1. Measurement & Analysis of Poverty: Some Concepts & Applications Prof. Dr. Devendra Chhetry November 20, 2007

2. Preliminaries … Development has long been conceptualized in terms of economic growth and measured in terms of GDP/GNP growth Poverty has long been conceptualized in terms of inadequacy of income/consumption in order to have a decent standard of living, and measured in terms of poverty incidence. Concepts and measurements of “development” and “poverty” began by borrowing several notions from economics. Assuming the economic growth will invariably take care of poverty, many developing countries introduced economic growth as the main agenda in their development plans in the 1950s.

3. Preliminaries Within a few years international communities realized that economic growth does not automatically reduce poverty. A series of dialogues then started among eminent scholars for the refinement in the notion of development. In this process, the notion of poverty also got refined. Now, development and poverty are considered as multi- dimensional concepts and their measurements are also practiced over several dimensions. Economic dimension  Monetary poverty Non-economic dimension  Non-monetary poverty

4. Nepal’s Development Efforts in 1990s Development practices in Nepal have taken new course since the restoration of democracy in 1990. Development programs are directed more and more towards eradication of poverty, building people's capabilities by improving health and education facilities, narrowing down the gender inequalities and empowering women People's participations in development works are more valued. Decentralization is recognized as a part of development and initiated this process. NGOs are encouraged in development works. Economic reforms were initiated and private sectors were encouraged to take part in the development process.

5. Some Major Outputs Due to all these and many others, Nepal made significant progress in different development fronts and poverty reduction. Succeed in improving female’s education and narrowing down the gender disparity in education sector. Succeed in improving demographic parameters: for the first time life expectancy of females exceeded that of males; reversed the increasing trend in the percentage share of children (0-to-14) in total population The above & many others improvements in socio-demographic factors will have several entwined implications on people's life. Succeeded in reducing monetary poverty: incidence of poverty has declined from 42 percent in 1995/96 to around 31 percent in 2003/04.

6. Economic Growth Since 1990 Economic growth remained sluggish since 1995. One major reasons behind it may be the failure to capture by our National Accounts System the rapidly increasing economic activities. Due to conflict, migration from rural to market centers has increased tremendously, and the migrants have initiated new service oriented economic activities or enlarged the boundary of already existed economic activities for their survival in the market centers. Rapid progress in the area of transport and communication as well as in the expansion of small service oriented business enterprises has taken place in market centers within a short span of time. The migrants have also placed tremendous pressure on housing construction, real estate business expansion, and on available dwelling unit habitation.

7. Monetary Poverty: Absolute & Relative Monetary poverty when defined in absolute terms in relation to the basic needs - like food, shelter, cloth and basic services – is known absolute poverty. Monetary poverty when defined in relative terms, relative to the overall mean or median of per capita income/consumption, is known as relative poverty. To identify the numbers of people living in poverty requires a threshold value that demarcates the whole population into poor and non-poor; the poverty line is the clear dividing line. The two definitions of poverty require two types of poverty lines: 1.Absolute poverty line 2.Relative poverty line

8. Estimation of Absolute Poverty Line A number of methods for estimating absolute poverty line exist. Nepal has been using the cost-of-basic needs (CBN) method. Four steps are 1.Estimation of nutritional needs: Involves in estimating minimum per capita daily calorie requirements for survival. 2.Identification of food basket: Involves in identifying a food basket that meets the daily calorie requirements 3.Estimation of the cost of food basket: Involves in cost estimation of food basket (food poverty line). 4.Estimation of non-food poverty line: Involves in estimating minimum cost of non-food items and basic services Absolute poverty line = food poverty line + non-food poverty line

9. Measures of Absolute Poverty Currently, three measures of absolute poverty are in use: Headcount Index Poverty Gap Index Squared Poverty Gap Index. Assume that there are N individuals whose household per capita expenditure (y) are arranged in ascending order as follows. y 1  y 2  y 3 ...,  y q  y q+1  y q+2 ...  y N Suppose z is the poverty line and y 1  y 2  y 3 ...,  y q < z, showing that there are exactly q individuals below the poverty line. To understand these measures consider the following set up.

10. Headcount Index Headcount index (H) is defined as the ratio of the number of individuals below the poverty line to the total number of individuals. That is H = q/N. This measure is also known as Headcount ratio Poverty rate Incidence of poverty H is simply the proportion of individuals below the poverty line and its value ranges in the interval 0 to 1.

11. 1.It does not tell how poor the poor are: whether they are close to the poverty line, are far below it, or are distributed in some other manners. 2.It violates the monotonicity axiom. This axiom says that, other things remaining the same, a reduction in income of someone below the poverty line must increase the poverty measure. 3.It also violates the weak transfer axiom. This axiom says that, other things remaining the same, a transfer of income from a richer poor person to a poorer poor person must lead to a reduction of poverty measure. Weaknesses of Headcount Index

12. Average Poverty Gap The average poverty gap is defined by Poverty lineMean consumption of poor It measures the mean cost per poor to eliminate poverty Its value ranges over the interval (0, z)

13. Depth of Poverty Sen suggested to divide the average poverty gap by z to obtain a slightly different measure, called the depth of poverty (G) Depth of poverty is the average poverty gap expressed as a fraction of the poverty line. Average poverty gap or G satisfies the monotonicity axiom; however the measure does not satisfy the transform axiom.

14. Poverty Gap Foster-Greer-Thorbecke (1984) have introduced a class of poverty measures P(  ), where  is index  0 and P(  ) is defined by In particular, P(0) = headcount ratio. P(1), called the poverty gap. P(1) can also be written as Headcount ratio Depth of poverty

15. Squared Poverty Gap where C is the coefficient of variation of consumption distribution among the poor, a measure of inequality. P(2) is called the Squared Poverty Gap It takes into account of inequalities among the poor. It is thus particularly sensitive to the severity of poverty. It satisfies the axiom of weak transfer.

16. Some Remarks on Poverty Measures The inequality P(0) > P(1) > P(2) is always true. P(0) is easy to understand and interpret than others P(1) & P(2) seem to have less significance in advocacy Sometimes P(1) & P(2) are unavoidable to policy makers as can be seen below, where the value of P(0) is almost same in both regions but intensity of poverty and severity of poverty is much higher in hills. RegionP(0)P(1)P(2) Hills0.410.1360.061 Tarai0.420.0990.034 Use of P(0) alone may have dangerous implication. Measures fail to explain the causes of poverty.

17. Definition of Poor Two definitions of poor arise. 1.One is the standard of living concept, where one is perceived as poor if his/her total expenditure falls below the given poverty line. 2.Other is the right to a minimum level of resources concept, where one is perceived as poor if his/her income falls below the poverty line. Now-a-days standard of living concept is preferred to right to a minimum level of resources concept.

18. Poverty Measuring Practices in Nepal Nepal has made four attempts to measure poverty 1.NPC based on the 1976/97survey- Employment, Income Distribution and Consumption Patterns Survey 2.NRB based on the 1984/85 survey – Multipurpose Household Budget Survey 3.CBS based on the 1995/96 survey – Nepal Living Standard Survey I 4.CBS based on the 2003/04 survey – Nepal Living Standard Survey II 1976/771984/851994/952003/04 36.241.441.830.9 Not Comparable Comparable

19. Poverty Related Indicators 1995/962003/04 Average household size5.685.27 PCE in current prices (Rs) for all Nepal723515224 PCE in real (1995/96) prices (Rs) for all Nepal723510318 Per capita remittance amount for all Nepal (nominal Rs)6252100 Share of farm income in household income in %61.047.8 Average annual growth rate in real PCE over the period is 4.5 percent, which is nearly three times higher than the average annual growth rate in per capita GDP or in per capita private consumption.

20. Poverty Analysis A poverty profile is a set of poverty outcomes usually presented in tabular forms. In monetary poverty, two different sets of analyses can be undertaken through poverty profiles. 1.Temporal/cross-sectional comparison of poverty measures across population or household groups defined by place of residence or socio-demographic characteristics. 2.Temporal/cross-sectional comparison of socio-economic, demographic and living conditions of poor and non-poor. Such analyses are useful for policy makers. 1.Identifying locations where the poverty reduction works are most urgently needed as well as in identifying the appropriate poverty reduction programs 2.Understanding disadvantages of the poor as compared to non-poor as well as disparity.

21. Temporal changes in poverty indices 1995/962003/04% change Headcount Index (%) Nepal41.830.9-26 Urban21.69.6-56 Rural43.334.6-20 Poverty Gap Index (%) Nepal11.87.6-36 Urban6.52.2-67 Rural12.18.5-30 Squared Poverty Gap Index (%) Nepal4.72.7-42 Urban2.70.7-73 Rural4.83.1-37 Poverty has declined very sharply Poverty continued to remain as a rural phenomenon Rural-urban disparity ratio (RUDR) in the measures of poverty widened RUDR in P(0) has increased from 2 in 1995/96 to 3.6 in 2003/04. Likewise, from 1.9 to 3.9 in P(1) and from 1.8 to 4.4. More efforts would require in the reduction of rural poverty in order to meet MDG

22. Comparison of poverty across household groups Poverty increases as the number of children in households increases Household with # of children0123+ Poverty rates (%)13.729.341.654.0 Poverty increases as the household size increases Household size1234567+ Poverty rates (%)7.211.011.719.324.933.541.4 Poverty decreases rapidly as the education level of household head increases Years of schoolingIlliterate  5 6 -78 -1011+ Poverty rates (%)42.028.223.38.41.6 Poverty decreases moderately as the farm land ownership increases Land ownership in hectare< 0.20.2 to 1.01.0 to 2.0> 2 Poverty rates (%)39.338.127.323.8

23. Poor/non-poor comparison in Age Composition NLSS I (1995/96) (%)NLSS II (2003/04) (%) PoorNon-poorOverallPoorNon-poorOverall 0 – 1447.840.544.1 48.235.739.6 15 – 6448.554.851.5 47.859.255.7 65+3.74.74.4 4.05.14.7 Increasing trend in %tage of children over the past consecutive census years 1961, 1971, 1981 and 1991 has been reversed in 2001. Child dependency ratio of the poor has increased from 98% in 1995/96 to 101% in 2003/04, while that of non-poor has decreased from 74 to 60 percent Overall %tage of children decreased from 44 in 1995/96 to 40 in 2003/04. On the contrary, this %tage for the poor does not decrease.

24. Characteristics of the Poor IndicatorsPoorNon-poor Average household size6.34.9 Average number of children 0-14 per household3.01.8 Average nominal annual household income in ‘000 Rs41.293.7 Child (6-to-14) literacy rate in %32.663.4 Adult (15+) literacy rate in %28.954.8 Child Women Ratio0.790.41 Child dependency ratio10160 Poor households are characterized by large household size with excessively large number of children, and low income level. Children of poor households are characterized by low literacy rate. Adult couples of poor households are characterized by having high fertility and high burden of children.

25. Disadvantages of Currently Married Reproductive Women IndicatorsPoorNon-poorOverall % of CMRW using contraceptives2743 38.3 % of births took place in hospital421 13.9 % of births attended by health personnel729 19.8 % of women consulting for pre-natal care4069 57.1 % of women consulting for post-natal care61812.9 Low use of family planning methods and low use of safe delivery facilities during pre-delivery, delivery and post delivery period.

26. Summary & Discussion Measurement of monetary poverty began more than a century ago with a number of conceptual, practical and technical problems. Some of them been partially resolved; more works still yet to be done. However, the measures are useful for monitoring and evaluation of poverty over time & space. Monetary poverty analyses clearly demonstrate that poverty is not a state of static reality, but a dynamic reality that varies across different population groups defined either by their place of residence or by their socio-economic and demographic conditions. The reasons for drastic reduction in monetary poverty should not be understood only within the economic domain, but they also need to be understood in a broader perspective.

27. Thank You