0-354.0 1,902-5,806

Lower respiratory illness/child/asthmatic 0.010-0.024 4,756-11,414

Asthma attacks/person 0.33-1.96 541,200-3,214,400

Respiratory symptoms/person 0.8-2.56 1,312,000-4,198,400

Chronic bronchitis/100,000 30.0-93.0 492-1,525

Restricted activity days/person 0.29-0.58 475,600-951,200

Number of hospital admissions per health endpoint (as obtained from literature) was multiplied by the % decrease in endpoint hospital admissions due to a 10 μg/m3

reduction in PM10

Applied methodApplied methodData for morbidity assessment in Data for morbidity assessment in

Lebanon Lebanon –– WTP approachWTP approachValues from US studies; to be adjusted for country health care rates

1 1990 dollar value (USEPA, 1997)2 Lebanese valuation is obtained by multiplying the US valuation by the health care ratio (0.24)3 Lebanese valuation is obtained by multiplying the US valuation by the per capita GNP ratio (0.1)

Endpoint US Valuation (US$ / case)1

Hospital admission2

COPD 8,100

Pneumonia 7,900

All respiratory 6,100

Respiratory illness or symptom3

Chronic bronchitis 260,000

Acute bronchitis 45

Acute asthma 32

Acute respiratory symptoms 18

Upper respiratory symptoms 19

Lower respiratory symptoms 12

Restricted activity day3

Work loss days 83

Mild restricted activity days 38

Applied MethodApplied MethodAssumptions Assumptions

• There is no threshold below which PM10 is harmless or not a cause of mortality.

• No difference in susceptibility or exposure between different populations.

• Reviewed studies are of similar quality and need not be weighted for difference in methodology or sample size.

• Where an age-specific DRF is unavailable, the estimate for all age groups will be applied to the baseline number of deaths in each age group.

• The estimations are not restricted to a particular or an average value but ranges of values are considered in order to ensure a broader perspective of the subject.

ResultsResultsMortality benefits for 10 Mortality benefits for 10 μμg/mg/m33

reduction in PMreduction in PM1010

• Human capital approach– By multiplying average productivity years (25 to 69) by

average monthly salary (400 USD in 1998)– Average benefit per case is 0.055 million USD

Age group

# of lives saved

Average productivity years

Benefit(MUS$/yr)

40-59 2-102 20 0.2-9.7

60-69 3-122 5 0.07-2.9

Total 0.27-12.6

Average per case1 0.0551 Average total economic benefit divided by the average total number of

lives saved

HC Approach• WTP approach

– Values from US studies were adjusted for country GNP using a ratio of 0.1

– Average benefit per case ranges between 0.06 and 1.35 million USD

ResultsResultsMorbidity benefits for 10 Morbidity benefits for 10 μμg/mg/m33

reduction in PMreduction in PM1010• Cost of illness

– By multiplying # of cases avoided by corresponding health endpoint cost

– High benefits can be generated

Endpoint Avergae Stay(Days) 1

Average cost(US$/day) 1

Economic benefit(MUS$/yr)

COPD 6.6 261 0.06-0.9

Pneumonia 10 207 0.03-0.4

Emergency visit - 76 0.05-1.9

Total 0.14-3.21 Based on survey data from the American University Hospital

and insurance companies (Djoundourian et al., 1998)

ResultsResultsMorbidity benefits for 10 Morbidity benefits for 10 μμg/mg/m33

reduction in PMreduction in PM1010• WTP approach

– US values were adjusted using a health care ratio of 0.24 (obtained by comparing cost of COPD & pneumonia in Lebanon & US; higher than income ratio → health care is expensive in Lebanon)

Endpoint US Valuation (US$ per case)1 Lebanese Valuation (US$ per case)

Hospital admission2

COPD 8,100 1,944

Pneumonia 7,900 1,896

All respiratory 6,100 1,464

Respiratory illness or symptom3

Chronic bronchitis 260,000 26,000

Acute bronchitis 45 5

Acute asthma 32 3

Acute respiratory symptoms 18 2

Upper respiratory symptoms 19 2

Lower respiratory symptoms 12 1

Restricted activity day3

Work loss days 83 8

Mild restricted activity days 38 4

1 1990 dollar value (USEPA, 1997b)

2 Lebanese valuation is obtained by multiplying the US valuation by the health care ratio (0.24)

3 Lebanese valuation is obtained by multiplying the US valuation by the per capita GNP ratio (0.1)

ResultsResultsSummary table Summary table

Endpoint Number of cases avoided

Total Economic benefit(MUS$/yr)

COI1 WTP2

Mortality3 11-617 0.27-12.6 3.5-157.9

All COPD 31-441 0.06-0.9 0.98-13.9

All Pneumonia 13-189 0.03-0.4 0.05-0.7

Emergency visits 609-25,578 0.05-1.9 NA4

Total 0.41-15.8 4.53-172.5

Percent of GDP5 0.003-0.1 0.03-1.06

Percent of adjusted GDP6 0.03-1 0.3-10.61 COI = cost of illness; 2 WTP = willingness to pay; 3 Human capital approach; 4 NA = not available; 5 World Bank, 19986 Adjusted GDP assuming that the construction and transportation sectors are the main sources of particulate emissions in urban areas and

accounting for source/sector contribution to GDP and percent of urban population exposed as compared to total country population.

Benefits of reducing PM concentration in the air can be significant; they are dominated by mortality valuation although the number of

mortality cases is relatively small.


Thank YouThank You

Economic Benefits of Economic Benefits of Reducing Particulate and Reducing Particulate and

Sulfate Emissions from the Sulfate Emissions from the Cement Industry, LebanonCement Industry, Lebanon

0

1

2

3

4

5

1993 1994 1995 1996 1997 1998 1999 2000

Cem

ent d

eliv

erie

s (M

illio

n to

ns)

0 200 400 600 800 1000 1200 1400

USA, 1995

Switzerland, 1995

Norway, 1995

New Zealand, 1995

Lebanon, 1999

Lebanon, 1995

Latvia, 1995

Jordan, 1995

Iceland, 1995

Germany, 1995

Egypt, 1999

Canada, 1995

Australia, 1995

Cou

ntry

Cement production (106 ktons/capita/year )

Case DescriptionStudy area - Lebanon

• Cement industrial complex, Chekka, Lebanon

• 60% of country cement production• Over years, production peaked in

1995; Lebanon had one of the highest per capita production worldwide

• Contributes 77.2% of country’s industrial emissions

• Chronic public complaints• Media coverage suggested a

correlation between emissions & adverse health impacts

Applied MethodApplied MethodHealth & air quality assessmentHealth & air quality assessment

• Health– Questionnaire related to health outcomes– 159 randomly selected households in 2 groups

• Group 1: 0 to 4 km from Chekka, 80 households• Group 2: 4 to 7 km from Chekka, 79 households• More focus on children than on adults

• Air quality– Continuous 5-min measurement of CO, NO2 & SO2 (4

locations, 4 to 7 days)– Gravimetric sampling of PM10 & less + analysis for

priority metals

Applied MethodApplied MethodHealth impact & economic valuationHealth impact & economic valuation

• Health impact assessment– Benefits due to reduction of PM10 & SO2 in the form of SO4

2-

were estimated based on studies developed for the US– Potential lives saved = % change in mortality or respiratory

systems avoided due to reduction x local occurrence rates x the exposed population number

• Economic valuation– Mortality: HC approach– Morbidity: COI approach, using US values adjusted for:

• Inflation (1990 → 1998): factor of 1.2547• Income: ratio of 0.12 (based on GNP comparison)• Health care: ratio of 0.24 (based on comparison of cost of COPD &

pneumonia in Lebanon & US)

ResultsResultsHealth & air quality Health & air quality

• Higher prevalence of health complaints among Group 1

• Exceedence of 24-hr AAQS at all locations

Outcomes Group1 (%)

Group2 (%)

Chronic bronchitis 15 11

Asthma 10 6

Other respiratory problems 15 6

Hospitalization during the past year

Respiratory 11 9

Non-respiratory 59 30

Cough with colds 74 43

Cough apart from colds 13 6

Chronic cough 10 6

Phlegm & congestion with colds 36 20

Phlegm & congestion apart from colds 8 3

Chronic phlegm & congestion 6 1

Episodes of increased phlegm & congestion for 1 week/year 23 5

Chest colds every year 19 4

Wheezing 53 32

Chest illness that kept child off his activities in the past 3 years 25 11

Asthma 4 1

Bronchitis 10 8

Chest illness before age 2 years 9 5

Hospitalization due to chest illness before age 2 3 3

Constituent Concentration range

AAQS

CO (ppm) < 2 9 (8-hour average)

NO2 (ppm) 6.4 - 10.1 0.053 (annual average)

TSP (μg/m3) 67 - 316 260 (24-hour average)b

75 (annual average)b

PM10 (μg/m3) 36.8 – 173.8 b 150 (24-hour average)

50 (annual average)b

SO2 (ppm) 0.45 - 0.7 0.14 (24-hour average)

0.03 (annual average)

ResultsResultsHealth benefits Health benefits –– PMPM1010 reductionreduction

• For 10 μg/m3

reduction in PM10– Mortality

• Size of exposed population was multiplied by death rate

• A 0.1 to 4.6% decrease in mortality was applied

• 0 to 34 lives saved– Morbidity

• DRFs from industrialized regions were multiplied by exposed population

Endpoint Change(Pearce &

Croward, 1996)

Total cases avoided

Respiratory hospital admissions (RHA)/100,000

6.6-17.3 1 – 16

Emergency department visits/100,000

116.0-354.0 33 – 319

Lower respiratory illness (LRI)/ asthmatic child

0.010-0.024 10 – 78

Asthma attacks/person

0.33-1.96 9,652 –176,577

Respiratory symptoms/person

0.8-2.56 23,400 –230,630

Chronic bronchitis (CB)/100,000 (incidence)

30.0-93.0 8 – 84

Restricted activity days (RAD)/person

0.29-0.58 8,482 –52,253

Morbidity Results

ResultsResultsHealth benefits Health benefits –– SO4

2- reduction reduction • For 10 μg/m3 reduction in

SO42-

– Mortality• Expressed by the equation

(Chestnut and HaglerBailly Consulting, 1995): LΔS = a x P x ΔSLΔS = annual # of lives saved P = exposed peopleΔS = change in concentrationa = constant (8-112)x10-6

• 2 to 101 lives saved– Morbidity

• Similar equation with different ‘a’ values

Endpoint a Number of cases avoided

CB (0.5 - 2.0)x10-4 14-181

RHA (1.3 - 1.8)x10-5 3-17

CHA (cardiac hospital admissions)

(1.0 - 1.7)x10-5 2-16

ASD (asthma symptom days)

(3.3 - 9.9)x10-1 4,536-41,919

RAD* (4.7 - 14.6)x10-2 8,523-81,550

LRS* (6.6 - 23.0)x10-2 11,969-128,469

Morbidity Results

* Cases of RAD and LRS are computed using adults 18 years of age and above as the exposed population, which constitutes around 62 percent of the total population (US Bureau of the Census, 2001)

Applied MethodApplied MethodMortality related benefitsMortality related benefits

• Multiplying average productivity years (20 to 35 years) by the average income (US$313/month)

• Economic benefit for a 10 μg/m3 reduction in PM10 & SO4

2- is equivalent to 0 to 0.027% & 0 to 0.08 % of the GDP, respectively

PM10 SO42-

Number of lives saved

Economic benefit (US$/year)

Number of lives saved

Economic benefit (US$/year)

Total economic benefit 0 – 34 0 – 4,469,640 2-101 150,240 – 13,277,460

Equivalent percent of GDP 0 – 0.027 0 – 0.08

Average per case* 131,460 131,640* Average total economic benefit divided by the average total number of lives saved

Applied MethodApplied MethodMorbidity related benefits (1)Morbidity related benefits (1)

• Multiplying predicted # of cases avoided due to reduction in pollutant concentrations by cost of corresponding health endpoints

• To avoid double counting in adding up costs, overlapping categories were subtracted assuming the following:– Incidence is proportional to age distribution: 62% of the

population is 18 years and older;– Each RHA averages 6.8 days & each CHA averages 6.9 days;– All days in the hospital and all asthma symptom days are also

RADs & therefore are subtracted from the latter;– RADs are also acute respiratory symptom days & therefore a

fraction of RADs is subtracted from LRSs;– 28% of acute respiratory symptoms are lower respiratory tract

Applied Method Applied Method Morbidity related benefits (2)Morbidity related benefits (2)

Endpoint US valuation (US$/case)a

Lebanesevaluation

(US$/case)

PM10 SO4

Cases avoided

Total cost (US$) Cases avoided

Total cost (US$)

RHA (Respiratory hospital admissions)

6,100 1,837b 1 – 16 1,837 – 29,392 3 – 17 5,511 – 31,229

CHA (Cardiac hospital admissions)

6,100 1,837b - - 2 – 16 3,674 – 29,392

CB (Chronic bronchitis) 260,000 32,622c 8 – 84 260,976 – 2,740,248 14 – 181 456,708 – 5,904,582

ASD (Asthma symptom days)

32 4c 9,652 –176,577

38,608 – 706,308 4,536 –41,919

18,144 – 167,676

Net RAD (Restricted activity days)

38 4.8c 2,493 –11,685

11,966 – 56,088 5,689 –55,420

27,307 – 266,016

Net LRS (Lower respiratory symptoms)

12 1.5c 4,712 –18,296

7,068 – 27,444 9,582 –105,635

14,373 – 158,453

Total 320,455 – 3,559,480 525,717 – 6,557,348

Per capita benefit (US$/capita)

3.5 – 122 5.8 – 224

Percent of per capita incomed

0.09 – 3.16 0.15 – 5

a 1990 Dollars (USEPA, 1997)b Lebanese valuation is obtained by multiplying the US valuation by the health care ratio (0.24), after adjustment for inflation.c Lebanese valuation is obtained by multiplying the US valuation by the per capita GNP ratio (0.1), after adjustment for inflation.d GDP = 16.6 billion US$; per capita capita = US$ 3860 (World Bank, 2001).

EEnd of nd of SSessions essions 14 14 && 1515

Thank YouThank You

A health-based socio-economicassessment of drinking waterquality: the case of LebanonM. El-Fadel, R. Maroun, L. Semerjian and H. HarajliDepartment of Civil and Environmental Engineering, American

University of Beirut, Beirut, Lebanon

Keywords Water, Health, Economics, Lebanon

Abstract Water-related diseases are a human tragedy, resulting in millions of deaths each year,preventing millions more from leading healthy lives, and undermining development efforts byburdening the society with substantial socio-economic costs. This problem is of great significance indeveloping countries, where polluted water, water shortages, and unsanitary living conditionsprevail. This paper presents a case study on a health-based socio-economic assessment of drinkingwater quality in Lebanon, based on relevant valuation approaches and available country-specificdata. The assessment revealed that the potential health and economic benefits due to water andsanitation improvements can be significant (0.15-3.35 percent of GDP).

IntroductionHistorically, the provision of urban environmental services in general, andpiped water in particular, has been the responsibility of the public sector.However, due to the rapid increase in urban populations, governments are oftenfacing major difficulties in meeting the citizens’ basic needs (Bennet, 1998).This issue is of major concern in developing countries, where existingconditions of the water supply infrastructure is poor, services are inferior, andfinancial resources for the construction and maintenance of infrastructure areinadequate (Gidman et al., 1995). In this context, theWorld Health Organization(WHO) estimates that 20-30 percent of urban residents in Latin America,Africa, and Asia lack access to potable water. Experience during the lastdecade confirms that the solution to these problems is not merely to expandcapacity, but rather to better manage service delivery to meet user’s demand,via the establishment of public-private partnerships (PPP) (Gidman et al., 1995).Four groups can be identified to play a role in PPPs:

(1) the government at the national, regional, or local level;

(2) the formal private sector;

(3) non-governmental organizations/community-based organizations; and

(4) the informal private sector.

The Emerald Research Register for this journal is available at The current issue and full text archive of this journal is available at

http://www.emeraldinsight.com/researchregister http://www.emeraldinsight.com/1477-7835.htm

Special thanks are extended to the United States Agency for International Development for itssupport to the Environmental Engineering and Science Programs at the American University ofBeirut.

Drinking waterquality

353

Management of EnvironmentalQuality: An International Journal

Vol. 14 No. 3, 2003pp. 353-368

q MCB UP Limited1477-7835

DOI 10.1108/14777830310479441

In recent years, the role of the informal private sector has been gaining muchattention, because it is seen to be most accountable to the low-incomepopulation. More specifically, the informal private sector is widely involved inthe provision of water through what is commonly known as “water vending”,which is defined as the act of providing water through tanker trucks or mobilewater vendors, stand-pipes, and water shops, with the exclusion of the “bottledwater” industry (World Health Organization and United Nations ChildrenFund, 2000). In most developing countries where public water and sanitationnetworks are not trusted or are altogether absent, consumers resort toalternative sources of freshwater, such as small-scale water vendors. As such,water vending is an old tradition worldwide, and in some African cities, forinstance, it has become the major mode of access to drinking water (see Table I).

The proportion of the population served through vendors and tanker trucksvaries significantly between different urban and rural areas, with urbanpopulations being the largest consumers (see Table II). In Lebanon, watersupply through vendors is becoming substantial, particularly in the southern

CityPublic connection

(%)Private standpipes

(%)Other private providers

(%)

Kampala-Uganda 36 5 59Dar Es Salam-Tanzania 31 0 69Conakry-Guinea 29 3 68Nouakchott-Mauritania 19 30 51Cotonou-Benin 27 0 73Ouagadougou-Burkina Faso 23 49 28Bamako-Mali 17 19 64

Source: Collignon and Vezina (2000)

Table I.Mode of access todrinking water inselected Africancities in 1999

Country Year Source of water Urban population (%) Rural population (%)

Angola 1996 Tanker truck 25.2 0.8Cambodia 1998 Vendor 16 3.5Chad 1997 Vendor 31.5 0.5Ecuador 1990 Tanker truck 16 7Eritrea 1995 Tanker truck 30.5 1.4Jordan 1997 Tanker truck 1 10.6Libya 1995 Tanker truck 6.8 13.9Mauritania 1996 Vendor 53 0.9Mongolia 1996 Vendor 16 1Niger 1998 Vendor 2.4 1.9Syria 1997 Tanker truck 4.1 11.3

Source: World Health Organization and UNICEF United Nations Children Fund (2000)

Table II.Percentage ofselected populationconsuming vendedwater

MEQ14,3

354

suburbs of Beirut, as a response to the inadequate public water supply system.Yet, it is difficult to estimate the exact numbers and locations of thesesmall-scale entrepreneurs due to the absence of official registries, and to theillegal status of the majority (Al-Safir Newspaper, 2002).

Recently, water vending is being acknowledged more by researchers andpolicy makers, and linked with an increased interest in cost recovery andprivatization. However, the main concern regarding this service is thequality of the supplied water and its associated health risks. In fact, scarceand unclean water supplies represent critical public health problems inmuch of the world. Polluted water, water shortages, and unsanitary livingconditions are associated with various short- and long-term health risks.Water-related diseases are indeed a human tragedy, whereby about 2.3billion cases are estimated with over 12 million deaths per year.Furthermore, some 60 percent of all infant mortality is linked to infectiousand parasitic diseases, most of them are water-related (Hinrichsen et al.,1997). In most countries, the main risks to human health associated withthe consumption of polluted water are microbiological in nature. However,the importance of chemical contamination should not be underestimated(World Health Organization, 1997). Table III summarizes majorwater-related diseases prevalent worldwide.

The present study assesses the problems associated with unclean andinadequate water supply in Lebanon. The quality of vended water at a typicalarea (Beirut southern suburbs) where the informal sector plays a pivotal role inproviding drinking water to the local population is presented. Water-relatedhealth impacts in terms of increased mortality and morbidity rates areassessed, and corresponding socio-economic burdens are estimated.

The Lebanese contextHistorically, Lebanon has always been distinguished in the region to be blessedwith relatively adequate water resources; however, available resources areunevenly distributed geographically and seasonally (El-Fadel et al., 2000),leading to problems of water shortage that are prominent in over-populatedcities, such as the capital Beirut and its suburbs (Yamout, 2002). Beyond theimpact of population growth, water demand has been rising in response toindustrial development, increased reliance on irrigated agriculture, massiveurbanization, and rising living standards. On the other hand, high“unaccounted-for water” (UFW) volumes, estimated at an average of 40percent, further aggravate the problem of water shortage as well as waterquality (Environmental Resource Management, 1995). Recent attempts atestimating water demands, availability, and deficit for Beirut reveal asignificant water deficit in the city (see Table IV).

Evidently, the water supply situation in Beirut cannot meet the demand. Thecombination of weak water utilities and water scarcity at the source leads tointermittent supply in most areas (10 hours of supply every other day), and to a


355

Disease

Geographicextent

Number

ofcasesa

Deathsper

year

Water-bornediseases

Diarrhealdisease

Worldwide

500million/year

3-4million

Cholera

South

America,Asia,Africa

384,000/year

20,000

HepatitisA

Worldwide

600,000–3million/year

2,400-12,000

Paratyphoidandtyphoid

80percentin

Asia,20

percentin

Latin

America,

Africa

16million

currently

600,000

Polio

66percentin

India,34percentin

NearEast,Asia,

Africa

82,000

currently

9,000

Water-baseddiseases

Ascariasis

Africa,Asia,Latin

America

250million

currently

60,000

Clonorchiasis

Southeast

Asia

28million

currently

Nonereported

Dracunculiasis(guinea

worm)

78percentSudan,22percentinotherSub-Saharan

Africa,andfew

casesin

India

andYem

en153,000/year

Nonereported

Paragonim

iasis

Far

East,Latin

America

5million

currently

Nonereported

Schistosomiasis(bilharzias)

Africa,NearEast,rain

forest

beltin

Central

Africa,Western

Pacific,Cam

bodia,Laos

200million

currently

20,000

Water-relatedvector

diseases

Dengue

Tropical

environments,concentrated

inAsia,

Central

andSouth

America

50-100

million/year

24,000

Filariasis

Africa,Eastern

Mediterranean,Asia,South

America

120million

currently

Nonereported

Malaria

Africa,Southeast

Asia,India,South

America

300-500million/year(clinical)

2million

Onchocerciasis

Sub-Saharan

Africa,Latin

America

18million

currently

Nonereportedb

RiftValleyFever

(RVF)

Sub-Saharan

Africa

NAc

1percentof

cases

Notes:a

Number

ofcasesisreportedas

incidence

(“per

year”)–thenumber

ofnew

casesoccurringin

ayear–or

asprevalence

(“currently”)–

thenumber

ofcasesexistingat

apointin

time

bNodeathsbutcauses270,000reportedcasesof

blindnessannually

cNA¼not

available

Source:Muller

andMorera(1994);World

HealthOrganization(1996,1998)

Table III.Major water-relateddiseases

MEQ14,3

356

lack of piped water supply for a large number in poor urban areas (Saghir et al.,2000). It is estimated that about 22 percent of the total population are not evenconnected to the public water supply system, with the highest proportion ofunconnected households being located in the southern suburbs (Council forDevelopment and Reconstruction, 1998; Central Administration of Statistics,1997). In response to the shortage/absence of public piped-water supply,consumers in some quarters of Beirut, especially in the southern suburbs, relyon their own resources to meet their water requirements. Many resort topumping water from private wells, or purchasing water from small-scaleprivate water providers, more commonly known as water vendors. While it isdifficult to estimate the exact numbers and locations of these small-scaleentrepreneurs in the absence of official registries, the majority operates withoutany regulatory authorization (Al-Safir Newspaper, 2002).

Water quality indicatorsRecent statistics reported by the Central Laboratory, affiliated with theMinistry of Health, revealed the microbiological contamination of 24 percent of403 samples collected from water vending companies and 40 percent of 863samples collected from the potable water network in various regions acrossLebanon[1]. Moreover, of 450 samples collected from groundwater and springs,37 percent were microbiologically contaminated (see Table V).

As part of the present study, water quality from 65 water-vending shopslocated in the Beirut southern suburbs was examined. The analysis revealedthe poor quality of some samples, particularly with respect to microbiologicalindicators (see Figure 1).

The statistical significance of the sample size is limited, as the number ofsamples and the frequency of sampling need to be increased and spread over awider geographical area to draw more accurate conclusions and ascertain

1995 2000 2005 2015

Beirut City population 600,000 608,000 612,000 628,000Northern suburbs population 546,000 591,000 637,000 695,000Southern suburbs population 698,000 820,000 939,000 1,258,000Total population 1,844,000 2,019,000 2,188,000 2,581,000Domestic water demand (Ml/d) 371 418 466 580Non-domestic water demand (Ml/d) 121 134 149 174Water losses (Ml/d) 266 241 212 208Total consumption & loss (Ml/d) 758 793 827 962Minimum water available (Ml/d)a 315 575b 575b 575b

Maximum water available (Ml/d)a 415 675b 675b 675b

Maximum water deficit (Ml/d) 443 218 252 387

Notes: a Water resources vary throughout the yearb Includes the planned construction of a water supply reservoir south of the citySource: Council for Development and Reconstruction (1998)

Table IV.Predicted maximum

population, waterdemand,

availability, anddeficit for Beirut

City


357

trends within some statistical confidence. Nevertheless, these data arealarming, owing to the fact that polluted water is associated with various shortand long-term health risks, which translate into a socio-economic burden.

Health impactsSimilar to worldwide trends, Lebanon suffers from adverse health impacts as aresult of water pollution. Data pertaining to water-related mortality andmorbidity in the country are limited due to the absence of a proper diseasereporting mechanism. Available data are restricted to prevalent knownwater-related diseases, including diarrhea, typhoid and paratyphoid, andhepatitis A. In terms of mortality, a study conducted by the United NationsDevelopment Program (UNDP) in 1990 stated that each child under five isexposed, on average, to 3.5 incidents of diarrhea each year, causing the death of750 children per year (United Nations Development Program, 1995). While morerecent data are unavailable today, this value may be an over-estimation,especially now that substantial efforts for the improvement of water supply andsanitation have been ongoing. As for morbidity, the average annual number ofreported incidents of dysentery, hepatitis A, and typhoid and paratyphoid forthe years 1995 to 2000, as compiled by the Directorate of Preventive Medicine ofthe Ministry of Health (MoH), were 529, 287, and 809, respectively (see Figure 2).

Economic valuation of water-related health impactsThe health costs of water pollution and the benefits of improvements in waterand sanitation in Lebanon are assessed based on international experience and

Water sourceBottled water Network water Groundwater Total

Samples exhibitingmicrobiological pollution 97 (24 percent) 345 (24 percent) 450 (37 percent) 892 (36 percent)Total number of samplesanalyzed 403 863 1,215 2,481

Source: Al-Safir Newspaper (2002)

Table V.Results of wateranalysis by thecentral laboratory

Figure 1.Percentage/number ofsamples exceedingUSEPA drinking waterguidelines

MEQ14,3

358

available country-specific data. Refined estimates can be obtained withelaborate and updated local data such as statistics about mortality andmorbidity and their corresponding unit costs.

Health-related costs of water pollutionGenerally, the economic valuation of health impacts proceeds by conductingepidemiological studies in order to establish dose-response relationships(DRRs) linking environmental variables with observable health effects.However, in the case of water pollution, establishing dose-response functions(DRFs) is complicated and less advanced than is the case of evaluating healthimpacts from air pollution, for instance. This is due to the fact that “it is notambient water quality per se that affects health but access to clean drinkingwater and adequate sanitation along with the household’s level of income andeducation” (The World Bank Group, 1998).

Valuing the health costs of water pollution depends on country-specificfactors such as the cost of labor, labor productivity, capital and medical care,life expectancy, people’s value of health and life, and their willingness to acceptrisk. Hence, concrete valuation is limited by the availability of such data as wellas the level of uncertainty in the adopted approximations.

Valuation techniques. Several methods are available for valuing mortalityand morbidity costs associated with water pollution. These methods cangenerally be grouped in two categories:

(1) methods that measure only the loss of direct income such as lost wages;and

(2) approaches that attempt to capture the willingness to pay (WTP) ofindividuals for avoiding or reducing the risk of death or illness.

The first category does not include inconvenience, suffering, losses in leisureand other less tangible impacts to individual and family well being. They mayalso underestimate the health cost of people who are not members of the laborforce. Therefore, these methods indicate only the lower bound of social costs.

Figure 2.Average number of

reported cases per yearfor the period 1995-2000


359

Mortality valuation. Typical methods used for valuing mortality outcomesinclude the human capital approach and the willingness to pay (WTP)approach:

(1) The human capital approach. The human capital approach, also knownas the forgone-earnings approach, measures the loss of productivity (netpresent value of productivity) resulting from an individual’s death orinjury (Lesser et al., 1997). This approach considers individuals as unitsof human capital that produce goods and services for society. As such,the value of a premature death is obtained by multiplying the number oflife years lost by the average wage rate in the country in question. In thecase of Lebanon, the total socio-economic cost due to prematuremortality was estimated based on two main assumptions including:. The average Lebanese monthly salary is US$400[2].. Productivity age ranges between 20 and 65 years.

As such, considering that the mortality rate is 750 cases per year (UnitedNations Development Program, 1995), all pertaining to children less thanfive years of age, the total economic cost of premature death caused bywater pollution is estimated at US$162 million[3].

(2) The WTP approach. Unlike the human capital approach that measurestangible changes in productivity, the WTP approach captures intangibleaspects. It involves asking people directly through surveys – contingentvaluation studies – or assessing from market behavior, their WTP forreduced risks of increased mortality (or their willingness to accept theincreased risk of increased mortality). The total value of the benefit orcost in question is then estimated by averaging theWTP of all concernedmembers of society. In case original data on the WTP are lacking, onecan use income adjusted mortality values from other countries aftercompensating for income difference. In this study, individual WTPvalues estimated in the USA have been adopted (see Table VI), wherebyUS mortality values estimated for the year 1990 have been adjusted forthe Lebanese context based on income ratio as depicted in Figure 3 witha ratio of about 0.1 (World Bank, 1999).

Hence, on adjusting the range of WTP estimates (US$0.6-13.5 million) to theLebanese context, the value of a statistical life in Lebanon would range betweenUS$0.06-1.35 million. The corresponding total cost of mortality would rangefrom US$45 to 1,012.5 million, constituting 0.2 to 6.1 percent of the GDP inLebanon for the year 2000. This range encompasses the value calculated usingthe human capital approach (US$162 million).

Morbidity valuation. To assess the morbidity outcomes, the methodstypically used include the cost of illness (COI) approach, and the WTPapproach. The choice of the method to be used depends on the availability of

MEQ14,3

360

data. In this study, the cost of water-related diseases was estimated using theCOI approach, whereby the direct cost of morbidity is measured in terms ofmedical expenditure for treating an illness (physician care, drugs, andhospitalization costs), and in terms of lost wages during days missed fromwork, and other days when activities are significantly restricted due to illness.The diseases were restricted to dysentery, hepatitis A fever, typhoid andparatyphoid, being the only water-related illnesses reported to the Ministry ofHealth. The actual medical expenditure on hospitalization requirements forthese diseases and the associated number of days of restricted activity wasobtained through surveys of medical personnel. Transportation costs wereapproximated based on average round trip to hospital and associated cost offuel, while productivity loss was estimated by assuming equal age distributionbetween the reported cases and the Lebanese population (see Figure 4).

The total cost per case was found to vary with type and severity of diseaseas well as the class of hospital admission, as depicted in Table VII.

StudyaValuation/case(MUS$/year)b Studya

Valuation/case(MUS$/year)

R.S. Smith (1974) 7.2 Herzog and Schlottman (1987) 9.1R.S. Smith (1976) 4.6 Leigh (1987) 10.4V.K. Smith (1976) 4.7 Gerking et al. (1988) 3.6Viscusi (1978) 4.1 Moore and Viscussi (1988) 2.5Olson (1981) 5.2 Moore and Viscussi (1988) 7.3Viscusi (1981) 6.5 Gaten (1988) 13.5Marin et al. (1982) 2.8 Cousineau et al. (1988) 3.6Butler (1983) 1.1 Jones-Leec (1989) 3.8Leigh and Folson (1984) 9.7 Kneisner and Leeth (1991) 0.6-7.6Smith and Gilbert (1984) 0.7 Miller and Guriac (1991) 1.2Dillingham (1985) 0.9-3.9 Viscussi et al. (1991) 2.7Gegax et al.c (1985) 3.3

Notes: a Labor market estimateb 1990 dollar valuec Contingent valuation estimateSource: El-Fadel and Massoud (2000)

Table VI.Summary of

mortality valuationestimates in the

USA

Figure 3.Transfer of mortalityand morbidity values

across countries


361

Accordingly, the total COI incurred by society from the reported water-relatedcases is estimated to range between US$613,295 and US$2,664,502 (seeTable VIII). This provides a conservative estimate of the real costs as itexcludes the value of pain, suffering, diet, and behavior modification inaddition to the side effects of medications.

Benefits of improved water supply and sanitation. Providing clean water andensuring proper sanitation facilities have been shown to reduce the prevalenceof water-related diseases. Several studies examined the health benefitsassociated with water and sanitation interventions. Despite the mix of both

Number ofreported cases

Cost of illness per case(US$/case)

Total cost of illness(US$)

Dysentery 529 254-1,054 134,366-557,566Hepatitis A 287 389-1,822 111,643-522,914Typhoid 809 454-1,958 367,286-1,584,022Total 613,295-2,664,502% GDP 0.003-0.02

Table VIII.Total COI incurreddue to water-relatedillnesses

COI per admission class (US$/case)Social security Third class Second class First class

Dysentery 254-423 387-599 487-777 658-1,054Hepatitis A 389-836 544-1,076 683-1,371 902-1,822Typhoid 454-972 609-1,212 748-1,507 967-1,958

Table VII.Variation of thetotal COI per casewith type of diseaseand class ofadmission

Figure 4.Age distribution of theLebanese population

MEQ14,3

362

positive and negative results towards the improvements in one or morecomponents of water supply and sanitation, the overwhelming evidence is infavor of positive impacts, with the exception of hookworm infection. Expectedreductions in morbidity rates from improved water and sanitation for selectedwater-related diseases are summarized in Table IX. The mechanisms throughwhich improved water and sanitation can promote health vary from onedisease to another as depicted in Table X.

Economic valuation of health benefits associated with the provision of cleanwater supply and ensuring proper sanitation facilities can be estimated usingsimilar health costs derived in the previous section. This is calculated by firstmultiplying the reported number of cases by the expected percent reduction inprevalence for each disease (see Table IX) to obtain the number of casesavoided. Then, the economic benefit is estimated by multiplying the number ofcases avoided by the COI per case. Table XI outlines the application ofeconomic valuation of water-related health benefits in the context of Lebanon.

Intervention

DiseaseImproved

drinking waterWater for

domestic hygieneWater for

personal hygiene Sanitation

Ascariasis + ++ 2 ++Diarrhea + ++ ++ ++Dracunculiasis ++ 2 2 2Hookworm infection 2 2 2 ++Schistisomiasis 2 ++ ++ ++Trachoma 2 + ++ 2

Notes: + ¼ strong impact; ++ ¼ stronger impact; 2 ¼ little or no impactSource: Esrey et al. (1991)

Table X.Potential relationsbetween water and

sanitationinterventions andmorbidity from

selected diseases

Disease Percent reduction in morbidity

Diarrheal diseases 26-50Typhoid 80Paratyphoid 40Infective hepatitis 10Ascariasis 29-40Cholera 90Dracunculiasis 78Onchoserciasis 20?Hookworm infection 4Schistosomiasis 60-77Trachoma 27-60Guinea worm 100

Source: Esrey et al.(1991); Dougherty and Hall (1995)

Table IX.Expected reductions

in morbidity forselected diseasesfrom water and

sanitationimprovements


363

Hence, based on the reported percent reduction in cases, high economic benefitsare expected to result from improving the quality of the water supply andsanitation, ranging between US$25 and US$558 million per year, constituting0.15 to 3.35 percent of the GDP in Lebanon for the year 2000. In other words,around 55 percent of the incurred costs calculated above (US$45-1,0125 million)can be avoided. However, the true relation between the degree of improvementin water supply and sanitation and the expected reduction in diseaseprevalence and mortality still needs to be ascertained.

Limitations of the studyLimitations in this study can be classified into three categories:

(1) limitations in the available data;

(2) limitations in the health impact assessment; and

(3) limitations in the economic assessment.

Limitations in the available data, particularly data pertaining to morbidity andmortality rates, are mainly attributed to the absence of population-based vitaland disease registries. Limitations in the health impact assessment are mainlydue to the absence of DRFs for water-related health effects and the assumptionthat all reported diarrhea, typhoid, paratyphoid and hepatitis-A fever cases arewater-related. In contrast, underestimation in the number of cases is expected,since many remain unreported. The limitations in the economic assessmentresult from uncertainties in mortality valuations, the adjustment ofinternational values by the income ratio between countries, and theassumptions adopted in the absence of better estimates of specific datarelated to productivity loss, transportation costs, etc.

Summary and conclusionsScarce and unclean water supplies represent critical public health problems inmuch of the world, particularly in developing countries. In Lebanon, watershortages, weak public utilities, and poor management of available waterresources have forced the public to rely on water sources and supply practiceswhich are often polluted, posing significant health risks, welfare and financiallosses. In some quarters of Beirut, especially in its southern suburbs,consumers rely on small-scale private water providers, more commonlyreferred to as water vendors. An investigation of the quality of vended water inthis area revealed serious microbiological contamination.

Various valuation methods were used in the economic assessment of thehealth impacts of water pollution. The human capital and WTP approacheswere applied to estimate mortality costs, and the COI approach was used toestimate morbidity costs. The WTP values were adjusted by the income ratiobetween the USA and Lebanon. Estimated health costs ranged between US$45and US$1,025 million per year (0.2-6.1 percent of the GDP in the country for theyear 2000), dominated by mortality costs. Economic benefits of improvement in

MEQ14,3

364

Param

eter

Diarrhoeal

diseases

Typhoid/paratyphoid

Infectivehepatitis

Mortality

Total

Number

ofcasesa

529

809

287

750

2375

Percentreductionb(%

)26-50

60c

1055

d

Number

ofcasesavoided

e138-264

485

29412

Costper

case

(US$)f

254-1,054

454-1,958

389-1,822

(0.06-1.35)£

106

Econom

icbenefit(US$million/year)

0.035-0.278

0.220-0.949

0.011-0.053

25-557

25-558

Percentof

GDP

(0.2-1.7)£

1023

(1.3-5.7)£

1023

(0.06-0.3)

£10

23

0.15-3.34

0.15-3.35

Notes:aRefer

toFigure

2bRefer

toTableIX

c Averagepercentreduction

dReductionin

childmortality

from

water-related

diseasesas

reportedin

Esrey

etal.(1991)

e Number

ofcases£

percentreduction

f Refer

toTableVIII

Table XI.Estimated economicbenefit derived fromwater and sanitation

improvements


365

water and sanitation were approximated based on reduction in prevalence andmortality as reported in epidemiological studies. They are expected to reducethe incurred costs by 55 percent.

Based on these estimates, actions to improve the water and wastewaterinfrastructure are imperative. Economic health benefits should be incorporatedwithin the cost analysis of water and sanitation infrastructure projects. Thismust be coupled with proper quality/quantity water resources managementand planning policies to overcome the water problems and constraintscommonly encountered by governments. Strategies should emphasize properresource allocation and water quality (for example instituting water qualitymonitoring systems, protecting watersheds, regulating wastewater discharge,imposing sanctions for abuse and pollution) in addition to resourcedevelopment and increasing water quantity. While these strategies aremedium to long-term, immediate efforts should be directed towards managingthe water vending sector as an intermediary solution to water deprivation inlow-income areas.

Notes

1. Note that these samples were collected from houses and not from the network itself,indicating that the source of pollution could have occurred at any spot between the point ofdelivery from the network to the point of collection.

2. Based on the per capita GDP in Lebanon for the year 2000 (Banque Audi, 2000).

3. At the 2002 dollar value < 1,500 LP.

References

Al-Safir Newspaper (2002), “Official examinations confirm water pollution”, in Al-SafirNewspaper, 14 February (in Arabic).

Banque Audi (2000), Lebanon Facts and Figures 1998-2000, Banque Audi.

Bennet, E.B. (1998), Public-private Cooperation in the Delivery of Urban Infrastructure Services(Water and Waste), Yale-United Nations Development Program-Public PrivatePartnerships (UNDP-PPP).

Central Administration of Statistics (1997), Mount Lebanon in 1996: Census of Buildings andHouseholds, Central Administration of Statistics, Beirut, Lebanon.

Collignon, B. and Vezina, M. (2000), “Independent water and sanitation providers in Africancities: full report of a ten-country study”, Water and Sanitation Program, UNDP-WorldBank, Washington, DC.

Council for Development and Reconstruction (1998), Awali-Beirut Water Conveyor Project: PhaseI Interim EA Report, Project No. 1026, Council for Development and Reconstruction,Beirut, Lebanon.

Dougherty, T.C. and Hall, A.W. (1995), “Environmental impact assessment of irrigation anddrainage projects”, Irrigation and Drainage Paper 53, FAO, UK, available at:wwwfaoorg/docrep/V8350E/v8350e00htm

El-Fadel, M. and Massoud, M. (2000), “Particulate matter in urban areas: health based economicassessment”, The Science of Total Environment, Vol. 257 No. 2-3, pp. 133-46.

MEQ14,3

366

El-Fadel, M., Zeinati, M. and Jamali, D. (2000), “Water resources in Lebanon: characterization,water balance, and constraints”, Journal of Water Resources Development, Vol. 16 No. 4,pp. 619-42.

Environmental Resource Management (1995), Lebanon: Assessment of the State of theEnvironment, Mediterranean Environmental Technical Assistance Program (METAP)and Ministry of Environment (MoE), Beirut, Lebanon.

Esrey, S.A., Potash, J.B., Roberst, L. and Shiff, C. (1991), “Effects of improved water supply andsanitation on ascariasis, diarrhea, dracunculiasis, hookworm infection, schistosomiasisand trachoma”, Bulletin of the World Health Organization, Vol. 69 No. 5, pp. 609-21.

Gidman, P., Blore, I., Lorentzen, J. and Schuttenbelt, P. (1995), Public-private Partnerships inUrban Infrastructure Services, UMP Working Paper Series 4, UNDP/Habitat/World Bank,Nairobi.

Hinrichsen, D., Robey, B. and Upadhyay, U.D. (1997), Solutions for a Water-Short World,Population Reports, Series M, No. 14, Population Information Program, Johns HopkinsSchool of Public Health, Baltimore, MD.

Lesser, J.A., Dodds, D.E. and Zerbe, R.O. Jr (1997), Environmental Economics and Policy,Addison-Wesley, Reading, MA.

Ministry of Health (2000), Compilation of Lebanese Epidemiological Newsletter, EpiNews1995-2000.

Muller, R. and Morera, P. (1994), “Helminthoses”, in Lankinen, K.S., Berstrom, S., Makela, P.H.and Peltomaa, M. (Eds), Health and Disease in Developing Countries, Macmillan Press,London.

Saghir, J., Schiffler, M. and Woldu, M. (2000),Urban Water and Sanitation in the Middle East andNorth Africa Region: The Way Forward, World Bank, Middle East and North Africa.

United Nations Development Program (1995), Online, available: www.undp.org.lb/programme/governance/advocacy/hdr97/chp31.pdf

World Bank (1999), Lebanon at a glance, available at: www.worldbank.org/data/countrydata/aag/lbn_a ag.pdf

The World Bank Group (1998), Pollution Prevention and Abatement Handbook 1998: TowardCleaner Production, The World Bank Group, Washington, DC.

World Health Organization (1996), The World Health Report 1996: Fighting Disease, FosteringDevelopment, World Health Organization, Geneva.

World Health Organization (1997), Guidelines for Drinking-water Quality: Surveillance andControl of Community Supplies, Vol. 3, 2nd ed., World Health Organization, Geneva,available at: www.who.int/ water_sanitation_health/GDWQ/PDF_docs/ gdw3.pdf

World Health Organization (1998), Division of Control of Tropical Disease homepage, availableat: wwwwhoch/ctd/

World Health Organization and UNICEF (United Nations Children Fund) (2000), Global WaterSupply and Sanitation Assessment 2000 Report, World Health Organization, Geneva,available at: wwwwhoint/water_sanitation_health/ Globassessment/GlobalTOChtm

Yamout, G. (2002), “An optimization model for water supply multi-sectoral allocation in theGreater Beirut area”, MS thesis, Department of Civil and Environmental Engineering,American University of Beirut, Lebanon.


367

Further reading

Cairncross, S. and Kinnear, J. (1991), “Water vending in urban Sudan”, Water ResourcesDevelopment, Vol. 7 No. 4, pp. 267-73.

Kjellen, M. (2000), “Complementary water systems in Dar es Salaam, Tanzania: the case of watervending”, Water Resources Development, Vol. 16 No. 1, pp. 143-54.

Seckler, D., Barker, R. and Amarasinghe, U. (1999), “Water scarcity in the twenty-first century”,Water Resources Development, Vol. 15 No. 1/2, pp. 29-42.

United States Environmental Protection Agency (2001), National Primary Drinking WaterStandards, EPA 816-F-01-007, Office of Water, available at: wwwepagov/safewater

MEQ14,3

368

Leaded gasolinephase-out:

Lebanon

389

Environmental Management andHealth, Vol. 12 No. 4, 2001,

pp. 389-406. # MCB UniversityPress, 0956-6163

Socio-economic benefits ofleaded gasoline phase-out

The case of LebanonZ. Hashisho and M. El-Fadel

Department of Civil & Environmental Engineering,American University of Beirut, Lebanon

Keywords Leaded gasoline, Lead health effects, Health economics, Blood lead levels, Lebanon

Abstract Lead emissions from vehicles using leaded gasoline is a serious environmentalproblem in urban areas. While leaded gasoline has been completely phased out in many developedcountries, it is still the predominant fuel grade in most developing countries. This paper presentsan estimation of the health and economic benefits and costs of the transition from leaded tounleaded gasoline in Lebanon based on relevant dose-response functions and available country-specific data. Comparing the potential costs of the phase-out and the predicted benefits, it wasconcluded that such action is economically highly justified.

1. IntroductionLead is a toxic heavy metal. Even at low exposure level, it is associated withserious health hazards such as cardiovascular problems in adults andneurological disorders in children. Lead emissions from leaded gasolinecombustion is a significant source of atmospheric lead, accounting for morethan 90 percent where no lead phase-out measures have been implemented(Lovei, 1998). The reduction in blood lead levels (BLL) will reportedly reducehealth hazards associated with lead exposure, which is directly translated intoa lower health expenditure. In the USA, annual health benefits from reducingthe population's BLL by 1�g/dl were estimated at 6.9 and 9.9 billion USD forchildren and adults respectively (Schwartz, 1994). Similarly, in Russia, theannual environmental benefit from the complete phase-out of leaded petrol isestimated at about 1.44 billion USD (SCEP, 1997). As a result, many countrieshave implemented (Antigua and Barbuda, Argentina, Austria, Bahamas,Bolivia, Brazil, Canada, Columbia, Costa Rica, Denmark, Dominican Republic,El Salvador, Finland, Germany, Guatemala, Haiti, Honduras, Hungary, Japan,Mexico, Nicaragua, Norway, Saba, Slovak Republic, St Eustasius, Thailand, theUSA) or are implementing ± EU countries, Egypt ± a lead phase-out program(Lovei, 1999). Even some countries, such as China and India, have switched tounleaded gasoline in major urban cities as a direct solution for the problem oflead pollution.

The research register for this journal is available athttp://www.mcbup.com/research_registers

The current issue and full text archive of this journal is available athttp://www.emerald-library.com/ft

The authors would like to thank Dr I. Nuwayhid, Faculty of Health Sciences, AmericanUniversity of Beirut, for providing data on blood lead levels in Lebanon. Special thanks areextended to the United States Agency for International Development for its support to theEnvironmental Engineering and Science Programs at the American University of Beirut.

http://www.mcbup.com/research_registers

http;//www.emerald-library.com/ft

EMH12,4

390

Lead particles emitted to the atmosphere from the combustion of leadedgasoline are dispersed into various environmental media from where they canenter the human body, mainly through inhalation and ingestion[1]. While theformer is the major exposure route for adults, ingestion of lead-contaminatedwater, food, dust, and soil, can be more significant in children. In the humanbody, lead is accumulated mainly in the mineralizing tissue for long periods,thus allowing for latent exposure, even after the cessation of the externalexposure, as a result of decalcification processes (WHO, 1995).

1.1 Physiologic effectsAlthough health effects of exposure to high doses of lead have been well knownfor at least some 5,000 years[2], it was not until the 1970s that it was establishedthat exposure to low levels of lead can have significant implications on humanhealth. In fact, advancements in edpidemiology, toxicology, and laboratorytechnologies allowed for better detection, identification, and characterization ofthe effects of lead exposure in systems not known to be vulnerable, and atlevels lower than previously acknowledged.

In this context, the health impacts of lead exposure are diversified andhighly dependent on several factors, including age, individual susceptibility,health conditions, and exposure level. Although the exposure dose to lead fromvehicle emissions might be low, it is aggravated by the chronic and cumulativenature of exposure. Irrespective of the route, the physiologic effects of leadexposure are the same and include biochemical effects (such as anemia, andinterference with enzyme synthesis), and neurobehavioral effects (such as IQdeficiency, mental retardation, hyperactivity) in children. In adults, lead isstrongly correlated with increased probability of hypertension andcardiovascular diseases. Despite the long known adverse effects of exposure tolead, many countries are still adding lead to gasoline because it is the cheapestway to boost the octane content in gasoline and lubricate vehicle engines.

1.2 Standards and regulationsStandards for blood lead levels have been continuously reviewed and reducedas emergent studies showed the inadequacy of previous standards (ATSDR,1995; IOMC, 1998; WHO, 1995) (see Table I). While a firm consensus on themaximum permissible concentration of lead in blood has not been reached, the

Table I.Summary of standardsand regulations forlead in air and blood(IOMC, 1998)

Air (�g/m3) Blood (�g/dl)

WHO 0.5-1.0 20US 1.5a 10b

EU 2.0c NAd

Australia 1.5 10Canada 5.0 10

Notes: a Annual average; b Action level for children; c Quarterly average; d NA = notavailable


Lebanon

391

amount of lead considered to cause lead poisoning continues to be reduced asnew toxicological and epidemiological evidences are becoming available. In1980, the World Health Organization (WHO) recommended a health-basedblood lead level (BLL) of 40�g/dl for workers and a limit of 30�g/dl for womenof childbearing age (WHO, 1980). Subsequent WHO standards have been basedon keeping levels in the vast majority of the population below 20 �g/dl (WHO,1995). The US Centers for Disease Control (CDC) have repeatedly revised itsstandard (Figure 1). Today, although a 10�g/dl is adopted, several studies haverevealed some health effects associated even with lower levels. In fact, theremay be no safe level below which no negative health effects are observed(ATSDR, 1995).

This paper assesses the problem of lead emissions from the combustion ofleaded gasoline in Lebanon. Available information on BLL, and gasolineconsumption and characteristics are presented. The benefits of the switch fromleaded to unleaded gasoline in urban areas is evaluated using available dose-response relationships developed by the US Environmental Protection Agency(EPA) for selected health endpoints. The potential costs of the transition tounleaded gasoline in addition to the benefits from health and car-relatedsavings are estimated by relying on previous phase-out initiatives withadjustment for income difference.

2. The Lebanese contextThe Lebanese vehicle fleet mainly comprises of passenger cars and ischaracterized by relatively old (14 years average age), deteriorated and poorlymaintained vehicles, resulting in high pollutant emission rates (Kaysi andSalvucci, 1993). In the absence of direct governmental intervention, leadedgasoline (maximum lead content of 0.66g/l) is the predominant grade of fuelused. In 1997, total gasoline import was about 1,319 thousand tons, of which530 thousand, were of the regular grade, 631 thousand of the premium grade,and 158 thousand of the unleaded grade (UNDP, 1999).

Figure 1.Historical recommended

action level for bloodlead level in children in

the USA

EMH12,4

392

Measurements of blood lead levels have been conducted in adults and children(Nuwayhid, 1999). One study conducted in 1993 measured BLL in maleworking adults (Figure 2). The distance commuted to work along withoccupation type and smoking habit were associated to a mean BLL of 15.8�g/dl(Nuwayhid, 1999), which is comparable to values reported in European andAmerican countries before phase-out programs were implemented (Figure 3).

Another study measured BLL in more than 70 working children (10-17 yearsold) and compared it to children who are studying but working part-time andchildren who are only studying (Figure 4). Mean BLL was 9.75�g/dl forstudents, 11.36�g/dl for working students and 13.54�g/dl for workers.

Measurements of BLL among 284 healthy babies and children (1-3 years old)presenting to the AUB medical center for routine checkup have also beenconducted in 1998. Mean BLL was 6.6�g/dl; 14 percent had a BLL of 10 mg/dl

Figure 2.Blood lead levels forworking adults inLebanon

Figure 3.BLL in Lebanese adultmales in comparison tointernational reportedtrends


Lebanon

393

or more. This was associated with manual job of the father, living in an areawith traffic jams, and low income.

Measurements of BLL among school children revealed that 15-18 percent oftested children have BLL greater than or equal to 10�g/dL. Highest BLL wereabout 17-18 �g/dL (Figure 5).

As is the case with adults, children's BLL are comparable to values reportedin countries in the 1970s and 1980s when leaded gasoline was the predominantfuel grade (Figure 6).

3. Impact assessmentAlthough the association between the phase-out of leaded gasoline and thedrop in blood lead levels is evident, it is difficult to quantify the impacts oflead exposure. As such, precise assessment of health outcomes of the phase-out of leaded gasoline is difficult to make. This is mainly due to factors such

Figure 4.BLL in 10-17 year old

working andnon-working children in

Lebanon

Figure 5.BLL in 1-3 year oldhealthy children in

Lebanon

EMH12,4

394

as the lack of a dose-response relationship for some exposure outcomes, orthe uncertainty associated with available ones. In this context, dose-responserelationships have been derived for only few outcomes associated with BLL.This section presents the methodology used to estimate harmful healthendpoints associated with elevated BLL. Dose-response functions andsummary of health impacts for each affected group are presented.

3.1 MethodologyThe impact assessment of the predominant use of leaded gasoline in Lebaneseurban areas was conducted using dose-response functions for children andadults derived by the USEPA (USEPA, 1997) based on available epidemiologicstudies.

3.1.1 Potential health impacts in children. Lead induces neurobehavioraldeficiency in young children, including hyperactivity, behavioral and attentiondifficulties, as well as mental, motor and perceptual skill deficits. Limitedeffects can be quantified and expressed in IQ tests and scores due to lack ofdata. Currently, available data permit the quantifying of the relationshipbetween BLL and IQ points loss, mental retardation, and mortality. A highlysignificant relationship (p < 0.0001) suggests that a 1�g/dl increase in BLLresults in a decrease of 0.245 � 0.039 IQ points. Accordingly, the total IQ pointsloss for each group can be expressed by equation (1). Note that the populationof children up to age six was divided by seven in order to avoid multiplecounting, assuming even distribution by age.X

IQ � 0:25��E�x� � P

7�1�

Figure 6.BLL in Lebanesechildren in comparisonto international trendsbefore lead phase out


Lebanon

395

where�E�x� = change in the mean of BLL distributionP = total population of children up to age six

In addition, lead increases the incidence of mentally retarded children (IQscores < 70). Assuming that the population of children have blood leaddistributions defined by some geometric mean and geometric standarddeviation, and that the population has a normalized IQ point distribution with amean of 100 and a standard deviation of 15, the proportion of populationexpected to have an IQ < 70 is determined by:

�PIQ<70� � ��znoÿcontrol� ÿ ��zcontrol� � ��znoÿcontrol� ÿ 0:02775 �2�

��z� � 1��2�p

�zÿ1

eÿz2

2 dz �3�

where� (z) = standard distribution function� PIQ < 70 = change in the probabilty of children having IQ < 70zcontrol = standard normal variate

For an IQ score of 70, with mean IQ score of 100 and standard deviation of 15, Zis computed as:

z � 70ÿ 100

15� ÿ2

z�noÿcontrol� �70ÿ ÿ100� 0:25��PbB

�15

�4�

�PbB = change in the average BLL, �g/dl

Multiplying the probability change �P (IQ<70) by the exposed population ofchildren yields the number of children with mental retardation due to increasein BLL. As in equation (1), this relation is applicable to children up to age six,and the population is divided by seven to avoid multiple counting.

Maternal blood lead levels were correlated with several adverse healthimpacts on fetuses including decreased gestational age, induced birth weight,late fetal death, and increased rate of infant mortality. To estimate the changesin infant mortality, gestational age was linked to maternal BLL and infantmortality was linked to gestational age. Accordingly, the risk of infantmortality decreased by 10-4 for each 1�g/dl decrease in maternal BLL.

3.1.2 Potential health impacts in adults. Quantified health effects in adults arerelated to the effects of lead on blood pressure, including increased probability ofhypertension, initial coronary heart disease (CHD), strokes, and premature

EMH12,4

396

mortality. Elevated BLL has been related to high blood pressure in adults. Formen aged between 20 and 74 years a dose-response relationship for hypertension(defined as diastolic blood pressure above 90 mmHg) is expressed as[3]:

�Pr�HYP� � 1

1� e2:744ÿ0:793�ln x1� ÿ1

1� e2:744ÿ0:793�ln x2� �5�

where� Pr(HYP) = change in the probabilty of hypertension[4]x1 = BLL in the control scenario, �g/dlx2 = BLL in the no-control scenario, �g/dl

In addition to increased probability of hypertension, increased blood pressureincreases the probability of initial occurrence and reoccurrence of coronaryheart disease (CHD), initial cerebrovascular accidents (CA), initialatherothrombotic brain infarctions (BI), and premature mortality. Accordingly,changes in BLLs are related to changes in blood pressure which are in turnrelated to changes in the probability of occurrence of these effects. Thus formen, the change in diastolic blood pressure can be expressed as:

�DBPmen � 1:4� In

�x1

x2

��6�

where� DBPmen = changes in men's diastolic blood pressure, mmHg

For women, results of several studies suggested that the effect on bloodpressure of a decrease of BLL from 10 to 5�g/dl is about 60 percent of the effectof the same change observed in men.

�DBPwomen � �0:6� 1:4� � ln

�x1

x2

��7�

where� DBPwomen = changes in women's diastolic blood

The general form of the change of probability of occurrence of the above effectsin adults is expressed as:

�Pr � 1

1� e�ÿ��ln DBP1� ÿ1

1� e�ÿ��ln DBP2� �8�

where�Pr = change in the probabilityDBP1 = mean diastolic blood pressure in the control scenario, mmHgDBP2 = mean diastolic blood pressure in the no-control scenario, mmHg�= empirical constant� = Does-response coefficient


Lebanon

397

Table II presents a summary of dose-response coefficients used in theprobability distribution to estimate lead health effects in adults. Note that thedose-response coefficients for CHD and strokes account only for non-fatalcases. It was reported that two-thirds of the CHD and 70 percent of the strokesare non-fatal. Fatal CHD and strokes were accounted for in the mortalitycoefficients.

3.2 Summary of health impactsUsing the above dose response relationships, the impact of vehicular leademissions on adult and children in major urban regions in Lebanon isestimated. Due to limitation in data availability, several assumptions have beenmade:

(1) Population of major urban areas only are considered (about 1,639,000(ERM, 1995)) since it is expected that in such locations vehicularemissions would constitute a significant source of lead pollution withhigh exposure potential.

(2) Age distribution of the studied population is assumed to be the same asthat of the whole Lebanese population distribution (Figure 7).

(3) Population within each age group is assumed to be evenly distributedby age.

(4) Blood lead levels are assumed to be independent of age in children andadults.

(5) Based on average levels reported in several studies, average BLL inwomen is assumed to be 64 percent of that in men (COWI, 1998a;Wietlisbach et al., 1995).

Table II.Summary of dose-

response coefficientsfor two-year

probability distribution

�

Health effect Age group � meanStandarddeviation

CHDmena 40-59 4.996 0.03036 0.003586

60-64 5.19676 0.023531 0.02865-74 4.90723 0.02031 0.00901

CHDwomen 45-74 6.9401 0.03072 0.00385BImen 45-74 8.58889 0.04066 0.00938BIwomen 45-74 9.07737 0.04287 0.00754CAmen 45-74 9.9516 0.04840 0.00711CAwomen 45-74 10.6716 0.0544 0.00637Mortalitymen

b 40-45 5.3158 0.03516 0.1659655-64 4.89528 0.01866 0.0053365-74 3.05723 0.00547 0.00667

Mortalitywomen 45-74 5.40374 0.01511 0.00419

Notes: a ten-year probability; b 12-year probability

Source: USEPA (1997)

EMH12,4

398

(6) The phase-out of leaded gasoline will result in a 77 percent reduction inBLL of children and 78 percent reduction in BLL of men and women(Pirkle et al., 1994).

(7) BLL in pregnant women is assumed to be the same as that in non-pregnant ones.

The results of the health impact assessment for children and adults arepresented in Tables III and IV respectively.

4. Cost-benefit analysis of lead phase-outThe costs and benefits of switching from leaded to unleaded gasoline inLebanese urban areas are assessed based on international experience. Refinedestimates can be obtained with elaborate and updated local data and detailedanalysis of country-specific characteristics such as blood lead levels, ambientair lead monitoring, statistics about mortality and morbidity as well as theircorresponding unit costs, and maintenance savings unit costs.

Figure 7.Age distribution of theLebanese population

Table III.Impact of vehicularlead emissions onchildren in Lebaneseurban areas

Number of casesEffect Average Range

Total IQ point loss (points) 42,689 35,176-48,495Mental retardation (cases) 167 NAChild mortality (cases) 31 NA

Notes: NA = Not applicable


Lebanon

399

4.1 Costs of lead phase-outWorldwide experience showed that phasing out lead from gasoline istechnically feasible, and the costs are generally modest. In general, the costs ofremoving lead from gasoline are grouped into: first, costs of refineryadjustments and gasoline additives; and second, costs of the distributionsystem adjustments.

4.1.1 Unleaded gasoline and lead additives. In gasoline importing countries,phase-out of leaded gasoline can be easily accomplished without substantialcost. In fact, the incremental cost of switching to unleaded gasoline isdetermined by the difference in price between unleaded and leaded gasoline ontheir main import markets. Frequently, unleaded gasoline prices have beenlower for some octane grades (for example, 95 RON) than leaded gasoline, dueto structural over-capacities in production. The cost of switching from leaded tounleaded gasoline for importers is expected to be very low, consisting mainly ofthe costs of lubricant additives and additional transportation costs to accessnew suppliers if necessary (Lovei, 1997).

World-wide experience and hypothetical estimates indicate that annualizedinvestment expenditures and added operating costs associated with theremoval of lead from gasoline are typically about 0.01-0.02 USD per liter.Although these estimates were made for specific refineries, the numbers arerelatively comparable and within the same order of magnitude. Hence, theyprovide a reasonable estimate of the expected costs of actions and measuresassociated with lead phase-out programs.

In Lebanese urban areas, annual gasoline consumption is estimated at 0.65million tons (equal to the 1997 import of leaded gasoline multiplied bypercentage of vehicles in urban areas). Taking gasoline density to be equal to0.785kg/l, the annual cost difference between unleaded and leaded gasoline isabout 7.6-15.1 million USD for a unit cost of 0.01-0.02 USD/l.

Table IV.Impact of vehicular

lead emissions onadults in Lebanese

urban areas

Number of casesEffect Age group Mean Minimum Maximum

Hypertension men 20-74 1,688.3 NA NACHD men 40-59 54.6 15.3 133.0

60-64 24.1 10.0 482.665-74 30.7 3.3 152.4

CHD women 45-74 46.8 12.2 122.2CA men 45-74 42.7 7.7 162.5CA women 45-74 22.9 4.6 77.3BI men 45-74 26.5 3.8 127.4BI women 45-74 16.8 3.0 63.3Mortality men 40-54 50.9 21.1 0.01

55-64 36.1 6.8 121.265-74 12.9 5.4 78.0

Mortality women 45-74 25.0 5.3 75.8

Notes: NA = Not applicable

EMH12,4

400

4.1.2 Distribution system adjustments. Typically, adjustments to thedistribution system do not entail major expenditure, particularly in developingcountries (where labor is significantly cheaper), since the same system can beused for distributing unleaded gasoline after phasing out leaded gasoline.These costs are attributed to the once-and-for-all cleaning of tank lorries, pipes,pumps, and underground storage tanks which can be performed by one to twoworkers for one work day (Naranong, 1995). In the transition period duringwhich leaded and unleaded gasoline are both distributed, unleaded gasolinecan replace one grade of leaded gasoline without the need for additionaldistribution infrastructure, except for different pump nozzle sizes for leadedand unleaded grades. Overall, distribution costs are minor and can be neglectedin comparison to the cost of refinery adjustments, when assessing the cost oflead removal from gasoline.

4.2 Benefits of lead phase-outGenerally, the benefits associated with phasing out lead from gasoline can begrouped into: health-related benefits; and car-related benefits. In the USA, thebenefits of lead phase-out from gasoline are estimated to exceed its costs bymore than ten times (Schwartz, 1994). Similarly, estimated benefits of leadremoval exceeded the costs by three to six times in Nizhny Novgorod, RussianFederation (Lovei, 1997).

4.2.1 Health benefits. Valuing the benefits of lead phase-out depends oncountry-specific factors such as the cost of labor, labor productivity, capital andmedical care, life expectancy, people's values of health and life, and theirwillingness to accept risk. Therefore, concrete valuation of the health benefitsof the phase-out of leaded gasoline is difficult to make due to the lack ofcountry-specific data. The situation is worsened by the uncertainty associatedwith the parameters and functions used in estimating the health outcomes oflead exposure. Nevertheless, this process provides a rough estimate of themagnitude of the health benefits associated with switching to unleadedgasoline.

4.2.1.1 Valuation techniques. Several methods are available for valuing thehealth benefits of removing lead from gasoline. Typical methods used forvaluing mortality outcomes include the human capital approach and thewillingness to pay (WTP) approach. In case original data on the WTP arelacking, one can use income adjusted mortality values from other countriesafter compensating for income difference (see Figure 8).

To assess the morbidity outcomes, methods typically used include the costof illness (COI) approach, and the WTP approach. The choice of the method tobe used depends on several factors, particularly the availability of data for themethod used. In this study, COI values estimated for the US population havebeen used to estimate morbidity outcomes of lead exposure. To estimatemortality costs, WTP values estimated in the USA have been used. USmorbidity and mortality values estimated for the year 1990 (USEPA, 1997)have been adjusted for inflation to the year 1998 (by multiplication with an


Lebanon

401

inflation factor of 1.2547 (CJR, 1999) and for the Lebanese context based onincome ratio as depicted in Figure 8, where a ratio of about 0.12 has beenestimated based on a gross national per capita product (GNP) of 29,340 USD forthe USA (World Bank, 1999b) and 3,560 USD for Lebanon (World Bank, 1999a)for the year 1998. Note that the income ratio method was not applied in theestimation of the phase-out costs since they are based on unit costs, which relyto a large extent on the international market.

. Valuing benefits of lead phase-out for children. For children, IQ reductionis expected to result in reduced present value of expected lifetime earningsand increased educational expenditure on an infant who becomesmentally disadvantaged or is in need of compensatory education. Table Vpresents the estimated benefits resulting from decreased BLL in children.

. Valuing health benefits of lead phase-out for adults. Benefits fromreduced morbidity in adults can be approximated by medical costs(physician care, drugs, and hospitalization costs) and costs of restrictedactivity or work loss days (Table VI). This provides a conservativeestimate of the real costs as it excludes the value of pain, suffering, diet,and behavior modification in addition to the side effects of medications.

4.2.1.2 Application to Lebanon. Using the methodology outlined above andapplied on estimated reduction in adverse health outcomes from Tables III andIV, the monetized benefits of phasing out leaded gasoline are presented inTables VII and VIII. While these costs exhibit wide variations and are forselected age groups and for some health effects, they provide a conservative

Figure 8.Transfer of mortalityand morbidity values

across countries

Table V.Monetized benefits

from reducingchildren's BLL in the

USA

Cost per caseHealth outcome (1990 USD)a (1998 USD)b

Lost IQ points (per point) 2,957 3,710IQ < 70 (per case) 42,000 52,697Infant mortality (per case) 4,800,000 6,022,560

Notes: a (USEPA, 1997); b Inflation factor = 1.2547 (CJR, 1999)

EMH12,4

402

estimate of the benefits expected from switching to unleaded gasoline. Clearly,mortality costs constitute the major contributor to the estimated benefits fromthe reduction in BLL.

4.2.2 Other benefits of lead phase-out. In addition to the human healthbenefits, shifting from leaded to unleaded gasoline results in important costsavings associated with reduced car maintenance from lead induced corrosionof exhaust systems and engines. Switching from leaded to unleaded gasolinemay increase engine life by as much as 150 percent (Faiz et al., 1996). Thesavings can be attributed to (Walsh and Shah, 1997):

Table VI.Monetized benefitsfrom reducing adultBLL in the USA(USEPA, 1997)

Cost (1990 USD per case)a Cost (1998 USD per case)b

Health outcome Men Women Men Women

Hypertension 681 - 854CHD 52,000 52,000 65,244 65,244Stroke 200,000 150,000 250,940 188,205Mortality 4,800,000 4,800,000 6,022,560 6,022,560

Notes: a USEPA (1997); b Inflation factor = 1.2547

Source: CJR (1999)

Table VII.Monetized benefitsfrom reducingchildren's BLL inLebanese urban areas

Cost per case Cost (1998 USD) � 1,000Health outcome (1998 USD) Average Range

Total IQ loss 450 19,217.5 13,414.7-25,175.5Mental retardation 8,023 1,069.3 NAChild mortality 730,754 22,427.0 NATotal ± 42,713.8 36,911.0-48,671.8

Note: NA = Not applicable

Table VIII.Monetized benefitsfrom reducing adultBLL in Lebanese urbanareas

EffectCost per case(1998 USD)

Average cost(1998 USD) � 1,000

Cost range(1998 USD) � 1,000

Hypertension men 104 175.0 NACHD men 7,916 867.5 226.3-6,079.7CHD women 7,916 370.8 96.4-967.7CA men 30,448 1,301.2 233.6-4,947.1CA women 22,836 521.9 104.2-1,766.4BI men 30,448 808.2 114.6-3,880.6BI women 22,836 384.2 69.4-1,444.4Mortality men 730,754 66,799.7 21,796.8-145,578.9Mortality women 730,754 18,286.6 3,836.7-55,379.5Total 89,514.2 26,653.1-220,219.3

Note: NA = Not applicable


Lebanon

403

. longer intervals between spark plug changes (every other year insteadof every year);

. longer intervals between oil and filter changes (once a year instead oftwice a year);

. reduced need for muffler replacement (twice per five years instead ofonce per year);

. reduced need for exhaust pipe replacement (none opposed to once everyfive years);

. reduced need for carburetor servicing.

Car maintenance savings from switching from leaded to unleaded gasoline inCanada were estimated at about 0.017 USD per liter of gasoline (1980 prices)corresponding to about 27 USD per year per car (Walsh and Shah, 1997).Savings estimates for the USA are about 0.003-0.024 USD per liter of petrol(Hirshfeld and Kolb, 1995). If similar unit cost savings are assumed forLebanon, the estimated total benefits in car maintenance savings fromswitching to unleaded gasoline in Lebanese urban areas range from 2.5 to 19.9million USD after accounting for inflation till the year 1998. However, theseestimates seem to over-predict real values since part of these savings areassociated with labor costs which vary among countries, particularly betweenLebanon and the USA. On the other hand, using the income ratio approachwould provide the minimum cost savings because car maintenance involvesboth labor and spare parts. The latter is typically at the same price, if not moreexpensive, in non-manufacturing countries. Accordingly, the minimum carmaintenance savings from switching to unleaded gasoline range from 0.3 to 2.4million USD.

In addition to cost savings from reduced maintenance costs, the phase-out oflead from gasoline improves fuel economy in three ways:

(1) increasing the energy content of petrol through more intense processing;

(2) reducing the fouling of oxygen sensors in mis-fuelled late-modelvehicles; and

(3) reducing the fouling of spark plugs.

Monetized estimates of the cost savings attributable to improved petrol energydensity from shifting from leaded to unleaded petrol are estimated at 0.0024USD per liter of petrol (COWI, 1998b). For the case of Lebanon, using unleadedgasoline in urban areas [5] will result in a cost saving from improved fuelefficiency of about 2 million USD, which when multiplied by income ratioresults in 0.25 million USD. These values are conservative estimates since theincome ratio approach reflects minimal savings in this case.

Generally, the costs of modifying the vehicle fleet to operate on unleadedpetrol or the cost of the addition of lubricants may be comparable to thesavings in maintenance costs and the increased fuel economy from phasing outlead (COWI, 1998b). Comparing the estimated benefits and costs of the phase-

EMH12,4

404

out of leaded gasoline, it is evident that such action is economically justified.However, several measures should be adopted to accelerate the transition fromleaded to unleaded gasoline.

5. LimitationsLimitations in this study can be classified into three categories:

(1) limitations in the available data;

(2) limitations in the health impact assessment; and

(3) limitations in the economic assessment.

Limitations in the available data are mainly attributable to BLL values as theyreflect only recent exposure at a single point in life, and to the relatively smallsample size. Limitations in the health impact assessment are mainly incurredfrom the absence of dose-response functions for several health effects of leadexposure and the uncertainties in the available ones, as well as in theassumptions adopted in the absence of better estimates of specific data, such asthe assumption of even age distribution within each age group, independenceof BLL of age in children and adults, population estimation, relation betweenmen's and women's BLL values, etc. The limitations in the economicassessment result not only from uncertainties in the mortality and morbidityvalues estimated in the USA, but also in the extrapolation of these values bythe multiplication by the income ratio between countries. The application ofincome ratio in the estimation of car maintenance savings limits the usefulnessof this method to lower bounds rather than real estimates.

6. Summary and conclusionsLead emission constitutes a significant environmental health concern,particularly in urban congested areas. Although vehicles are not the onlysource of lead emissions, they are often the main one in these areas. In Lebanon,leaded gasoline is the predominant fuel grade used (88 percent). Availableinformation reveals that average BLL in Lebanese babies one to three years old(6.6�g/dl), children 10-17 years old (9.75 to 13.54�g/dl), and adults (15.8�g/dl)are comparable to those previously reported in other countries before thephase-out of leaded gasoline, and are expected to drop with the implementationof such a program.

Based on previous lead phase-out initiatives, BLL are expected to decrease inexcess of 75 percent in children and adults. Using dose-response relationshipsdeveloped on the basis of epidemiologic studies, selected health benefits from adecline in BLL are estimated for children and adults in Lebanese urban areas.In children, a 77 percent reduction in BLL will prevent the loss of 42,689 IQpoints, the occurrence of 167 cases of mental retardation, and 31 cases ofpremature mortality. In adults a 78 percent reduction in BLL will prevent theoccurrence of 1,688 cases of hypertension in adult males, 157 cases of coronaryheart disease, 44 cases of brain infarctions, 66 cases of cerebrovascular


Lebanon

405

accidents, and 125 cases of premature mortality in both males and females. Thecost of illness and willingness to pay values derived for the USA are used toestimate morbidity and mortality cost savings respectively after adjustment forincome difference between the USA and Lebanon. Estimated cost savings areabout 42.7 million USD for children and 89.5 million USD for adults. Thesebenefits will translate to nearly 40 USD per capita per year for the entirepopulation. Additional cost savings can be accrued from reduced carmaintenance and improved fuel efficiency. The comparison of the expected costsavings from phasing out leaded gasoline with the potential costs indicates thatsuch action is economically highly justified.

Notes

1. While organic lead such as gasoline lead additives can be easily absorbed through theskin, dermal absorption of inorganic lead, the predominant form of vehicular leademissions, is negligible (WHO, 1995).

2. Symptoms described by Hippocrates at about the fourth century BC for lead intoxicationare the same as those classified today.

3. Elevated BLL is associated with hypertension in both men and women; however, no dose-response function for hypertension in women is currently available.

4. Due to lack of averaging time, the number of annual cases of hypertension was estimatedby dividing the total number of cases by 55 (74 ± 20 +1).

5. Costs and savings for cars outside urban areas were not included to restrict the cost-benefitanalysis to urban areas.

References

ATSDR (Agency for Toxic Substances and Disease Registry) (1995), Case Studies inEnvironmental Medicine: Lead Toxicity, US Department of Health and Human Services(US DHHS), Atlanta, GA.

CJR (Columbia Journalism Review) (1999), ` CJR dollar conversion calculator'', CJR. http://www.cjr.org/resources/inflater.asp

COWI Consulting Engineers and Planners (1998a), UN/ECE Task Force to Phase-out LeadedPetrol: Country Assessment Report, Danish Environmental Protection Agency (DEPA),Copenhagen.

COWI Consulting Engineers and Planners (1998b), UN/ECE Task Force to Phase-out LeadedPetrol: Main Report, Danish Environmental Protection Agency (DEPA), Copenhagen.

Earth Summit Watch (1997), ` Four in '94: accessing national actions to implement agenda 21'',http://www.earthsummitwatch.org/4in94.html

ERM (Environmental Resources Management) (1995), Lebanon: Assessment of the State of theEnvironment, Ministry of the Environment, Beirut.

Faiz, A., Weaver, C.S. and Walsh, M.P. (1996), Air Pollution from Motor Vehicles, World Bank,Washington DC.

Hirshfeld, D. and Kolb, J. (1995), ` Phasing out lead from gasoline: feasibility and costs. A study ofthe refining sector in Romania'', in series of Implementing the Environmental ActionProgramme for Central and Eastern Europe, World Bank, Washington, DC, cited in Lovei,M. (1998), Phasing out Lead from Gasoline: Worldwide Experience and Policy Implications,World Bank, Washington DC.

EMH12,4

406

IOMC (Inter Organization Programme for the Sound Management of Chemicals) (1998), GlobalOpportunities for Reducing the Use for Leaded Gasoline, United Nations, Geneva.

Kaysi, I. and Salvucci, F. (1993), Passenger Transportation Options for a Revitalized Beirut, AUB/MIT Collaborative Program on Science, Technology, and Development, AmericanUniversity of Beirut, Beirut.

Lovei, M. (1997), Phasing out Lead from Gasoline in Central and Eastern Europe, Health Issues,Feasibility, and Policies, World Bank, Washington DC.

Lovei, M. (1998), Phasing out Lead from Gasoline: Worldwide Experience and Policy Implications,World Bank, Washington DC.

Lovei, M. (1999), Eliminating a Silent Threat: World Bank support for the Global Phaseout of Leadfrom Gasoline, World Bank, Washington, DC.

Naranong, A.A. (1995), ` An analysis of potential policies for reducing lead in gasoline inBangkok'', PhD Dissertation, Faculty of Graduate School, Vanderbilt University, TN.

Nuwayhid, I. (1999), ` Faculty of health sciences'', American University of Beirut, Beirut, personalcommunication.

Pirkle, J.L., Brody, D.J., Gunter, E.W., Kramer, R.A., Paschal, D.C., Flegal, K.M. and Matte, T.D.(1994), ` The decline in blood lead levels in the United States. The National Health andNutrition Examination Surveys (NHANES)'', Journal of American Medical Association,Vol. 272 No. 4, pp. 284-91.

SCEP (State Committee for Environmental Protection of the Russian Federation) (1997), WhitePaper: Lead Contamination of the Environment in the Russian Federation and its Effect onHuman Health, Moscow.

Schwartz, J. (1994), ` Societal benefits of reducing lead exposure'', Environmental Research,Vol. 66 No. 1, pp. 105-24, cited in Lovei, M. (1997), Phasing out Lead from Gasoline inCentral and Eastern Europe, Health Issues, Feasibility, and Policies, World Bank,Washington DC.

UNDP (United Nations Development Programme) (1999), Climate Change, UNDP, Beirut.

USEPA (United States Environmental Protection Agency) (1997), Benefits and Costs of the CleanAir Act, Final Report to Congress on Benefits and Costs of the Clean Air Act, 1970 to 1990,USEPA, Office of Air and Radiation, EPA 410-R-97-002.

Walsh, M. and Shah, J. (1997), Clean Fuels for Asia. Technical Options for Moving TowardUnleaded Petrol and Low-Sulfur Diesel, World Bank, Washington DC.

WHO (World Health Organization) (1980), Recommended Health Based Limits in OccupationalExposure to Heavy Metals: Report of a WHO Study Group, Geneva, WHO, TechnicalReport Series N0. 647, cited in WHO (1995), Environmental Health Criteria 165: InorganicLead, United Nations Environment Programme, International Labour Organization, andWHO, Geneva.

WHO (World Health Organization) (1995), Environmental Health Criteria 165: Inorganic Lead,United Nations Environment Programme, International Labour Organization, and WHO,Geneva.

Wietlisbach, V., Rickenbach, M., Berode, M. and Guillemin, M. (1995), ` Time trend anddeterminants of blood lead levels in a Swiss population over a transition period (1984-1993)from leaded to unleaded gasoline use'', Environmental Research, Vol. 68 No. 2, pp. 82-90.

World Bank (1999a), Lebanon at a Glance, World Bank, http://www.worldbank.org/data/countrydata/aag/lbn_aag.pdf.

World Bank (1999b), United States at a Glance, World Bank, Washington, DC, http://www.worldbank.org/data/countrydata/aag/usa_aag.pdf

30 Noise Control Eng. J. 50 (1), 2002 Jan–Feb © 2002 Institute of Noise Control Engineering

1. INTRODUCTION

Aircraft traffic is a major source of community concerndue to aircraft noise emissions, particularly in locations closeto airports and aircraft flight tracks. The effects of noise onhumans can range from minor annoyance, sleep disruption,and speech interference, to hearing damage.1-3 Noise impactsare typically dependent on the duration and level of the noiseexposure. As a result, noise exposure criteria have beendeveloped by many organizations using various noiseindicators (Table 1).

While noise is best known for its disruptive effects suchas loss of sleep, decrease in productivity, and loss of hearing,these effects are very hard to quantify in monetary terms.However, since noise is associated with a place, its socialcost has been commonly related to a loss in property value.10

In this context, the hedonic price method has been extensivelyused to evaluate social cost due to noise pollution. In thismethod, the evaluation of the cost of noise can depend onseveral variables ranging from locational traits (access tocentral business district), land use characteristics (zoningcodes), and site peculiarities (width of the frontal road).11

While hedonic methods have become the best-suitedtechniques for estimating noise damage cost,12 they requiresignificant amounts of hard to acquire data to conduct ameaningful cost estimation. Another method for the evaluationof noise cost is through the determination of the willingnessto pay (WTP) for environmental benefits. This methodconsists of asking people if they are willing to pay for areduction of noise in their neighborhood for example, or howmuch they want as a compensation for deterioration inenvironmental conditions such as increased noise.13 Simplermethods based on international field surveys relating socialcosts to travel distance have also been reported.10 In this study,

the hedonic price method was applied and compared withinternational survey results. The WTP approach was deemedinappropriate because of the illegal nature of major settlementsin the immediate vicinity of the airport where it is not feasibleto conduct an economic survey to define the WTP.

A. Hedonic price method

The cost of noise can be best estimated by the decline in aproperty value affected by noise generated from aircraft traffic.This is reinforced by the fact that people are ready to paymore for a property in order to avoid high noise levels.14 Mostestimates are developed using hedonic pricing methods whichassume that an item’s value (property or house) is composedof a number of factors (area, age, location, neighborhood,environmental quality, etc.). The contribution of eachparameter can be obtained by regression analysis.10 Toestimate the implicit cost of noise due to aircraft traffic, thedecline in the price of a property is modeled as a function ofambient noise.15 For this purpose a noise depreciation indexor the percent reduction in a house price per A-wtd soundpressure level (in dB) above a reference background value isused. Using this method, the annual cost of noise impactscan be estimated from Eq. (1).12

C I P (N N ) Hn NDI v ai

i

0 i= ◊ ◊ - ◊Â (1)

Where INDI

is the noise depreciation index, Pv is the annual

average house rent in the area, Nai is the noise for the ith section

of the noise contour, N0 is the background noise, H

i is the

number of residences in the ith section.Research has shown that the I

NDI has an average value of

0.62 percent and an A-wtd background sound pressure levelof 50 dB is typical in urban areas.10 International surveysconstitute a more simplified process that can be used for noisevaluation. This method is based on field surveys in which thecost of noise is evaluated as a function of passengers andkilometers traveled as shown in Table 2.

In this paper, both approaches were compared in

a) American University of Beirut, Faculty of Engineering and Architec-ture, Bliss Street, P.O. Box 11-0236, Beirut, Lebanon; Fax: 961-1744-462; E-mail: [email protected].

Case History: An assessment of the economic impact of airport noise emissionsnear Beirut International Airport

M. El-Fadela) and M. Chahinea)

(Received 2001 January 15; revised 2001 October 04; accepted 2001 October 12)

This paper presents a socio-economic assessment of noise impacts from aircraft traffic oncommunities near the Beirut International Airport. Mathematical modeling was conducted tosimulate aircraft-induced noise emissions and delineate noise contour zones. An economicvaluation of noise-impacted areas was performed using model-delineated noise contour zonescoupled with population and home rental statistics. The same aircraft traffic data were alsoused to evaluate the effects of a different runway configuration to minimize the impacted areasand corresponding noise cost. When the current runways were used a total social cost of noiseimpact was evaluated at 16.9 million USD per year or 0.0038USD/passenger/km traveled. It isestimated that an 87 percent reduction in social cost could be accomplished by changing therunway configuration. © 2002 Institute of Noise Control Engineering.

Primary subject classification: 67.1; Secondary subject classification: 76.1.1.3

31Noise Control Eng. J. 50 (1), 2002 Jan–Feb

conducting the economic valuation of aircraft noise impactsat the Beirut International Airport (BIA), which is located onthe southern periphery of the city of Beirut, about 8 km fromthe city center, occupying a coastal site bordered by mountainsas shown in Fig. 1. For this purpose, noise modeling was firstperformed to delineate noise-impacted areas by aircraft trafficand an economic valuation of noise impacts was thenperformed using the hedonic approach and a comparison withinternational surveys. The same aircraft traffic data were alsoused to evaluate the effects of changing the runwayconfiguration on minimizing the impacted areas andcorresponding noise cost.

2. AIRCRAFT NOISE MODELING

A. Existing conditions

The airport has two intersecting runways, the easternrunway (21-03) for take-off, and the western runway (36-18)for landing. A new maritime runway extending into the seawas completed in summer 2000. This maritime runway wouldreplace the presently used western runway as shown in Fig.2. This study focuses on delineating, through modelsimulations, the change in noise impacted areas due to theintroduction of the new runway using an economic valuationapproach that correlates noise-impacted areas withcorresponding house rental statistics.

At present, the airport serves a number of regional andinternational airlines, in addition to two local airlines. The

Table 1 – Worldwide criteria for noise exposure.

A. World Health Organization annoyance criteria in residential areas 4

Impact Characterization Daytime Leq

(dB) Nighttime Leq

(dB) Approximate DNL (dB)

Serious Annoyance 55 45 55

Moderate Annoyance 50 40 50

B. World Bank Group 5

Receptor Time Period Time Period Leq

Equivalent DNL (dB)

Residential, institutional, educational Daytime 55 55

Nighttime 45

Industrial, commercial Daytime 70 —

Nighttime 70

C. International Organization for Economic Co-operation and Development6

Land Area Time Period Time Period Leq

(dB) Equivalent DNL (dB)

Urban Daytime 55 55

Nighttime 45

Rural Daytime 50 50

Nighttime 40

D. US National Bodies

D.1 US Department of Transportation, Federal Highway Administration Noise Abatement Criteria7

Activity Leqa (dB) Description of Category

Category

A 57 (exterior) Land on which serenity and quiet are of extraordinary significance and serve an important public need and where the

preservation of those qualities is essential if the area is to continue to serve its intended purpose

B 67(exterior) Picnic areas, recreational areas, playgrounds, active sports areas, parks, residences, motels, hotels, schools, libraries, and hospitals

C 72(exterior) Developed land, properties and activities not included in categories A or B

D — Undeveloped land

E 52(interior) Residences, motels, hotels, schools, libraries, public meeting rooms, churches, auditoriums, and hospitals

a A-wtd equivalent noise level

D.2 US Environmental Protection Agency 8

Level requisite to protect health and welfare with an adequate margin of safety DNL 55 dB

D.3 National Research Council 9

Residential areas DNL of 55 dB

32 Noise Control Eng. J. 50 (1), 2002 Jan–Feb

most common types of aircraft using the airport are the AirbusA310 and A320. Other types of commercial jet aircraft includethe Boeing B707, B727, B737, B767, B777, and the TupolevTU54 as summarized in Table 3. There are still a few dailyoperations of the aging Boeing 707 for cargo shipments. Theaverage number of daily operations is 78 landings anddepartures. The estimated current number of passengers usingBIA is 2.2 million per year and is expected to reach 6 millionpassengers by the year 2015.16

B. Noise simulation

The Integrated Noise Model (INM) developed by theFederal Aviation Administration (FAA) was used to simulateaircraft noise emissions. Day-night noise level (DNL)contours, which are the most widely used metric for the

analysis and development of compatible land use,17 wereadopted as indicators to delineate noise-impacted areas dueto aircraft traffic. The DNL metric is a member of a group ofexposure metrics obtained from the noise level expressed as:

LE = 10 log (W

1E

1 + W

2E

2 ) – 10 log (T), (2)

where LE is the A-weighted equivalent noise level, W

1 and

W2 are weighting factors for day and night time periods, E

1

and E2 are noise exposure ratios for day and night time periods,

and T is averaging time over a reference time of one second.The area around the airport was divided into six DNL

values, which would produce five A-weighted noise contourranges: 55-60, 60-65, 65-70, 70-75, and 75-80 dB. Aircraftfleet composition and daily arrival and departure scheduleswere obtained from airport authorities. Figure 3 depicts thearrival and departure noise contours at standard INM landingand take-off weights and flight operation procedures forpresent airport activities. A total developed area of 27 km2

Table 2 – Noise cost generated from aircraft traffic.10

Average cost$/ passenger/km traveled

Country 1995 1999

Canada 0.0039 0.0043

Germany 0.0049 0.0054

Italy 0.0079 0.0087

Holland 0.0099 0.0108

Sweden 0.0014 0.0015

Switzerland 0.0017 0.0019

France 0.0030 0.0033

United Kingdom 0.0018 0.0020

Average 0.0043 0.0047

Fig. 1 – Location of the Beirut International Airport relative toBeirut City (note that North is down in the photograph).

NMed

iterr

anea

n S

ea0 100 250 500 m

Fig. 2 – Runway configuration at the Beirut International Airport.


was affected by noise generated from aircraft traffic. Notethat noise contours generated from departing aircraft aremostly above the sea.

3. ECONOMIC VALUATION

The area surrounding the airport and falling under theaircraft flight path was subdivided into three zones, shown inFig. 4, corresponding to economic affluence based on a surveyof household average annual income and house rentalstatistics. Table 4 summarizes the total number of residences

and average annual rents in areas exposed to aircraft-inducednoise with corresponding noise exposure levels. The averageannual rent was adjusted for an average house lifetime of 30years and a mortgage interest of 6 percent. Using the hedonicprice method (Eq. 1), the corresponding total social cost dueto noise was estimated at 16.9 million USD per year.

Using annual travel activities (number of passengers andthe travel distance to each destination as obtained from airportauthorities)20 and the social cost incurred by applying thehedonic price method, the equivalent average cost will be0.0038 $/passenger/km traveled which is consistent with therange reported in the literature for international cost-basedsurveys (see Table 2).

4. EFFECT OF RUNWAY CONFIGURATION

Day-night average sound levels (DNL) contours weregenerated for the same aircraft traffic, but using the newrunways configuration whereby the western runway wasreplaced by the maritime runway. The simulated noisecontours are depicted in Fig. 5 and a comparison (with respectto the old runway) of the affected areas is shown in Fig. 6.While the total noise-impacted areas decreased by 5.5 percent(Fig. 6), using the hedonic price method (Eq. 1) the estimatedcost of noise was reduced by nearly 87 percent or 2.2 million

Table 3 – Average daily traffic at BIA.

Arrival Departure

Aircraft type Daya Eveningb Nightc Daya Eveningb Nightc

A310 9 1 2 10 0 2

A320 6 5 4 8 1 6

B707 2 0 0 2 0 0

B727 1 0 0 0 0 1

B737 0 0 3 0 0 3

B767 0 0 1 0 0 1

B777 2 0 0 2 0 0

MD82 1 0 1 1 0 1

TU54 1 0 0 1 0 0

Total 22 6 11 24 1 14

Grand Total 39 39

a 7:00 A.M.–7:00 P.M. b 7:00 P.M.–10:00 P.M. c 10:00 P.M.–7:00 A.M.

Med

iterr

anea

n S

eaN

0 1 2 3 Km

Fig. 3 – Day-night-level noise contours for the year 2000 trafficand existing runway.

Med

iterr

anea

n S

ea

N

0 1 2 3 Km

Fig. 4 – Zone subdivision of noise impacted areas.


USD per year. This sharp reduction is mainly due to the factthat when the new maritime runway is used, the most affluentand heavily populated zones A and B are no longer affectedby aircraft noise, as shown in Table 5.

Currently a night surcharge of 184 USD is applicable foreach movement (landing or take-off) at BIA.21 Based on thedaily average traffic, the yearly charges are 2.1 million USDper year. However this surcharge is for the usage of the runwayand apron lights only, and is not related to the noise categoryof the aircraft or its weighted noise impact increase.

Other operational measures that can be adopted to reduceaircraft noise include flight scheduling control andenforcement of environmental regulations. At present, about30 percent of the daily flights occur at night (between 10:00PM and 7:00 AM). Noise during the night is known to causea high level of annoyance. The DNL metric adds a 10 dBnight operation penalty to account for this fact. Most airportsin Europe and the US restrict the number of operations duringthe night, and many airports are charging additional fees foraircraft landing at night in accordance with its acoustic noise

category.12 At the BIA, international airline companies arethe main night operators. This can be attributed in part to thegeographic location of Beirut, which dictates its use as a transithub for the region. Another reason may be that the airlines inquestion are avoiding airports that ban night operations, andaccordingly they switch to airports with less stringentstandards on night operations. In addition, existingenvironmental standards are not generally enforced at the BIA.Many aircraft operating at the airport are not in compliancewith international noise standards. While Chapter 2 aircraft(such as the Boeing 747-200) are being phased out in severalairports22 and Chapter 1 (such as the Boeing 707) aircraft arebanned from most airports, they are still in use at the BIA.These aircraft are known to contribute to excessive noisepollution and they must be replaced.

5. SENSITIVITY ANALYSIS

Important parameters that directly affect noise contourswere examined to evaluate 1) how the results might change ifimportant parameters were varied; and 2) mitigationalternatives such as modifying aircraft operations or types.The fleet mix, the breakdown of day and night operations,and airport operations are the parameters considered in thissensitivity analysis. For this purpose the current traffic of 78operations per day (landing and departure), and the currentrunway configurations were used in two additionalsimulations. In the first scenario, all Chapter 1 aircraft (B707and B737) were replaced with Chapter 2 aircraft (B767-300)of comparable weight and capacity. Chapter 1 aircraft, whichare banned at most international airports, are still in operation

Table 4 – Residences within noise contours.

Zone Total number Average annual Percent of area in noise contour (A-wtd. level contours)

of residences18 rent (USD)19

55-60 dB 60-65 dB 65-70 dB 70-75 dB 75-80 dB 80-85 dB

A 42,849 18,150 26.3 — — — — —

B 54,173 10,890 17.9 10.7 1.2 — — —

C 18,876 3,630 19.4 11.9 9.4 9.4 2.5 1.3

Med

iterr

anea

n S

ea

N

0 1 2 3 Km

Fig. 5 – Day-night-level noise contours for the year 2000 trafficand new runway configuration.

20

16

12

8

4

0 – – –

– – –

Y 2000 old runways

Y 2000 new runways

Are

a (k

m2 )

55-60 dB 60-65 dB 65-70 dB 70-75 dB

15.5

11.9

6.96.0

3.6 2,7

1.0 0.9

Fig. 6 – Variation in noise impacted areas with new maritime run-way.


(12 daily operations) at the BIA. It is expected that theseaircraft would be replaced in the coming two years by Chapter2 and 3 airplanes. In the second scenario, all night operationswere replaced by day operations. Night operations (between10:00 P.M. and 7:00 A.M.) are known to cause the highestlevel of annoyance in communities near an airport, and theDNL metric accounts for this by applying a 10 dB night-timepenalty. Finally, in the third scenario, the departure and arrivaltrack paths were swapped: the western runway was used fortaking-off, while the eastern runway was used by approachingaircraft.

The DNL metric was used as an indicator in this sensitivityanalysis, when simulating the areas of noise contours. Theexclusion of Chapter 1 aircraft resulted in a 5 to 16 percentdecrease in noise-affected areas and the total social cost wasevaluated at 15.5 million USD as summarized in Table 6.Banning night operations would decrease the affected areasmore significantly reaching 60 to 69 percent and a total costof 6 million USD. This is mainly due to the high number ofnight operations (31 percent) and the 10 dB weighting factorapplied in the DNL metric. On the other hand, the change inthe runways operations resulted in minimal increase in theoverall impacted area (total social cost of 17.2 million USD).However, in the basic simulation a major part of the noisecontours from departing aircraft was over the sea, while thesecontours would be mainly above zone A residential-commercial area if the departure and arrival track paths areswapped.

6. STUDY LIMITATIONS

Despite recent improvements in INM simulation accuracy,the results should be considered carefully. Several elementsthat may attenuate the exposure of individuals to aircraft noisearound airports (i.e. topography of the area, existence ofbuildings, vegetation, and other types of sound barriers nearthe airport) are not yet fully considered in the INM.23 Inaddition, atmospheric conditions (i.e. humidity, winddirection, turbulence, etc.) are not simulated in the INMcalculations although it is well established that theseconditions have a direct effect on noise propagation.24 Thevalidity of the noise contours generated is a function of theaccuracy of the input data. While these data may be known tosome extent, unpredictable airline and airport decisions canlargely affect these variables. In addition, noise contours weredeveloped for an average 24-hour day, and actual dailyconditions may differ from the average conditions considered.

In calculating the cost of noise using the hedonic pricemethod (Eq. 1), the Noise Depreciation Index (I

NDI) value was

taken as 0.62 percent based on previous studies which maynot be representative of the characteristics of the study area.The cost of noise may have also been reduced if thebackground noise level was taken into account. On the otherhand, an A-weighted background noise level of 50 dB wasassumed as an average value for day and nighttime. Extensivenoise measurements are necessary to better assess temporalvariations, validate model simulations, and define a moreaccurate value for the background noise level. Also, in thehedonic economic valuation method, it is not usual, if notrare, that the buyer is informed, the relations between variablesare linear, or the variables are independent. In fact, noise may

Table 5 – Percent of area within noise contours when using new runways.

Zone Total number Average annual Percent of area in noise contour (A-wtd. level contours)

of residence18 rent (USD)19

55-60dB 60-65 dB 65-70 dB 70-75 dB 75-80 dB 80-85 dB

A 42,849 18,150 — — — — — —

B 54,173 10,890 6.0 3.6 0.6 — — —

C 18,876 3,630 15.6 8.1 7.5 8.8 1.9 1.3

Table 6 – Effects of Changing the Fleet Mix, Removing Night Operations, and Changing Runways Operation.

Fleet mix:No Chapter 1a No night operations Change runways operationNoise contour zone Do nothing

(A-wtd)(dB) area(km2) Area(km2) Difference(%) Area(km2) Difference(%) Area(km2) Difference(%)

55-60 23.43 22.32 5 8.80 62 23.60 -1

60-65 11.91 10.04 16 3.64 69 12.16 -2

65-70 4.90 4.61 6 1.72 65 5.15 -5

70-75 2.12 1.86 12 0.84 60 2.24 -6

Total Cost

(Million USD) 16.9 15.5 6.0 17.2

a Chapter 1 aircraft: licensed before 1970 (ICAO classification)


correlate directly with variables that can increase an asset’sworth such as distance from the airport. Last but not least,the social cost does not include occupational related expensesdue to employee potential illness, absenteeism, or reducedefficiency due to exposure to elevated and prolonged noiselevels inside the airport.

7. SUMMARY AND CONCLUSIONS

An economic assessment of aircraft noise impacts wasconducted at the Beirut International Airport. The FAA INMwas applied to simulate aircraft induced noise and to delineatenoise-affected areas for the current runway configuration, andwhen a new maritime runway is utilized. The hedonic pricemethod was adopted to estimate the social cost generated fromaircraft noise (in both simulations) and the cost was comparedto an international survey relating social cost to travel distance.Based on the hedonic price method, the social cost (excludingpotential occupational costs) attributed to aircraft traffic noisereached 16.9 million USD/year or 0.0038 $/passenger/kmtraveled which is within the range reported in internationalsurveys for many countries. The estimated cost of noiseimpacts was reduced significantly when the new maritimerunway is introduced (87 percent reduction). While there is anight surcharge on aircraft movements, this surcharge is notrelated to the noise category of the aircraft. Further noise/cost reduction could be attained by phasing out Chapter 2aircraft (such as Boeing 747-200) and Chapter 1 (such as theBoeing 707) aircraft.

8. ACKNOWLEDGMENTS

The study was funded by the Lebanese National Councilfor Scientific Research and the University Research Board atthe American University of Beirut. Special thanks areextended to the United States Agency for InternationalDevelopment for its continuous support of the EnvironmentalEngineering and Sciences Programs at the AmericanUniversity of Beirut.

9. REFERENCES1 K. Kryter, “Community annoyance from aircraft and ground vehicle noise,”

J. Acoust. Soc. Am. 72(4), 1222-1242 (1982).2 S. Lang, “Kids near airports don’t read as well because they tune out

speech,” Cornell Science News (April 28, 2997).3 A. Suter, “Noise and its effects,” Administrative Conference of the United

States (Nov. 1991).4 Guidelines for Community Noise, Edited by Birgitta Berglund, Thomas

Lindvall, and Dietrich Schwela (World Health Organization, Geneva,Switzerland, 1999).

5 Pollution Prevention and Abatement Handbook, General EnvironmentalGuidelines (World Bank Group, 1998)

6 Environmental Criteria for Sustainable Transport, Report on Phase 1 ofthe Project on Environmentally Sustainable Transport (EST) (Organizationfor Economic Co-Operation and Development, OCDE/GD(96)136, Paris,France, 1996).

7 Highway Traffic Noise Analysis and Abatement Policy and Guidance(Federal Highway Administration, Office of Environment and Planning,Noise and Air Quality Branch, Washington DC, 1995).

8 Information on Levels of Environmental Noise Requisite to Protect PublicHealth and Welfare with an Adequate Margin of Safety (US EnvironmentalProtection Agency, Office of Noise Abatement and Control (ONAC), ReportEPA550/9-74-004, Washington D.C., 1974).

9 Guidelines for Preparing Environmental Impact Statements on Noise(National Research Council, Assembly of Behavioral and Social Sciences,Committee on Hearing, Bioacoustics and Biomechanics (CHABA),Washington DC, 1977).

10 D. Levinson, D. Gillen, A. Kanafani, and J. M. Mathieu, The Full Cost ofIntercity Transportation—A Comparison of High Speed Rail, Air andHighway Transportation in California (Institute of Transportation Studies,University of California at Berkeley, 1996).

11 T. Morioka, T. Fujita, and N. Yoshida, “Performance and shortcomings oftypical environmental pollution control programs for automobile traffic inKobe City and surrounding areas. Social cost evaluation of noise pollutionby Hedonic Price Method,” The Science of the Total Environment 189/190, 99-105 (1996).

12 P. Morell and C. Lu, Social Costs of Aircraft Noise and Engine Emissions—A Case Study of Amsterdam Airport Schiphol (In Preprint CD ROM,Transportation Research Board, 79th Annual Meeting, Washington, DC,January 2000.

13 K. Saelensminde, “Stated choice valuation of urban traffic air pollutionand noise,” Transportation Research Part D 4, 13-27 (1999).

14 D. Haling and D. Cohen, “Residential noise damage costs caused by motorvehicles,” In Transportation Research Record 1559, TRB, NationalResearch Council, Washington, D.C., 84-93 (1996).

15 M. Delucchi and S. L. Hsu, “The external damage cost of noise emittedfrom motor vehicles,” J. Trans. Stat. 1(3) 1-24 (1998).

16 Dar Al-Handassah Shair and Partners, “Beirut international airport:Feasibility study,” (Council for Development and Reconstruction, Beirut,Lebanon, January 1994).

17 A. Filho, J. Braaksma, and J. Phelan, “Interpreting airport noise contours,”Transportation Research Record 1475, TRB, National Research Council,Washington, DC, 66-69 (1995).

18 Team International in association with IAURIF and SOFRETU, “GreaterBeirut Transportation Plan. Household Survey Results,” (Council forDevelopment and Reconstruction, Beirut, Lebanon, November 1994).

19 Lebanon Opportunities, Real Estate Average Market Prices per m2, April1999–April 2000.

20 1998 Passenger Traffic, Beirut International Airport. http://www.beirutairport.gov.lb/airport/statistics/statistics.htm. Accessed July 15, 2000.

21 Airport Charges, Beirut International Airport. http://www.beirutairport.gov.lb/airport/parking/parking.htm. Accessed July 15, 2000.

22 W. Meyer and W. Willkie, “A noise contour comparison of stage 3 Hushkitoptions for the Boeing 727-200” (In Preprint CD ROM, TransportationResearch Board, 79th Annual Meeting, Washington, DC, January 2000).

23 FAA, Integrated Noise Model (INM) User’s Guide; version 6.0 (1999)(Federal Aviation Administration, Office of Environment and Energy, AEE99-03, 1999).

24 TRC (Transportation Research Circular), Aircraft Noise Modeling(Transportation Research Board, National Research Council, No. 473, May1997).

Ž .The Science of the Total Environment 257 2000 133]146

Particulate matter in urban areas: health-basedeconomic assessment

M. El-FadelU, M. MassoudDepartment of Ci il & En¨ironmental Engineering, American Uni ersity of Beirut, Beirut, Lebanon

Received 1 February 2000; accepted 31 March 2000

Abstract

The interest in the association between human health and air pollution has grown substantially in recent years.Based on epidemiological studies in several countries, there is conclusive evidence of a link between particulate airpollution and adverse health effects. Considering that particulate matter may be the most serious pollutant in urbanareas and that pollution-related illness results in financial and non-financial welfare losses, the main objective of thisstudy is to assess the economic benefits of reducing particulate air pollution in Lebanese urban areas. Accordingly,the extent and value of health benefits due to decreasing levels of particulate in the air are predicted. Health impactsare expressed in both physical and monetary terms for saved statistical lives, and productivity due to different typesof morbidity endpoints. Finally, the study concludes with a range of policy options available to mitigate particulate airpollution in urban areas. Q 2000 Elsevier Science B.V. All rights reserved.

Keywords: Particulate matter; Health impacts; Dose]response function; Economic assessment

1. Introduction

Ž .Ambient particulate matter PM is composedof a heterogeneous mixture of particles varying insize and chemical composition. Particles differ in

U Corresponding author. American University of Beirut,Faculty of Engineering and Architecture, 850 Third Avenue,New York, NY 10022, USA. Fax: q1-961-1-744-462.

Ž .E-mail address: [email protected] M. El-Fadel .

sources, size ranges, formation mechanisms, andchemical composition and are characterized byvarious physical and chemical properties. Whilephysical properties affect the transport and depo-sition of particles in the human respiratory sys-tem, chemical composition determine their im-pact on health. A wide range of natural andanthropogenic emission sources contribute to PMconcentrations in the atmosphere such as wind-blown soil dust, marine and biogenic aerosol,road traffic and off-road vehiclesrmachines, sta-

0048-9697r00r$ - see front matter Q 2000 Elsevier Science B.V. All rights reserved.Ž .PII: S 0 0 4 8 - 9 6 9 7 0 0 0 0 5 0 3 - 9

( )M. El-Fadel, M. Massoud r The Science of the Total En¨ironment 257 2000 133]146134

tionary combustion processes, industrial and con-struction processes, and combustion of agricul-tural waste. Particles can be emitted directly fromsuch sources and are commonly referred to asprimary particulates, or formed in the atmo-sphere from gaseous precursors and are calledsecondary particulates. The chemical complexityof PM requires that sources of a large number ofprimary and secondary components be consideredŽ .De Nevers, 1995 .

Suspended particulate matter is ubiquitouslyrecognized as the most important air pollutant interms of human health effects considering thatmany epidemiological studies substantiate sig-nificant associations between concentration of PM

Žin the air and adverse health impacts USEPA,.1997; Vedal, 1997 . Fine particulates are likely to

be the most significant contributors to theobserved health effects, owing to their ability toaccumulate and reach the lower regions of therespiratory system. While the effects of PM varyconsiderably depending on its composition andsize distribution, generally, exposure to inhalablePM can cause an increase in cardiac and respira-tory mortality, a decrease in levels of pulmonarylung function in children and adults with obstruc-tive airways disease, an increase in daily preva-lence of respiratory symptoms in children andadults, an increase in functional limitations asreflected by school absenteeism or restricted ac-tivity days, and an increase in physician and emer-gency department visits for asthma and other

Žrespiratory conditions COMEAP, 1998; Miche-.lozzi et al., 1998 .

The best evidence that particulate air pollutionis causally associated with human adverse healthimpacts is provided by the mass of existing epi-demiological data. A number of time-series stud-

Žies using various measures of PM TSP, PM ,102y.1PM , COH, BS, SO have been widely re-2.5 4

ported in the literature. A relatively large numberof these studies adopted PM as an indicator.10The severity of health disorders is directly related

1Abbre¨iations: TSP, total suspended particualtes; PM ,10PM-10 mm in aerodynamic diameter; PM , PM of the 2.52.5mm size and less; COH, coefficient of haze; SO2y, sulfate.4

to the concentration of particulates in ambientair, which is often expressed in dose]response

Ž .functions DRFs that correlate mortality andmorbidity outcomes of susceptible populationgroups with ambient concentrations of a certainair pollutant. DRFs can also be derived for lesserhealth impacts, such as respiratory hospitaladmissions, emergency admissions, bed disabilitydays, restricted activity days, asthma attack, acuterespiratory symptoms, chronic bronchitis, lowerrespiratory illness, and others. Table 1 representsa summary of health impact DRFs for PM ,10PM , BS, and TSP as derived from an array of2.5worldwide literature-based sources. Sensitivegroups that appear to be at a greater risk forparticulate air pollution include the elderly, thosewith pre-existing respiratory conditions and car-dio-pulmonary diseases such as asthma, smokers,and children.

Global annual deaths as a result of air pollu-tion are estimated at more than 2.7 million, with

Žcities accounting for approximately 33% WHO,.1997 . Approximately 1.4 billion urban residents,

mostly in developing countries, may be exposed toair with borderline or unacceptable levels of par-

Žticulates ALA, 1998; Gamble, 1998; AEAT,.1999 . This paper assesses the health impacts of

particulate air pollution in Lebanese urban areas.The economic benefits due to decreasing levels ofambient air particulates are estimated. Healthimpacts are expressed in both physical and mone-tary terms for saved statistical lives, and produc-tivity due to various morbidity endpoints. Thestudy concludes with a range of policy optionsand mitigation measures to minimize TSP levelsin the air.

2. TSP measurements in Lebanon

Air samples collected from several locations inŽ .Beirut Fig. 1 revealed that TSP concentration

ranges from 102 to 291 mgrm3 with an average3 Ž .value of 166 mgrm Fig. 2 . In addition to vehi-

cle-induced emissions, movement of motor vehi-cles on dusty roads and on-going constructionactivities are generally the major potential sourcesfor particulates. Anthropogenic sources coupled

( )M. El-Fadel, M. Massoud r The Science of the Total En¨ironment 257 2000 133]146 135

Table 1aŽ .Summary ranges of worldwide health impacts DRFs for PM , PM , BS, and TSP Vedal, 199710 2.5

Change in PM Percent increase Percent increase Morbidity typebŽ . Ž .concentration in mortality % in morbidity %

3Increase of 10 mgrm in PM 0.1]4.6 0.2]2.9 Pneumonia hospital admissions100.8]11.5 COPD hospital admissions0.2]6.4 Respiratory hospital admissions0.6]1.2 Cardiac hospital admissions0.4]6.0 Emergency cases of asthma0.3]0.4 Bronchitis hospital admissions1.1]24.9 LRI symptoms0.4]13.0 URI symptoms1.6]17.6 Cough symptoms

3Increase of 10 mgrm in PM 0.4]3.7 0.41]24.6 Respiratory hospital admissions2.53.7]20.9 Asthma hospital admissions

3Increase of 10 mgrm in BS NR 0.07]18.2 Respiratory hospital admissions0.3]5.3 Asthma hospital admissions1.2]16.5 COPD hospital admissions

3Increase of 100 mgrm in TSP 3.3]8.3 NR NR

aAbbre¨iations: COPD, chronic obstructive pulmonary disease; LRI, lower respiratory illness; URI, upper respiratory illness; NR,not reported.

b Morbidity } the incidence of respiratory andror cardiovascular symptoms and diseases.

with the nature of the dry Lebanese climate,particularly during the summer, results in high

Ž .dust levels in the atmosphere ERM, 1995 . Whilethe measurements serve to give a general indica-tion of particulates at various urban junctions, theclear implication is that anthropogenic activitiescontribute substantially to these levels.

2.1. Air quality standards

Based on clinical, toxicological and epidemio-logical evidence, guideline values of ambient par-ticulate concentrations were established by de-termining concentrations with the lowest observedadverse effect and adjusted by an arbitrary mar-gin of safety factor to allow for uncertainties inextrapolation. Generally, the most frequently usedreference guidelines for PM are those set by the

Ž .World Health Organization WHO , the Euro-Ž .pean Union EU , and the United States Environ-

Ž . Ž .mental Protection Agency USEPA Table 2 .While WHO guidelines are based on health con-siderations only, standards determined by the EU

and USEPA reflect the technological feasibility ofattainment as well. Many countries adopted theseguidelines or else they established their own am-bient air quality standards.

In Lebanon, ambient air quality standards havebeen proposed within the 1994 Urgent Draft Lawconcerning the determination of the specifica-tions and levels for the prevention of air, water

Ž .and soil pollution MOE, 1996 . However, it hasnot been approved to date. These standards seemto be political or administrative settings only.They do not provide a scientifically representative

Žpicture of particulate standards Staudte et al.,.1997 . The limit values are lower than some inter-

national standards and seem to be unreachableunder current emission practices where no con-trol is enforced.

3. Economic valuation of health impacts

Valuing the health impacts of air pollution is amajor problem facing policy makers. It comprisesthe actual identification and measurement of


Fig. 1. Air quality sampling locations in Beirut.

Fig. 2. Average particulates concentrations at various locations in Beirut.


Table 2aŽ .Reference standards and guidelines of average ambient particulate concentration USEPA, 1996; WHO, 1997; RECCEE, 1998

3 3 bŽ . Ž .International Long-term mgrm Short-term mgrmstandard PM BS TSP PM BS TSP10 10

EU limit values NA 80 150 NA 250 300EU guide values NA 40]60 NA NA 100]150 NAUSEPA 50 NA 260 150 NA 75WHO guidelines NA 40]60 60]90 NA 100]150 150]230WHO guidelines NA 50 NA 70 125 120

for EuropeLebanon NA NA NA 80 NA 120

a NA, not available.b24 h.

health impacts and the estimation of monetaryvalues for associated premature mortality andmorbidity. Generally, health damage studiesproceed by establishing average levels of ambientconcentration of a pollutant and then relatingthose concentrations to health effects throughDRFs. Local or country-specific epidemiologicalstudies are the most appropriate indicators ofhealth impacts associated with air pollution in agiven region. These studies establish DRFs linkingenvironmental variables to observed health ef-fects. However, given the time and cost involvedin such studies as well as the problems encoun-tered with data availability, DRFs established inother countries can be adopted assuming thathuman reaction is similar in different locations.Consensus DRFs are possible if there are numer-ous reliable studies that appear to converge, asappears to be the case for PM. The next step is torelate the DRF to the population at risk and then

Žapply unit economic values USGAO, 1994;Calthrop and Maddison, 1996; Pearce and

.Crowards, 1996; Hartman et al., 1998 .Methods used in various studies to value health

costs associated with environmental pollution canŽ .generally be grouped into two categories: 1 those

that measure only the loss of direct income suchŽ . Ž .as lost wages or cost of illness COI ; and 2

approaches that attempt to capture the willing-Ž .ness to pay WTP individuals for avoiding or

reducing the risk of death or illness. The firstcategory does not include inconvenience, suffer-ing, losses in leisure and other less tangible im-pacts to the individuals well being. They may also

underestimate the health cost of people who arenot members of the labor force. Thus, thesemethods indicate only the lower bound of thesocial cost and understate the total cost to indi-

Ž .viduals Larssen et al., 1997 . A summary of vari-ous valuation methods of health effects resultingfrom air pollution is represented in Table 3.

4. Health assessment

The first step in a health assessment is toestablish a DRF between an increase in PM andadverse health effects. Considering that the datawere acquired from studies worldwide, severalassumptions were employed in this assessmentincluding:

v there is no threshold below which PM is10harmless or not a cause of mortality;

v there is no difference in susceptibility or expo-sure between different populations;

v reviewed studies are of similar quality andneed not be weighted for differences inmethodology or sample size;

v where an age-specific DRF is unavailable, theestimate for all age groups will be applied tothe baseline number of deaths in each agegroup; and

v the estimations are not restricted to a particu-lar or an average value but ranges of valuesare considered in order to ensure a broaderperspective of the subject.


Table 3aEconomic valuation methods for the health effects of environmental pollution

Valuation method Description

Mortality effectsHuman capital The human capital approach places a value on a premature death. It is based on the

economic productivity of an individual and values life according to the net presentvalue of productivity of an individual. As such, individuals are considered as units ofhuman capital that produce goods and services for society. The value of each unitof human capital is equivalent to the present value of the future output, in the formof earning, that might have been generated had the individual not died prematurely.In the absence of WTP estimates, this approach provides the best availablealternative for valuing loss in productivity.

WTP Unlike the human capital approach that measures tangible changes in productivity,the WTP captures intangible aspects. The WTP consists of asking people directlywhat they would be willing to pay for reduced risk of increased mortality.

Morbidity effectsCOI The COI approach applies mostly to morbidity and is consistent with or similar to the human

capital approach. The direct cost of morbidity can be divided into two categories:Ž . Ž .1 Medical expenditure for treating illness; and 2 lost wages during days spent in bed,days missed from work, and other days when activities are significantly restricted to illness.Most COI estimates fall short of being full estimates because of insufficientinformation. Even full COI estimates will understate total WTP because they do notinclude the value of avoiding the pain and suffering associated with the illness thatnecessitates hospital admission.

WTP The WTP approach is a more theoretically sound measure of morbidity effects. Itestimates what people would be willing to pay to avoid illness and can be inferred

Ž .using two different methods: 1 The averting behavior method, which is based on thenotion that the time and money spent by an individual to avoid exposure to airpollution or avoid illness is indicative of a lower bound value hershe attaches to

Ž .avoid it. 2 The contingent valuation method, which uses survey information todetermine what people are willing to pay to avoid a certain symptom or illness.Unit valuations that rely exclusively on the contingent valuation method includechronic bronchitis, respiratory related diseases, minor restricted activity days, andvisibility.

aAbbre¨iations: COI, cost of illness; WTP, willingness to pay.

4.1. Mortality characterization

Total lives saved due to a reduction in PM

concentration in the air is dependent on thebaseline number of deaths in the country, whichis calculated by multiplying the size of the ex-

Table 4Percent distribution of death in Lebanese households

a bSex Age group

0]9 10]19 20]39 40]59 60]69 )70 Unknown Total

Ž .Male 51 5.54 2.67 9.60 18.31 21.35 37.48 5.07 100.00Ž .Female 49 6.50 2.91 4.00 13.30 17.18 51.13 5.00 100.00

Ž .Total 100 5.93 2.77 7.31 16.27 19.66 43.03 5.03 100.00Total number of deaths 808 375 921 2131 2593 5935 677 13 440

a Ž .US Bureau of the Census 1999 .b Ž .MSA 1996 .


Table 53Distribution of predicted lives saved per year due to 10 mgrm reduction in PM10

Sex Age group

0]9 10]19 20]39 40]59 60]69 )70 Unknown Total

Ž .Male 51 0]17 0]8 1]30 1]58 1]67 3]118 0]16 6]314Ž .Female 49 0]20 0]9 0]12 1]40 1]52 3]155 0]15 5]303

Ž .Total 100 0]37 0]17 1]42 2]98 2]119 6]273 0]31 11]617

posed population with the death rate. Based on adeath rate of 8.2 deathsr1000 person per yearŽ .MSA, 1996 and a total population correspond-ing to the Lebanese urban areas of approximately

Ž .1.64 million ERM, 1995 , the baseline number ofdeaths is calculated to be approximately 13 440.Multiplying the value of total deaths with thepercent distribution of death in Lebanese house-hold by age and sex, results in the total number of

Ž .deaths by age group Table 4 .The calculation of the number of lives saved by

age group involves multiplying the values of base-line number of deaths in each age group with thepercent change in the number of cases due to aspecific reduction in PM , which is assumed to10be 10 mgrm3 in this study. As mentioned previ-ously, where age-specific DRFs are not available,the estimate for all age groups will be applied tothe baseline number of deaths in each age group.Based on epidemiological time series studiesŽ .Vedal, 1997; Gamble, 1998 , the decrease inmortality due to 10 mgrm3 reduction in PM10ranges between 0.1 and 4.6%. Accordingly, thepredicted total number of lives saved in Lebanese

Ž .urban areas ranges between 11 and 617 Table 5with average values depicted in Fig. 3. Note thatwhile it is preferable to use DRFs from country-specific studies, these are not available for mostdeveloping countries.

4.2. Morbidity characterization

The assessment was performed on the effect ofdecreasing 10 mgrm3 of PM in the air with10pneumonia, COPD and emergency visits as end-points. Similar to mortality assessment, morbiditycalculations are performed in two steps. First, thetotal number of hospital admissions of each healthendpoint is determined by multiplying the total

number of hospital admissions in urban areaswith the percent of health endpoint hospitaladmissions. The data used in order to performthese calculations are summarized in Table 6.Second, multiplying the value obtained in the firststep by the percent decrease in health endpointhospital admissions due to a 10 mgrm3 reductionin PM provides the total number of cases10avoided. Accordingly, the ranges of pneumoniaand COPD cases avoided per 10 mgrm3 reduc-tion in PM are 15]214 and 35]498, respectively10Ž .Table 7 . Similarly the number of emergencyvisits avoided ranges between 609 and 25 578cases.

Considering that there are no adequate datafor all morbidity effects, morbidity DRFs per per-son due to a 10 mgrm3 reduction in PM can be10used to compute the morbidity cases avoided inurban areas. The assumption again is that suchDRFs can be transferred across societies. In thiscase, DRFs are multiplied by the total urban

Ž .population 1.64 million given that the impact ismeasured per person. Thus, Table 8 representsthe predicted values of morbidity effects.

Fig. 3. Predicted average number of lives saved due to 10mgrm3 reduction in PM .10


Table 6Data used to perform morbidity health assessment

Total hospital admissions per Lebanon 400 000ayear Beirut 133 000

Other urban areas 53 200

Type of hospital admission Emergency visits in Beirut 145 000bper year Emergency visits in other urban areas 58 000

Ž . Ž .a Respiratory and cardiac hospital admissions % 15Ž . Ž .b Respiratory admissions % of a 37Ž . Ž .c COPD admissions % of b 37Ž . Ž .d Pneumonia admissions % of b 63

Percent decrease in Pneumonia hospital admissions 0.2]2.93morbidity due to 10 mgrm COPD hospital admissions 0.8]11.5

creduction in PM Emergency visits 0.3]12.610

a Ž .Total number of hospital admissions in Lebanon is obtained from the Central Bank of Lebanon 1998 Annual Report,whereby one-third of the total hospital admissions occurred in Beirut. The hospital admissions in other urban areas is assumed tobe proportional to the ratio of inhabitants in Beirut and other urban areas.

b Total number of emergency room visits per day is extracted from a project conducted by senior medical students at theAmerican University of Beirut in 1993. The percentage of patients admitted to hospitals for respiratory and cardiovascular diseases

Žwas estimated by contacting three major hospitals in Beirut. Respiratory diseases include COPD bronchitis, chronic bronchitis,. Žemphysema, asthma, bronchiectasis, and chronic airway obstruction all pneumonias pneumococcal, other bacterial, infectious

.diseases, and bronchopneumonia , acute laryngitis and tracheitis, acute upper respiratory infections, and acute bronchitisŽ .Djoundourian et al., 1998 .

c Literature-based time series studies.

5. Economic assessment

5.1. Mortality calculations using human capitalapproach

Air quality-related mortality typically occurs ata late age due to a long period of chronic expo-sure to inferior air quality. Therefore, most stud-ies measure compensation of mortality risks forindividuals who are, on average, approximately 40years of age. In the present study, the total socio-economic cost due to premature mortality in ur-ban areas was predicted assuming that:

Table 73Morbidity cases avoided per year due to 10 mgrm reduction

in PM10

Endpoint Cases Casesoccurringryear avoidedryear

COPD 4334 31]441Pneumonia 7379 13]189Emergency visits 203 000 609]25 578

v Based on the per capita GDP in Lebanon forthe year 1998, the average Lebanese monthly

Žsalary is approximately US$400 Audi Bank,.1998 .

v Productivity age ranges between 25 and 69years.

v The two age groups 40]59 and 60]69 arerepresentative death ages with correspondingaverage lost productivity years of 20 and 5,respectively.

v The estimations are not restricted to a partic-ular or average value but ranges of values areconsidered in order to ensure a broader per-spective of the subject.

Based on the health assessment, the ranges oflives saved per 10 mgrm3 reduction in PM for10the age groups 40]59 and 60]69 are 2]10 and3]122, respectively. Multiplying average produc-tivity years by the average income yields the total

Ž .economic benefits Table 9 . Therefore, it is ben-eficial to control particulate emissions, as it con-stitutes a significant productivity source relative


Table 83 Ž .Change in morbidity effects per 10 mgrm reduction in PM Pearce and Crowards, 199610

Endpoint Change Total cases avoided

Respiratory hospital 6.6]17.3 108]284admissionsr100 000

Emergency department 116.0]354.0 1902]5806visitsr100 000

aLower respiratory illnessr 0.010]0.024 4756]11 414asthmatic child

Asthma attacksr 0.33]1.96 541 200]3 214 400person

Respiratory symptomsr 0.8]2.56 1 312 000]4 198 400person

Chronic bronchitisr 30.0]93.0 492]1525100 000

Restricted activity 0.29]0.58 475 600]951 200daysrperson

a Ž . Ž .Taking 29% of the total urban population 1.64 million between 0 and 14 US Bureau of the Census, 1999 .

Ž .to country economic resources up to 1% of GDP .While the change in PM seems small in terms ofhealth risk, they signify a substantial number ofavoidable deaths due to the size of the populationimpacted.

5.2. Morbidity calculations using cost of illnessapproach

The predicted number of cases avoided due to10 mgrm3 reduction in PM multiplied by the10cost of the corresponding health endpoint resultsin the total economic benefits due to morbidity

Ž .avoidance Table 10 . Similar to mortality, mor-bidity calculations imply a considerable increasein economic benefits.

5.3. Mortality calculations using willingness to payapproach

Generally, WTP estimates are lacking in mostcountries. Consequently, the value of a statisticallife from several US studies is adjusted by the percapita GNP ratio which is approximately 0.1 based

Ž .on country-specific data World Bank, 1999 .Table 11 represents a summary of mortality valu-ation estimates based on the individual WTP forsmall reductions in mortality risk.

Each study provides an estimate of the meanWTP to avoid a statistical premature death. Mul-tiplying the per capita GNP ratio by the range of

Ž .WTP estimates 0.6]13.5 MUS$ , the value of astatistical life in Lebanon would range between0.06 and 1.35 MUS$. It is reasonable to obtainsuch a relatively high range considering that theWTP captures the value that an individual assignsto measurable and less tangible effects.

5.4. Morbidity calculations using willingness to payapproach

Willingness to pay to avoid a day of specificmorbidity endpoint has been estimated by only asmall number of studies. However, total benefitsassociated with any reduction in pollutant con-centrations is determined largely by the benefitassociated with the corresponding reduction inmortality risk because the dollar value associatedwith mortality is significantly greater than anyother valuation estimate. In the case of hospitaladmissions, the COI substitutes for the WTPestimates due to lack of information regardingthe latter. These COI estimates are likely tosubstantially understate the total WTP to avoidan illness or a particular hospital admission. Theaverage ratio of health care in Lebanon to that in


Table 93Mortality related economic benefits due to a 10 mgrm reduction in PM10

Age group likely Number of Average productivity Economic benefitŽ .to be affected lives saved years MUS$ryear

40]59 2]102 20 0.2]9.760]69 3]122 5 0.07]2.9Total 0.27]12.6

aAverage per case 0.055

aAverage total economic benefit divided by the average total number of lives saved.

Table 103Economic benefits due to reduced morbidity per 10 mgrm reduction in PM10

Endpoint Average stay Average cost Economic benefita aŽ . Ž . Ž .days US$rday MUS$ryear

COPD 6.6 261 0.06]0.9Pneumonia 10 207 0.03]0.4Emergency visit ] 76 0.05]1.9Total 0.14]3.2

a Ž .Based on survey data from the American University Hospital and insurance companies Djoundourian et al., 1998 .

the US is estimated at approximately 0.24. Thisratio was obtained by comparing the cost of COPD

Žand pneumonia in Lebanon Djoundourian et al.,. Ž .1998 and the US USEPA, 1997 . On the other

hand, the per capita GNP ratio of Lebanon andŽ .the US is 0.1 World Bank, 1997 , reflecting that

in Lebanon the cost of health care is relativelyvery expensive in comparison to income. Conse-quently, the COI estimates in the US are multi-plied by the ratio of health care in the two coun-tries, resulting in approximate estimates forLebanon. Similarly, in the case of work loss days,mild restricted activity days, chronic bronchitis,and respiratory-related illnesses or symptoms,which rely exclusively on the contingent valuationmethod, the per capita GNP ratio in the twocountries is used to obtain a value for LebanonŽ .Table 12 .

Table 11Ž .Summary of mortality valuation estimates USEPA, 1997

Study type ValuationrcaseaŽ .MUS$ryear

Labour market 0.6]13.5Contingent valuation 1.2]3.3

a1990 Dollar value.

5.5. Limitations and uncertainty

In evaluating epidemiological studies as a

Table 12Health effect unit valuation

Endpoint US valuation LebaneseaŽ .valuation US$ per case valuation

Ž .Us$ per case

bHospital admissionCOPD 8100 1944Pneumonia 7900 1896All respiratory 6100 1464

cRespiratory illness or symptomChronic bronchitis 260 000 26 000Acute bronchitis 45 5Acute asthma 32 3Acute respiratory symptoms 18 2Upper respiratory symptoms 19 2Lower respiratory symptoms 12 1

cRestricted acti ity dayWork loss days 83 8Mild restricted activity days 38 4

a Ž .1990 Dollar value USEPA, 1997 .b Lebanese valuation is obtained by multiplying the US

Ž .valuation by the health care ratio 0.24 .c Lebanese valuation is obtained by multiplying the US

Ž .valuation by the per capita GNP ratio 0.1 .


whole, many issues arise, primarily associated withcausality. While the mortality effects attributed toPM have generally been consistent, disagreementremains as to whether these effects can be at-tributed entirely to fine particles. The fact thatnegative correlations between PM exposure anddifferent health endpoints have been indicated incertain studies raises uncertainty about the truedose]response relationship. Associations foundin a particular study may reflect chance, bias, orcause. In addition, epidemiological studies wereconducted in various cities under a broad rangeof environmental conditions, by a number of dif-

Žferent investigators Frey, 1998; Rabl and.Spadaro, 1999 . Real world exposures involve

combinations of potentially toxic materials thatmay be inhaled together or sequentially underdifferent conditions. Possible interactions thatmay result from the inhalation of a mixture oftoxicants can include simply additive, synergisticor antagonistic effects. Further uncertainties inthe evidence of a causal relationship in theobserved association between adverse health im-pacts and increase in airborne particles includeŽ .Larssen et al., 1997; Lipfert and Wyzga, 1997 :

v lack of an accepted biologically demonstratedmechanism;

v lack of quantitative support from experimen-tal animal andror human clinical studies;

v confounding, and difficulty of separating ef-fects of co-occurring pollutants;

v misclassification of personal exposure to am-bient particles;

v estimates relate to all PM regardless of10source; and

v differences in socio-demographic factors andthe health status of the exposed population.

The problems in extrapolating DRFs are alsoexacerbated by potential inaccuracy in estimatingeconomic health impacts. The economic costs ofmortality and morbidity were predicted on thebasis of epidemiological studies reviewed in theliterature. This is attributable to the fact thatLebanon, similar to many countries, lacks popula-tion-based vital and disease registries. The associ-

ated uncertainty includes both selection of scien-tific studies and statistical uncertainty from theoriginal studies. Similarly, range of estimates formonetized benefits is based on the quantifieduncertainty associated with the health effects esti-mates and the unit valuations applied to them.Moreover, due to the lack of affluence and publicperception, the WTP approach cannot be per-formed and thus, increasing the uncertainty bytransferring estimates across countries.

Uncertainty about the true dose]response rela-tionship of PM and health endpoints and theuncertainty regarding the extrapolation should notdelay the implementation of control measures,particularly that the true association may likely bestronger than that observed in epidemiologicalstudies. Moreover, even a small effect such asincrease in total mortality associated with a 10mgrm3 increase in daily PM would have a large10impact at the population level.

6. Air quality management

Particulate emission reduction will have multi-dimensional benefits considering the adverse ef-fect of particulate air pollution on health and theenvironment. In addition to the improved healthstatus of the population, a decrease in pollution

Žlevels will reduce work absence arising from.health problems , and the costs of health insur-

ance. Thus, new, stricter and enforced standardswill provide increased health protection due tosaved lives, lower hospital admissions and emer-gency room visits, reduced risk of symptoms asso-ciated with chronic bronchitis and asthma, andreduced risk of respiratory symptoms in children.In many developing countries, there is a lack ofinstitutional capacity and technical expertise toadequately address environmental issues. Newlegislation may fail to meet its objective unlessseveral combinations of measures are simultane-ously implemented alongside.

Proper air quality management requires an in-tegrated approach that encompasses coordinationand consensus building across sectors, identifica-tion of technically feasible abatement options,


and introduction of policies and instruments tosupport implementation. A clear set of objectivesand priorities for the environmental policy needto be determined and related to overall develop-ment and growth goals. Identifying a target groupof serious polluters that can be regulated effec-tively constitutes the basis of any environmentalmanagement program. Improving environmentalperformance of identified polluters requires theadoption and enforcement of updated environ-mental standards. These standards have to be ofrealistic nature in order to attain higher levels ofcompliance and polluters should be given ade-quate time before standard implementation is

Ž .effective Califano, 1996 .Incentives and economic instruments such as

charge systems, fiscal and financial instruments,property rights, market creation, and liability sys-tems are essential for medium term application ofa comprehensive environmental strategy. For in-stance, a wide range of market-based instrumentsare applicable at the level of the transportationsector such as taxes on emissions from fuel andprivate vehicle ownership. Moreover, measures tomitigate the negative effects of pollution mayfocus on separating pollution sources and recep-tors, reducing the pollution activity, and its char-acteristics, and controlling emissions with airquality control devices. Implementing alternativesof emissions reduction vary across pollutionsources. Mitigation measures that should beadopted to reduce air pollution, particularly PMinclude:

v Improving fuel quality or introducing fuel al-ternatives and compulsory vehicle testing andmaintenance at state controlled and certifiedgarages.

v Adopting proper construction measures suchas site enclosure, on-site mixing and unloadingoperations, adequate maintenance and repairof construction machinery, minimal trafficspeed on-site, and proper water spraying whennecessary.

v Installing proper end-of-pipe control tech-nologies at industrial facilities such as elec-trostatic precipitators and baghouses.

7. Summary and conclusions

Epidemiological studies conducted in severalcountries show consistent associations of expo-sure to ambient particulates with adverse healtheffects including increased mortality, hospitaliza-tion for respiratory or cardiovascular diseases,and respiratory symptoms and decreased lungfunction. Based on epidemiological time seriesstudies, dose]response functions were identifiedbetween an increase in PM and adverse healtheffects. Accordingly, mortality and morbidityeconomic valuations were performed forLebanese urban areas.

Overall, the assessment showed that potentialhealth and economic benefits of reducing PMconcentration in the air can be significant. Asummary of the average economic benefit for themain health endpoints is represented in Table 13.These benefits can be dominated by mortalityvaluation. While the number of mortality cases isrelatively small, the wide range of monetary valuecan result in large monetary benefits. The rangesare indicative of the uncertainties inherent insuch an exercise within a specific methodologyand across methodologies. Despite such limita-tions, the striking feature of these estimates isthat high benefits could result from reductions inconcentrations of PM in the air.

Implementation of monitoring and setting ofŽenforceable regulations i.e. ambient air quality

.and emission standards are necessary to initiatea comprehensive air quality management pro-gram. Monitoring activities and regulations mustbe developed, taking into consideration localsocio-economic and technical characteristics.Moreover, it is essential to develop a strategy thatcan be used to evaluate the willingness to pay torefine the economic valuation of exposure to PM.

Acknowledgements

The authors wish to express their gratitude toMr E. Bou Zeid and Mr H. Sbayte at the Depart-ment of Civil and Environmental Engineering,


Table 133 aEconomic benefit due to 10 mgrm reduction in PM10

Ž .Endpoint Number of cases Total economic benefit MUS$ryearavoided COI WTP

b cMortality 11]617 0.27]12.6 3.5]157.9All COPD 31]441 0.06]0.9 0.98]13.9All pneumonia 13]189 0.03]0.4 0.05]0.7Emergency visits 609]25 578 0.05]1.9 NATotal 0.41]15.8 4.53]172.5

dPercent of GDP 0.003]0.1 0.03]1.06ePercent of adjusted GDP 0.03]1 0.3]10.6

aAbbre¨iations: COI, cost of illness; WTP, willingness to pay; NA, not available.b Human capital approach.c Determined by multiplying the value of a statistical life in Lebanon by the number of cases avoided between the age groups

40]59 and 60]69 in order to compare with the human capital approach value.d Ž .World Bank 1999 .eAdjusted GDP assuming that the construction and transportation sectors are the main sources of particulate emissions in urban

areas and accounting for sourcersector contribution to GDP and percent of urban population exposed as compared to total countrypopulation.

American University of Beirut, for conductingfield measurements used in this study. Specialthanks are extended to the United States Agencyfor International Development for its continuoussupport to the Environmental Engineering andScience programs at the American University ofBeirut.

References

Ž .Atomic Energy Authority Technology AEAT . Quantificationof the health effects of air pollution in the UK for the

wreview of the national air quality strategy http:rrwww.x Ž .aeat.co. html 1999 .

Ž .American Lung Association ALA . Particulate matter airpollution. American Lung Association fact sheet.w x Ž .http:rrwww.lungusa.orgrairrpmy factsheet.html 1998 .

Ž .Audi Bank. Lebanon facts and figures 1996]1998 1998 .wCalifano, R.J. Risk assessment of air emissions. http:rrwww.

x Ž .pirnie.com.html 1996 .Calthrop E, Maddison D. The dose]response function ap-

proach to modeling the health effects of air pollution.Ž .Energy Policy 1996;24 7 :599]607.

Central Bank of Lebanon. Annual Report, Beirut, LebanonŽ .1998 .

ŽCommittee on the Medical Effects of Air Pollutants COM-.EAP . The quantification of the effects of air pollution on

whealth in the United Kingdom. http:rrwww.x Ž .doh.gov.uk.html 1998 .

De Nevers N. Air pollution control engineering. New York:McGraw Hill, Inc, 1995.

Djoundourian S, Nuwayhid I, Chaaban F. The economic andsocial impacts of mobile source pollution in Beirut,Lebanon: A country case study, METAP III, Med Policies

Ž .Initiatives 1998 .Ž .Environmental Resources Management ERM . Lebanon As-

sessment of the State of the Environment, Council ofŽ .Development and Reconstruction, Beirut, Lebanon 1995 .

Frey HC. Methods for quantitative analysis of variability anduncertainty in hazardous air pollutant emissions. Air andWaste Management Association’s 91st Annual Meetingand Exhibition, June 14]18, 1998, San Diego, California,

Ž .98]RP105B.01 1998 .Gamble JL. Effects of ambient air pollution on daily mortal-

ity: a time series analysis of Dallas, Texas, 1990]1994. Airand Waste Management Association’s 91st Annual Meet-ing and Exhibition, June 14]18, 1998, San Diego, Califor-

Ž .nia, 98]MP26.03 1998 .Hartman RS, Wheeler D, Singh M. The cost of air pollution

abatement. World Bank, Policy Research Working PaperŽ .a1398 1998 .

Larssen S, Gronskei KE, Hanegraaf MC et al. In: Shah JJ,Nagpal T, Brandon C, editors. Urban air quality manage-ment strategy in Asia guidebook. Washington, DC: WorldBank, 1997.

Lipfert FW, Wyzga RE. Air pollution and mortality: theimplications of uncertainties in regression modeling andexposure measurement. J Air Waste Manage Assoc 1997;47:517]523.

Michelozzi P, Forastiere F, Fusco D et al. Air pollution anddaily mortality in Rome, Italy. Occup Env Med1998;55:605]610.

Ž .Ministry of Environment MOE . Proposed Law number 1r52.Standards and regulations for limiting air, water and soil


Žpollution. Ministry of Environment, Beirut, Lebanon In. Ž .Arabic 1996 .

Ž .Ministry of Social Affairs MSA , Population and housingŽ .survey, Beirut, Lebanon 1996 .

Pearce D, Crowards T. Particulate matter and human healthŽ .in the United Kingdom. Energy Policy 1996;24 7 :609]619.

Rabl A, Spadaro JV. Damages and costs of air pollution: anŽ .analysis of uncertainties. Env Int 1999;25 1 :29]46.

Regional Environmental Centre for Central and Eastern Eu-Ž .rope RECCEE . Reduction of SO and particulate emis-w x Ž .sions. http:rrwww.rec.org 1998 .

Staudte M, Rau M El-Fadel M. Urban air quality monitoringprogram for the greater Beirut area, Transtec-FitchnerConsortium, Ministry of Environment, Beirut, LebanonŽ .1997 .

US Bureau of the Census, International Data Basew x Ž .http:rrwww.census.govrgci-binripcr idbsum 1999 .

Ž .United States Environmental Protection Agency USEPA .Air quality criteria for particulate matter. USEPA, Office

Ž .of Air and Radiation, USEPA 600rP-95r001aF 1996 .

Ž .United States Environmental Protection Agency USEPA .The benefits and costs of the clean air act, 1970 to 1990,

Ž .EPA 410-R-97-002, Office of Air and Radiation 1997 .Ž .United States General Accounting Office USGAO . Air pol-

lution: EPA’s progress in determining the costs and bene-fits of clean air legislation. Resources, Community, andEconomic Development division, GAOrRCED-94-20Ž .1994 .

Vedal S. Ambient particles and health: lines that divide. CritRev J Air Waste Manage Assoc 1997;47:551]581.

Ž .World Health Organization WHO . WHO guidelines for airwquality. Fact Sheet No. 187 http:rrwww.who.intrinf-

x Ž .fsrenrfact187.html 1997 .wWorld Bank, Lebanon at a Glance http:rrwww.

xworldbank.orgrdatarcountrydataraagrlbnyaag.pdfŽ .1999 .

World Bank. Pollution prevention and abatement handbookw x Ž .} part III http:rrwww-esd.worldbank.org.html 1997 .




Sessions 16 &17The Value of Life and Health

GROUP EXERCISES


1 of 14

SESSIONS 15-16

GROUP EXERCISE 1

Impact of Water Quality on Health in Syria (World Bank, 2004)

Case description Syria is a relatively water scarce country. According to the World Bank statistics, renewable freshwater resources per capita are around 2,700 cubic meters. While this is higher than average for the countries of the Middle East and North Africa, it is less than a third of the world average. Moreover, water availability is unevenly distributed across the country in relation to population centers and irrigated agricultural land, resulting in local pressures on water resources, declining groundwater tables, and water quality problems. In terms of water and sanitation services, Syria reported that only 64 percent of the rural population had access to an improved water source in 2000. For urban areas, the water coverage rate was 94 percent, but water is being pumped from distant sources for some major urban centers due to local water quality and scarcity problems. For sanitation, nearly 20 percent of the rural population was reported as lacking access to hygienic sanitation facilities. The aim of this study is to estimates environmental damage costs associated with the health impacts of low quality potable water, inadequate sanitation and hygiene, and the economic impacts of water resources pollution. 1. Calculate the number of DALYs from child diarrheal disease deaths: Given: Live births per year = 401,000 thousand Child mortality = 20.2 per 1,000 live births Child diarrheal disease deaths = 13.0% of child mortality rate DALYs per child death = 35 discounted years of life lost

DALYs from child diarrheal disease deaths = __________________________________________

__________________________________________________________________________________

2. Calculate the number of DALYs from child diarrheal disease morbidity: Given: Child population (0-4 yrs) = 2.106 million 4.5% of children under 5 suffered from diarrhea in the last 24 hrs A severity weight of 0.2 assigned to diarrhea (DALYs lost from one day of diarrhea = 0.2/365) DALYs from child diarrheal disease morbidity = _______________________________________

__________________________________________________________________________________


2 of 14

3. Calculate the monetary loss from the DALYs calculated above for both mortality and morbidity using the human capital approach. Perform the calculations at 50% and 100% of the GDP. Report the value in terms of SYP/year and in terms of % of GDP.

Given: GDP (2001) = 920 billion SYP GDP/capita (2001) = 55,000 SYP Mortality (based on 50% of GDP): _______________________________________________ SYP

___________________________________________ % GDP

(based on 100% of GDP) _______________________________________________ SYP

___________________________________________ % GDP

Morbidity (based on 50% of GDP): ______________________________________________ SYP

___________________________________________ % GDP

(based on 100% of GDP) _______________________________________________ SYP

___________________________________________ % GDP

Total: _______________________________________________________________________ SYP

___________________________________________________________________ % GDP

4. Calculate the Cost of Illness for severe cases of diarrhea treated in public and private clinics Given: Reported cases of diarrhea in public clinics = 130,000 Reported cases of diarrhea in private clinics = 390,000 Cost of doctor visit per treatment = 200 SYP/case Cost of medication per treatment = 600 SYP/case Assumed that 1 day is lost by caregiver per case of severe diarrhea Value of one day lost to caregiver = 175 SYP Cost of treating severe diarrhea =_________________________________________________ SYP

Cost of lost time due to care giving = ______________________________________________ SYP

Total cost of treating severe diarrhea cases= _______________________________________ SYP


3 of 14

5. Calculate the Cost of Illness for mild cases of diarrhea treated by Oral Rehydration Therapy Given: 8.1 million mild cases of diarrhea per year 43.5 % of mild cases treated by ORT at home Unit cost of ORT treatment = 75 SYP/case Total cost of ORT treatment = ___________________________________________________ SYP 6. Calculate the Cost of Illness for mild cases of diarrhea treated by private doctors and with

medication Given: 8.1 million mild cases of diarrhea per year 50 % of mild cases treated by private doctor Cost of doctor visit per treatment = 200 SYP/case Cost of medication per treatment = 500 SYP/case Total cost of treatment by private doctors and medication = __________________________ SYP 7. Calculate the total COI in terms of SYP/year and in terms of % of GDP. Given: GDP (2001) = 920 billion SYP GDP/capita (2001) = 55,000 SYP Total cost of diarrheal illness = __________________________________________________ SYP

Total cost of diarrheal illness = _______________________________________________ % GDP


4 of 14

GROUP EXERCISE 2

Impact of Water Quality on Health in Tunis and Sfax

(Sarraf et al., 2004)

Case description Sub-standard quality and an inadequate quantity of potable water for drinking and hygiene purposes, inadequate sanitation facilities and sanitary practices and inadequate personal, food and domestic hygiene have a cost to society. It is well known that these factors are associated with waterborne illnesses and mortality. The most common of these illnesses is diarrhea. The impact assessment presented below is linked mainly to mortality and morbidity in children younger than five years due to diarrheal diseases. The aim of this study is to estimates environmental damage costs associated with the health impacts of low quality potable water, inadequate sanitation and hygiene, and the economic impacts of water resources pollution. 1. Calculate the number of DALYs from child diarrheal disease deaths: Given: Annual child deaths from all causes= 5,392 per year Child diarrheal disease deaths = 10.0% of child mortality DALYs per child death = 35 discounted years of life lost DALYs from child diarrheal disease deaths = __________________________________________

__________________________________________________________________________________

2. Calculate the number of DALYs from child diarrheal disease morbidity: Given: Child population (0-14 yrs) = 2.9 million Diarrheal episode per child per year = 2.8 Average duration per episode = 96 hrs A severity weight of 0.2 assigned to diarrhea DALYs from child diarrheal disease morbidity = _______________________________________

__________________________________________________________________________________

__________________________________________________________________________________


5 of 14

3. Calculate the monetary loss from the DALYs calculated above for both mortality and morbidity using the human capital approach. Perform the calculations at 50% and 100% of the GDP Report the value in terms of DT/year and in terms of % of GDP.

Given: GDP (2001) = 24 billion DT GDP/capita (2001) = 2,634 DT

Mortality (based on 50% of GDP): _______________________________________________ DT

___________________________________________ % GDP

(based on 100% of GDP) _______________________________________________ DT

___________________________________________ % GDP

Morbidity (based on 50% of GDP): _______________________________________________ DT

___________________________________________ % GDP

(based on 100% of GDP) _______________________________________________ DT

___________________________________________ % GDP

Total: _______________________________________________________________________ DT

___________________________________________________________________ % GDP

4. Calculate the Cost of Illness for mild cases of diarrhea treated by Oral Rehydration Therapy Given: 2.52 million mild cases of diarrhea per year 94.8 % of mild cases treated by ORT at home Unit cost of ORT treatment = 2 DT/case Total cost of ORT treatment = _________________________________________________DT


6 of 14

5. Calculate the Cost of Illness for severe cases of diarrhea treated by private doctors and with medication

Given: Child population (0-4 yrs) = 0.9 million Percentage of severe diarrhea cases of children (0-4 yrs) = 5.75% If the average duration of a severe diarrhea case is 10 days, the number of cases per child per year is 1 (lower bound) If the average duration of a severe diarrhea case is 7 days, the number of cases per child per year is 1.5 (upper bound) Cost of doctor visit per treatment = 16 DT/case Cost of medication per treatment = 15.5 DT/case Assumed that 1 day is lost by caregiver per case of severe diarrhea Value of one day lost to caregiver = 11.5 DT Cost of treating severe diarrhea (include lower and upper bound) =________________________

_____________________________________________________________________________ DT

Cost of lost time due to care giving (include lower and upper bound) = =____________________

_____________________________________________________________________________ DT

Total cost of treating severe diarrhea cases (include lower and upper bound) =

_____________________________________________________________________________ DT

6. Calculate the total COI in terms of DT/year and in terms of % of GDP. Given: GDP (2001) = 24 billion DT GDP/capita (2001) = 2,634 DT Total cost of diarrheal illness = ___________________________________________________ DT

Total cost of diarrheal illness = _______________________________________________ % GDP


7 of 14

SESSIONS 15-16

GROUP EXERCISE 3

Impact of Air Quality on Health in Syria (World Bank, 2004)

Case description Significant sources of air pollution in Syria include power stations, residential furnaces, industry and traffic. Excessive emissions from traffic are in part due to Syria’s ageing vehicle fleet that is 15 to 20 years old. There is substantial research evidence from around the world that outdoor urban air pollution has significant negative impacts on public health and results in premature deaths, bronchitis, respiratory disorders, and cancer. The air pollutant that has shown the strongest association with these health endpoints is particulate matter (PM), and especially fine particulates of less than 10 microns in diameter (PM10) or smaller. The gaseous pollutants (SO2, NOx, CO, and ozone) are generally not thought to be as damaging. This study therefore focuses on PM10, the smallest measure of PM for which data is available in Syria. The aim of this study is to estimates environmental damage costs associated with the health impacts of poor air quality, particularly elevated levels of PM10. There are three main steps to quantifying the health impacts from air pollution. Step 1. The pollutant needs to be identified and its concentration measured. Monitoring data from nine cities was used: Damascus, Aleppo, Homs, Hama, Lattakia, Dier-Azzour, Al-Raka, Al-Sweida, and Tartous. For each city, four to ten monitoring sites had data available from the Syrian Atomic Energy Commission (SAEC) and the Higher Institute of Applied Sciences and Technology (HIAST). All data is for 2001 with the exception of Lattakia and Hama where data was collected in 1994 and 1992 respectively. Step 2. Calculate the number of people exposed to the pollutant. City population estimates were taken from the Central Bureau of Statistics (2001). It was assumed that 100 percent of the city’s population is exposed to air pollution. Using expert advice from HIAST, the number of people living or spending most of their time near each monitoring site was estimated. The remainder of the city’s population not living near a monitoring site were assumed to be exposed to the average PM10 levels measured in the residential and background location monitoring sites. Some health outcomes affect only certain segments of the population such as adults or children. As only total population data is available at the city level the number of adults and children in each city had to be estimated. This was done by applying the percentage of Syria’s population that is under fifteen years of age to the city population data. A sample of Steps 1 and 2 for Damascus are presented in the table below.


8 of 14

Monitoring site Commercial center

Major Roundabout

Heavy traffic & industrial workshop

Residential Residential Other pop

Crude death rate (per 1,000) 4.8 4.8 4.8 4.8 4.8 4.8

Annual average PM10 (μg/m3) 222 304 437 120 102 111

Exposed total population (millions) 0.15 0.02 0.08 0.35 0.28 1.83

Exposed adult population (≥15 yrs) (millions)

0.10 0.01 0.05 0.22 0.18 1.16

Exposed children population (≤14 yrs) (millions)

0.05 0.01 0.03 0.13 0.10 0.66

Step 3. Estimate the health impacts from this exposure based on epidemiological information. For this, the study relied upon scientific literature. Scientific studies estimate a dose-response coefficient linking PM10 concentrations with mortality and morbidity outcomes. The health endpoints considered as well as the dose-response coefficients are presented in the table below. The dose-response coefficients are taken from Lvovsky et al (2000). Health categories Units Impacts per 1μg/m3 Annual cases in Syria

Premature mortality % change in crude mortality rate 0.084 3,513

Chronic bronchitis Per 100,000 adults 3.06 16,970

Hospital admissions Per 100,000 population 1.2 10,454

Emergency room visits Per 100,000 population 23.54 205,073

Restricted activity days Per 100,000 adults 5,750 31,887,775

Lower respiratory illness in children Per 100,000 children 169 535,054

Respiratory symptoms Per 100,000 adults 18,300 101,486,312


9 of 14

Step 4. Quantify the health impacts and estimate the value of this damage. 1. Calculate the number of DALYs from premature mortality and morbidity associated with elevated

PM levels by completing the table below (Fill in the grey cells). Given the number of DALYs per case adopted from Lvovsky et al. (2000) and the annual cases in Syria.

Health categories DALYs/ 10,000 cases Annual cases Syria Annual DALYs Syria

Premature mortality 100,000 3,513

Chronic bronchitis 12,037 16,970

Hospital admissions 264 10,454

Emergency room visits 3 205,073

Restricted activity days 3 31,887,775

Lower respiratory illness in children 3 535,054

Respiratory symptoms 3 101,486,312

Total DALYs lost per year

2. Calculate the monetary loss from the DALYs calculated above for both mortality and morbidity.

For mortality estimations, use the HCA approach for lower bound calculations and the WTP approach for upper bound calculations. For morbidity estimations, use only HCA approach:

Given: GDP/capita (2001) = 55,000 SYP GDP (2001) = 920 billion SYP WTP = 320,000 SYP/year (adopted from studies in Europe and US and then adjusted for GDP per capita differentials for Syria) Mortality (HCA-lower bound): __________________________________________________ SYP

Mortality: (WTP- upper bound): __________________________________________________ SYP

_____________________________ - ______________________ % GDP

Morbidity: ______________________ SYP

______________________ % GDP


10 of 14

3. Calculate the Cost of Illness for health impacts caused by urban air pollution by completing the table below.

Given: Hospital admissions:

Two days of hospitalization Two work days lost

ER visits: Cost of ER visits Half a day of work lost

RADs: 1 work day lost per 10 RADs

Chronic bronchitis: Monthly doctor visit for 25% of individuals with CB Twice a year doctor visit for 65% of individuals with CB Emergency doctor visit once a year for 30% of individuals Average 6 day hospitalization for 2.5% of individuals 5 working days lost pr year for 35% of individuals Costs discounted at 10% for 15 years to reflect chronic nature of illness Data based on studies from US and Europe

Unit costs Chronic

bronchitis Hospital admissions

Emergency room visits

RAD Total cases

Annual cases 16,970 10,454 205,073 31,887,775 32,120,273

COI (million SYP/yr)

Hospitalization 2,000 SYP/day 43

Doctor visits 400 SYP/visit 244

ER visits 200 SYP/visit 9

Lost work days 200 SYP/day 50

Total COI (million SYP/yr) 345

4. Calculate the total COI in terms of SYP/year and in terms of % of GDP. Total COI = _____________________________________________________________ (SYP/yr)

______________________________________________________________ % GDP


11 of 14

GROUP EXERCISE 4

Impact of Air Quality on Health in Tunis and Sfax (Sarraf et al., 2004)

Case description There is substantial research evidence from around the world that outdoor/urban air pollution have significant negative impacts on public health and result in premature deaths, chronic bronchitis, respiratory disorders, and even cancer. The most significant air pollutant in terms of impacts on health is most commonly found to be particulate matter, especially fine particulates (PM10 or smaller). No study that statistically links urban air pollution and health, based on local health and ambient air monitoring data, has been carried out in Tunisia. However, applying findings from international studies to the local air pollution situation in Tunisia can produce an estimate. The aim of this study is to estimates environmental damage costs associated with the health impacts of poor air quality, particularly elevated levels of PM10. There are three main steps to quantifying the health impacts from air pollution. Step 1. The pollutant needs to be identified and its concentration measured. The annual average of PM10 considered, is a result of monthly averages of 9 months recorded by CITET (Centre International des Technologies de l’Environnement de Tunis) and reported in 1998. Step 2. Calculate the number of people exposed to the pollutant. City population estimates were taken from WDI, World Bank (2001). It was assumed that 80% of the population is exposed to air pollution. The population was broken-down by age-groups applying percentages of Tunisia’s population. A sample of Steps 1 and 2 for Tunis and Sfax are presented in the table below.

Parameter Tunis Sfax

Crude death rate (per 1,000) 5.6 5.6

Annual average PM10 (μg/m3) 65 65

Exposed total population (80% of Total population) (millions) 1.3 0.6

Exposed adult population (≥15 yrs) (millions) 0.9 0.4

Exposed children population (≤14 yrs) (millions) 0.4 0.2

Step 3. Estimate the health impacts from this exposure based on epidemiological information. For this, the study relied upon scientific literature. Scientific studies estimate a dose-response coefficient linking PM10 concentrations with mortality and morbidity outcomes. The health endpoints considered as well as the dose-response coefficients are presented in the table below. The dose-response coefficients are taken from Lvovsky et al. (2000).


12 of 14

Health categories Units Impacts per 1μg/m3 DALYs per 10,000 cases



Hospital admissions Per 100,000 population 1.2 264

Emergency room visits Per 100,000 population 23.54 3

Restricted activity days Per 100,000 adults 5,750 3

Lower respiratory illness in children Per 100,000 children 169 3

Respiratory symptoms Per 100,000 adults 18,300 3

Step 4. Quantify the health impacts and estimate the value of this damage. 1. Calculate the number of DALYs from premature mortality and morbidity associated with elevated

PM levels by completing the table below (Fill in the grey cells). Given the number of DALYs per case adopted from Lvovsky et al. (2000) and the annual cases in Tunis and Sfax.

Health categories DALYs per 10,000 cases

Cases in Tunis

Cases in Sfax

Total cases Tunis & Sfax

Annual DALYs Tunis & Sfax

Premature mortality DALY 100,000 391.4 195.7

Chronic bronchitis 12,037 1,771 885

Hospital admissions 264 998 499

Emergency room visits 3 19,585 9,793

Restricted activity days 3 3,327,627 1,663,814

Lower respiratory illness in

children

3 42,805 21,402

Respiratory symptoms 3 10,590,535 5,295,267

Morbidity DALY

Total DALYs lost per year

2. Calculate the monetary loss from the DALYs calculated above for both mortality and morbidity.

For mortality estimations, use the HCA approach for lower bound calculations and the WTP approach for upper bound calculations. For morbidity estimations, use only the HCA approach:


13 of 14

Given: GDP (2001) = 24 billion DT GDP/capita (2001) = 2,634 DT WTP = 6,440 DT (adopted from studies in Europe and US and then adjusted for GDP per capita differentials for Tunisia) Mortality (lower bound): ________________________________________________________ DT

Mortality: (upper bound): ________________________________________________________ DT

_________________________ - _______________________ % GDP

Morbidity: _________________________________________________________________ DT

_______________________________________________________________ % GDP

3. Calculate the Cost of Illness for health impacts caused by urban air pollution by completing the

table below. Given: Treatment costs used are adopted from averages in Lebanon and Morocco and then adjusted for GDP per capita differentials for Tunisia Hospital admissions

Two days of hospitalization Two work days lost

Emergency Room (ER) visit estimation is based on: Cost of consultation Half a day of work lost

Restricted Activity Days (RADs) 1 work day lost per 10 RADs

Chronic bronchitis Average 6 day hospitalization for 2.5% of patients Monthly doctor visit for 25% of patients Twice a year visit for 65% of patients Emergency doctor visit once a year for 30% of patients 5 working days lost per year for 35% of patients Costs discounted at a rate of 10% for 15 years to reflect chronic nature of illness Data based on studies from US and Europe


14 of 14

Unit costs Chronic

bronchitis Hospital admissions


RAD Total cases

Annual cases 2,656 1,498 29,378 4,991,441 32,120,273

Annual COI (million DT / year)

Hospitalization DT 250 / day 0.8

Doctor visits DT 25 / visit 2.4

ER visits DT 65 / visit 0.4

Lost work days DT 30 / day 1.2

Total COI (million DT/yr) 4.8

4. Calculate the total COI in terms of DT/year and in terms of % of GDP. Total COI = _________________________________________________________________ (DT/yr)

_________________________________________________________________ % GDP



Sessions 16 &17The Value of Life and HealthThe Value of Life and Health



1. Impact of water quality on health in Syria2. Impact of urban air quality on health in Syria3. Impact of water quality on health in Tunis and

Sfax4. Impact of urban air quality on health in Tunis

and Sfax5. Impact of water quality on health in Egypt6. Impact of urban air quality on health in Egypt7. Impact of water quality on health in Morocco8. Impact of urban air quality on health in Morocco9. Impact of water quality on health in Lebanon10. Impact of urban air quality on health in Lebanon

CaseCase--study 1:study 1:Impact of Water Quality on Impact of Water Quality on

Health in SyriaHealth in Syria

Value of Life and HealthValue of Life and HealthCaseCase--study 1:study 1: Impact of Water Quality on Health Impact of Water Quality on Health

in Syriain Syria

Background

• Syria is water scarce• Renewable freshwater

resources = 2,700 m3

• Less than a third of the world average

• Water availability unevenly distributed

• Local pressure on water resources

• Declining groundwater tables

• Water quality degradation

According to WHO/UNICEF (2000)•64 % of rural population had access to an improved water source•94 % of urban areas covered with water supply network•20 % of rural population lacking access to hygienic sanitation facilities

Value of Life and Health Value of Life and Health CCasease--study 1:study 1: Impact of Water Quality on Health Impact of Water Quality on Health

in Syriain SyriaBackground

The lack of safe drinking waterInadequate hygiene and sanitation

water pollution

Impact on human health and quality of life through diarrheal diseases (mainly children)



in Syriain SyriaMethodologyMortality•DALY/ HCA approach

Morbidity•DALY/ HCA approach

+•Cost of illness approach

• Severe diarrhea• Medication costs• Lost time for caregivers

• Mild diarrhea• Medication costs

• Oral rehydrationtherapy

• Private doctor visits

Value of Life and Health Value of Life and Health CaseCase--study 1:study 1: Impact of Water Quality on Health Impact of Water Quality on Health

in Syriain SyriaResults• DALYs- Mortality

– 13% of all deaths in children under five attributable to diarrheal disease (MoH)

– Death of a child under 5 represents a loss of 35 DALYs (Global Burden of Disease)

Parameter ValueChild population (0-4 yrs) 2.106 millionLive births per year 401 thousandChild mortality 20.2 per 1,000 live birthsAnnual child deaths (all causes) 8,100 per yearChild diarrheal disease deaths 13.0% of child mortality rateChild diarrheal disease mortality rate 2.6 per 1,000Annual child diarrheal disease deaths 1053DALYs per child death 35 discounted years of life lostDALYs from child diarrheal disease deaths 36,856


in Syriain SyriaResults• DALYs – Morbidity

– Based on household surveys by MoH• 8.6 million cases of diarrhea per year• 4.5% of children under 5 suffered from diarrhea in the last 24 hrs

– A severity weight of 0.2 assigned to diarrhea• DALYs lost from one day of diarrhea = 0.2/365

Parameter ValueChild population (0-4 yrs) 2.106 millionDiarrheal prevalence in children (0-4 yrs) in last 24 hrs 4.5%Total diarrhea days per year 34.6 millionDALY (disability severity weight) 0.2DALYs from child diarrheal disease morbidity 18, 954 per year


in Syriain Syria

Parameter ValueMortality

Total number of DALYs 36,900Price per DALY 27,500-55,000 SYP/yrMonetary loss 1,015 - 2,030 million SYP/yr% of GDP 0.11 – 0.22 %

MorbidityTotal number of DALYs 19,000Price per DALY 27,500-55,000 SP/yrMonetary loss 523 – 1,045 million SYP/yr% of GDP 0.06 – 0.11 %


in Syriain SyriaResults• COI- Severe cases of diarrhea

– Severe cases often treated in health clinic– 130,000 diarrhea cases per year in public clinics (MoH)– Ratio of private to public clinics according to NEAP = 3:1– Assumed that 1 day is lost by caregiver per case of severe diarrheaParameter ValueReported cases of diarrhea, public 130,000Reported cases of diarrhea, private clinics 390,000Total cases of reported diarrhea 520,000Cost of doctor visit per treatment 200 SYPCost of medication per treatment 600 SYPTotal cost of treatment 800 SYPCost of treating severe diarrhea 416 million SYPValue of one day lost to caregiver 175 SYPCost of lost time due to care giving 91 million SYP


in Syriain Syria

Results• COI- Mild cases of diarrhea treated by Oral

Rehydration Therapy– 8.6 million cases -0.5 million severe cases = 8.1

million mild cases of diarrhea per year– 44 % of mild cases treated by ORT at home– Average length of diarrheal episode = 4 days

Parameter ValueNumber of diarrhea cases per year in children (0-4 yrs) 8.1 millionPercent of cases treated with ORT 43.5 %Cases treated with ORT 3.53 millionUnit cost of ORT treatment 75 SYP/caseTotal cost of ORT treatment 264 million SYP


in Syriain SyriaResults• COI- Mild cases of diarrhea treated by private

doctors and with medication

Total COI = 3.6 billion SYP/year= 0.4% of GDP

Parameter ValueNumber of diarrhea cases per year in children 0-4 8.1 millionPercent of cases treated by private doctors 50%Cost of doctor visit per treatment 200 SYPCost of medication per treatment 500 SYPTotal cost of treatment 700 SYPCost of treating non-severe diarrhea 2,835 million SYP

CaseCase--study 2:study 2:Impact of Urban Air Quality Impact of Urban Air Quality

on Health in Syriaon Health in Syria

Value of Life and Health Value of Life and Health CaseCase--study 2:study 2: Impact of Urban Air Quality on Impact of Urban Air Quality on

Health in SyriaHealth in SyriaBackground• Sources of air pollution in Syria

– Power stations– Residential furnaces– Industry– Ageing vehicle fleet that is 15-20 yrs old



Methodology


Health in SyriaHealth in SyriaMethodology- Identify pollutant and measure its

concentration• Monitoring data from 9 cities used

• 4-10 monitoring sites per city from– Syrian Atomic Energy Commission– Higher Institute of Applied Science and Technology

• Data for 2001 except Lattakia (1994) and Hama (1992)

• In all cities, one monitoring station collected both PM10and TSP– PM10 data inferred from remaining monitoring data by

calculating the ratio PM10:TSP

-Damascus -Aleppo -Tartous-Homs -Hama -Al-Sweida-Lattakia -Deir-Azzour -Al-Raka


Health in SyriaHealth in SyriaMethodology- Determine exposed population• City population estimates taken from Central Bureau of

Statistics (2001)

• Assumed that 100% of the city’s population exposed to air pollution

• Using expert advice from HIAST– Number of people living or spending most of their time near each

monitoring site estimated– Remaining population assumed to be exposed to average PM10

levels measured in residential and background locations

• Population broken-down by age-groups applying percentages of Syria’s population



Methodology-Sample for Damascus

Monitoring site Commercial center

Major Roundabout

Heavy traffic & industrial workshop

Residential Residential Other pop

Crude death rate (per 1,000) 4.8 4.8 4.8 4.8 4.8 4.8

Annual average PM10 (μg/m3) 222 304 437 120 102 111

Exposed total population (millions)

0.15 0.02 0.08 0.35 0.28 1.83

Exposed adult population (≥15 yrs) (millions)

0.10 0.01 0.05 0.22 0.18 1.16

Exposed children population (≤14 yrs) (millions)

0.05 0.01 0.03 0.13 0.10 0.66



Methodology- Estimate health impacts from exposure

• Relied on dose-response coefficients reported in the literature by Lvovsky et al. (2000)

• Limitation: reported DRR from developed countries

Health categories Units Impacts per 1μg/m3

Premature mortality % change in crude mortality rate 0.084Chronic bronchitis Per 100,000 adults 3.06Hospital admissions Per 100,000 population 1.2Emergency room visits Per 100,000 population 23.54Restricted activity days Per 100,000 adults 5,750Lower respiratory illness in children Per 100,000 children 169Respiratory symptoms Per 100,000 adults 18,300



Mortality•DALY/ HCA approach•DALY/ WTP approach



• Chronic bronchitis• Hospital admissions• Emergency room visits• Restricted activity days

Methodology- Valuate health impacts from exposure


Health in SyriaHealth in SyriaResults- Mortality and Morbidity• Health outcomes converted to DALYs• DALYs adopted from Lvovsky et al. (2000)• Data calculated per city and then aggregated


Annual cases Syria

Annual DALYsSyria

Premature mortality 100,000 3,513 35,126Chronic bronchitis 12,037 16,970 20,427Hospital admissions 264 10,454 276Emergency room visits 3 205,073 62Restricted activity days 3 31,887,775 9,566Lower respiratory illness in children 3 535,054 161Respiratory symptoms 3 101,486,312 30,446Total DALYs lost per year 96,062


Health in SyriaHealth in SyriaResults• DALYs –Valuation

– Mortality• Low estimate at

GDP/capita (2001) = 55,000 SYP

• High estimate at WTP adopted from studies in Europe and US and then adjusted for GDP per capita differentials for Syria

• Adjusted WTP then modified to reflect approximate number of DALYs lost due to air pollution relative to DALYslost in WTP studies

– 10 DALYs per case usually considered for air pollution

– Morbidity• Low estimate at

GDP/capita (2001)= 55,000 SYP


Total number of DALYs 35,100Price per DALY 55,000 – 320,000 SYP/yrMonetary loss 1,931 – 11,232 million

SYP/yr% of GDP 0.21 – 1.22 %

MorbidityTotal number of DALYs 60,900Price per DALY 55,000 SP/yrMonetary loss 3,350 million SYP/yr% of GDP 0.36 %


Health in SyriaHealth in SyriaResults- COI approach- Basis for cost estimates• Chronic bronchitis

– Monthly doctor visit for 25% of individuals with CB– Twice a year visit for 65% of individuals with CB– Emergency doctor visit once a year for 30% of individuals– Average 6 day hospitalization for 2.5% of individuals– 5 working days lost pr year for 35% of individuals– Costs discounted at 10% for 15 years to reflect chronic nature of illness– Data based on studies from US and Europe

• Hospital admissions– Two days of hospitalization– Two work days lost

• ER visits– Cost of ER visits– Half a day of work lost

• RADs– 1 work day lost per 10 RADs



Results- COI approach

Total COI = 1.1 billion SYP/year= 0.12% of GDP

Unit costs Chronic bronchitis

Hospital admissions


RAD Total cases

Annual cases 16,970 10,454 205,073 31,887,775 32,120,273

COI (million SYP/yr)Hospitalization 2,000 SYP/day 43 42

Doctor visits 400 SYP/visit 244

ER visits 200 SYP/visit 9 41

Lost work days 200 SYP/day 50 4 21 638

Total COI (million SYP/yr) 345 46 62 638 1,090Cost per case (SYP) 20,331 4,400 300 20 33.9


Health in Tunis and Health in Tunis and SfaxSfax


in Tunis and in Tunis and SfaxSfaxBackgroundThe lack of: safe drinking water

& sanitation& water pollution


Mortality and MorbidityMortality and Morbidity


in Tunis and in Tunis and SfaxSfaxMethodologyMortality•DALY/ HCA approach



• Severe diarrhea• Medication costs• Private doctor visits• Lost time for caregivers

• Mild diarrhea• Oral rehydration therapy


in Tunis and in Tunis and SfaxSfaxResults• DALYs- Mortality

– 10% of all deaths in children under five attributable to diarrheal disease – Death of a child under 5 represents a loss of 35 DALYs (Global Burden

of Disease)

Parameter ValueChild population (0-4 yrs) 0.9 millionChild mortality 30 per 1,000 live birthsAnnual child deaths (all causes) 5,392 per yearChild diarrheal disease deaths 10.0% of child mortality rateChild diarrheal disease mortality rate 2.9 per 1,000Annual child diarrheal disease deaths 539DALYs per child death 35 discounted years of life lostDALYs from child diarrheal disease deaths 18,865


in Tunis and in Tunis and SfaxSfaxResults• DALYs – Morbidity


Parameter ValueChild population (0-14 yrs) 2.9 millionDiarrheal episode per child per year 2.8Total number of episodes per year 8.12 millionAverage duration per episode 96 hrsTotal diarrhea hrs per year 780 millionTotal diarrhea duration in years per year 89,000DALY (disability severity weight) 0.2DALYs from child diarrheal disease morbidity 17,808 per year


in Tunis and in Tunis and SfaxSfax


Total number of DALYs

18,865Price per DALY 1,315-2,630 DT/yrMonetary loss 24.81 – 49.61 million DT/yr% of GDP 0.10 – 0.20 %

MorbidityTotal number of

DALYs17,808

Price per DALY 1,315-2,630 DT/yrMonetary loss 23.41 – 46.83 million DT/yr% of GDP 0.10 – 0.20 %


in Tunis and in Tunis and SfaxSfaxResults• COI- Severe cases of diarrhea

– If the average duration of a severe diarrhea case is 10 days, the number of cases per child per year is 1 (lower bound)

– If the average duration of a severe diarrhea case is 7 days, the number of cases per child per year is 1.5 (upper bound)

– Assumed that 1 day is lost by caregiver per case of severe diarrhea

Parameter ValueChild population (0-4 yrs) 0.9 millionPercentage of severe diarrhea cases of children < 5 5.75Number of severe diarrhea cases per year (lower bound) 0.9 millionNumber of severe diarrhea cases per year (lower bound) 1.3 millionCost of doctor visit per treatment 16 DTCost of medication per treatment 15.5 DTTotal cost of treatment 31.5 DTCost of treating severe diarrhea 28.31- 40.95 million DTValue of one day lost to caregiver 11.5 DTCost of lost time due to care giving 10.3 - 13.93 million DT


in Tunis and in Tunis and SfaxSfaxResults• COI- Mild cases of diarrhea treated by Oral

Rehydration Therapy– 94.8 % of mild cases treated by ORT at home

Total COI = 43.4 – 59.66 million DT/year= 0.16-0.25 % of GDP

Parameter ValueNumber of diarrhea cases per year in children (0-4 yrs) 2.52 millionPercent of cases treated with ORT 94.8 %Cases treated with ORT 2.4 millionUnit cost of ORT treatment 2 DT/caseTotal cost of ORT treatment 4.78 million DT

CaseCase--study 4:study 4:Impact of Urban Air Quality Impact of Urban Air Quality on Health in Tunis and on Health in Tunis and SfaxSfax


Health in Tunis and Health in Tunis and SfaxSfaxBackground

• No air pollution impact studies has been undertaken in Tunisia. The present study considers international studies results adjusted for Tunisian conditions.


Health in Tunis and Health in Tunis and SfaxSfaxBackground

• The annual average of PM10 is a result of monthly averages of 9 months recorded by CITET (Centre International des Technologies de l’Environnement de Tunis) and reported in 1998

Tunis Sfax

Crude death rate (per 1,000) 5.6 5.6


Exposed total population (80% of Total population) (millions) 1.3 0.6











Health in Tunis and Health in Tunis and SfaxSfaxThe Results• Estimate the DALY per Health Category

• DALY estimations relied on dose-response coefficients reported in international studies


DALYs per 10,000 cases



Hospital admissions Per 100,000 population 1.2 264

Emergency room visits Per 100,000 population 23.54 3

Restricted activity days Per 100,000 adults 5,750 3

Lower respiratory illness in children Per 100,000 children 169 3

Respiratory symptoms Per 100,000 adults 18,300 3


Health in Tunis and Health in Tunis and SfaxSfaxThe Results • Calculate the number of DALYs lost due to urban air

pollutionHealth categories DALYs per

10,000 cases

Cases in Tunis

Cases in Sfax

Total cases Tunis & Sfax

Annual DALYsTunis & Sfax

Premature mortality DALY 100,000 391.4 195.7 587.1 5,871Chronic bronchitis 12,037 1,771 885 2,656 3,197

Hospital admissions 264 998 499 1,497 40

Emergency room visits 3 19,585 9,793 29,378 9

Restricted activity days 3 3,327,627 1,663,814 4,991,441 1,497

Lower respiratory illness in children 3 42,805 21,402 64,207 19

Respiratory symptoms 3 10,590,535 5,295,267 15,885,802 4,766

Morbidity DALY 9,528

Total Total DALYsDALYs lost per yearlost per year 15,39915,399


Health in Tunis and Health in Tunis and SfaxSfaxThe Results • Valuate the lost DALYs

–Low estimate at GDP/capita (1999) = 2,630 DT

–High estimate at WTP adopted from studies in US and then adjusted for GDP per capita differentials for Tunisia = 6,440 DT

Lower Bound

Upper Bound

Total DALYs lost per year 15,399

Price per DALY (DT)GDP / capita WTP

2,630 6,440Total Annual monetary loss (million DT)

40.5 99.2

The Average Annual DALY Losses Due to Urban The Average Annual DALY Losses Due to Urban Pollution represent 0.2Pollution represent 0.2--0.4% of the GDP0.4% of the GDP


Health in Tunis and Health in Tunis and SfaxSfaxThe Results Cost of Illness (COI) assumptions

• Treatment costs used are adopted from averages in Lebanon and Morocco and then adjusted for GDP per capita differentials for Tunisia

• Chronic bronchitis– Average 6 day hospitalization for 2.5% of patients– Monthly doctor visit for 25% of patients– Twice a year visit for 65% of patients– Emergency doctor visit once a year for 30% of patients– 5 working days lost per year for 35% of patients– Costs discounted at a rate of 10% for 15 years to reflect chronic nature of illness– Data based on studies from US and Europe


• Emergency Room (ER) visit estimation is based on: – Cost of consultation– Half a day of work lost

• Restricted Activity Days (RADs)– 1 work day lost per 10 RADs



Unit costs Chronic bronchitis

Hospital admissions ER visits RADs Total

Annual cases 2,656 1,498 29,378 4,991,441

Annual COI (million DT / year)

Hospitalization DT 250 / day 0.8 0.7

Doctor visits DT 25 / visit 2.4

ER visits DT 65 / visit 0.4 1.9

Lost work days DT 30 / day 1.2 0.1 0.4 15.0

Total Cost of Treatment (million DT per year) 4.8 0.8 2.3 15.0 22.9

Cost per case (DT) 1,807 534 78 3

Results Cost of Illness (COI) calculations

The Annual Cost of Illness Due to Urban Pollution The Annual Cost of Illness Due to Urban Pollution representsrepresents

22.9 million DT or 0.09 % of the GDP22.9 million DT or 0.09 % of the GDP


Health in EgyptHealth in Egypt


in Egyptin EgyptBackground


water pollution

Impact on human health and quality of lifeIntestinal work infections, Intestinal work infections, schistosomiasesschistosomiases, , diarrhealdiarrheal diseasesdiseases

17,000 children 17,000 children die annually from die annually from diarrhealdiarrheal diseasesdiseases



in Egyptin EgyptMethodologyMortality•DALY/ HCA approach



in Egyptin EgyptResults• DALYs- Mortality

– 20% of all deaths in children under five attributable to diarrhealdisease


– Base data are from WDI, World Bank 2001

Parameter ValueChild population (0-4 yrs) 8.15 millionAnnual child deaths (all causes) 88,020 per yearChild diarrheal disease deaths 20.0% of child mortalityAnnual child diarrheal disease deaths 17,604DALYs per child death 35 discounted years of life lostDALYs from child diarrheal disease deaths 616,140 per year


in Egyptin EgyptResults• DALYs – Morbidity

– Estimates based on children only because of their high incidence rate– A severity weight of 0.2 assigned to diarrhea

• Base data are from WDI, World Bank 2001Parameter ValueChild population (0-14 yrs) 22 millionDiarrheal episodes per child per month 1Total episodes per year 264 millionAverage duration per episode 10 hrsTotal duration per year (hrs) 2,640 millionTotal duration per year (yrs) 301,370DALY (disability severity weight) 0.2DALYs from child diarrheal disease morbidity 60,274 per year


in Egyptin Egypt


Total number of DALYs 616,140Price per DALY 2,400-4,800 LE/yrMonetary loss 1,478 - 2,957 million LE/yr% of GDP 0.49 – 0.98 %

MorbidityTotal number of DALYs 60,274Price per DALY 27,500-55,000 LE/yrMonetary loss 145– 289 million LE/yr% of GDP 0.05 – 0.1 %


in Egyptin Egypt

ResultsMore than 657,000 657,000 DALYsDALYs lost lost each

year due to diarrheal diseases amounting to a damage cost of

0.50.5--1.1 % of GDP per year1.1 % of GDP per year


on Health in Egypton Health in Egypt



Background



Methodology


Health in EgyptHealth in EgyptMethodology- Identify pollutant, measure its concentration,

determine exposed population

– Annual average concentrations of PM10 in Greater Cairo– Rough estimates of PM10 from Alexandria– No study in Egypt statistically linking urban pollution and health

based on local health data

Parameter Egypt Greater Cairo AlexandriaPopulation (million) 63 14.9 3.3

Adult population ≥15 yrs) (millions) 41 9.7 2.1

Children population (≤14 yrs) (millions) 22 5.2 1.2

Crude death rate (per 1,000) 7 7 7


Exposed total population (millions) 11.92 2.64





Methodology- Estimate health impacts from exposure

• Relied on dose-response coefficients from international studies (Lvovsky et al., 2000)

• Limitation: reported DRR from developed countries









Health in EgyptHealth in EgyptResults- Mortality and Morbidity• Health outcomes converted to DALYs• DALYs adopted from Lvovsky et al. (2000)

Health categories DALYsper 10,000 cases

Annual cases Greater Cairo

DALYsGreater Cairo

Annual cases Alexandria

DALYsAlexandria

Annual DALYsSyria

Premature mortality 100,000 18,924 189,242 1,552 15,523 204,765

Chronic bronchitis 12,037 64,092 77,148 5,257 6328 83,476

Hospital admissions 264 38,621 1,020 3,168 84 1,103

Emergency room visits 3 757,611 227 62,146 19 246

Restricted activity days 3 120,434,571 36,130 9,879,048 2,964 39,094

Respiratory symptoms 3 383,296,114 114,989 31,440,000 9,432 124,421

Lower respiratory illnessin children

3 1,899,367 570 155,802 47 617

Total DALYs lost per year 419,326 34,397 453,722


Health in EgyptHealth in EgyptResults• DALYs –Valuation

–Mortality• Low estimate at

GDP/capita (2001) = 4,800 LE

• High estimate at WTP = 28,000 LE – adopted from studies in

Europe and US and then adjusted for GDP per capita differentials for Egypt

– Adjusted WTP then modified to reflect approximate number of DALYs lost due to air pollution relative to DALYs lost in WTP studies» 10 DALYs per case usually

considered for air pollution–Morbidity

• Low estimate at GDP/capita (2001)= 4,800 LE


Total number of DALYs 200,000Price per DALY 4,800 – 28,000 LE/yrMonetary loss 960 – 5,600 million LE/yr% of GDP 0.32– 1.86 %

MorbidityTotal number of DALYs 250,000Price per DALY 4,800 LE/yrMonetary loss 1,200 million LE/yr% of GDP 0.4 %



ResultsMore than 450,000 450,000 DALYsDALYs lost lost each

year due to urban air pollution amounting to a damage cost of

0.7 0.7 -- 2.3 % of GDP per year2.3 % of GDP per year


Health in MoroccoHealth in Morocco


in Moroccoin MoroccoBackground


water pollution




in Moroccoin MoroccoMethodologyMortality•DALY/ HCA approach



• Severe diarrhea• Medication costs• Lost time for caregivers

• Mild diarrhea• Medication costs

• Oral rehydrationtherapy


in Moroccoin MoroccoResults• DALYs- Mortality

– 20% of all deaths in children under five attributable to diarrheal disease (National Survey on the Health of Mother and Child, ENSME, 1997)


Parameter ValueChild population (0-4 yrs) 3.038 millionChild mortality rate 46 per 1,000 live birthsAnnual child deaths (all causes) 27,951 per yearChild diarrheal disease deaths 20.0% of child mortality rateChild diarrheal disease mortality rate 9.2 per 1,000Annual child diarrheal disease deaths 5,590DALYs per child death 35 discounted years of life lostDALYs from child diarrheal disease deaths 186,868


in Moroccoin MoroccoResults• DALYs – Morbidity

– Based on the National Survey on the Health of Mother and Child

• 9% of children under 5 suffered from diarrhea in the last 24 hrs– A severity weight of 0.2 assigned to diarrhea

• DALYs lost from one day of diarrhea = 0.2/365

Parameter ValueChild population (0-4 yrs) 3.038 millionDiarrheal prevalence in children (0-4 yrs) in last 24 hrs 9%Total diarrhea days per year 99.8 millionDALY (disability severity weight) 0.2DALYs from child diarrheal disease morbidity 64,887 per year


in Moroccoin Morocco


Total number of DALYs 196,000Price per DALY 6,150-12,300 Dh/yrMonetary loss 1,205 - 2,411 million Dh/yr% of GDP 0.34 – 0.68 %

MorbidityTotal number of DALYs 55,000Price per DALY 6,150-12,300 Dh/yrMonetary loss 338 – 677 million Dh/yr% of GDP 0.10 – 0.19 %


in Moroccoin MoroccoResults• COI- Severe cases of diarrhea

– Severe cases often treated in health clinic– Cost of treatment obtained Moroccan doctors’ consultation– Assumed that 1 day is lost by caregiver per case of severe diarrhea– Value of a work day based on average rural income = Dh1,500/ month

Parameter ValueCases of diarrhea treated in public hospitals (20% of cases) 4,191,780Cases of diarrhea treated in private clinics (7% of cases) 1,357,338Total cases of reported diarrhea 5,549,118Cost of doctor visit per treatment 70 DhCost of medication per treatment 100 DhTotal cost of treatment 170 DhCost of treating severe diarrhea (< 5 yrs) 943 million DhValue of one day lost to caregiver 60 DhCost of lost time due to care giving 333 million Dh



Results• COI- Mild cases of diarrhea treated by Oral

Rehydration Therapy– 30% of mild cases treated by ORT at home– Average length of diarrheal episode = 5 days

Parameter ValueTotal number of diarrheal days/yr in children < 5yrs 99.8 millionAverage duration of a diarrheal case 5 daysNumber of diarrhea cases per year in children (0-4 yrs) 19.9 millionPercent of cases treated with ORT 30 %Cases treated with ORT 5.99 millionUnit cost of ORT treatment 60 Dh/caseTotal cost of ORT treatment 359 million Dh



Results

Total cost of treatment= 1,635 million Dh

= 0.46 % of GDP


on Health in Moroccoon Health in Morocco


Health in MoroccoHealth in MoroccoBackground• Sources of urban air pollution in Morocco

– Ageing diesel vehicle fleet (in 2000, 808,000 diesel vehicle, of which 74% are 10 years or older)

– Low quality oil products– Industry not converted to cleaner technologies:

• Thermal energy centers and oil refineries• Chemical and para-chemical industries• Textile and leather industries• Agro-industries• Electrical and electronic industries• Metal and metallurgical industries

– Power stations



MethodologyMeasure the concentrationof atmospheric pollutants

Identify populationsvulnerable to pollution

Assess the effect on health withthe dose-response coefficients

Valuate Morbidity

Assess mortality risk

Step 1

Step 2

Step 3

Step 4

Step 5


Health in MoroccoHealth in MoroccoMethodologyStep 1: Measure the concentration of atmospheric pollutants• Air quality monitoring data are available from:

– Casablanca– The urban prefecture of Rabat-Sale– Cities of Safi– Fes, Marrakech and Tangiers (by analogies with other

Moroccan cities) • 2 to 7 monitoring sites per city reported in:

– Casa Airpol, 2000: Etude de la pollution atmosphérique et de son impact sur la santé des populations à Casablanca

– REEM, 2001 (Status of Environment Report in Morocco)– Ministry of Public Health,1998: Etude de la pollution

atmosphérique et de son impact sur la santé de la population de Safi


Health in MoroccoHealth in MoroccoMethodologyStep 1: Measure the concentration of atmospheric pollutants

• Lower estimate for PM10 in Casablanca is based on a conversion factor between PM2.5 and PM10 equivalent to 0.35

• Lower estimates for PM10 in Rabat-Sale and Safi is based on expert judgements• Higher estimates for PM10 in Casablanca, Rabat-Sale and Safi is based on the average PM

concentration converted to PM10 on the basis of a ratio of 0.5• For Fes, Marrakech and Tangiers air pollution is mainly due to the transport sector, as in

Rabat. The lower value of PM10 for the city of Rabat was allocated to these cities (70).

City PM (μg/m3)Average level

PM3 (μg/m3)Average level

PM10 (μg/m3)Low level

PM10 (μg/m3)High level

Casablanca 244 33 94 122

Rabat - Sale 246 - 70 123

Safi 277 - 70 139

Fes, Marrakech & Tangiers - - 70


Health in MoroccoHealth in MoroccoMethodologyStep 2: Identify populations vulnerable to pollution

• Population data were taken from the Ministry of Public Health (Santé en chiffres, 2001)• The demographic distribution of the six cities was estimated by extrapolation from national

averages• Certain impacts affect more particularly a certain portion of the population (for example, the elderly

and children under five years of age).

Cities Casablanca Rabat-Sale Safi Fes, Marrakech & Tangiers

Crude death rate (per 1,000) 4.9 4.9 4.9 4.9

Low level PM10 (μg/m3) 94 70 7070

High level PM10 (μg/m3) 122 123 139

Exposed total population (millions) 3.31 1.44 0.43 1.72

Exposed adult population (≥15 yrs) (millions) 2.16 0.94 0.28 1.12

Exposed population (≤14 yrs) (millions) 1.15 0.50 0.15 0.60

Exposed children population (≤5 yrs) (millions) 0.27 0.15 0.05 0.18


Health in MoroccoHealth in MoroccoMethodologyStep 3: Assess the effect on health with the help of dose-response coefficients

• Relied on dose-response coefficients reported in the literature by Lvovskyet al. (2000)


Premature mortality % change in crude mortality rate 0.084

Chronic bronchitis Per 100,000 adults 3.06

Hospital admissions Per 100,000 population 1.2

Emergency room visits Per 100,000 population 23.54

Restricted activity days Per 100,000 adults 5,750

Lower respiratory illness in children Per 100,000 children 169

Respiratory symptoms Per 100,000 adults 18,300



Cases / city

Health categories

Casablanca Rabat-Sale Safi Fes, Marrakech, Tangiers

Total cases

Low High Low High Low High Low High

Premature mortality 1281 1662 416 731 125 248 496 2,318 3,137

Chronic bronchitis 6217 8069 2020 3549 607 1205 2410 11,254 15,233

Hospital admissions

3734 4846 1213 2131 365 724 1447 6,759 9,148


73242 95059 23794 41810 7151 14201 28392 132,579 179,462

Restricted activity days 11682529 15162431 3794846 6668087 1140556 2264818 4528061 21,145,992 28,623,397

Lower respiratory illness in children 182462 236812 59290 104180 17820 35285 70745 330,317 447,022

Respiratory symptoms 37180919 48256086 12077511 21221913 3629944 7208032 11411045 64,299,419 88,097,076



• DALY approach is used to compare and assess the number of years lost due to disabilities


Annual cases Morocco Annual DALYs MoroccoLow High Low High

Premature mortality 100,000 2,318 3,137 23,180 31,370 Chronic bronchitis 12,037 11,254 15,233 13,546 18,336

Hospital admissions 264 6,759 9,148 178 242

Emergency room visits 3 132,579 179,462 40 54

Restricted activity days 3 21,145,992 28,623,397 6,344 8,587 Lower respiratory illness in children 3 330,317 447,022 99 134

Respiratory symptoms 3 64,299,419 88,097,076 19,290 26,429 Total Morbidity 39,497 53,782

Total DALYs lost per yearTotal DALYs lost per year 62,677 62,677 85,152 85,152


Health in MoroccoHealth in MoroccoResultsStep 4: Assess mortality risk

• Two common approaches were used:– Human Capital (HC) approach

as a lower bound estimate1 DALY = GDP/capita= 12,308 Dh

– Willingness To Pay (WTP)approach as a higher bound estimate; adopted from studies in US and then adjusted for GDP differentials for Morocco = 63,324 Dh

Lower Bound

Upper Bound

Total Annual Mortality DALYs 23,180 31,370

One DALYmonetary value (Dh)

GDP / capita WTP

12,308 63,324

Total Annual Mortality DALYs value (million Dh) 285.3 1,986.5

Average Annual Mortality Average Annual Mortality DALY value (million Dh)DALY value (million Dh) 1,135.91,135.9

The Average Annual Mortality Losses Due to Urban The Average Annual Mortality Losses Due to Urban Pollution Represents 0.32% of the GDPPollution Represents 0.32% of the GDP



MethodologyStep 5: Valuate Morbidity

Two approaches used1. Assess on DALY at GDP per capita to account for

people’s suffering associated with respiratory illnesses

2. Medical Cost of Treatment of respiratory diseases such as chronic bronchitis and lower respiratory illnesses in children



MethodologyStep 5: Valuate Morbidity

Assess DALY at GDP per capita to account for people’s suffering associated with respiratory illnesses

Lower Bound

Upper Bound

Total Annual Morbidity DALYs 39,497 53,782

One DALYmonetary value (Dh) 12,308

Total Annual Morbidity DALYs value (million Dh) 486.1 661.9

Average Annual Morbidity DALY Average Annual Morbidity DALY value (million Dh)value (million Dh) 574.0574.0


Health in MoroccoHealth in MoroccoMethodologyStep 5: Valuate Morbidity - Cost of Treatment

• Chronic bronchitis estimates are based on:– Average 6 day hospitalization for 2.5% of patients– Monthly doctor visit for 25% of patients– Two visit per year for 65% of patients– Emergency consultation once a year for 30% of patients– 5 working days lost per year for 35% of patients– Costs are converted into annual numbers and discounted rate of 10% over a

period of 15 years to take into account the nature of chronic bronchitis– Data based on studies conducted in the US and Europe


• Emergency Room (ER) visit estimation is based on:– Cost of consultation– Half a day of work lost

• Restricted Activity Days (RADs)– 1 work day lost per 10 RADs


Health in MoroccoHealth in MoroccoMethodologyStep 5: Valuate Morbidity - Cost of Treatment

Unit costsChronic bronchitis

Hospital admissions ER visits RADs

Low High Low High Low High Low High

Annual cases 11,254 15,233 6,759 9,148 132,579 179,462 21,145,992 28,623,397

Annual COI (million Dh / year)

Hospitalization Dh 1200/ day 16.7 22.6 16.6 22.4

Doctor visits Dh 70/ visit 28.6 38.7

ER visits Dh 300/ visit 8.7 11.8 40.6 55.0

Lost work days Dh 115/ day 19.1 25.8 1.8 2.5 7.4 10.0 243.1 327.9

Total Cost of Treatment (million Dh per year) 73.1 98.9 18.4 24.9 48.0 65.0 243.1 327.9

Cost per case (Dh) 6,569 2,741 362 11.5

Total cost of treatment ranges between 383 and 517 million Dh/yeTotal cost of treatment ranges between 383 and 517 million Dh/yearar



ResultsStep 5: Valuate Morbidity

The average The average annual morbidity annual morbidity losses due to losses due to urban pollution urban pollution represents 0.29% represents 0.29% of the GDPof the GDP

Lower Bound

Upper Bound

Total Annual Morbidity DALYs value (million Dh) 486.1 661.9

Total Annual Cost of Treatment (million Dh) 382.6 516.7

Total Annual Morbidity Cost (million Dh) 868.7 1,178.6

Average Annual Morbidity Cost (million Dh) 1,023.651,023.65


Health in LebanonHealth in Lebanon

Value of Life and HealthValue of Life and HealthCCasease--study 9:study 9: Impact of Water Quality on Health Impact of Water Quality on Health

in Lebanonin LebanonBackground Substandard quality and inadequate quantity of potable water

Inadequate sanitation facilities and sanitation practicesInadequate personal, food and domestic hygiene




in Lebanonin LebanonMethodologyMortality•DALY/ HCA approach



– Treatment costs• Doctor and medical

facilities visits• Medicines• Oral Rehydration Therapy

– Time cost for caregivers of severe cases


in Lebanonin LebanonResults• DALYs- Mortality

– 10% of all deaths in children under five attributable to diarrheal disease (based MoH, 1996 and CBS/Unicef, 2001)


Parameter ValueChild population (0-4 yrs) 0.44 millionChild mortality rate in 2000 30 per 1,000 live birthsAnnual child deaths (all causes) 2,640 per yearChild diarrheal disease deaths 10.0% of child mortality rateChild diarrheal disease mortality rate 3 per 1,000Annual child diarrheal disease deaths 264DALYs per child death 35 discounted years of life lostDALYs from child diarrheal disease deaths 9,240 per year


in Lebanonin LebanonResults• DALYs – Morbidity

– Based on Lebanese Mother and Child Health Census by MoH• Average duration of diarrhea case is 4 days• 2.0 million cases of diarrhea per year• 5.0% of children under 5 suffered from diarrhea in the last 24 hrs


Parameter ValueChild population (0-4 yrs) 0.44 millionDiarrheal prevalence in children (0-4 yrs) in last 24 hrs 5.0 %Total diarrhea days per year 8.0 millionTotal diarrhea duration in years per year 22,000DALY (disability severity weight) 0.2DALYs from child diarrheal disease morbidity 4,400 per year




Total number of DALYs 9,240Price per DALY 1,950-3,900 US$/yrMonetary loss 18.01 – 36.2 million US$/yr% of GDP 0.11 – 0.22 %

MorbidityTotal number of DALYs 4,400Price per DALY 1,950-3,900 US$/yrMonetary loss 8.5 – 17.1 million US$/yr% of GDP 0.05 – 0.10 %

Results• DALYs – Valuation

– Valuation per DALY:• High, 100% of Lebanese GDP per capita, 3,900 US$/yr• Low, 50% of Lebanese GDP per capita, 1,950 US$/yr


in Lebanonin LebanonResults• Cost of illness approach

– According to MoH, 1996:• 48% of diarrhea cases are treated by doctors or medical facilities; and• 9% by pharmacies

– According to CBS/Unicef, 2001:• 44% of diarrhea cases are treated by Oral Rehydration Therapy (ORT)

– Average cost (Beirut and small town) of doctor or medical facility visit is 30 US$/visit (information from doctors in Lebanon)

– Cost of medicines for diarrhea treatment is 12 US$/case (information from doctors in Lebanon)

– Cost of ORT is 1.5 US$/case– Assumed that 1 day is lost by caregiver per case of severe diarrhea– Value of 1 day lost by caregiver (based on low-skilled wage rate) is

10.0 US$/day


in Lebanonin LebanonResults• COI- Valuation of treatment costs

Parameter ValueTotal diarrhea cases per year (all children 0-4 years) 2.01 million

Percent of cases treated (doctor, medical facilities) 48%

Total cases treated (doctor, medical facilities) per year 0.96 millionCost of doctor/medical facilities visit 30 US$/visit

Total cost of doctor and medical facilities per year 28.9 million US$Percent of cases treated at pharmacy 9%

Total cases required medicines (cases treated at doctors and pharmacy, 57%) per year

1.14 million

Cost of medicines per case 12 US$

Total cost of medicines per year 13.8 million US$Percent of cases treated by ORT 44%

Total cases treated by ORT 0.88 millionCost of ORT per case 1.5 US$

Total cost of ORT per year 1.3 million US$


in Lebanonin LebanonResults• COI- time cost of care giving for severe cases

– According to CBS/Unicef, 2001, 0.76 million of severe diarrheacases per year

– 1 day is lost by caregiver per case of severe diarrhea

Parameter ValueNumber of severe diarrhea cases per year 0.76 millionValue of day lost by caregiver 10.0 US$Total cost of lost time by caregivers 7.63 million US$



Results• COI- Total cost of diarrheal illness

Total COI = 51.6 million US$/year= 0.31% of GDP

Parameter ValueTotal cost of doctor and medical facilities visits per year

28.9 million US$

Total cost of medicines per year 13.8 million US$Total cost of ORT per year 1.3 million US$Total cost of lost time by caregivers 7.63 million US$Total cost of diarrheal illness 51.63 million US$


on Health in Lebanonon Health in Lebanon



Background• Sources of air pollution in Lebanon

– Power stations– Industries– Vehicle induced emissions

Urban air pollutionespecially PM10and Lead (Pb)

Negative impacts on public health

- Premature deaths- Chronic bronchitis- Respiratory disorders- Cancer- Hypertension- IQ loss



Methodology


Health in LebanonHealth in LebanonMethodology- Identify pollutant and measure its

concentration

• Impacts of PM10 concentrations in Greater Beirut and Tripoli areas were considered

• Data for pollutant concentrations:– PM10 in Greater Beirut area from El-Fadel et. al (2002).

Monitoring data after the ban of vehicle diesel fuel. Annual average of 55 μg/m3

– Annual average PM10 concentration in Tripoli was assumed equivalent to 55 μg/m3.


Health in LebanonHealth in LebanonMethodology- Determine exposed population

• City population estimates taken from WDI, World Bank (2001)

• Assumed that 80% of the population exposed to air pollution

• Population broken-down by age-groups applying percentages of Lebanon’s population



Methodology-Data for Greater Beirut and Tripoli

Key Parameter Unit Lebanon Greater Beirut area

Tripoli

Total population Million 4.3 1.3 0.35

Adult population (≥15 yrs) Million 2.95 0.89 0.24

Children population (≤14 yrs) Million 1.35 0.41 0.11

Exposed population (80% of total) Million 1.04 0.28

Exposed adult population (≥15 yrs) Million 0.7 0.2

Exposed children population (≤14 yrs) Million 0.3 0.1

Crude death rate Per 1,000 6 6 6

Annual average PM10 μg/m3 55 55



Methodology- Estimate health impacts from exposure• Relied on dose-response coefficients reported in the

literature by Ostro (1994) and Lvovsky et al. (2000)











Health in LebanonHealth in LebanonResults- Mortality and Morbidity• Health outcomes converted to DALYs• DALYs adopted from Lvovsky et al. (2000)• Data calculated per city and then aggregated


Cases Greater Beirut

DALYs Greater Beirut

Cases Greater Tripoli

DALYs Greater Tripoli

DALYs Total

Premature mortality 100,000 288 2,883 78 776 3,659

Chronic bronchitis 12,037 1,201 1,445 323 389 1,835

Hospital admissions 264 686 18 185 5 23

Emergency room visits 3 13,465 4 3,625 1 5

Restricted activity days 3 2,256,407 677 607,494 182 859

Lower respiratory illness in children

3 30,349 9 8,171 2 12

Respiratory symptoms 3 7,181,260 2,154 1,933,416 580 2,734

Total DALYs lost per year 7,191 96,062 1,936 9,127


Health in LebanonHealth in LebanonResults• DALYs –Valuation

– Mortality• Low estimate at

GDP/capita (2001) = 3,857 USD

• High estimate at WTP adopted from studies in Europe and US and then adjusted for GDP per capita differentials for Lebanon

• Adjusted WTP then modified to reflect approximate number of DALYs lost due to air pollution relative to DALYs lost in WTP studies

– 10 DALYs per case usually considered for air pollution

– Morbidity• Low estimate at

GDP/capita (2001)= 3,857 USD


Total number of DALYs

3,650Price per DALY 3,857 – 21,000 USD/yrMonetary loss 14.2 – 76.6 million USD/yr% of GDP 0.08 – 0.46 %

MorbidityTotal number of

DALYs5,480

Price per DALY 3,857 USD/yrMonetary loss 21.37 million USD/yr% of GDP 0.12 %



•Cost of illness approach• Chronic bronchitis• Hospital admissions• Emergency room visits• Restricted activity days (RADs)


Value of Life and HealthValue of Life and HealthCaseCase--study 10:study 10: Impact of Urban Air Quality on Impact of Urban Air Quality on

Health in LebanonHealth in LebanonResults- COI approach- Basis for cost estimates• Chronic bronchitis

– Monthly doctor visit for 25% of individuals with CB– Twice a year visit for 65% of individuals with CB– Emergency doctor visit once a year for 30% of individuals– Average 6 day hospitalization for 2.5% of individuals– 5 working days lost pr year for 35% of individuals– Costs discounted at 10% for 15 years to reflect chronic nature of illness– Data based on studies from US and Europe


• ER visits– Cost of doctor visits– Half a day of work lost

• RADs– 1 work day lost per 10 RADs

Value of Life, Health, Risk & SafetyValue of Life, Health, Risk & SafetyCaseCase--study 10:study 10: Impact of Urban Air Quality on Impact of Urban Air Quality on

Health in LebanonHealth in LebanonResults- COI approach

Unit costs* Chronic bronchitis

Hospital admissions


RADs Total cases

Annual cases 1,524 871 17,090 2,863,901 2,883,386

COIHospitalization 600 US$/day 1,148 1,045

Doctor visits 50 US$/visit 2,742

ER visits 100 US$/visit 383 1,709

Lost work days 60 US$/day 1,339 105 513 17,183

Total COI (Million US$/yr) 5.61 1.15 2.22 17.18 26.16Cost per case (US$) 3,681 1,320 130 6.0 9.1

Total COI = 26.16 Million US$/year= 0.16% of GDP

* Cost of illness based on information from doctors in Lebanon

End of Sessions 16 & 17

Thank You




Session 18aEconomic Assessment of

Environmental Degradation due to the July 2006 Hostilities



Session 18aEconomic Assessment of Environmental Economic Assessment of Environmental

Degradation due to the July 2006 Hostilities Degradation due to the July 2006 Hostilities

OUTLINEOUTLINE

• Introduction• Oil Spill• Demolition, Military, and Medical Waste• Water Degradation• Quarries• Air Pollution• Forest Fires


• The 34-day hostilities in Lebanon started on July 12, 2006 and continued until August 14, 2006– killed close to 1,200 people– left more than 4,400 injured– displaced more than a quarter of the population– damaged severely the country’s infrastructure – had a devastating impact on the country’s fragile environment

and people’s health• destruction of infrastructure left enormous amounts of debris and

rubble• bombing of a power plant in Jiyeh caused the spill of about 12,000-

15,000 tons of oil into the Mediterranean Sea• widespread fires and oil burning deteriorated the air quality,

especially in Southern Beirut

INTRODUCTIONINTRODUCTION• This study aims at assessing the cost of environmental

degradation caused by the 2006 hostilities in Lebanon• Valuation methods used

Impacts Method used1.Oil spill - Impact on birds and turtles1

- Beach resorts, hotels, restaurant, marinas, fishing, ...Restoration cost modelMarket price2

2. Waste- Impact of demolition waste on environment- Impact of UXOs on health- Impact of UXOs on agriculture- Impact of medical waste

Cost of transport and disposalDALYs3

Market priceCost of disposal

3. Water Cost of alternative sources

4. Quarries Hedonic price method

5. Air Not estimated

6. Forests- Impact on forests- Impact on national reforestation program

Market price4, substitute goods5, cost-based methods6

Restoration costs


• The overall estimated cost of environmental degradation

Impacts US$ million(Min)

US$ million(Max)

US$ million(Average)

% of GDP1

Waste 206.8 373.5 290.2 1.4Oil spill 166.3 239.9 203.1 1.0Water 131.4 131.4 131.4 0.6Quarries 15.4 175.5 95.5 0.5Forests 7.0 10.8 8.9 0.0Air n.e. n.e.Total environmental cost caused by hostilities 526.9 931.1 729.0 3.6%

Based on an estimated GDP for 2006 of US$20.5 billion (Economic Intelligence Unit, 2006)

OIL SPILLOIL SPILL

Oil SpillOil Spill

• Bombing of Jiyeh power utility, located 30 km south of Beirut, led to burning and spilling of heavy oil into the Mediterranean Sea– About 12,000 to 15,000 tons of oil– Air and naval blockade limited

mitigation before two months• According to MEDSLIK from the

Oceanography Centre, University of Cyprus– spilled oil moved northward and onto

the shoreline, with heaviest impacts occurring between Jiyeh and Beirut, then between Byblos and Chekka, and onto the Palm Islands offshore

• Other areas showed patchy impacts• Oil reached the shoreline at Tartus,

Syria

Oil SpillOil Spill

• Initial shoreline assessment by the MOE indicated that– heavy pooled oil existed in coastal coves and harbors– sand and gravel beaches south of Beirut and around Byblos to

the north showed surface and buried oil– several observations of oil on the bottom

during this spill, probably a result of • oil burning• heavy oil concentrations mixing with

sediment to form oil mats on the bottom

• This type of oil has relatively low impact potential on fisheries and invertebrates, due to the low content of (acute) toxic hydrocarbons– Environmental problems are mainly caused by the oil’s physical

properties, such as the tendency to stick to objects and surfaces

Oil SpillOil Spill

• Major impact on marine biodiversity, including – shoreline biota– subtidal bottom communities

birds– marine reptiles– marine mammals– Fish– Nature reserves, particularly the Palm Islands Nature Reserve in

the North• These impacts were impossible to valuate within the

scope of the study• Impact of the oil spill on water quality

– laboratory analysis did not show contamination of groundwater through seawater intrusion, which could have occurred in densely fissured zones

Oil SpillOil Spill

• Estimates the environmental damages caused by the oil spill on the coastal zone– users’ forgone benefits through the differences between

the expected and actual benefits derived from activities on the coast

• Expected benefits: the level of environmental benefits which would have been enjoyed, had the oil spill not occurred

• Actual benefits: those currently provided after the outbreak of the conflict.

– A 3-year time frame was adopted for the analysis, during which the losses will gradually subside

• conservative time frame, as it does not capture potential effects not yet probed or that may occur over an extended period of time

Oil SpillOil Spill• In the absence of primary surveys several assumptions were made

to arrive at damage estimates

• The assumptions for 2006 rely on the baseline information and, consequently, vary from one activity to another

• For 2007 and 2008, conservative assumptions based on the experts’best knowledge at the time of valuation

Impacts on 2006 2007 2008Jul‐Aug1 Sep ‐Dec Jan‐Dec Jan‐Dec

% of expected incomeCommercial fishing 0 50a 5‐10 0‐5Shore‐side fishing 0 50a 5‐10 0‐5Hotels 0 10‐20 5‐10 0‐5World Heritage Site 0 25‐50 5‐10 0‐5Beach resorts and chalets 0 25‐50 5‐10 0‐5Nature Reserve 0 75‐100 5‐10 0‐5Restaurants 0 75‐100 5‐10 0‐5Sport Services 0 75‐100 5‐10 0‐5

Damages caused by the oil spill

1all losses during July-August are assumed to be caused by the hostilities themselves

Oil SpillOil Spill

• The following impacts were valuated– Hotels and furnished apartments

– Beach resorts, chalets and public beaches

– Marinas sports activities

– Palm Islands Nature Reserve

– Byblos World Heritage Site

– Restaurants

– Fishing

– Oil fuel burnt and spilled in Jiyeh

– Oil spill clean-up operations

Oil SpillOil SpillHotels and Furnished ApartmentsHotels and Furnished Apartments

• A drop in the occupancy rate of hotels and furnished apartments along the coast– In 2006 significant reduction due to visual signs of oiled beaches and

contaminated water– The Syndicate of Hotel Owners (2006) lists 54 licensed hotels located

on the coast • about 3,500 rooms• room rates US$40 to US$300/night, averaging to US$100/night• additional hotel revenue is US$50/day for meals, phone and laundry• the average hotel income is about US$150/person/day

– 97 furnished apartment establishments are located on the coast• 2,800 apartment units• daily price per apartment on average assumed that the net price of one furnished

apartment is about US$220/night

• With a damage cost of 5-10% of expected income in 2007 and 0-5% in 2008, total forgone income due to the oil spill ranges between US$23-60 million, with an average of US$41 million

Oil SpillOil SpillBeach Resorts and ChaletsBeach Resorts and Chalets

• The Lebanese coast hosts about 68 beach resorts• Many beach resorts made low income, while others closed for the

whole season• According to discussions with the Syndicate of Maritime

Establishments– about 500 daily visitors/beach resort during peak season – around 300 visitors/beach/day during the rest of the season.– daily spending per visitor averages US$20/day– the expected income of beach resorts in September 2006 was

estimated at US$12 million, while the expected seasonal income is about US$55.4 million

• In September 2006, the hostilities and the oil spill altogether caused a decline in the expected income of about 80% in September 2006.

– The oil spill alone likely contributed a loss of about 25-50% to the expected income in September 2006

– As in other cases, considerably lower share of 5-10% is assumed for 2007 – 0-5% for 2008

• Total forgone income due to the oil spill falls between US$5-13 million


• Twenty-five chalet complexes can be found on the Lebanese coast, all located north of Jiyeh– The high season for renting chalets covers May-October– Each chalet complex has about 200 chalets– Chalet rent is about US$1,000/month– Thus, the monthly income from renting chalets averages to

US$5 million– The chalets closed during hostilities and re-opened at the

beginning of September 2006• expected income after re-opening

– US$10 million in 2006– US$30 million in each of 2007 and 2008

– Assuming that the oil spill contributes to the income decline in a similar way as in the case of beach resorts, total forgone income to chalets is about US$4 - 9 million


• About 15 public beaches in Lebanon, covering a total length of 10-12 km– peak season from July to September– oil spill affected only 9 beaches– as entrance is free, it is assumed that the individual benefit is

about half of that enjoyed by visitors to beach resorts, i.e. US$10/day

• Expected monthly benefits from using public beaches during high-season = US$2.6 million– Public beaches closed during hostilities and re-opened

beginning of September 2006• Expected income after re-opening is estimated at US$2.6 million in

2006 and US$7.8 million in each of 2007 and 2008• Assuming oil spill contributes to the decline in benefits from

public beaches in a similar way as in the case of beach resorts, the present value of total forgone income is about US$0.7-1.5 million


• Beach resorts and chalet complexes organize weddings and other social events from May to October– Social events count about 300 participants and cost US$40/person– Beach resorts can organize events during warm months

• 4 events/week during 4 months– Chalets complexes can arrange such events during half of the year

• 3 events/week during 6 months– About 6,000-6,700 events per season, providing an income of about

US$71-80 million per year• Income decline in 2006 is considered to be due to hostilities• Assuming that in 2007 and 2008 the oil spill contributes to the

decline in the income from events in a similar way as in the case of beach resorts– the present value of forgone income to events is about US$3-11 million.

Overall, forgone income to beach resorts, chalets, public beaches and events falls within US$13-35 million, with an average of

US$24 million

Oil SpillOil SpillMarine Sports ActivitiesMarine Sports Activities

• Marinas offer recreational services to public such as– Boating– Diving– Water-skiing– Docking– Maintenance of private boats

• In 2006, oil pollution of seawater, boatsand its effects on health prevented most marinas to resume their boat rental and water sports activities

• Based on field interviews, the total revenue of marinas is estimated about US$1 million per year, assume that– Income during May-June is equal to September-October, i.e.

about 25% of total annual income– As recreational activities resumed in September 2006, the

expected income for the rest of the year was US$250,000


• Assume that– Income during May-June is equal to September-October, i.e. about 25%

of total annual income– As recreational activities resumed in September 2006, the expected

income for the rest of the year was US$250,000

• Assume that– oil spill caused about

75-100% drop in incomein September-October 2006.

– loss is about 5-10%of annual income in 2007 and 0-5% in 2008

• Loss from recreational activities in marinas ranges between US$0.23-0.38 million, with an average of US$0.3 million

Min(‘000 UsD)

Max(‘000 UsD)

Expected income:- in 2006 (Sept-Oct) 250 250- in 2007 (May-Oct) 1,000 1,000- in 2008 (May-Oct) 1,000 1,000Forgone income due to oil- in 2006 (Sept-Oct) 188 250- in 2007 (May-Oct) 50 100- in 2008 (May-Oct) 0 50PV of forgone income 238 377


• Pollution of private leisure boats docked in marinas and fishingboats docked in fishing ports – limited the owners’ benefits from using their boats in the period following

the hostilities until the end of 2006 – imposed additional costs of cleaning the boats

• Loss of the recreational benefit from private leisure boats equal to– annual depreciation of the boat – cost of upkeep and docking in marinas

• Assuming that only 890 boats actually oiled (50% of 1,775)– Average price for common boat (6-12 m) = US$30,000– Considering a lifetime of about 20 years

• annual value of a boat = US$1,500.– Assuming an annual cost of US$300/m/season and the average size of

a boat of 9 m• Annual cost of upkeep and docking is about US$2,700

• The total loss to owners of private leisure boats = US$3.7 million.


• To estimate the loss due to oiled fishing boats, annual maintenance costs used as proxy– 20 oiled fishing boats– annual maintenance costs = US$2,700

• Accordingly, loss due to oiled fishing boats amounts to US$54,000

Total loss to private owners of leisure and fishing boats = US$3.8 million

Overall losses to marinas’ sports activities = US$4 to US$4.2 million, with an average of US$4.1

million

Oil SpillOil SpillPalm IslandsPalm Islands’’ Nature ReserveNature Reserve

• The loss to tourism in 2006 estimated by the difference between– the expected number of tourists (averaging 22,500)– actual arrivals (about 1,700)

• Forgone benefits– losses in revenues from boat transportation of individuals and groups to the islands – rentals of chairs and umbrellas

• The tourist season is about 13 weeks (July-September) of which only three remained after the end of blockade

• Considering that tourists are evenly distributed in time throughout the season, and assuming that the oil spill will contribute to the forgone income between

– 75-100% in 2006– 5-10% in 2007 – 0-5% in 2008

• Loss in tourism due to oil spill is estimated at about US$15,400-27,600NATURE RESERVE Min Max NotesForgone annual income (13 weeks) 72.4 91.1Forgone income due to the oil spill % of expected income:- in 2006 (3 weeks) 12.5 15.8 75-100% - in 2007 (13 weeks) 3.6 9.1 5-10%- in 2008 (13 weeks) 0 4.6 0-5%PV of forgone income (‘000 USD) 15.4 27.6

Oil SpillOil SpillPalm IslandsPalm Islands’’ Nature ReserveNature Reserve

• A long term monitoring program is foreseen for the reserve and other ecologically significant sites affected by the spill– 7-10 years duration– US$1.2-1.7 cost

• Part of this cost is directly related to the oil spill damage, while the rest being an expression of WTP for future information

• Assumed that 50% of the total impact assessment and monitoring cost is due to the oil spill damage, ie. US$600,000-850,000

• The overall impact of the oil spill on the Palm Islands Nature Reserve and other ecologically sensitive areas amounts to US$0.7-1.2 million

• Assigning a monetary value to the loss of biodiversity was not possible due to lack of such studies

Oil SpillOil SpillByblos World Heritage SiteByblos World Heritage Site

• Oil spill heavily contaminated– the harbor– two medieval towers at the entrance– other ancient ruins located below the archaeological Tell in

Byblos

• Absence of information on the willingness-to-pay loss of historical value using restoration cost method– UNESCO team recommended a procedure to clean the oiled

archaeological remains• Assuming stones cleaned manually with a specially prepared solution• Total cleanup cost of operations = US$100,000 as the minimum bound

of the damage caused by the oil spill

• Loss in recreational value estimated in terms of forgone benefits due to decrease in number of visitors during 2006-2008

Oil SpillOil SpillByblos World Heritage SiteByblos World Heritage Site

• Visits to Byblos take place throughout the year and are organized both by tour operators and private individuals

• Annual income from all visits to Byblos = US$144,000• Assuming the oil spill contributes

25-50% of the September-December income in 2006, 5-10%in 2007 and 0-5% in 2008

• Damage to tourism inByblos and other historical townsranges between US$15,300-42,800.

Min MaxExpected annual income 144.0 144.0Forgone income due to oil spill- in 2006 (Sept-Dec 9.0 24.0- in 2007 (Jan-Dec) 7.2 14.4- in 2008 (Jan-Dec) 0.0 7.2PV of forgone income 15.3 42.8

Tours•Average number of visitors by tours is about 300/year•Fee is about US$30/person if meals are excluded•The annual income of tour operators from organizing visits to Byblos is about US$72,000

Private individuals• There are twice as many visitors to

Byblos by private cars as those coming through tour operators

• The average spending is US$15/person

•The annual income from individual trips would be about US$72,000

Total estimated damages to Byblos US$115,300Total estimated damages to Byblos US$115,300--142,800142,800

Oil SpillOil SpillRestaurantsRestaurants

• Oil spill affected negatively the activity of these restaurants, mainly due to fears of negative impacts of contaminated fish on human health

• According to the Syndicate of Restaurant Owners– 170 restaurants specialized in fish– Annual turnover ranges between US$200,000 and US$ 600,000

• Assuming average turnover of a fish restaurant is US$400,000/year (US$33,000/month) expected income during September – December 2006 would be about US$133,000 per restaurant

RESTAURANTS Min MaxNo. of fish restaurants 170 170Annual turnover (000 US$/rest./ yr.) 400 400Monthly turnover (000 US$/rest/mth) 33.3 33.3Expected income in Sept-Dec 2006 (000 US$/rest.)

133.3 133.3

Forgone income due to the oil spill- in 2006 (million US$) 17.0 22.7- in 2007 (million US$) 3.4 6.8- in 2008 (million US$) 0 3.4PV of forgone income to oil spill (million US$)

19.5 31.1

• In 2006, 75-100% of the expected income loss• In 2007-2008, successful cleanup of oil

and gradual return to normal life• potential effects on human health

reduce the restaurants’ expected profitsby 5-10% in 2007 and by 0-5% in 2008

Present value of forgone benefits ranges between Present value of forgone benefits ranges between US$19.5US$19.5--31.1 million with average at US$25.3 million31.1 million with average at US$25.3 million

Oil SpillOil SpillFishingFishing

• Fishing supports about 30,000 fishermen who catch on average 8,000 ton of fish per year

• Oil spill caused– direct damages to the boats and gears and ultimately a partial decline of fish supply– indirect damages whereby the actual fish contamination or the perception of its effects

on health reduced the overall demand for fish consumption• In commercial fishing, fish catch varies largely across seasons

– 30% of annual catch in spring– 42% in summer– 22% in autumn – 8% in winter

• Annual income from fishing is about US$31 million• Applying the seasonal catch factor to the total income, the expected fish

income during September-December 2006 was US$7.4 million. • It is estimated that the hostilities and oil spill caused the income of

fishermen to drop by 45%.• Assuming that only 50% of this drop is owing to oil spill, the associated

damage cost in 2006 is about US$1.3 million• Assuming oil spill causes a 5-10% decline in 2007 and 0-5% in 2008, the

present value of damages to commercial fishing falls between US$3-6 million

Oil SpillOil SpillFishingFishing

• Recreational fishing– Impacts assumed to be similar to commercial ones– Oil affected 2,600 anglers– Value of shore-side fishing includes the consumption and recreational value of fish– Using an average catch of 2 kg/day for a minimum of 50 days and an average price of

US$4/kg (FAO, 2006), the consumption value of fish is US$1 million/year• Assuming recreational value of anglers in Lebanon is similar to the value of

recreation on public beaches (US$10/day), the recreational value of anglers is US$1.3 million/year

• Overall, the annual value of shore-side fishing is about US$2.3 million. • Considering that fish catch varies seasonally in the same proportion as in

the case of commercial fishing– expected fish income during September-December 2006 was estimated at about

US$0.7 million• Estimating the impact of the oil spill on shore-side fishing uses the same

percentages adopted for commercial fishing for 2006-2008,– the present value of forgone benefits ranges between US$260,000-472,000

Overall, the impact of the oil spill on commercial and shore-side fishing amounts to US$3.2-6.5 million, with an average of US$5 million

Oil SpillOil SpillOil Fuel Burnt and Spilled in JiyehOil Fuel Burnt and Spilled in Jiyeh

• In addition to environmental damages the loss of an estimated 44,000 tons of stored IFO 150 at Jiyeh electrical power plant represent an economic loss– Loss in resources due to the spill and burning of the

Jiyeh fuel = US$ 20 million (US$450/ ton)– Cost of hiring three floating tankers to replace burnt

tanks = US$ 4 million– Transfer of fuel from different plants to Jiyeh power

plan and soil test of soil in burnt tanks’ location = US$ 15 million

• In total, the burning and spilling of Jiyeh fuel oil due to the hostilities resulted in a direct economic loss estimated at US$ 39 millionUS$ 39 million.

Oil SpillOil SpillOil Spill CleanOil Spill Clean--up Operationsup Operations

• MOE estimated the cost of oil spill clean-up in the range of US$137-205 million (US$13,800/t)

• Approximately US$4 million was provided as equipment and materials by international community– Considering depreciation from use with this oil type – Considering that much of the material was provided

as expendable supplies • 25% of the total supplied, or USD 1 million, will be available

for use after this incident and therefore is deducted from the total mitigation cost


• Estimated amount of oiled waste generated by Phase I clean up operation– 1,030 m3 of liquid waste– 6,250 m3 of polluted waste

• sand, garbage, debris and equipment• Cost of oil waste removed during Phase I based on the waste

management options considered by the MOE– Cost of liquid waste re-processed at Zahrani refinery = US$

92,000– Cost of treating non-liquid oil-polluted waste = US$ 47 million

• 25% includes low-to-medium contaminated sand – 25% of low-to-medium contaminated sand will be re-used in cement, construction

or asphalt industries– At a unit cost of US$10/m3 and a transport cost of US$80,000, the total cost of

transporting and treating the low-to medium contaminated sand is estimated at US$96,000

• 75% represents heavily contaminated sand and pebbles– 4,700 m3 of heavily contaminated sand and pebbles to be shipped under Basel

convention at a cost of US$10,000/ m3


• The estimated cost of transporting and treating the oiled waste resulting from Phase I cleaning operation is US$ 47.1 million

• Phase II of oil spill clean up will generate 4,500 m3 of solid waste

• Additional costs include– 1 million USD for Phase II of the clean-up and monitoring

operations in Palm Islands Nature Reserve– 0.5 million USD for sampling and analysis of water, sediments,

biota and fish species in 9 sites along the Lebanese coast over a period of three years

Overall, the cost of oil clean-up, treatment of oiled waste and monitoring the Lebanese coast is estimated at US$

63.5 million

Oil SpillOil SpillSummarySummary

• Overall damage and clean-up cost due to the oil spill is conservatively estimated at US$203 million, or 1.0% of GDP in 2006

• Value represents lower bound of real costs– does not capture several damage costs

• effects on health (skin diseases)• effects on ecosystem services (loss in habitat for spawning)• effects on marine biodiversity

– fails to cover the cost of many future clean-up operations– tends to reflect only partially the real cost of the oil spill for many

impacts, as a result of the conservative assumptions adopted forvaluation

• Overall estimate and breakdowns should be regarded with care, as many of the assumptions are subjective and debatable due to lack of accurate data

Oil SpillOil SpillEstimated costs of damage and cleanEstimated costs of damage and clean--up up

due to the oil spilldue to the oil spillParameter Estimated Cost (million $)

Min Max MeanDamage- Hotels 22.8 59.6 41.2- Beach resorts, chalets, public beaches 13.2 34.8 24.0- sports activities 4.0 4.2 4.1- Nature Reserve 0.7 1.2 1.0

0.1 0.1 0.1- Restaurants 19. 5 31.1 25.3- Commercial fishing 3.0 5.9 4.4- Sea‐shore fishing 0.3 0.5 0.4- Cost of oil fuel burnt 39.1 39.1 39.1

Sub‐total 102.8 176.4 139.6Oil spill clean up- Expenses already made 14.9 14.9 14.9- Oiled waste 48.2 48.2 48.2- Monitoring expenses 1.5 1.5 1.5

Sub‐total 63.5 63.5 63.5Total 166.3 239.9 203.1

CONSTRUCTION, DEMOLITION CONSTRUCTION, DEMOLITION AND MILITARY WASTEAND MILITARY WASTE

Construction & Demolition WasteConstruction & Demolition Waste

• Construction and Demolition (C&D) waste concentrated in three areas:– Southern Suburbs of Beirut– Districts of the South– Baalbek El Hermel region

• Typical C&D debris constituents:– Primary inert fractions

• Asphalt, brick, glass, plastic pipes, etc.– High organic based fractions

• Ceiling tiles, insulation-treated cellulose, plywood, etc.– Composite materials

• Carpeting, gypsum wallboard, electrical fixtures, etc.

Construction & Demolition WasteConstruction & Demolition WasteGeneral QuantitiesGeneral Quantities

• Beirut Southern Suburbs• 100-200 residential buildings completely

destroyed• 28 partially blasted• 70 damaged

• Districts of the South and Baalbek El Hermel– 11,140 units destroyed– 1,249 units partially destroyed– 81,000 units lightly damaged

• Actual quantities of rubble and demolition waste

Regions Quantities (million m3)Beirut Southern Suburbs 1.43

South 3.32

Bekaa 1

Total 5.75

Construction & Demolition WasteConstruction & Demolition WasteHandling, transport & disposalHandling, transport & disposal

• Beirut Southern Suburbs– Waste disposed at 4 sites

• 2 in low-lying areas by the sea• 1 in Choueifat• 1 temporary along the airport

road– Waste dumped haphazardly– Slope reached 1:1– Where sea encroachment

occurs, the bulky C&D waste gives a good angle of stability


• The South– Waste used to fill depressions– Waste dumped on nearby lands and

in valleys (Khyam)– Waste dumped in an abandoned

pond in Aytaroun– Associated damage difficult to

quantify• Ecosystem damage and visual intrusion• Hydrology and hydrogeology• Opportunity cost associated with land-

use


• Baalbek El Hermel– Waste used to rehabilitate

depressions– Waste dumped in abandoned

quarries– Sorting of asbestos mats,

construction steel, and concrete bricks at dumpsites

– Asbestos only in Baalbek El Hermel

• Limited short-term exposure


Estimated cost of loading and transport of C&D wasteEstimated cost of loading and transport of C&D wasteDescription Rate

Waste Haulinga

Dozer charging rateb ($/day) 400

Filling capacity of 3 dozersc (Truck/day) 30

Daily volume of C&D waste loaded (m3/day) 540

Cost of Loading each Truck per m3 of DW ($/m3) 0.07Waste Transporta

Truck charging rateb ($/day) 250

Daily number of round trips 6

Loading capacity per truck (m3) 18

Daily Volume of DW Transported per Truck (m3/day) 108

Cost of Transport per m3 of DW ($/m3) 2.31Total Unit Cost ($/m3) 2.38Cost in Beirut S. Suburbs ( millions $) 3.4

Cost in the South (millions $) 7.9

Cost in Baalbek El Hermel (millions $) 2.4

Total Cost (million $) 13.7

a Based on field surveys, expert opinion, and GIS analysis

b Range accounts for degree of intervention and thickness damaged

Construction & Demolition WasteConstruction & Demolition WasteRoad depreciationRoad depreciation

• In South and Baalbek El Hermel– Damage due to military aggression– No damage from waste transport

• In Beirut- cost of road depreciationDescription RateAverage road length (km)a 2-3Average road width (m)a 6-8Average road area (m2) 12,000-24,000Cost of road refurbishment ($US/ m2)a

40 cm of compacted gravel10 cm of asphalt

20-30b

Total cost of road depreciation (US$) 240,000-720,000

a Based on field surveys, expert opinion, and GIS analysisb Range accounts for degree of intervention and thickness damaged

Construction & Demolition WasteConstruction & Demolition WasteTraffic DelaysTraffic Delays

• In South and Baalbek El Hermel, delays not encountered– Traffic management and rerouting away from

city center– Dumpsites located in outskirts

• In Beirut, 1 to 3 hrs of traffic delays– Difficult to differentiate between delays due to

waste transport and delays due to bombarded roads

Construction & Demolition WasteConstruction & Demolition WasteTraffic DelaysTraffic Delays

Estimated cost of traffic delaysEstimated cost of traffic delaysDescription Rate

Average extra time spent in traffic (hr/day)a 2

Average hourly wage (US$/hr)b 2.5

Number of working days per monthc 22

Fraction of lost productive timea 0.5

Duration of waste removal (months) 6-8

Opportunity cost of time (US$/ person/6-8 month) 330-440

Average daily number of affected commutersd 115,150

Opportunity Cost of Time (US$ million) 38-51Fuel consumption per hr in traffic (L/hr)a 1

Unit cost of fuel ($US/L)a 0.8

Cost of fuel spent per person per month ($US/month) 35.2

Number of affected vehiclesd 60,000

Cost of gasoline spent per person per 6-8 months ($US/person)

211-282

Total cost of gasoline spent per 6-8 months ($US million) 13-17

Total cost of traffic delay (US$ million) 51-68

a Based on field surveys and expert opinion;

b Based on a GDP of 5,300 US$/capita;

c Peak travel delays are assumed to occur 22 working days/month;

d Based on DMJM+HARRIS, 2003-Refer to Annex 3

Construction & Demolition WasteConstruction & Demolition WasteCost of land for waste disposalCost of land for waste disposal

• Assumptions– All waste in each area is disposed in one

equivalent landfill– Height of landfill = 25 m– Unit cost of land adopted is average to low

Region Waste volume(000m3)

Landfill height (m)

Area of waste

(000m2)

Landfill area

(000m2)

Cost of land1

($/m2)Cost of

land (million $)

Beirut 1,430 25 57.2 74.4 1,000 74.4

South 3,320 25 132.8 172.6 10 1.72

Baalbek 1,000 25 40.0 52.0 15 0.781 Based on real estate information and expert opinion

Construction & Demolition WasteConstruction & Demolition WasteDepreciation of land surrounding dumpsitesDepreciation of land surrounding dumpsites

• Waste disposal represents to surrounding neighborhood– Health hazard– Visual intrusion

• Damage assessment was not possible

Construction & Demolition WasteConstruction & Demolition WasteSummarySummary

Parameter Damage cost (million US$)Beirut South Baalbek Total

Waste hauling and transport 3.4 7.9 2.4 13.7Road maintenance 0.2-0.7 - - 0.2-0.7Traffic delays 51 - 68 - - 51 - 68Land for disposal 74.4 1.7 0.8 76.9Land depreciation - - - -

Subtotal 129-146.5 9.6 3.2 142-159

Estimated total damage cost of C&D waste

Military Waste (Unexploded Ordnances)Military Waste (Unexploded Ordnances)IntroductionIntroduction

• 864 cluster bomb strike locations in South Lebanon

Military Waste (Unexploded Ordnances)Military Waste (Unexploded Ordnances)IntroductionIntroduction

• One million UXOs on 34 million square meters

• Demining– Costs 5.5 million USD per year– Will need a period of two years

• Impact of UXOs– Death and injuries– Preventing access and exploitation of

agricultural lands

Military Waste (Unexploded Ordnances)Military Waste (Unexploded Ordnances)Deaths and InjuriesDeaths and Injuries

• Casualties from August 14 2006 to April 03 2007 (MACCSL, 2007)– 29 deaths – 195 injuries

• UXO casualties by the end of the two-year demining period was projected

102

43

2316

24

8 6 4 4 4 4 4 4 4 3 3 3 2 2 2 2 2 2 2 20

20

40

60

80

100

120

Num

ber o

f cas

ualti

es

Current P ro jec ted


• DALY methodology applied– Disability weight for death = 1– Disability weight for injuries resulting from UXO (leg

or arm amputation) = 0.3• DALY approaches

– Human Capital Approach• 5,300 USD as GDP per capita in 2006

– Value of Statistical Life (VSL)• 42,000 based on the VSL divided by a time horizon of 25

years and discount rate of four percent • Total damage cost of casualties resulting from

US$ 14 and 109 million over a period of two years.


Age group Current nb of casualtiesa

Current and projected nb. of casualtiesb

DALYs per

casec

DALY($ VSL)

Current & projected economic loss (million

US$/ age-group)

MORTALITY0-12 2 2.5 33 5,300-42,000 0.43 – 3.40

13-18 4 4.9 36 5,300-42,000 0.94 – 7.43

19+ 23 28.2 20 5,300-42,000 2.99 – 23.72

SubTotal 29 35.6 4.36 – 34.55MORBIDITY

0-12 24 29.5 9.9 5,300-42,000 1.55 – 12.25

13-18 39 47.9 10.8 5,300-42,000 2.74 – 21.72

19+ 132 162.1 6 5,300-42,000 5.15 -40.84

SubTotal 195 239.4 9.44 – 74.81Total 224 275 13.80 – 109.35

Estimated damage cost of UXOsEstimated damage cost of UXOs

a MACCL, 2007b Based on Figure 3.8 and the assumption that percent distribution of projected vs. current casualties is the samec Murray and Lopez, 1996

Military Waste (Unexploded Ordnances)Military Waste (Unexploded Ordnances)Access to Agricultural LandsAccess to Agricultural Lands

• Limited access to agricultural lands in the South will impact on agricultural production and farmer livelihoods for at least two years

• Farmers may respond by – burning their orchards to get rid of UXOs, losing

plantations in the process– doing nothing and wait for their lands to be cleared

from UXOs– migrating to urban areas and adding to the poverty

situation in urban belts. • It is difficult to assign a monetary value on these

types of behaviors.


• Assessing the productivity loss due to lack of access to agricultural lands.

Crop typeCultivated area (du) a Production rate b

Total production Value b Total value

South Nabatiye Total (Tonne/du) (Tonnes) (USD/tonne) (USD)

Cereals 37,638 59,525 97,163 0.28 26,781 297 7,965,069

Legumes 2,096 5,869 7,966 0.54 4,270 565 2,413,839

Fruit trees 123,304 20,768 144,073 1.26 181,973 746 135,786,268

Olives 89,340 116,124 205,464 0.29 58,759 1,268 74,525,973

Oleaginous trees 5,806 3,836 9,642 0.10 931 2,083 1,939,556

Vegetables 20,753 12,141 32,894 3.19 104,871 251 26,276,441

Raw tobacco 14,625 40,026 54,652 0.12 6,395 2,988 19,110,223

Total 268,017,000


• Scenarios adopted and associated costs

Scenario Damage cost

1) 25% of the agricultural area (evenly distributed among crop categories in

the 2 Mohafazas) will not be accessible for a whole year

67 million US$ per year







Military Waste (Unexploded Ordnances)Military Waste (Unexploded Ordnances)SummarySummary

• Scenarios adopted and associated costs

Damage cost (million $US)Minimum Maximum

Casualties 14 109Loss in agricultural opportunities 40 94De-mining 11 11

Total 65 214

Medical WasteMedical WasteSummarySummary

• Hostilities caused– 1,200 deaths– 4,400 injuries

• Around 200-250 tons of medical waste generated• All generated waste is assumed to be infectious

requiring sterilization• Handling cost of medical waste estimated at 0.015-0.045

million USD– Sterilization cost = 60 USD/ton– Disposal cost at an operational landfill = 15-120 USD/ton

• Cost of disposal of unwanted pharmaceuticals could not be estimated due to lack of data

WATER DEGRADATIONWATER DEGRADATION

Water DegradationWater DegradationImpacts on Water ResourcesImpacts on Water Resources

• Strikes on industrial facilities– Damage of Choueifat Industrial Area

• Pollution of Ghadir stream with waste residue, contaminated soil, ash

• Groundwater contaminated with heavy metals and toxic benzene

• Strikes on water and wastewater infrastructure– Affected mostly the Beirut Southern Suburbs and the South– Caused a high risk of cross-contamination and a disruption of

water and sanitation services– Imposed additional costs of securing clean water

• Destruction of bridges over the Litani River– Obstruction of flow– Risk of flooding in neighboring areas– Excessive erosion and destruction of stream banks

• Destruction of irrigation canals

Water DegradationWater DegradationDamage to Water ResourcesDamage to Water Resources

• Impact on water quality– Requires establishment of dose-response

functions– Absence of pre-conflict monitoring– Absence of data regarding water use changes

and waterborne disease incidences– No monetary estimate could be made


• Impact on water quantity– 52 water reservoirs damaged– 100s of kms of water and wastewater

networks– About 150,000 people directly affected

• Assumptions– Water reservoirs provide 50% of total water

supply– Average daily water consumption

• 1 liter/capita for drinking• 79.5 liter/capita for other uses


• Reservoir restoration– 82,900 people were gradually served by 48

reservoirs during Sept-Dec 2006 – 62,100 people were expected to be supplied

with water gradually during Jan-Dec 2007.– It is expected that 5,000 people will be served

by June 2007


• Sample calculations:– Additional costs of getting water during September-December 2006

Sep Oct Nov Dec Total

Population affected (‘000) 83 62 41 21

Cost of bottled water a (USD/liter) 0.7 0.7 0.7 0.7 …

Cost of water tanksa (USD/liter) 0.06 0.06 0.06 0.06 …

Number of days 30 31 30 31 …

Cost of drinking waterb (million USD) (1) 0.2 0.1 0.1 0.1 0.5

Cost of water for other usesc (million USD) (2) 11.3 8.7 5.6 3.0 28.6

Cost if hostilities had not occurredd (million USD) (3) 0.1 0.09 0.06 0.03 0.3

Additional cost (million USD) (1) + (2) - (3) 11.4 8.8 5.6 3.0 28.8

Notes: a market price observed during the field visit (April 2007); b based on a drinking water consumption of about 0.5 liter/capita/day; 16% of population relies on bottled water and the remaining on water tanks; c based on a consumption of 79.5 liters/capita of water for other uses; all population relies on water tanks; d assumes that 50% of the daily consumption would have been satisfied by water reservoir.


• Total additional cost of getting water = 99.4 million USD– 64.4 million USD for the 62,100 people

expected to gradually receive water during Jan-Dec 2007

– 6.2 million USD for the remaining 5,000 residents expected to be served by June 2007.

• Additional cost of repairing the water infrastructure = 33 million USD

QUARRIESQUARRIES

QuarriesQuarriesMethodologyMethodology

• Pressure on quarrying activity to supply the needed aggregate and sand for reconstruction

• Adopted methodology– estimate the amount of aggregate and sand needed

for the reconstruction, based on the amount of debris and demolition waste;

– estimate the distribution of quarrying activities by Mohafazah;

– estimate the impact of quarrying activities during operation on the surrounding environment

– non-rehabilitating quarries after completion of exploitation on the surrounding environment

QuarriesQuarriesAggregate and Sand QuantitiesAggregate and Sand Quantities

• Based on the amount of debris and rubble generated by the hostilities

Location Aggregate and sand (m3)

Beirut

Demolition waste generated a 1,430,000

Of which agg & sand (35-50%) b 501,000 715,000

Of which concrete (35-50%) b 501,000 715,000

Equivalent in agg & sand c 356,000 508,000

Sub-total (i) 857,000 1,223,000

South and Bekaa


Of which agg & sand (30-40%) b 1,296,000 1,728,000

Of which concrete (40-60%) b 1,728,000 2,592,000

Equivalent in agg & sand c 1,227,000 3,568,000

Sub-total(ii) 2,523,000 3,568,000

Average sub-total (i)+(ii) 4,086,000

Total Aggregate & Sand (adding 15% loss of raw material at quarry) 4,700,000

QuarriesQuarriesAggregate and Sand QuantitiesAggregate and Sand Quantities

• Based on the amount of debris and rubble generated by the hostilities

Location Aggregate and sand (m3)

Beirut


Of which agg & sand (35-50%) b 501,000 715,000

Of which concrete (35-50%) b 501,000 715,000

Equivalent in agg & sand c 356,000 508,000

Sub-total (i) 857,000 1,223,000

South and Bekaa


Of which agg & sand (30-40%) b 1,296,000 1,728,000

Of which concrete (40-60%) b 1,728,000 2,592,000

Equivalent in agg & sand c 1,227,000 3,568,000

Sub-total(ii) 2,523,000 3,568,000

Average sub-total (i)+(ii) 4,086,000

Total Aggregate & Sand (adding 15% loss of raw material at quarry) 4,700,000

QuarriesQuarriesDistribution of Activities by MohafazaDistribution of Activities by Mohafaza

• Based on the number of short administrative extensions granted by the Ministry of Interior and Municipalities according to the council of ministers decision #6 dated January 4, 2007 – South & Nabathieh 32%– North & Akkar 24%– Bekaa &Baalback Hermel 31%– Mount Lebanon 14%

• Assuming the scale of excavation is evenly distributed, the distribution of aggregate and sand is as follows:– South & Nabthieh 1,494,000 m3

– North & Akkar 1,105,000 m3

– Bekaa &Baalback Hermel 1,447,000 m3

– Mount Lebanon 654,000 m3

– Total 4,700,000 m3

QuarriesQuarriesEnvironmental Impacts of Quarries in Environmental Impacts of Quarries in

Mount LebanonMount Lebanon• Threats to the environment

– Destruction of natural vegetation and habitat– Air pollution from dusts– noise pollution– traffic from trucks carrying aggregates– Deterioration of road condition– Irreversible long-term visual/aesthetic impact

• Hedonic Price Method previously conducted– 4 quarrying sites in Mount Lebanon

• the depreciation in real estate price resulting from quarrying activities for the Nahr Ibrahim quarry = 4 USD/m3 of extracted aggregate and sand

• the change in property price due to the non-rehabilitation of the three other quarries at the end of operation

– Land prices: 0.13 - 40.0 USD/m3 of extracted aggregate and sand– Apartment prices: 0.4 - 5.0 USD/m3 of extracted aggregate and sand

QuarriesQuarriesEnvironmental Impacts of Quarries in Environmental Impacts of Quarries in

Mount LebanonMount LebanonQuarry 1 Nahr Ibrahim (Impact during operation)Estimated quarry area (m2)a 96,830Estimated excavated volume (m3)b 4,115,000Land area affected by the quarry (m2)c 2,000,000Decline in Land price (US$/m2) in 2002c 7.0Decline in Land price (US$/m2) in 2006d 8.2Total decline in land price 2006 16,400,557Decline in land price US$/ m3 4.0

Quarry 3 Abu Mizan (Impact after closure)Estimated quarry area (m2) in 3 locationa 276,930Estimated excavated volume (m3)b 11,769,525Land Area affected by the quarry (m2)c 175,000Decline in Land price (US$/ m2) in 2002c 8Decline in Land price (US$/ m2) in 2006d 9Total decline in land price 2006 1,537,552Decline in land price US $/ m3 0.13

Quarry 2 Shnanaayer (Impact after closure)Estimated quarry area (m2) in 2 locations a 48,370Estimated excavated volume (m3)b 2,056,000Land Area affected by the quarry (m2)c 600,000Decline in Land price (US$/m2) in 2002c 125Decline in Land price (US$/m2) in 2006d 146Total decline in land price 2006 87,860,125Decline in land price US $/ m3 43Apartments affected by quarry (m2)c 36,000Decline in apartment price (US$/ m2) in 2002c 225Decline in apartment price (US$/ m2) in 2006d 264Total decline in apartment value 2006 9,488,894Decline in apartment value US$/m3 5

Quarry 4 Antelias (Impact after closure)Estimated quarry area (m2) in 1 location a 51,577Estimated excavated volume (m3) b 2,192,000Land Area affected by the quarry (m2) c 100,000Decline in Land price (US$/ m2) in 2002 c 50Decline in Land price (US$/ m2) in 2006 d 59Total decline in land price 2006 5,857,342Decline in land price US$/m3 3Apartments affected by quarry (m2) c 7,500Decline in apartment price (US$/ m2) in 2002 c 100Decline in apartment price (US$/ m2) in 2006 d 117Total decline in apartment value 2006 878,601Decline in apartment value US$/m3 0.40

QuarriesQuarriesRelative price of land and apartments Relative price of land and apartments

by Mohafazaby Mohafaza• Land and apartments prices compiled to derive

a price of land per Mohafazah– 90 districts in Mount Lebanon– 15 districts in the Bekaa– 41 districts in the North– 10 districts in the South.

Mohafazah Relative price of land (USD/m2)

Relative price of apartments(USD/m2)

Mount Lebanon 1 1Bekaa 0.19 0.58North 0.68 1.03South 1.11 0.99

QuarriesQuarriesImpact of quarrying activities for Impact of quarrying activities for

reconstructionreconstruction• Computed as if all impacts were to happen in

2006• Alternatively, the impact should have been

– spread over a few years and than discounted back to 2006 (using a 4% discount rate).

– taking into account the inflation rate (varying between 3.5 and 4.8%)

– Hence the overall impact is likely to have been the same as if it was computed for 2006 only

• Estimated overall damage15 - 175 million USD (average = 95.5 million USD)

QuarriesQuarriesImpact of quarrying activities for Impact of quarrying activities for

reconstructionreconstructionSouth & Nabatieh min max avg

Needed aggregate (million m3) 1.5

Impact during quarrying operation on land price (million US$) 6.6

Impact of non rehabilitating quarries on land price (million US$) 0.2 71.0 35.6

Impact of non rehabilitating quarries on apartment price (million US$) 0.6 6.8 3.7

Sub Total 46.0

North & Akkar min max avg





Sub Total 22.1

Bekaa & Baalback Hermel min max avg





Sub Total 9.2

Mount Lebanon min max avg





Sub Total 18.3

AIR POLLUTIONAIR POLLUTION

Air PollutionAir PollutionSources of PollutionSources of Pollution

• Dust from reconstruction sites and quarrying activities• Increased emissions from transport sector due to

reduced average speed in affected roads or highways• Emissions from burning of petroleum products (mainly

heavy fuel oil, kerosene, gasoline, and diesel)• Emissions from forest fires• Emissions from damaged industrial facilities• Emissions from exploded weapons and ammunitions.• Other sources of air pollution such as those generated

by waste disposal and burning of dead carcasses, rotten vegetables/fruits, municipal and health care waste

Air PollutionAir PollutionImpact AssessmentImpact Assessment

• Air pollution from site clearing and removal, hauling, transport, and disposal of demolition wastes in the Beirut Southern Suburbs– Estimated using the Fixed Box Model– Total Suspended Particulates (TSP) concentration range from

190 μg/m3 under typical scenario to 860 μg/m3 under worst-case scenario

– Both values exceed the Lebanese, EU, USEPA, and WHO 24-hour standards

• Emissions from the transport sector due to decrease in average vehicle speed – Estimated to increase by a factor of 6 to 7, particularly at

hotspots


• Air pollution from the burning of around 60,000 m3 (55,764 tonnes) of fuel oil at the Jiyeh Thermal Power Plant over a period of 12 days. – Generated pollutants

• sulfur dioxide, nitrogen oxides, carbon monoxide,soot, particulate matter, semi-volatile organic compounds including polycyclic aromatic hydrocarbons (PAHs) and dioxins and furans, volatile organic compounds, such as benzene, and other compounds resulting from incompletecombustion of the oil and oil products.

– Quantities of released pollutants were calculated– The generated plume trajectory was estimated using the ALOFT-FT (A Large

Outdoor Fire Plume Trajectory- for Flat Terrains) model– The model indicated that

• Particle concentrations at their highest concentrations near the pool of fire, reaching approximating 34 mg/m3 (vertical elevation 0 m).

• Concentrations drop to 217 –295 µg/m3 at 1 to 4 km downwind and vertical elevation of 695 m

• concentrations at 20 km downwind indicate a range of particulate concentrations between 21 and 29 µg/m3 (vertical elevations 780 m and 350 m respectively).


Pollutant Emission Factors Estimated Emissions

Sulfur dioxide (SO2) 40 g/kg 2.2 Gga

Nitrogen oxides (NOx) 5 g/kg 0.3 GgParticles 15 g/kg 0.8 GgSoot 5 g/kg 0.3 GgOrganic carbon 8 g/kg 0.5 GgPolyaromatic hydrocarbons (PAHs) 0.8 g/kg 0.04 Gg

Polychlorinated Dibenzodioxins (PCDDs)2.6 µg TEQ/TJb

6 mg TEQ

Volatile organic compounds (VOCs) 7 g/kg 0.4 Gg

Carbon Monoxide (CO) 5 g/kg 0.3 GgSource : UNDP, 2007 a 1 Gg = 1,000 tonnes; b TEQ/TJ = Toxic Equivalents/ Terajoule

Emission factors and estimated emissions from the Jiyeh oil fire


• Air pollution from the burning of 40,000 tonnes of kerosene at the Rafiq Hariri International Airport.

• Pollutants released include– nitrogen oxides, particulate matter,

formaldehyde, volatile organic compounds, and polycyclic aromatic hydrocarbons

– Quantities of released pollutants were calculated

– Generated plume trajectory was estimatedusing the ALOFT model• Particulate matter concentrations are at their highest concentrations

near the pool of fire, reaching almost 3.1 mg/m3 (vertical elevation 0 m)

• Concentrations drop to 30.3 µg/m3 at 3 km downwind and vertical elevation of 725 m.

• Concentrations at 20 km downwind indicate a range of particulateconcentrations between 1 µg/m3 and 3.2 µg/m3 (vertical elevations 260 m and 725 m respectively)


Emission factors and estimated emissions from the Airport tanks fire

Pollutant Emission Factors Estimated EmissionsNitrogen oxides (NOx) 11 g/kg 441 tonnesVolatile organic compounds (VOCs) 0.133 g/kg 5.3 tonnesCarbon Monoxide (CO) 2.8 g/kg 112 tonnesSulfur Dioxide (SO2) 4 g/kg 160 tonnesPM10 1.4 g/kg 56 tonnesPolychlorinated Dibenzodioxins (PCDDs) 4.3 x 10-9 172 mgMethane (CH4) 0.02 0.8 tonnes


• The impact of the burning of– 1,000 ha of forests in Mount Lebanon– 800 ha in South Lebanon– Main pollutants released

• particulate matters, carbon monoxide, total hydrocarbons or volatile organics, and nitrogen oxides

– Emissions estimated

Pollutant Emission Factors (kg/Mg)

Estimated Emissions (Mg)

PM10 8.5 88.1CO 70 725.6VOCs as methane 12 124.4NOx 2 20.7

Air PollutionAir PollutionHealth Impact AssessmentHealth Impact Assessment

• Much of the emissions are short-term and their impacts are hard to quantify

• PM the most significant remaining pollutant• Impacts include:

– Increase in cardiac and respiratory mortality– Decrease in levels of pulmonary lung function in children and adults with

obstructive airway disease– Increase in daily prevalence of respiratory symptoms in children and

adults– Increase in functional limitations as reflected by school absenteeism or

restricted activity days– Increase in physician and emergency visits for asthma and other

respiratory conditions• A task force formed by the NSCR, AUB, and USJ is monitoring

PM10 levels in the Southern suburbs during the reconstruction period.– Generally, the recorded levels are high, exceeding national and

international air quality standards

Air PollutionAir PollutionEconomic ValuationEconomic Valuation

• Data available for long-term exposure to increased ambient concentrations of PM10

• Extrapolation for short-term exposure is difficult

FOREST FIRESFOREST FIRES

Forest FiresForest FiresIntroductionIntroduction

• Impacts of hostilities on forests– direct impacts

• accidental fires, resulting from direct bombing or fallen flares

• deliberate fires from burning the land to clear unexploded ordnances

– indirect impacts• the occurrence of summer forest fires raging

unchecked because attention was focused on humanitarian ai

• the limited accessibility to and utility from unburnt forest sites where UXOs had not been cleared

Forest FiresForest FiresImpacts on forests Impacts on forests

• The best available estimate of burnt forest area during the 34 day hostilities = 2,930 ha– Assuming that most fires occur during the three-month summer

season, the area burnt during one month = 400 ha– The area burnt due to the hostilities = 2,530 ha

• Value of damages caused by forest fires depends on the value of the forest benefits lost.– Intensive fires may cause a complete loss of benefits – Lighter fires may cause only partial losses– The degree of damage and the period over which the impacts of

fires persist depend on the intensity of fires• No accurate information on these issues has been

reported for Lebanon• 10-20 years considered time-frame for forest regeneration


• Based on Sattout 2005 the total economic value of one hectare of forests is least 465 USD per year

• This is a high estimate because – it represents the

forests’ gross benefit, which is higher than their net benefit

– the valuation is based on the actual instead of the sustainable rate of extraction

– the limited forest area in Lebanon contributes to obtaining very high averages per hectare of forests

Types of values Quantity Value(000 $)

Value ($/ha)

Use valuesFirewood (m3) 82,300 2,011 15

Charcoal (m.t.) 11,400 2,011 15

Honey and wax n.s. 12,928 96

Pine nuts (t) 600 13,000 96

Medicinal and aromatic plants n.s. 17,717 130

Fodder for grazing (mil. FU) 9.6 1,022 8

Carob (t) 2,000 625 5

Hunting (no. hunters) 600,000 12,769 95

- Legal hunting 200,000 6,384 48

- Illegal hunting 400,000 6,384 48

Recreation in reserves (no. visits) n.s. 287 2

Non-use valuesBiodiversity conservation (ha) n.s. 919 7

TEV 63,300 465


• Because of high intensity of fires, it is assumed

– Forest benefits are completely lost in 2006

– Benefits will gradually recover within 10-20 years

– Benefits recover linearly– A discount rate of 4% is used

• The Present Value of losses on 1 hectare of burnt forest ranges between 2,200 – 3,700 USD

• The total damage on 2,540 ha of forests ranges between 5.6 – 9.4 million USD

• The cost of cleaning of burned broadleaves forest = 600 USD/Ha

0

50

100

150

200

250

300

350

400

450

500

1 3 5 7 9 11 13 15 17 19 21Years

annual lo

sses

of benefits

(US$/h

a)

The total damage on 2,540 ha of forests ranges between 6.4 – 10.2 million USD

Forest FiresForest FiresImpacts on protected areasImpacts on protected areas

• Al-Shouf Cedars Biosphere Nature Reserve affected– Decline in tourism activities and sales of local

products caused a loss of 150,000 USD– This impact affects the conservation effort of

protected areas in the near future

Forest FiresForest FiresImpacts on the National Reforestation Impacts on the National Reforestation

ProgramProgram• The MOE’s reforestation program impacted in several ways:

– direct impacts: • direct shelling and bombing leading to partial or total burning of the site• lack of access due to scattered UXOs

– indirect impacts• wilting of newly planted saplings because watering schedules were

interrupted• termination of all contracts• halting of maintenance of other areas (~ 360 ha) for 1.5 years until re-

contracting, thus losing saplings and a 5-year equivalent of forest benefits• At least 5 sites in the cazas of the South and Nabatiyeh were

impactedSite Caza Area (ha) Post-conflict statusMarwanieh South 15 Not assessed yet

Rihane South 20 At least 60% burnt

Markaba Nabatiyeh 15 At least 50% burnt

Khirbet Silem Nabatiyeh 15 Not assessed yet

Zawtar el Charkieh Nabatiyeh 15 100% burnt


ProgramProgram• The replacement cost method used

– the long-term benefits provided by forests will be higher than the costs of reforestation

– Damages to saplings on burnt areas• Each reforestation site has an average density of 750 saplings/ha and an

average total cost of US$6/sapling• All the saplings in South Lebanon died as a result of the hostilities, this loss

= 360,000 USD– Cleaning the burnt sites

• The MOE will incur the costs of cleaning broadleaved burnt sites• Assuming that burnt area covered by broadleaves is proportional to that at

national level (55%), i.e., about 44 ha• At a cost of US$600/ha, the cost of cleaning operations on burnt

broadleaved forests = 26,400 USD– Forgone forest benefits

• If the saplings planted in 2004 had continued growing, the resulting forest would have provided benefits earlier than any new planting will be able

• As the burnt stands were less than 2 years old when hostilities started, and assuming reforestation at these sites is undertaken relatively quickly (~5 years)

• The present value of the delayed forest benefits is probably relatively small


ProgramProgram• As burnt stands cannot be reforested for at least 5 years,

the forgone carbon credits are a potential damage to Lebanon in this period. – Average carbon increment is 1.3tC/ha of broadleaved and 0.8tC/ha

of conifers – 45% of forest area is coniferous and 55% is broadleaved– Assuming the same distribution of forest types on the 80ha and a

market price of US$42/tC, the annual damage due to carbon loss is about US$3,300.

– The present value of this loss for the next 5 years is about US$14,600.

• Overall, the replacement costs of burnt forests = US$401,000– This figure is conservative, as it does not include

• loss of access to forests because of UXOs• cost of cleaning UXOs• loss of forgone benefits during 5 years of halted reforestation, etc.


Thank YouThank You




Session 18bPOLICY IMPLICATIONS AND

CONCLUSIONS



Session 18bPOLICY IMPLICATIONS AND CONCLUSIONS

POLICY IMPLICATIONS AND POLICY IMPLICATIONS AND CONCLUSIONSCONCLUSIONS

Assess costs and benefits via environmental valuation

IntuitiveIntuitiveDecisionDecision--makingmaking

CalculatedCalculatedDecisionDecision--makingmaking


• EIA– Quantifies and describes the physical impact of projects and

policies– Documents complexity of an environmental issue– Fails to help the decision-maker who has little knowledge of how

environmental changes affect the utility of the individual

• Environmental valuation– Gives the ‘true’ value of environmental resources to the society– Tends to remove ambiguity and vagueness in the decision-

making process

• Environmental valuation should NOT be applied– To maximize benefits in order to justify a policy– To minimize the estimated externality values of a project to

ensure its approval


• Environmental impacts should be valued in monetary terms in order that they are given due and proper weight in the decision-making process

• The non-monetization of environmental impacts may mean that either they are under-valued or over-valued in the intuitive decision-making process

• Monetization will permit the comparison of various environmentalmanagement proposals

• Many studies revealed the inconsistency of intuitive decision-making compared with a more structured approach– The numerous cognitive psychological biases in intuitive decisions

renders rational choice problematic


Setting priorities for decisionSetting priorities for decision--making in making in environmental management environmental management

• Priority setting is essential because of– Limited money– Limitless problems– Limited political and public attention– Limited time– Limited managerial time and attention

• It is essential to clearly identify and set priorities for action (& investments)



• Important questions– What criteria and approaches can be used to rank

environmental problems (and thus set priorities)?

– What are the advantages and limitations of different economic methods for defining priorities? (e.g. BCA, CEA, other methods [e.g. multicriteria approaches])

– What are the principles of and key lessons in environmental priority-setting?



• General principles for setting priorities for environmental investments/ policies – Narrow down the range of problems to be addressed

(the initial screening)

– Choose clear selection criteria with respect to types of impacts (e.g. economic, ecological, social, equity, others) and ranking or evaluating alternatives (BCA/ CEA/ MCA?)

– Consider both benefits and costs of any action/ intervention whenever possible



• Priority Setting in the real world– In a “first best” world all costs and benefits can be

valued in monetary terms and an economic efficiency criterion used to rank actions

– In a “second best” world all benefits cannot be valued and a cost-effectiveness criterion may be necessary (based on cost information)

– In a “third best” situation with little information, time or resources, qualitative ranking approaches are the best recourse


A hypothetical exampleA hypothetical example-- setting priorities setting priorities for environmental interventions with for environmental interventions with

increasing levels of information increasing levels of information

• Collecting information takes time and costs money

• Public perceptions of what are priorities are not always well-informed

• Priorities often change with increased information• The challenge is to avoid wasting money

(resources), time, and political will


A hypothetical exampleA hypothetical example-- setting priorities for setting priorities for environmental interventions with increasing environmental interventions with increasing

levels of information levels of information

Given this information, what is the priority area for investment/ intervention??

Parameter Impacts on economic growthAir quality mediumWater quality highWaste management mediumCongestion highNoise low

POLICY IMPLICATIONS AND CONCLUSIONSPOLICY IMPLICATIONS AND CONCLUSIONSA hypothetical exampleA hypothetical example-- setting priorities for setting priorities for

environmental interventions with increasing levels environmental interventions with increasing levels of informationof information

Adding information on distributional impacts: what is the priority now?? Problems and Issues:• How to compare a waste management project with a noise reduction one? Both rate “high”.• Weighting (emphasis) among the various problems may depend on political considerations.• It is possible to use experts’ opinion (Delphi technique) to identify degrees of impacts, and their relative importance to society.

Parameter Impacts on economic growth

Distributional impacts (equity concerns)

Air quality medium high

Water quality high high

Waste management medium high

Congestion high medium

Noise low high


environmental interventions with increasing levels environmental interventions with increasing levels of informationof information

Problems and Issues:

•It is necessary to define the spacial and time limits of the analysis:

• short-term vs. Long-term, • on-site vs. off-site, • financial analysis vs.

economic analysisWhat does “high” or “low”mean?

•Pollution can have different impacts:

• Productivity• Health• Recreation• Ecology

Which are the most important?



Health effects

Air quality medium high high

Water quality high high high

Waste management

medium high medium

Congestion high medium low

Noise low high low

Adding health effects: do priorities change now??


environmental interventions with increasing levels of environmental interventions with increasing levels of informationinformation

• When information is qualitative, one can use multiple criteria approaches to rank problems:– Qualitative ranking of problems using Delphi-techniques where

quantitative data is unavailable.• Delphi is an expert-based, non- confrontational approach.

– Simple qualitative approaches rank problems by multiple criteria without using trade-off considerations

– Scoring and weighting of criteria offers a qualitative approach to evaluating the relative severity of problems

POLICY IMPLICATIONS AND CONCLUSIONSPOLICY IMPLICATIONS AND CONCLUSIONSA hypothetical exampleA hypothetical example-- setting priorities for setting priorities for environmental interventions with increasing environmental interventions with increasing

levels of informationlevels of information



Health effects

Annual management costs

Air quality medium high high 1000

Water quality high high high 800

Waste management

medium high medium 900

Congestion high medium low 1500

Noise low high low 1200

When economic data on costs are available to help identify priorities for action:Cost-effective analysis (CEA) is used

POLICY IMPLICATIONS AND CONCLUSIONS POLICY IMPLICATIONS AND CONCLUSIONS A hypothetical exampleA hypothetical example-- setting priorities for setting priorities for environmental interventions with increasing environmental interventions with increasing

levels of informationlevels of information


Distributional impacts

Health effects

Annual management

costs

Annual benefits

Annual Net Benefits (benefits

minus costs)

Air quality medium high high 1000 1300 300

Water quality high high high 800 900 100

Waste management

medium high medium 900 1150 250

Congestion high medium low 1500 1300 (200)

Noise low high low 1200 1100 (100)

Conduct the full Benefit Cost Analysis when both benefits and costs are known

POLICY IMPLICATIONS AND CONCLUSIONSPOLICY IMPLICATIONS AND CONCLUSIONSSelecting the appropriate valuation technique Selecting the appropriate valuation technique ––

a valuation flow charta valuation flow chart Environmental Impact

Measurable change in production

Change in environmental quality

Yes

Nondistorted market prices available?

Use change-in-productivity approach

Use surrogate market approaches, apply shadow prices to changes in production

Yes No

Habitat

Opportunity-cost approach

Replacement cost approach

Land value approaches


Air and water quality

No

Cost-effectiveness of prevention

Preventive expenditure

Replacement/ relocation costs

Health effects

Sickness Death

Medical costs

Loss of earnings

Human capital

CEA of prevention

Recreation


Travel cost

Aesthetic, Biodiversity, Cultural, Historical assets

Contingen Valuation


Hedonic wage approach



The Limits of Economic AnalysisThe Limits of Economic Analysis

• Areas where economic analysis is often weak include the following:– Incremental impacts– Uncertainty (especially with regards to the future)– Irreversible impacts– The value of genetic material (or biodiversity)– Preferences of future generations (and projects with

long time horizons)– Distributional effects across social sectors


• Much can be done with the available tools of economic analysis at both the macro and micro level

• Both environmental policies and investment projects can be analyzed using the tools of environmental economics

• The hard valuation areas (e.g. recreational demand, cultural values, genetic materials, biodiversity, others) are also often the same things that we care the most about in the environment


• Factors for limited use of environmental values in decision-making– Skepticism towards environmental valuation methods– Lack of environmental economists within government agencies– Absence of a legal requirement to undertake a CBA of projects

or policies– Uncritical acceptance of other methods such as

• Effect on production• Dose response• Opportunity cost approaches

– Suspicion of non-use values– Distorted perceptions of the valuation methods by non-

economist– Large variance associated with mean WTP and WTA values


Prevailing Situation• Environmental values are less routinely

incorporated into policy and project appraisal in a systematic way

• Environmental changes tend to be assessed through EIAs in the US and EU rather than through economic valuation and CBA

• The World Bank and the Asian Development Bank advocate the use of valuation methods to estimate the welfare effects of environmental changes


Prevailing Situation• Environmental valuation studies in different European

countries were undertaken spasmodically with varying degrees of influence on decisions and with marked variations between countries– Switzerland

• Highest number of academic/ scientific studies employing TCMs, HPMs, CVMs

– Germany• Proportionately fewer and more policy oriented

– UK• Shift away from TCMs to HPMs and CVMs due to nature of goods

being valued– Netherlands

• Demand for valuation studies by governments and organizations is low– Norway

• Benefit estimation studies provided support for environmental decision-making but had not played a crucial role in the process


Prevailing Situation• In some cases the environmental valuation process is formalized

and fairly explicit and institutionally incorporated in the decision-making process– US Forest Service

• Application of ‘unit day values’ of recreational opportunities and resources

– CERCLA• Type A assessment of natural resource damage from pollution spills

– Using an existing economic database• Type B assessment for major pollution incidents

– Requires a site specific investigation

– US Department of Interior authorized methods for environmental valuation

• Market price where applicable• Uniform Appraisal Standards for Federal Law Acquisition• Use values may be measured via

– TCM, HPM, unit values, CVM, and stated preference techniques• Non-use values may only be measured via

– CVM and SP


The future

• Environmental valuation will witness a search for – More accurate and robust semi- and non-parametric estimators– Improved understanding of the psychology of making choices

and decisions– The analysis of the non-stationarity of environmental values– The application of other theories and techniques from other

branches of economics• Bayesian perspectives• Game theory

• Desire to establish formal benefit transfer methods by governments and agencies

• Advocacy of the use of benefit transfer by organizations

Environmental valuation methods becoming• more institutionalized•more routinely included in CBAs


• Effort should be directed in the MENA region to – Increase awareness on environmental valuation– Build capacity on COED methodology– Institutionalize COED methodology in decision-

making process– Establish a database for environmental valuation

studies in the region

• Various databases on environmental valuation could be of help


Name of database

Web host Purpose of the database Number of studies

Regions covered

Available languages

Environmental Valuation Reference Inventory

Environment Canada on behalf of the EVRI Club1 http://www.evri.ca

To help policy analysts using the benefits transfer approach to estimate economic values for changes in environmental goodsand services or human health

1,500 International English,French

Envalue New South WalesEnvironment Protection Authority http://www.epa.nsw.gov.au/envalue

To help stakeholders value changes in environmental quality


Ecosystem Services Database

Gund Institute forEcological Economics, University of Vermonthttp://esd.uvm.edu

To provide a data and analysis portal to assist in the informed estimation of the economic values of ecosystem services


Review of Externality Data

European Commissionhttp://www.red-externalities.net

To assist policy makers in capturing the effects of externalities from new policies that have sustainable development as their core concern


Main features of selected valuation databases (McComb et al, 2006)


Name of resource Web host Database purpose Overview

New Zealand Non-market Valuation Database

Lincoln University, Canterbury, New Zealandhttp://oldlearn.lincoln.ac.nz.markval

To help researchers identify nonmarket valuation studies undertaken in New Zealand

Searchable database with 100 primary studies from New Zealand

ValuebaseSwe Beijer International Institute of Ecological Economics, and the Swedish Env. Protection Agency http://www.beijer.kva.se/valuebase.htm

To provide a survey of empirical economic valuation studies on environmental change in Sweden

Database with 200 primary studies from Sweden

Beneficial Use Values database

Department of Agricultural and Resource Economics, University of California, Davis http://buvd.ucdavis.edu/

A guide for decision makers, policy analysts, and others interested in valuation of water resources

Database of economic values for beneficial uses of water. Varietyof sources

Sportfishing Values database

Industrial Economics, Incorporated under contract to the U.S. Fish and Wildlife Servicehttp://www.indecon.com/fish/default.asp

To provide a detailed account of the contents of numerous recent non-market valuation studies

One hundred non-market valuation studies of sports fishing activity



Name of resource

Web host Database purpose Overview

Biodiversity Economics

IUCN-The World Conservation Union and WWF http://www.biodiversityeconomics.org/

To encourage and assist in the use of economics in support of biodiversity conservation and sustainable development

Library of 100 non-market environmental valuation studies and a host of others on incentives, and business and consumer relations

National OceanEconomics Project (NOEP)

U.S. National Oceanic and Atmospheric Administration http://noep.csumb.edu/

The creation and distribution to the public of a spatially and temporally consistent data set that will support a wide range of economic, scientific and resource management activities

Library of 200 non-market valuation studies and a database of market values from around the world

Ecosystem valuation

D. King (U. of Maryland) and M. Mazzotta (U.of Rhode Island), funded by the U.S. Department of Agriculture and National Oceanographic and Atmospheric Administration, U.S. http://www.ecosystemvaluation.org

For non-economists who need answers to questions about the benefits of ecosystem conservation, preservation or restoration

Clear, non-technical description of ecosystem valuation concepts, methods and applications

Environmental damagevaluation and cost-benefit news

Editor and Publisher: Kenneth Acks http://envirovaluation.org/

Newsletter on valuation of environmental damages

Legal, academic, and regulatory developments pertaining to the valuation of environmental amenities and disamenities



Thank YouThank You


LIST OF PARTICIPANTS

Name Country Position Institution Email Phone

1. Abboud, Mazen

Lebanon President Union of the Northern Associations for Development, Environment and Patrimony

[email protected] Phone: +961 5467128 Mobile:+961 3283642

2. Abdul Samad, Lama

Lebanon Environmental specialist

COSV ‐ Lebanon [email protected] [email protected]

Phone: +961‐5‐452 838 Mobile:+961‐3‐937 950

3. Akl, Georges

Lebanon Forest engineer Ministry of Environment [email protected] Phone: +961‐1‐976 555 Ext. 452 Mobile: +962‐3‐614 303

4. Al Daia, Roula

Lebanon Acting director Institute of the environment‐ Environmental economics program‐ University of Balamand

[email protected] Phone: 961‐930250 ext. 3966 Mobile: 961‐3‐152 726

5. Al‐Oran, Raeda Jordan Environmental officer

Ministry of Municipal Affairs, Regional Planning Department, Project Management Team of the Regional and Local Development Project (RLDP)

[email protected] Phone: +962 6 5235585 Mobile:+962 795059649

6. Asfour, Feras

Syria Head of Planning Section

Ministry of Local Administration and Environment

ferasenv@scs‐net.org Phone: +963‐11 4465905 Mobile: +963‐944‐380810

7. Batta, Shareef

Jordan/ West Bank

Director of internal auditing

Environmental Quality Authority/ Palestine

[email protected] [email protected]

Phone: +970 599674797‐Palestine Phone: +962 53992179‐ Jordan

8. Beainy, Nada

Lebanon Program Officer Italian Cooperation ‐ Lebanon [email protected] Phone: +961‐5‐451494 Mobile:+961‐3‐909 744



9. Bou Jawdeh, Issam

Lebanon Consultant Self‐Employed [email protected] Phone: +961‐4‐808 097

10. Chaabi, Ali

Tunisia Sub‐director Ministry of Agriculture‐ Office de developpment SYLVO‐PASTORAL DU NORD OUEST

[email protected] Phone: 216 78 655 810 Mobile: 216 98 500 920

11. El Hajj, Rana

Lebanon Project Coordinator AFDC [email protected] Phone: 961 1 752670 Mobile: 961 3 404625

12. El‐Shalakamy, Mohamed

Egypt Environmental Specialist & EPAP PMU staff

Egyptian Environmental Affairs Agency

[email protected] Phone:+202 22723172 Mobile: +201 01011235

13. Fanous, Ramzi

Lebanon Statistician Ministry of Environment [email protected] Phone: +961‐1‐976 555 ext. 459 Mobile: +961‐3‐594 283

14. Ghanimeh, Sophia

Lebanon Instructor and PhD candidate

Notre Dame University, Faculty of Engineering

[email protected] Mobile: +961‐3‐874 380

15. Ghannouchi, Sana

Tunisia Ingénieur principal Observatoire Tunisien de l'Environnement et du Développement Durable

[email protected] Phone: +216 22 519 725

16. Ghosn, Sabine

Lebanon Engineering Management Specialist

Ministry of Environment [email protected] Phone: +961‐1‐976 555 ext. 437 Mobile: +961‐3‐740 171



17. Hakimi, Abdurrahman

Libya METAP National Focal Point

Environment General Authority (EGA)

[email protected] Mobile: +218 21 3612836

18. Hassan, Mahgoub

Sudan National environmental expert

Ministry of Environment and Physical Development

[email protected] Mobile: +249912310284

19. Higazy, Mamdouh

Egypt Environmental specialist & EPAPII‐ PMU staff

Egyptian Environmental Affairs Agency

[email protected] Phone: +202 26355507 Mobile: +201 01583878

20. Lichaa El‐Khoury, Dany

Lebanon Environmental & land use planning expert

[email protected] Mobile:+961‐3‐858 943

21. Mawla, Darine

Lebanon Environmental Specialist

[email protected] Phone: +961‐4‐531173 Mobile: 961 3 682147

22. Mitri, Ghada

Lebanon Development Specialist


23. Nasr, Raoul

Jordan Agricultural Economist

Jordan University of Science and Technology, College of Agriculture

[email protected] Phone: +962‐2‐7201000 ext 22242 Mobile: +962‐79‐5580296

24. Omeira, Nada

Lebanon Member Green Line Association [email protected] Phone: +961‐1‐746 215 Mobile: +961‐3‐351 821



25. Rachid, Grace



26. Roukoz, Salim

Lebanon Engineer Ministry of Agriculture [email protected] Phone: +961‐1‐849 623 Mobile: +961‐3‐665 719

27. Saliba, Salah

Lebanon Coordination officer UNDP / UNRC [email protected] Phone: +961‐9‐932 650 Mobile:+961‐3‐318 093

28. Sattout, Elsa


[email protected] Mobile: +961‐3‐601767

29. Slika, Marwan

Syria Ministry of Local Administration and Environment, General Commission for Environment Affairs

[email protected] Phone: +963‐11 4465905 Mobile: 0932215790

30. Shatnawi, Malak

Jordan Head of section Cities and villages development bank

[email protected] Phone: +962 6 5682690 Mobile:+962 777787110

31. Al Duaij, Samia Kuwait Operation analyst METAP [email protected]


AUB and World Bank teams


32. Bou Fakhreddine Raja

Lebanon Environmental consultant

American University of Beirut rboufaldeen@sets‐lb.com Mobile: +961‐3‐575 229

33. Chaaban, Jad

Lebanon Assistant Professor American University of Beirut [email protected] Phone: +961‐1‐350 000 Ext. 4067

34. Dobardzic, Saliha

METAP Officer World Bank [email protected]

35. Doumani, Fadi

Lebanon Economist METAP consultant [email protected] Phone: +12022232623 Mobile: +12022151722

36. El‐Fadel, Mutasem

Lebanon Professor American University of Beirut [email protected] Phone: +961‐1‐350 000 Ext. 3470 Mobile: +961‐3‐228 338

37. Hindi, Khalil

Lebanon Professor American University of Beirut [email protected] Phone: +961‐1‐350 000 Ext. 3950

38. Jamali, Dima

Lebanon Associate professor American University of Beirut [email protected] Phone: +961‐1‐350 000 Ext. 3727

39. Lotayef, Dahlia

METAP Coordinator World Bank [email protected]

40. Maroun, Rania

Lebanon Environmental consultant

American University of Beirut rmaroun@sets‐lb.com Mobile: +961‐3‐396 318

41. Nuwayhid, Iman

Lebanon Professor American University of Beirut [email protected] Phone: +961‐1‐350 000 Ext. 4627

42. Salti, Nisreen

Lebanon Assistant Professor American University of Beirut [email protected] Phone: +961‐1‐350 000 Ext. 4068