1 IFPRI

Suresh BabuInternational Food Policy Research Institute

Poverty Measurement and Poverty Measurement and AnalysisAnalysis

2 IFPRI

Dietary IntakeMacro-nutrientsMicro-nutrients

Health Status

Access to food

Household income/ expenditures on food

Maternal & child care practices

Quality & quantity of care

Health environment & services

Access & quality of health, sanitation & water

Policies that encourage-food production-income generation-transfer food in-kind

Policies & programs that increase-caregivers access-caregivers resource control-caregivers knowledge, adoption & practice

Policies that improve-adequate sanitation-safe water supply-health care availability-environmental safety/ shelter

Political & legal InstitutionsPolitical commitment, legal structures for implementing food laws

Potential resources

Poverty/natural resources availability/ agricultural technology

Resource control, ownership, use

Resource use & pricing policies

Adult development and

Labor productivity

Immediate Causes

Underlying Causes

Basic Causes

Nutrition Security

Conceptual Framework of Linking Poverty & Nutrition

3 IFPRI

Nutrition Security

Dietary IntakeMacro-nutrientsMicro-nutrients

Health Status

Access to food

Household income/ expenditures on food

Maternal & child care practices

Quality & quantity of care

Health environment & services

Access & quality of health, sanitation & water

Immediate Causes

Labor productivity

Adult development

4 IFPRI

Policies that

encourage

-food production-income generation-transfer food in-kind

Policies & programs that increase

-caregivers access-caregivers resource control-caregivers knowledge, adoption & practice

policies that improve

-adequate sanitation-safe water supply-health care availability-environmental safety/ shelter

Underlying Causes

5 IFPRI

Political & legal institutions

Political commitment, legal structures for implementing food laws

Potential resources

Poverty/natural resources availability/ agricultural technology

Resource control, ownership, use

Resource use & pricing policies

Sources: Adapted from UNICEF (1998); Haddad (1999); and Smith and Haddad (2000)

Basic Causes

6 IFPRI

Introduction to Poverty AnalysisWhy poverty analysis?Who wants poverty information?Role of policymakers & policy analystsPoverty estimates for policy analysisPoverty estimates for policy evaluationCost-effectiveness of poverty analysis

7 IFPRI

What is Poverty?Poverty is “welfare level below a reasonable minimum.”Poverty has various dimensions

Income povertySecurity povertyEducation povertyHealth – Nutrition PovertyMultiple deprivation

Poor people’s perception of poverty level

8 IFPRI

What do we mean by “Poverty”?The primary focus is on individuals or groups suffering from multiple deprivations

Core Poor

Education poor

Health Poor

Security Poor

Income Poor

9 IFPRI

Poverty Lines and Poverty Measurement

Two Issues in Generating Poverty Estimates

Fixing a poverty line: Identification

Measuring poverty: Aggregation

10 IFPRI

Methods of Fixing Poverty LinesCost-of-basic-needs method (Food-share method)

Cost of basic food needsCost of basic non-food needs

Food-energy methodExpenditure level that meets the food energy requirementBased on calorie-income relationshipsFitting and tracing calorie-expenditure graph

11 IFPRI

Cost-of-Basic-Needs Method

Total Poverty Line = Z

Z=ZF + ZN

ZF = Food Poverty LineZN=Non-food Poverty Line

12 IFPRI

How to calculate the Food Poverty Line1. Calculate average household (HH) size2. Find minimum requirement of daily per-capita calories

for WHO3. Find the typical food bundle of the relative poor HH4. Calculate the calories of this food bundle5. Determine the cost of this food bundle

WHO’s average minimumZF = calorie requirement calories in average food bundle for relatively poor HH

Cost of the average food bundle

*

13 IFPRI

How to Calculate the Non-food Poverty Line1. Find typical Household (HH) on the food poverty line.2. Calculate the non-food expenditures of the HH.

xF = per capita expenditures on food XN = per capita expenditure on non-foodX = total per capita expenditure

ZN = E {XN|xF= ZF} for the poor (Non-food poverty line is the per capita non-food expenditure level when the

per capita food expenditure level is equal to the food poverty line)

ZN = E {XN|x= ZF} for the ultra (extreme) poor (The non-food poverty line is given by the per capita non-food expenditure

when the total expenditure is equal to the food poverty line. The food poverty line in essence becomes the total poverty line for the ultra poor)

Z = ZF + ZN

14 IFPRI

Minimum daily caloric requirements by sector and gender Urban Rural

Age categories Male Female Male Female0 to 1 year 820 820 820 820>1 to 2 years 1,150 1,150 1,150 1,150>2 to 3 years 1,350 1,350 1,350 1,350>3 to 5 years 1,550 1,550 1,550 1,550>5 to 7 years 1,850 1,750 1,850 1,750>7 to 10 years 2,100 1,800 2,100 1,800>10 to 12 years 2,200 1,950 2,200 1,950>12 to 14 years 2,400 2,100 2,400 2,100>14 to 16 years 2,600 2,150 2,600 2,150>16 to 18 years 2,850 2,150 2,850 2,150>18 to 30 years 3,150 2,500 3,500 2,750>30 to 60 years 3,050 2,450 3,400 2,750>60 years 2,600 2,200 2,850 2,450

Source: Caloric requirements are from WHO (1985, Tables 42 to 49).Notes: Requirements used are for men weighing 70 kilograms and for women weighing 60 kilograms. Urban

individuals are assumed to need 1.8 times the basal metabolic rate (BMR), while rural individuals are assumedto need 2.0 times the average BMR. Children under one year of age are assigned the average caloric need ofchildren either 3–6, 6–9, or 9–12 months old.

15 IFPRI

Poverty lines and spatial price indexes by region

Region

Food poverty

line

Reference poverty

line

Ultra poverty

line

Relative priceindex

Metropolitan 50.18 129.19 75.36 1.000

Lower urban 45.94 101.72 67.52 0.787

Lower rural 44.29 85.38 64.71 0.661

Upper urban 45.19 67.51 0.785

Upper rural 40.36 53.37 0.641

101.36

82.81

Notes: Poverty lines are monthly, per capita figures in Egyptian pounds. The Metropolitan poverty line is used as abase line to create the relative price index, which is simply the ratio of each region's reference poverty line tothe base line.

16 IFPRI

Issues in the Poverty LineDoes a poverty line exists?Can it be used & is it well accepted?Are international standards for setting poverty lines accepted in all countries?Can we use the same poverty line throughout a country?Can the nutritional basket underlying the poverty line be derived from surveys?

17 IFPRI

Measures of PovertyIncidence of Poverty: poverty rate

Use the headcount rate to calculate the poverty rate of the % of population below the poverty line

Depth of Poverty – how far a person is below the poverty linePoverty Gap – aggregation of depth of povertyPoverty Severity – aggregation with weights

18 IFPRI

Head-count Index of Poverty Proportion of population whose consumption (y) is less than the poverty line Z

Y1, Y2,..…Z, ..…Yn q

H =q/n H = Head-count index q = number of poor n = size of the populationEg: if n=100; q=50 then H=0.5 or 50%Problems

Insensitive to the depth of povertyH will not change when a poor persons welfare changes if he/she remains below the poverty line

19 IFPRI

Example of the Head-count Index Calculation

Income of 4 individuals in a sample: 1,2,3,4

Poverty Line Z = 3.0

H = q/n =3/4 = 0.75 or 75%

+ food energy method * BSS 1991 and BSS 1995 20 IFPRI

Year Sector BBS Graph* Fitting Method +

Ahmed et al. (1991)+

Ravallion & Sen (1994)

Rahman & Haque (1988)

Hossain & Sen (1992)

Sen & Islam (1993)

Muqtada (1986)

1973/1974

Rural

Urban

82.9

81.4 (5.6)

- - 65.3

62.5

71.3

n.a.

n.a.

63.2

55.9

37.81981/1982

Rural

Urban

73.8

66.0

71.8

65.3

- 79.1

50.7

65.3

n.a

n.a.

48.4

-

1983/1984

Rural

Urban

57.0

66.0

n.a.

n.a.

53.8

40.9

49.8

39.5

50.0

n.a.

n.a.

42.6

-

1985/1986

Rural

Urban

51.0

56.0

51.6

66.8

45.9

30.8

47.1

29.1

41.3

n.a.

n.a.

30.6

-

1988/1989

Rural

Urban

48.0

44.0

- 49.7

35.9

- 43.8

n.a.

n.a.

33.4

-

1991/1992

Rural

Urban

50.0

46.8

- 52.9

33.6

- - - -

Head-count of Absolute Poverty for Bangladesh

21 IFPRI

Poverty Gap Index (PGI)Aggregate short-fall of the poor relative to the poverty line Z

Y1, Y2,…, Yq; Yq ZPoorest Least poor

qPG = 1/n Σ [(Z-Yi)/Z] = mean proportionate

i=1 poverty gap across the whole population (zero gap for the non=poor)

22 IFPRI

Example of Poverty Gap CalculationIncome of4 individuals in a sample: 1,2,3,4

Poverty line = Z = 3; n=4

PG = [(3-1)/3 + (3-2)/3]/4 = [(2/3) + (1/3)]/4 = [(3/3)/4] = ¼ or 0.25

Poverty gap index does not capture differences in severity of poverty.

23 IFPRI

Why?Region A = (1,2,3,4)

Region B = (2,2,2,4)

Poverty line = Z = 3HA = 0.75 HB = 0.75PGA = 0.25 PGB = 0.25 Poverty gap will be unaffected by an income

transfer from a poor person to another poor person who remains below the poverty line

24 IFPRI

Squared Poverty Gap Index (SPG)Mean of the squared proportionate poverty gapReflects severity of povertySensitive to the distribution among the poor

q

SPG = 1/n Σ [(Z-Yi)/Z]2

i=1 Eg: Region A = (1,2,3,4) Region B = (2,2,2,4) with Z=3 SPGA = 0.14 SPGB = 0.08

Poverty in region A > Poverty in region B

25 IFPRI

Poverty AnalysisIncome/Consumption Poverty Profile

Correlates poverty with: Gender Age Residential location Ethnic characteristics Income source Employment sources Share of food/ non food consumption Education outcomes Malnutrition outcomes

26 IFPRI

Qualitative Analysis of Poverty

Role of informal sector?Social analysis of poverty?Institutional analysis of poverty reducing institutionsIntra-household distribution of resources

27 IFPRI

Use of Qualitative MethodsSubjective meaning of povertyIntra-household dimensions of povertyPoor people’s priorities for actionSocial, political, and cultural factors, gender roles, and traditional beliefsParticipants help in designing household surveysAssess the validity of HHS results at local level

28 IFPRI

Income or Consumption?Consumption reflects income as well as past savings, access to credit markets, and seasonal variation in incomeNo records of income or seasonal fluctuationsLarge informal sectorsConsumption data helps in deriving the poverty line

29 IFPRI

Measuring Income/ Consumption Poverty

Household data availability – toolsMeasurement of income povertyQuantitative analysis toolsQualitative analysis toolsIncome poverty dynamics – tools

30 IFPRI

Data Needs for Poverty Analysis

National level dataNational accounts – GDP, consumption, savings, investment, imports, exports, etc.Ministry of Finance, Central Statistical AgencyBudgets, price surveys, and data collectionMonthly, quarterly, and yearly

31 IFPRI

Data Needs for Poverty Analysis cont.

Local level dataConsumer and producer prices, climatic data, availability and use of markets and services CSA, local service providers, regional departmentsPrice and market surveysMonthly, yearly

32 IFPRI

Data Needs for Poverty Analysis cont.

Household – Individual level dataHousehold income, consumption, employment, assets, production, demography, etc.CSA, sectoral ministries, NGOs, academicsHousehold survey, rapid assessments, monitoring and evaluationYearly, 2-3 years, every 5 years

33 IFPRI

Data Sources for Poverty Analysis

Administrative dataPopulation CensusHousehold surveys – LSMS, I&E, Labor, DHS, RRAQualitative and Participatory Assessments – ethnographic, village studies, beneficiary assessments, etc.

34 IFPRI

Types of Household Surveys

Single-topic surveysMulti-topic surveysCensus dataPoverty monitoring surveysTimes series dataPanel data sets