economic value of the bwindi and virunga gorilla mountain ... · gorilla gorilla beringei beringei....

AWF Working Papers

July 2005

Richard Hatfield

Economic Value of the Bwindi and VirungaGorilla Mountain Forests

©JoseK

alpers

The African Wildlife Foundation,

together with the people of Africa,

works to ensure the wildlife and

wild lands of Africa will

endure forever.

www.awf.org

Arusha Center (Tanzania)African Wildlife FoundationPlot 27, Old Moshi RoadP.O. Box 2658ARUSHA, TANZANIATel: +255 27 2509616Fax: +255 27 2544453email: [email protected]

Nairobi Center (Kenya)African Wildlife FoundationBritak CentreMara Ragati RoadP.O. Box 48177, 00100NAIROBI, KENYATel: +254 20 2710367Fax: +254 20 2710372email: [email protected]

Kampala Center (Uganda)African Wildlife FoundationRuth Towers15A Clement Hill RoadP.O. Box 28217KAMPALA, UGANDATel: +256 41 344 510Fax: +256 41 235 824email: [email protected]

Washington D.C. Center (U.S.A.)African Wildlife Foundation1400 Sixteenth Street, N.W.Suite 120WASHINGTON, D.C. 20036, U.S.A.Tel: +1 202 939 3333Fax: +1 202 939 3332email: [email protected]

White River Center (South Africa)African Wildlife FoundationP.O. Box 2977WHITE RIVER 1240,SOUTH AFRICATel: +27 13 751 2483Fax: +27 13 751 3258email: [email protected]

Zambezi Center (Zambia)African Wildlife Foundation50 Independence AvenueP.O. Box 50844RidgewayLUSAKA, ZAMBIATel: + 260 1 257074Fax: + 260 1 257098email: [email protected]

About this paper series

The AWF Working Paper Series has been designed to disseminate to partners and the conservation community,aspects of AWF current work from its flagship African Heartlands Program. This series aims to share currentwork in order not only to share work experiences but also to provoke discussions on whats working or not andhow best conservation action can be undertaken to ensure that Africas wildlife and wildlands are conservedforever.

This paper is based on a study commissioned by the International Gorilla Conservation Program with Supportfrom the Howard G. Buffet Foundation and conducted by Richard Hafield and Dr. Delphinne Malleret-King in2002. An earlier edition was presented to the People in Parks: Beyond the Debate Spring Conference organizedby the International School of Tropical Forestry, Yale University in March 2004 as well as the Sixth BioEconWorkshop at Kings College, University of Cambridge in September 2004.The International Gorilla Conservation Program is a coaltion of the African Wildlife Foundation, World WildlifeFund for Nature and Fauna and Flora International.

Copyright: This publication is copyrighted to AWF. It may be produced in whole or part and in any form foreducation and non-profit purposes without any special permission from the copyright holder provided that thesource is appropriately acknowledged. This publication should not be used for resale or commecial purposeswithout prior written permission of AWF. The views expressed in this publication are the authors and they donot necessarily reflect those of AWF or the sponsor of this publication.

This working paper series is coordinated by Daudi Sumba, email [email protected] photo: Tourist viewing gorillas in Bwindi Impenetrable Forest National Park

AWF Working PapersJuly 2005

AWF Working Papers - July 2005

1

page 2

page 2

page 2

page 3

page 4

page 4

page 6

page 6

page 7

page 8

page 12

Table of Contents

Abstract

Background/Introduction

Study Area

Methods

Results

Forest Benefits

Forest Costs

Aggregate Benefits and costs and their Distribution

Conclusions

Discussion

References


The EconomicValue of Virunga and Bwindi Mountain Gorilla Protected Forests

2


Background/Introduction

The Virunga massif and Bwindi protected forest parksrepresent important local centres of diversity withinEast/Central Africa’s biodiversity-rich Albertine RiftValley. Particularly well-known as home to the world’sonly remaining natural populations of mountain gorilla(Gorilla beringei beringei), currently numbered at around700 individuals, the long-term viability of both forestsremains threatened despite the existence of long-runningconservation programs that have helped generatesubstantial gorilla-viewing tourism income.

Two broad factors account for this situation: first,forests tend to be undervalued on a global scale, sincethey provide significant benefits which to date havereceived little attention, either due to lack of knowledgeor difficulty in quantification. These include ecologicalservices such as climate control, waterflow regulation,soil retention, and the wider benefits of atmosphericpollution control; biodiversity value; aesthetic value;value to future generations; and ethical value. Second,the socio-political-economic context within which thetwo forests exist is not conducive to natural resourceconservation, being characterized by high humandensities; severe land pressure; and perceived unequaldistribution of tourism benefits - exacerbated by large-scale political conflict.

This study seeks to address these issues in three specificways: (1) to determine a more accurate estimate of the

AbstractThe Virunga and Bwindi afro-montane forests of Eastern/Central Africa are best known as home to the mountaingorilla Gorilla beringei beringei. A study was undertaken to estimate the economic value of the protected forests, and thedistribution of benefits and costs between local, national and international stakeholders. Results suggested that theforests are generating positive benefits - both tangible and intangible - relative to costs; but that benefits areoverwhelmingly accruing to the international community, with little-to-no benefit accruing to those countries containingthe protected areas. The implications are that the international community should be paying a greater share for thebenefits it enjoys; and that the real engine of development - and sustainable forest conservation - is likely to involveinvestment into local small-holder agricultural livelihoods.

Keywords: Virunga and Bwindi forests, mountain gorilla, total economic value, direct use value, indirect use value,non-use value, forest opportunity costs, benefit-cost distribution

true value of the forests, by conducting valuation ofmultiple forest benefits and costs; (2) to identify andexplore important economic drivers within locallivelihoods; and (3) examine the distribution of forestbenefits and costs between international, national andlocal levels. Results are used to:

• provide a baseline from which to further refineand/or expand on the current study

• draw policy conclusions• explore the economic impact of possible

interventions

Study Area

The forests of the Virunga volcanoes massif and Bwindirepresent two isolated and protected afromontane tractsof a once more extensive forest, now supporting one ofthe highest human population densities in Africa.Together they represent important “epicentres” withinthe Albertine branch of East Africa’s Great Rift Valley -an area listed in the top 20 of the Global 200 priorityareas for biodiversity.

Bwindi Impenetrable National Park (BINP) lies whollyin south-west Uganda. Created in 1991, it protects oneof the most diverse afromontane forests of the world,with a number of endemic trees, plants, and birds(Cunningham, 1996). The park consists of 32,100 ha.of rugged land - steep, narrow valleys bordered by hillcrests of 1200 m. in altitude in the north and 2600 m.in the south. Before being established as a nationalpark, the area was set up as forest reserve in 1932 and a



3

wildlife reserve in 1964. Throughout this period,timber was exploited and it is estimated that about 30%of the forest was cleared between 1954-91. Due to itsunique characteristics and its richness, it was declared aUNESCO World Heritage Site in 1995 (Wild andMutebi, 1996).Located approximately 35 kms. to the south-west ofBwindi, the Virunga forest cloaks the chain of sixspectacular volcanoes shared by three countries: Ugandato the northeast; Rwanda to the south; and theDemocratic Republic of Congo (DRC) to the northand west. Whilst remaining a single contiguous unit,the forest comprises three separate national parks:Rwanda’s Parc de Volcans (PNV); the Mikeno sectionof DRC’s Parc de Virunga Sud (PNVi-Sud); and Uganda’sMgahinga Gorilla National Park (MGNP). After severalexcisions, the PNV portion comprises 160 sq. kms. ofhigher altitude forest; DRC’s PNVi-S covers the largest

section - 250 sq. kms.; whilst MGNP comprises a smallsection of 27 sq. km. Whilst separately administered, arelatively large degree of co-operation exists betweenparks - facilitated by such conservation partners as theInternational Gorilla Conservation Program, andperhaps encouraged by common interest in the face ofpolitical instability.

Both forests are surrounded by some of the highesthuman population densities in Africa, with up to 400people per sq. km. around the Virunga volcanoes massif(Pers. Comm. 1), due to a combination of land pressureand rich volcanic soils apt for agricultural use. Thedominant land use activity surrounding both forests issmall-holder agriculture, with some 90% of the localpopulation earning a living from farming (Pers. Comm1).

MethodsThree main outputs were pursued:

• Forest benefits• Forest costs• Distribution of forest benefits and costs

Output 1: Forest BenefitsForest benefits fall into three categories, all of which

were investigated:direct use value (forest

products)indirect use value (mountain

gorilla-based tourism)non-use value (ecological

service value; biodiversity, orexistence, value; option (for futureuse) value; and bequest value)

A mixture of estimation methodswas utilized. Direct use value andwildlife damage cost was investigatedthrough formal face-to-facehousehold surveys carried out byfoot, following a randomized,structured transect and householdselection process. The travel-costmethod (for example, Brown andHenry, 1993) - which uses visitorexpenditure as a proxy for value - wasused to determine indirect use valuei.e., gorilla-based tourism, wheretourist expenditures on gorilla-

viewing were ascertained through questionnaire surveysover an eight day period. In addition, through analysisof demand this methodology allows estimation ofvisitors’ consumer surplus i.e., the additional value thataccrues to visitors from their gorilla-viewing experience,over-and-above expenditure - in essence, what they were

Map of Study Area



4

willing to pay as opposed to what they did pay - asignificant and commonly overlooked aspect ofconsumer benefit. Forest non-use value, together withforest opportunity cost, were estimated using contingentvaluation - an emerging principal method for determiningthe implicit (non-market) value and/or cost of “goodsand services” yielded by an environmental resource - inthis case, the protected forests. Contingent valuationascertained willingness-to-pay (WTP) amongst variousstakeholders (local, national, international) to eitherprevent deterioration of the resource - forest cover ormountain gorillas - or to improve the condition of forestcover, as appropriate.

WTP was determined using a double-boundeddichotomous choice (or double referendum) bidmethod, with an open-ended follow-up question:respondents were given a scenario and asked if they werewilling to pay a certain predetermined amount (bid)i.e., a discrete yes/no bid. Depending on the response,bids were either doubled or halved, and the questionrepeated. Respondents who answered no to both bidswere then asked to nominate their bid i.e., an open-ended amount. The main advantages of this methodare: (a) more accurate elicitation of WTP through useof two questions; (b) minimization of ‘strategic bias’i.e., deliberately understating WTP, since respondentsdo not decide the amounts (except in the open-endedscenario, where care must be taken during analysis todiscard obvious ‘protest’ bids); and (c) minimization of‘yea-saying’, or giving what is considered to be the corrector desired answer (for example, Carson, 1998; IIED,1997). Whilst possessing great flexibility and potentialutility, the main weakness of contingent valuation isthe reliance on peoples’ views rather than marketbehaviour. Consequently there are many sources ofpotential bias - particularly the realism of the scenario;and how the scenario is explained. Results were thenanalyzed through use of a logistic regression model.

Output 2: Forest CostsThe costs of the forest were assumed to be limitedto:

wildlife damageopportunity cost, that is, the value of lost

opportunities due to the existence of the forests

Two different methodologies were used to estimateforest costs amongst residents within 5 kms. of theforests: contingent valuation and estimation of incomefrom agriculture. This allowed direct comparison forconsistency of results. Both methodologies employedface-to-face household surveys using structured foottransects and questionnaires. For contingent valuation,local residents were asked what amount of annualcompensation they would be willing-to-accept (WTA)in order to offset the costs of the forests’ existence. Asin the case of WTP (above), the double referendum bidmethod was used and analyzed in a similar manner. Thesecond method, agricultural productivity, was used todetermine the cost of the protected Virunga-Bwindiforests in terms of the value of lost opportunitiesresulting from their existence – most likely conversionto smallholder agriculture. Household questionnairesurveys were carried out amongst a total of 735 farmerswithin a 5 km. band around each forest, with the numberof questionnaires conducted around each parkdetermined by individual park size.

An important assumption made here is that forestopportunity cost is borne by local residents i.e., thatthey would be the beneficiaries of forest conversion.This would not necessarily the case; it could be arguedthat the costs are borne at the national level, if decisionsconcerning the most desirable alternative use of theforests rest within government policy.

Output 3: Distribution of Forest Benefits and CostsIn addition to determining overall value for each benefitand cost, breakdown between international, nationaland local levels was carried out wherever possible.

ResultsResults are organized around the above four outputs:A. Forest benefitsB. Forest costsC. Distribution of benefits and costs betweeninternational, national and local levels

A. Forest BenefitsA.1 Direct Use Value: Forest ProductsWhile it was hoped that respondents would feelsufficiently secure to divulge (illegal) resource use withinthe protected areas, results were not forthcoming. Thisis an area for possible further research.



5

A.2 Indirect Use Value: Gorilla-Based TourismGorilla-based tourism levels vary considerably from yearto year, dependent on prevailing perceptions ofinsecurity in the region. Results from sampled touristexpenditures were applied to the latest available annualvisitor statistics (2000-2001). Annual gorilla viewingexpenditures amounted to an equivalent of USD7.75million - including USD2.78 million in gorilla trackingfees - and constituted 31% of total safari expenditure,on average. Annual ‘consumer surplus’ value accruingto international visitors amounted to a furtherUSD5.89 million. The secondary impacts to theeconomy from gorilla tourism expenditures were alsoconsidered, using previously determined income- andtax-multipliers, and indicated additional benefits ofUSD4.48 million in terms of secondary incomegeneration within the economy and USD3.10 millionin tax generation.

Fig. 1 above summarizes the benefits categories andlevels. Overall, gorilla tourism generates USD20.6million per year in benefits, with 53% accruing to thenational level; 41% to the international level; and 6%to the local level. The largest single benefit componentis international tourist consumer surplus (28% of totalbenefits) followed by national income generation (17%)and national tax impact (15%). International travelrevenue and gorilla tracking fees both captured 13% ofbenefits. As implied, local gains (direct and indirect)constituted the smallest proportion of benefits. Officialtourism statistics showed that for the year 2000-2001,gorilla-viewing operated at 41% of full capacity,suggesting potential for increasing revenues, with a

maximum attainable value of USD51.7 million per year.However this makes no allowance for seasonal demandvariations throughout the year. International visitorsmake up 84% of total visitor numbers to mountaingorilla parks, with 81% of all visitors being from threegeographical blocks: Europe (42%), U.S.A. (20%), andAustralasia/Japan (19%). The distribution of gorilla-tourism benefits across international, national and locallevels is summarized in Fig. 2 below.

A.3 Non-use ValueSamples were drawn and estimation carried out for fourpopulations:

• local residents surrounding the forests;• national citizens;• international citizens (non-gorilla tourists from

the three major geographical visitor blocks); and• ecological service value to agriculture based on

the experience of the former, now deforested,Gishwati forest in Rwanda.

Results gave an annual forest non-use value to localresidents of USD0.2 million, and USD1.0 million tonational citizens. Local ecological services were worthUSD0.2 million annually. By contrast, non-use valueto international citizens resulted in an annual estimateof USD186.5 million, motivated primarily by ethicaland intrinsic (existence) values. However, the latter twoestimates were based on relatively narrow samplepopulations and limited sample size due to time andresource constraints; both would benefit from greaterin-depth treatment in order to verify results, particularly

Fig 1: Annual benefits accruing from gorilla-basedtourism

InternationalTourists (Consumer

Surplus), 28%

InternationalRevenue, 13%

National (non-local)Revenue, 8%

Park Fees, 13%

National Income(non-local) Impact,

17%

Government TaxImpact, 15%

Local Revenue, 2%

Local IncomeImpact, 4%

Fig. 2: Distribution of annual gorilla tourism benefits(2000/2001)

8.7

11.2

0.7

20.6

0

5

10

15

20

25

International Level National Level Local Level Total

US$

mill

ion

per

year



6

given the significance of the result for international non-use value.In addition, international ecological service benefit inthe form of carbon sequestration was estimated throughsecondary sources, and valued at USD 0.7 million peryear.

B. Forest CostsB.1 Wildlife DamageData concerning loss from wildlife damage wasdetermined to be overstated and too inconsistent to beof use - for example it was commonly found that somerespondents claimed damage significantly over-and-aboveclaimed farm production. Again, this is an arearequiring more focused and carefully designed research.

B.2 Opportunity Cost Estimate 2: Agricultural ProductivityResults gave an overall average net income per hectareof USD436 per year, translating into an annual(opportunity) cost of USD15 million in terms of forgoneagricultural potential due the existence of the forests,assuming 50% of the existing park areas are cultivable.As depicted in Fig. 3, net income varied significantlybetween countries, being highest in Rwanda and DRC,as well locally between parishes and distance from parkboundary. However, no significant difference was foundregionally between Virunga and Bwindi.

On average, 83% of income is derived from crops withthe balance from livestock production. Woodlotsrepresent a third common land-use, however, valuationproved problematic and was therefore excluded.

B.3 Opportunity Cost Estimate 1: Contingent ValuationResults indicated an annual opportunity cost ofUSD13.4 million. This compares to the earlier,alternative estimate of annual opportunity cost ofUSD15.0 million estimated through the farmproduction survey. The proximity of the two resultsincreases confidence as to the true magnitude of forestcosts at the local level.

C. Aggregate Benefits and Costs, with Distributionbetween International, National and Local LevelsC.1 Aggregate Benefits versus CostsFig. 4 depicts the individual benefit and cost values overthe estimated categories. The overall aggregate total grosseconomic value (gross benefits less local forest costs) ofthe protected Virunga-Bwindi forests is USD196.4million on an annual basis. However, 90% of this totalis derived from the non-use value of mountain gorillahabitat to the international community. Of theremaining benefits, gorilla tourism is responsible for91%. However, it can be noted that the latter benefitsare tangible, in contrast to international non-use value.

Note: Overlap does exist between the following categories: ‘local indirectuse’ and ‘local use & non-use’; ‘International non-use’ and ‘internationalecological’. This is due to use of different (overlapping) methodologiesto ascertain the respective values (see text for details).

Fig. 5 shows that the distribution of benefits and costsamongst stakeholder levels is heavily skewed: annualbenefits of USD12.2 and USD9.4 million accrue tothe national and international levels respectively, whilea loss of USD11.7m is being absorbed at the local level.

Fig. 3: Net farm income by country and region

617.3

446.072

261.212

469

324

436.4

0

100

200

300

400

500

600

700

Rwanda DRC Uganda Virunga Region Bwindi Region Average

Country and Region

US$

per

hect

are

per

year

Net farm income per ha

Fig. 4: Gross annual forest benefits and costs

8,700,000

11,200,000

1,300,000

186,500,000

1,000,000

200,000

700,000

200,000

-13,400,000

-50,000,000 0 50,000,000 100,000,000 150,000,000 200,000,000

International Indirect UseTourism

National Indirect Use Tourism

Local Indirect Use Tourism

International Non-use

National Non-use

Local Use and Non-use

International Ecological

Local Ecological

Local Costs

Ben

efit

san

dC

osts

US$ millions per year



7

* Local tourism, use and non-use benefits less farming opportunitycosts** National direct and secondary tourism benefits plus non-use value,without consideration of National Park administration costs.*** International tourism consumer surplus and non-use value lesstourism travel cost

ConclusionsFive main conclusions emerge from the study:1. From an economic standpoint, the Virunga and

Bwindi forests should continue to be protectedsince, overall, they are generating significant benefitsover-and-above the costs borne.

2. There should be a commitment to capturing agreater part of the potential revenue currentlyoffered by mountain gorilla tourism, since tourismis the driving force of tangible benefits realized fromthe forests.

3. The international community is the majorbeneficiary of forest protection - mainly in terms ofnon-use value, but also significantly in terms oftourism benefits.

4. If it is assumed that the opportunity cost of theforests are being borne by the local community,then significant tangible benefits are being realizedat the national level (foreign exchange repatriationnot withstanding) whilst local communities areexperiencing losses of a similar magnitude level.Alternatively, if it is assumed that the opportunity

costs of small-holder agriculture are borneat the national level, meager tangible andnon-tangible benefits are being realizedboth at the national and community levels.5. The magnitude of international non-use value represents an appropriatepotential source of funding for addressingthese inequalities. However, the linksbetween forest protection, internationalinvestment, and local development wouldrequire emphasis in order to ensure thetwin outcomes of sustained forestconser vation and local/nationaldevelopment.

Limits and LimitationsSome of the most important limits andlimitations of the study have been alluded

to. These include:

1. Study Design GapsThe study set out to complete a full economic valuationof the Virunga-Bwindi protected forests. However, itwas recognized from the start that such a goal couldnot be completed given the reality of limited resources.As such, the goal remained to design an appropriatemethodology for better determining full economic value,and to complete the valuation as widely and fully aspossible. Accordingly, resources dictated that the studybe limited to the forests and areas immediatelysurrounding them. For this reason, the study remainsa preliminary baseline on which to build with furtherstudy. The most important gap to fill is a full economicvaluation of hydrological services throughout the waterbasins served by the Virunga-Bwindi systems – whichinclude not only agriculture principally, but also extendto the fisheries of Lake Edward. Such a valuation willgo a long way in better establishing the real worth ofthe forests at both local, national, and regional level –and will represent a crucial component of a full economicvaluation. Other gaps include determination ofprotected area administration costs in arriving at nettourism revenues.

2. Inadequate ResultsFour areas covered by the study require further work.At the local level both direct forest use value and directforest costs (wildlife damage and banditry) failed to yield

-11.7

12.2

195.9

-50

0

50

100

150

200

250

Local Net Loss National Gain International Net Gain

Distribution of Costs and Benefits

US$

Mill

ions

per

year

Fig. 5: Distribution of annual forest benefits and costs between stakeholder levels



8

positive results, due mainly to lack of revelation aboutwhat amount to illegal activities, and gross exaggeration,respectively. Proper determination would appear torequire a less rapid assessment carried out by trustedenumerators. Plans for this are underway through insitu organizations. Also required is in-depth study tomore accurately determine local tourism benefits. Thosearrived at here required interpolation based on limitedavailable data. Such a study should focus on theownership status of tourism facilities and economiclinkages between local suppliers and the tourism sector,as well as local multiplier effects. Lastly, given itssignificance, further work is needed to more accuratelydetermine, or reinforce, results for international non-use value since the sample used in this study was bothbiased (international non-gorilla visitors to East Africa)and limited in size.

3. Contingent Valuation MethodologyContingent valuation is especially hazardous indeveloping country context, where income is limitedand the concept of constructed scenarios is both alienan untrustworthy. Results for WTP (willingness-to-pay)are therefore likely to be underestimated. However, theresults reflected here can be considered encouraging,particularly if evaluated on the basis of percentage ofincome rather than outright dollar value. Traditionally,WTA (willingness-to-accept) compensation results tendto be overestimated particularly in poverty situations(Carson, 1998). However, the WTA results for forestopportunity cost were confirmed by the alternativeanalysis (small-holder agricultural value).

4. Opportunity CostForest opportunity cost was assumed to belong tocommunities surrounding the forest rather than nationalgovernments or economies. This is not necessarily thecase especially since, traditionally, protected areas tendto be viewed as state property. However, the case oflocal communities as beneficiaries is assuming increasingimportance as issues of social and economic justicebecome more prominent. The issue of whoseopportunity cost is discussed further below.

5. Agricultural ProductivityIt must be borne in mind that estimation of forestopportunity cost based on current conditions does nottake into account future improvements in agriculture

that would result in higher forest opportunity costs,thereby requiring forests to further justify theireconomic existence. However, this effect is temperedby the fact that forests services to agriculture are assumedto be significant, and therefore take on greater value asagricultural productivity increases. This aspect isdiscussed more below.

DiscussionThe following discussion considers future trends inbenefit-cost distribution and policy, and focuses onimportant economic factors that are likely to dictatefuture scenarios. Some were identified during the courseof the study from the primary data; others areconceptual in nature. Together they constitute a relevantset of economic criteria driving future benefit-costdynamics around the Virunga-Bwindi forests.

A. Long-term WorthLong-term worth was assessed by computing the netpresent value (NPV) of the annual set of values over a29-year horizon. Based on the common economicargument that the value of an asset in the future is worthless than its present-day value (mainly due to uncertainty),this is the process of expressing annual values over thenext 29 years in ‘today’s dollar’ terms, throughapplication of an appropriate ‘discount rate’ i.e. therate at which future benefits are discounted each yearinto the future. Choices of varying discount rate allowfor consideration of differing time preference (in broadterms, “priority time horizon”) amongst stakeholders,for example: a conservation organization; governmentplanning agency; private entrepreneur; or farm small-holder at the forest edge; where it can be expected thateach differs in attitude towards the value of futurebenefits relative to present-day needs. For example, netpresent values derived from this study’s annual set ofvalues can be represented for different stakeholder groupsas depicted in Figure 6: equivalent forest benefits worthUSD5,894 million to the conservationist over a 29-year period (assuming 0% (no) discount) may be worthonly USD1,962 million to a government planningagency (working with 10% discount rate), versus a mereUSD524 million to the rural small-scale farmer(surviving under a 40% discount rate), respectively (asimilar scenario also affects future costs). Choice oftime preference is therefore a major determinant ofresource value over time, as is the set of assumptions



9

underlying the choice. The underlying value of suchanalysis is in understanding divergent attitudes andresource use behaviour amongst stakeholders; and intaking such considerations into account.

Fig. 6: Long-term value of forest benefits and costs from differingperspectives

B. Realization of International Non-use ValueWhilst representing a very significant source of benefit(USD186.5 million per year according to this study),this value is only currently articulated to the tune ofapproximately USD3.0 million per year (Lanjouw, pers.comm.) – this being the investment value ofinternationally funded conservation enhancing programsrelated to the two forests.

One conclusion of this result is that significant levels ofuntapped funds exist, with potentially far-reachingimplications for fundraising.In addition, there are also implications for currentdonor funding levels and conditions. First, such resultssuggest that international donors should significantlyincrease current funding support levels. Second, thereis an accompanying - more controversial - implicationthat donor funding should be driven by a greater senseof obligation, rather than by discretion as is currentlythe case. Such assurances would enable moreappropriate levels of investment into conservation. Thechallenge is in developing the appropriate mechanisms.

C. Whose Opportunity Costs?When viewed at the country level - irrespective ofwhether who is bearing the opportunity costs of forests- benefits from gorillas are, on balance, only off-setting

the losses incurred.When viewed at a locallevel, populations aregaining little-to-nothingfrom the forests and, atworst, are bearing thecost of others’ benefit.Whilst at the very leastconservation should notimpose an economicburden on nationale c o n o m i e s / l o c a llivelihoods (‘do-no-harm’), it can be arguedthat benefits shouldref lect the value ofservices - in this case,

gorilla conservation - being provided to society-at-large,rather than costs of provision only. Given the currentdistribution of benefits and costs, this principal obligesthe major benefactor, the international community, toincrease in-country benefits whether local or otherwise.

D. Implications of Small-holder Farming LivelihoodConstraints on Protected Areas

During the course of the main study, major constraintsto improved farming livelihoods were identified andranked by respondents (Fig. 7):Constraints fall into two categories: those concerningproduction; and those concerning markets. Whilemarket constraints hold important implications forlonger-term development, production concerns directlyimpact shorter-term food availability.Some salient points can be highlighted from this andother secondary data that emerged from the study:• The linked constraints of declining soil fertility and

lack of land are critical to the future prospects offorest conservation. Fallow land – the principalmethod of maintaining soil fertility (see Fig. 8 below)- is becoming unavailable, especially in the Virungaregion (DRC; Rwanda; and Mgahinga, Uganda)where less than 10% of households utilize fallowland. Bwindi differs significantly, with 53% fallowuse.

Long-term (30-year) value of forest benefits & costs

-5000 -3000 -1000 1000 3000 5000 7000

0% (conservationist)

5% (gov't planner)

10% (Western businessman)

20% (E. African businessman)

40% (rural small-scale farmer)

US$ millions

Local community forest cost Local community benefits National level benefits International visitor benefits International community benefits



10

Note: chart show the proportion of farming households consideringeach ranked each constraint as high, medium of low per country.Original rankings (1-9) were divided into three categories: high(corresponds to ranks 1-3); medium (4-6); low (7-9). The fourth categoryis not ranked which is equivalent to no importance rather than forgotten.

• By universal principal, lack of available new landnecessitates that labour is substituted by capitalinputs (e.g., fertilizer and insecticide use;mechanization) in order to increase landproductivity, “intensification”. Mechanization levelswere found to be scarce to non-existent throughout,while use of fertilizer and insecticide around theparks is illustrated in Fig. 8. Results suggest that inRwanda capital investment into intensification isoccurring, while similar substitution is at an earlierstage around Bwindi (where land is more available,although apparently less fertile given respondents

concerns). The situations around Mgahinga and inthe DRC, however, much more serious – otherthings being equal. These results are reinforced bythe corresponding income levels (Fig. 4), all of whichsuggest that lack of capital represents a severeconstraint to improved livelihoods in the case ofthe Virunga-Bwindi region and that, by extension,pressure to convert forest to farmland will continueto grow over time.

• The situation is exacerbated by a lack of economicalternatives – for example, 81% of all working-agechildren in agricultural households are involved withfarming. In addition, further pressure is generatedby positive population growth rates.

• However, results suggest that the returns toinvestment into agriculture are relatively high, whereon average each dollar invested can be expected toreturn USD6.40 in direct income across the tworegions. This implies that, under currentproductivity levels, an annual investment ofUSD68.10 per hectare would negate the desire toconvert forest to farmland by yielding additionalincome equal to that expected over the longer termfrom converting the forest (i.e., USD436 per ha.).This latter statement, however, does not allow forthe fact that the immediate returns from forestconversion are likely to be higher.

When taken together, the above discussions undersections A, B and C appear to argue that internationalnon-use value represents an appropriate level ofuntapped funding for investment, either into locallivelihoods and/or national priorities – depending onwhether forest opportunity costs are deemed to becarried by local communities or the national economy,

or a combination of the two.

Two avenues are possible (a)increased benefits from tourism and(b) increased developmentinvestment into local livelihoods.

E. ScenariosTo order to explore possible futureoutcomes, the impact of a numberof important economic drivingfactors on international, national(non-local) and local benefits and

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

SoilFertility

PriceInputs

Lack ofLand

AccessMarkets

WildlifeRaids

Price ofCrops

Storage Water NoDemand

Other

Constraint

Not ranked Low Medium High

Fig. 7: Ranked constraints amongst small-holder farmers

0

10

20

30

40

50

60

Fertiliser Insecticide Other Fallow

Rwanda

DRC

Mgahinga

Bwindi

Uganda

Total

%

Fig. 8: Percentage of households using: fertilizers, insecticides, other inputs or fallow



11

costs was assessed by developing a sequence of scenarios.These specifically examined:• gorilla tourism monetary benefits accruing to the

international, national, and local levels;• opportunity cost of the forests to local residents;• value of tourism revenue at the local level;• the economic multiplier (spin-off) effect of tourism

revenue at the local level, based on suggestedpotential returns to agriculture.

The main economic ‘drivers’ considered were: differingstakeholder time preference; forest opportunity costdynamics; soil fertility trends; international gorillatourism demand; non-use value to international citizens;and returns to investment into local communities.

The five scenarios are ordered as a progressive sequence,the outcomes of which are largely self-explanatory. Inparticular, they illustrate the effects of different incomescenarios, primarily as they relate to the local level. Theresults are given below, followed by notes on eachscenario.

The outcomes suggested that a significant increase inthe share of tourism benefits at the local level, evencoupled with a sustained growth in tourism, is onlylikely to improve - rather than sustain - rural livelihoodsadjacent to the forests. Additional investment into thedevelopment of the main local livelihood - agriculture -is likely to prove the real engine of development, andby extension, sustainable forest conservation.

Note: All amounts are net present value, or worth, over a 29-yearperiod. It is important to note that international, national, and localtourism benefit are valued from the perspective of a policy maker and,accordingly, are based on a time preference rate of 10%, whilst theremaining values are from the standpoint of subsistence-level ruralcommunities, based on a time preference of 40%.

Scenario notes:Scenario 1:

• Current levels of benefits/costs with no changeover time.

Scenario 2:• Opportunity cost increasing at 4% annually due

to population growth.• Local time preference (discount rate) rising at

10% every 10 years due to increase inopportunity costs.

• Result: increase in opportunity cost tocommunities of USD5.0 million.

Scenario 3:• Tourism increasing at 5% per year until 75%

capacity, level thereafter.• Benefit gains at all levels.• Minimal gain from local perspective (USD1.0

million gain in local multiplier effect).• Opportunity cost and time preference still rising

due to minimal gains.

Scenario 4:• Same increase in tourism• Local share increased from 5to 25% of tourist expenditure atpark level (increase of USD1.6million per year at equal expenseof international and nationalshares).• No increase in opportunitycost and time preferenceimproving due to significantbenefits.• At local level conservationbenefits and multiplier effects(USD6.0 + USD28.0 million)now equal the opportunity costof the forests.

Fig. 9: Long-term benefits & costs: scenario outcomes

0

20

40

60

80

100

120

140

1 2 3 4 5Scenario

US$

mil

lion

s international tourism benefits

national tourism benefits

local tourism benefits

local worth of benefits

local worth of secondary benefits

local cost of forests



12

Scenario 5:• Same as Scenario 4, with additional annual

investment of USD2.0 million into localcommunities (matching local gorilla tourismreceipts under Scenario 4).

• Local level time preference improving due tosignificant benefits, further increasing value ofbenefits received.

• Substantial benefits flowing to local level whenviewed from either policy or local perspective.

References

Brown, G. Jr. and W. Henry (1993) ‘The viewing valueof elephants’ in ‘Economics and Ecology: New frontiersand sustainable development’ Ed. E. Barbier, Chapmanand Hall, London.

Carson, R. (1998) ‘Valuation of tropical rainforests:philosophical and practical issues in the use ofcontingent valuation’, Ecological Economics 24 (1998)pp15-29.

Cunningham, A.B. (1996) ‘People, parks and plant use:recommendations for multiple use zones anddevelopment alternatives around Bwindi ImpenetrableN.P., Uganda’, People and Plants working paper 4,UNESCO, Paris.

Hatfield R. and D. Malleret-King (2003) EconomicValuation of the Virunga and Bwindi Protected Forests,unpublished report, International Gorilla ConservationProgramme (IGCP), Nairobi.

IIED (1997) ‘Part 3: Economic Valuation Techniques’,in ‘Eco Valuation Manual’, IIED, London.

Pers. Comm. 2. Francois Bizimungu, then warden,ORTPN (Rwanda National Parks and Tourism Office),Ruhengeri, Rwanda, August 2002.

Pers. comm. 2. Annette Lanjouw, then ProgrammeDirector now Technical Advisor, IGCP, Nairobi Kenya,Dec. 2003.

Wild, R. and J. Mutebi (1996) ‘Conservation throughcommunity use of plant resources: establishing

collaborative management at Bwindi Impenetrable andMgahinga Gorilla National Parks, Uganda’, People andPlants Working Paper 5, UNESCO, Paris.

This Publication is funded by the NetherlandsMinistry of Foreign Affairs/Directorate Generalfor International Cooperation (DGIS).

www.awf.org

© 2005 African Wildlife Foundation

This document was produced by AWF’s Program teamFor more information, please contact [email protected]

economic value of the bwindi and virunga gorilla mountain ... · gorilla gorilla beringei beringei....

Documents