Economic Development, 12 th Ed. M. P. Todaro and S. C. Smith Slides for Chapter Five. Updated and Expanded Stephen C. Smith Fall 2017

Transcription

1 Economic Development, 12 th Ed. M. P. Todaro and S. C. Smith Slides for Chapter Five Updated and Expanded Stephen C. Smith Fall 2017

2 DistribuEon and Development: Eight CriEcal QuesEons 1. How can we best measure inequality and poverty? 2. What is the extent of relaeve inequality in developing countries; how is this related to the extent of poverty? 3. Who are the poor, and what are their economic characterisecs? 4. What determines the nature of economic growth that is, who benefits from economic growth, and why?

3 DistribuEon and Development: Eight CriEcal QuesEons 5. Are rapid economic growth and more equal income distribueon compaeble or confliceng objeceves?: Is rapid growth achievable only at a cost of greater income inequality or can lessening income dispariees contribute to higher growth rates? 6. Do the poor benefit from growth, and does this depend on the type of growth a developing country experiences? What might be done to help the poor benefit more? 7. What is so bad about extreme inequality? 8. What kinds of policies are required to reduce the magnitude and extent of absolute poverty?

4 A Capabilities-Based Social Welfare Framework Placing where we are at this point in our study of economic development in the context of Chapter 1 Generally, we may think of welfare as a function W: W = W(Y, E, H, M, other capabilities ), Where Y=standard of living, E=Education, H=Health, M=Empowerment and other factors reflecting important capabilities So far, these have been treated as averages; but also distribution matters In chapters 3 and 4, we focused attention on the first element, Y (income, or permanent income from wealth, or broadly living standards), in frameworks such as: Growth models Theories of constraints on growth like coordination failures Growth diagnostics There, we used average incomes Y as the outcome variable, focusing on investment Now, in considering distribution aspects of Y, we expand from averages to address: (Relative) Inequality of Income (Absolute) Income Poverty Thus in this chapter, our welfare framework will become: w = w(y, I, P) Then, having studied income poverty, we will be in a good position analytically to to study multidimensional poverty, returning to a broader capabilities framework, Later, we examine health and education capabilities and deprivation in chapter 8

5 5.1 Measuring Inequality Measuring Inequality Size distribueons (quineles, deciles) Lorenz curves Gini coefficients and aggregate measures of inequality FuncEonal distribueons

6 Desirable ProperEes for Inequality Measures Anonymity: measure should not depend on who has higher income; e.g. whether we believe the rich or poor to be good or bad people Scale independence: inequality measures should not depend on size of the economy want a measure of income dispersion PopulaEon independence principle: an inequality measure should not be based on the number of income recipients Transfer principle - all other incomes constant, if transfer income from a richer to a poorer person (not so much that the poorer person is now richer than the originally rich person), resuleng new income distribueon is more equal Gini coefficient saesfies all four properees; so does the coefficient of variaeon (CV), and some others

7 Table 5.1 Typical Size DistribuEon of Personal Income in a Developing Country by Income Shares QuinEles and Deciles

8 Figure 5.1 The Lorenz Curve

9 Figure 5.2 The Greater the Curvature of the Lorenz Line, the Greater the RelaEve Degree of Inequality

10 Figure 5.3 EsEmaEng the Gini Coefficient

11 Calculating the Gini Coefficient: Numerical Examples Expression: G = A/(A+B)

12 Calculating the Gini: In-Class Example Half the people have 1 The other half have 2 What is the Lorenz curve? How can you use this to compute the Gini?

13 A Brief Review: Desirable ProperEes for Inequality Measures Anonymity: not depend on who has higher income Scale independence: not depend on size of the economy What happens if everyone s income goes up by 25%? PopulaEon independence: not based on number of recipients Transfer principle - transfer income from a richer to a poorer person, resuleng distribueon is more equal What if transfers so much that these people trade places? Gini coefficient saesfies all four properees; but so does the CV and some others This can leave ambiguity in direceon of change in inequality: (As we will see, remains ambiguity when it comes to combining inequality concerns with other concerns)

14 Figure 5.4 Four Possible Lorenz Curves

15 Figure 5.5 FuncEonal Income DistribuEon in a Market Economy: An IllustraEon

16 Desirable ProperEes for Poverty Measures Desirable properees for poverty measures: Anonymity PopulaEon independence Monotonicity DistribuEonal sensievity Plus, the Focus Principle (Amartya Sen): A good poverty measure will be based only on the incomes (wellbeing) of the poor, not the nonpoor As we will see, P 2 has these properees

17 Monotonicity and DistribuEonal SensiEvity Monotonicity principle: If you add income to someone below the poverty line, all other incomes held constant, poverty falls There is a less decisive version, called weak monotonicity, which requires that poverty can be no greater than it was (the main version of the principle above is someemes called strong monotonicity for clarity) DistribuEonal sensievity principle: if you transfer income from a poor person to a richer person (even if the richer person is also below the poverty line), the resuleng economy is deemed strictly poorer. These principles are important; they clarify what is lacking in some of the most widely used measures Note: Looking ahead, we will see that the headcount raeo measure (fraceon below the poverty line) saesfies anonymity and populaeon independence, but not monotonicity (it saesfies only weak monotonicity ; and it fails on distribueonal sensievity The simple headcount (number who are poor) fails even to saesfy the populaeon independence principle.

18 5.2 Measuring Absolute Poverty Headcount Index: H/N Where H is the number of persons who are poor and N is the total number of people in the economy; H/N is the frac5on who are poor Does not meet desirably properees and accordingly is likely to cause inceneve problems:

19 Unintended Policy IncenEves of Using Headcount Measures When agencies told mission includes poverty reduceon Measuring poverty by headcount or fraceon poor creates inceneve to report improvements in these measures; this plausibly focuses efforts toward those closer to poverty line CircumstanEal evidence: Incumbent poliecians frame poverty progress in terms of headcount or fraceon staff would anecipate and want favorable results - by this measure - to report. Government policy exhibits urban bias, e.g. emphasizing job creaeon for the poor in ciees - who are closer to the poverty line. Reports: NGOs work near main roads, or district towns, easier-to-reach; if so, people assisted are less poor on average. MFIs have focused commonly on the richest of the poor MDGs: halving poverty sends a signal. Regions and countries were compared in UN MDG reports e.g based on progress on this measure without meneon of P2 or usually to any other indicator

20 5.2 Measuring Absolute Poverty Total poverty gap: Where Y p is the absolute poverty line; and Y i the income of the ith poor person H TPG = ( ) Yp Yi i=1

21 Figure 5.6 Measuring the Total Poverty Gap

22 5.2 Measuring Absolute Poverty Average poverty gap (APG): APG = TPG N Where N is number of persons in the economy TPG is total poverty gap Note: normalized poverty gap, NPG = APG/Y p

23 5.2 Measuring Absolute Poverty Measuring Absolute Poverty Average income shorqall (AIS): AIS = TPG H Where H is number of poor persons TPG is total poverty gap Note: Normalized income shorqall, NIS = AIS/Y p

24 5.2 Measuring Absolute Poverty The Foster-Greer-Thorbecke (FGT) index: P α = 1 N H i=1 " Y p Y % i $ # Y ' p & α N is the number of persons, H is the number of poor persons, and α 0 is a parameter When α=0, we get the headcount index measure When α=2, we get the P 2 measure

25 Measuring Poverty: Alternative Expression for P 1 - the normalized poverty gap Start with definition of P 1 from P α formula With some algebra we find: P 1 = (H/ N)*(NIS) (To find this, first multiply the basic P 1 formula by (H/H); then, take out constant Y p ; then apply definition of TIS and collect terms) Interpretation: P 1 is per capita poverty gap = headcount ratio (H/N) times normalized income shortfall (AIS) among the poor That is, poverty goes up whenever either the fraction of people in poverty goes up, or the fractional income deficits go up

26 Measuring Poverty: Numerical Example using P 2 Consider an 8-person economy with the following income distribution: (6,.6,.8,.8,2,2,6,6) The poverty line is 1. Question: What is the level of poverty? You want to answer using a measure that has the desirable properties for a poverty measure You can never go wrong by answering using P2 Using the formula (from equation 5.3 in the text, with α=2), we can calculate P 2 as: P 2 = (1/8)[ ] = (1/8)[ ] =.4/8 =.05

27 An Alternative Expression for P 2 Uses inequality among the poor P 2 = [H/N]{NIS 2 + (1 - NIS) 2 (CV P ) 2 } This form helps us see intuitively how each property is present BTW, similar in spirit to the Sen measure S : S = [H/N]{NIS + (1 NIS) (G P )}

28 Just an Aside - Measuring Poverty: Numerical Example with Alternative P 2 Formula Using P2: A Note (beyond scope of this class) Note: You can calculate the same result as above via the alternative formula in equation 5.4 (there may be round off error). The CV is. 1/.7 =.143; squared value.02. The NIS is.3. So, P2 =.5[.09 + (.49)(.02)] ~.05. Remember - you have the entire population, so do not make the sample size adjustment

29 Measuring Poverty: Just Continuing the Same Numerical Example, with P 1 You can always calculate P 1 using the full P α formula, summing individually; Or, using the simplified P 1 formula (H/N)*(NIS), we can calculate P 1 as: P 1 = (H/N)*[TIS/(H*Y p )] = (1/2)*[(1.2)*/(4*1)] = (1/2)*(.3) =.15

30 Measuring Poverty: Using P 2 The four key properties are present because for P 2 : Measure is Anonymous, by construction; The Population Principle holds because one divides the result by population; Monotonicity holds because adding income to any of the poor will decrease P 2 by lowering one of the squared income shortfalls; Distributional sensitivity holds because squaring poverty gaps ensures that income provided to a poorer person will reduce P 2 more than the same income provided to a less poor person.

31 Table 5.5 Regional Poverty Incidence, 2005

32 Table 5.5 Income Poverty Incidence at $1.25 per day (along with Gini index) in Selected Countries

33 Table 5.5 (conenued): Income Poverty Incidence at $2 per day (along with Gini index) in Selected Countries

34 Preview: Measuring Poverty - Income or MulEdimensional Indicators? Given that we are measuring poverty with income, we have good measures that, like P 2, saesfies desirable properees If must have a single indicator, income has advantages e.g. clarity, flexibility But in general is measuring income sufficient? Following Amartya Sen s capability approach, it is apparent that, in general, poverty needs to be conceptualized and so measured in a muledimensional way The new MPI is the best-known effort to do so we return to it soon

35 5.3 Poverty, Inequality, and Social Welfare What s So Bad about Extreme Inequality? DualisEc Development and Shiting Lorenz Curves: Some Stylized Typologies TradiEonal-sector enrichment (see Figure 5.7) Modern-sector enrichment (see Figure 5.8) Modern-sector enlargement (see Figure 5.9)

36 What s So Bad about Extreme Inequality? 3 Factors Extreme income inequality leads to economic inefficiency. Smaller fraction of population qualifies for credit e.g. for business or school Overall rate of saving in the economy tends to be lower: Highest rate of marginal savings usually found among the middle classes. Inequality may lead to an inefficient asset allocation e.g. overemphasis on higher education at the expense of quality universal primary education High inequality of land ownership inefficient because most efficient scales for farming are family and medium-size farms. The result can be a lower average income and a lower rate of economic growth when inequality is high. Moreover, these factors in turn often beget still more inequality Extreme disparities undermine social stability and solidarity High inequality strengthens political power of the rich, and thus concentrates economic and social bargaining power High inequality facilitates rent seeking Finally, extreme inequality is generally viewed as unfair. John Rawls thought experiment: uncertainty behind the veil of ignorance.

37 Table 5.2 Selected Income DistribuEon EsEmates

38 Poverty, Inequality, and Social Welfare Dualistic Development and Shifting Lorenz Curves: Some Stylized Typologies Traditional sector enrichment (see Figure 5.7) Modern sector enrichment (see Figure 5.8) Modern sector enlargement (see Figure 5.9) Note: there is an interaction effect An implication of Modern sector enlargement (but not other processes): Kuznets Inverted- U relationship That is, an inverted-u is consistent with modern sector enlargement growth, but not traditional or modern sector enrichment growth

39 Figure 5.7 Improved Income DistribuEon under the TradiEonal- Sector Enrichment Growth Typology

40 Figure 5.8 Worsened Income DistribuEon under the Modern- Sector Enrichment Growth Typology

41 Figure 5.9 Crossing Lorenz Curves in the Modern-Sector Enlargement Growth Typology

42 5.3 Poverty, Inequality, and Social Welfare Kuznets Inverted-U Hypothesis The inverted-u is consistent with modern sector enlargement growth, but not tradieonal or modern sector enrichment growth

43 Figure 5.10 The Inverted-U Kuznets Curve

44 Figure 5.11 Kuznets Curve with LaEn American Countries IdenEfied

45 Figure 5.12 Plot of Inequality Data for Selected Countries

46 Table 5.3 Income and Inequality in Selected Countries

47 Table 5.4 Income and Inequality in Selected Countries

48 Income and Inequality in Selected Countries, Most recent available (2016 WDI) Country Income Per Capita (US $, 2011) Gini Coefficient Survey Year for Gini Calculation Low Income Ethiopia Mozambique Nepal Cambodia 1, Zambia 1, Lower Middle Income India 1, Cameroon 1, Bolivia 3, Egypt 3,460.0 (32.1) (2005) Indonesia 3, Upper Middle Income Namibia 4, Bulgaria 7, South Africa 5, Argentina 11, Brazil 8, Mexico 9, Upper Income Hungary 12, Spain 27, Germany 43, United States 56, Norway 82,

49 Table 5.5 Regional Poverty Incidence, 2005

50 Table 5.5 Income Poverty Incidence at $1.25 per day (along with Gini index) in Selected Countries

51 Table 5.5 (conenued): Income Poverty Incidence at $2 per day (along with Gini index) in Selected Countries

52 5.4 Absolute Poverty: Extent and Magnitude Progress on Extreme Poverty Clear progress on $1.25-a-day headcount Less clear progress on $2.00-per-day headcount (see Figure 5.14) Incidence of extreme poverty is uneven RelaEonship between Growth and Poverty AssociaEon between growth and poverty reduceon When it is inclusive, growth reduces poverty Lower extreme poverty may also lead to higher growth

53 Figure 5.13 Global and Regional Poverty Trends,

54 Table 5.4 Regional Poverty Incidence, 2010

55 Regional Poverty Incidence, 2013 Region Headcount Ratio Poverty Gap (%) Squared Poverty Gap (%) Regional Aggregation at $1.25 per Day East Asia and the Pacific Europe and Central Asia Latin America and the Caribbean Middle East and North Africa Survey Data Coverage is too low, the result is suppressed South Asia Sub-Saharan Africa Total Regional Aggregation at $2 per Day East Asia and the Pacific Europe and Central Asia Latin America and the Caribbean Middle East and North Africa Survey Data Coverage is too low, the result is suppressed South Asia Sub-Saharan Africa Total

56 Rwanda Income Poverty Incidence at $1.25 per day (along with GINI index) in Selected Countries Country Year Per Capita Monthly Income (2011 PPP) Headcount Ratio (%) Poverty Gap (%) Squared Poverty Gap (%) Bangladesh Benin Brazil Burkina Faso China- Rural China-Urban Cote d Ivoire Guatemala Honduras India- Rural India- Urban Gini Index (%) Indonesia- Rural Indonesia- Urban Madagascar Mexico Mozambique Nicaragua Nigeria Pakistan Peru Philippines

57 Country Year Per Capita Monthly Income (2011 PPP) Headcount Ratio (%) Poverty Gap (%) Squared Poverty Gap (%) Bangladesh Benin Brazil Burkina Faso China- Rural China-Urban Cote d Ivoire Guatemala Honduras India- Rural India- Urban Indonesia- Rural Indonesia- Urban Income Poverty Incidence at $2.00 per day (along with GINI index) in Selected Countries Madagascar Mexico Mozambique Nicaragua Nigeria Pakistan Peru Philippines Gini Index (%)

58 5.4 Measuring Absolute Poverty The Newly Introduced MulEdimensional Poverty Index

59 The MulEdimensional Poverty Index (MPI) IdenEficaEon of poverty status through a dual cutoff: First, cutoff levels within each dimension (analogous to falling below a poverty line for example $1.25 per day for income poverty); Second, cutoff in the number of dimensions in which a person must be deprived (below a line) to be deemed mul5dimensionally poor. MPI focuses on deprivaeons in health, educaeon, and standard of living; and each receives equal (that is one-third of the overall total) weight.

60 MPI Indicators Health - two indicators with equal weight - whether any child has died in the family, and whether any adult or child in the family is malnourished weighted equally (each counts as one-sixth toward the maximum deprivaeon in the MPI) EducaEon - two indicators with equal weight - whether no household member completed 5 years of schooling, and whether any school-aged child is out of school for grades 1 through 8 (each counts one-sixth toward the MPI) Standard of Living, equal weight on 6 deprivaeons (each counts as 1/18 toward the maximum): lack of electricity; insufficiently safe drinking water; inadequate sanitaeon; inadequate flooring; unimproved cooking fuel; lack of more than one of 5 assets telephone, radio, TV, bicycle, and motorbike

61 InteracEon of the deprivaeons? Building the index from household measures up to the aggregate measure (rather than using already-aggregated staesecs), MPI approach takes account of muleplied or interaceve harm (complementarity) done when muleple deprivaeons are experienced by the same individual or family The MPI approach assumes an individual s lack of capability in one area can only to a degree be made up by other capabiliees capabiliees are treated as subsetutes up to a point but then as complements

62 CompuEng the MPI The MPI for the country (or region or group) is then computed A convenient way to express the resuleng value is H*A, i.e., The product of the headcount raeo H (the percent of people living in muledimensional poverty), and the average intensity of deprivaeon A (the percent of weighted indicators for which poor households are deprived on average) The adjusted headcount raeo HA is readily calculated HA saesfies some desirable properees. Important example - Dimensional monotonicity: If a person already idenefied as poor becomes deprived in another indicator she is measured as even poorer - not the case using a simple headcount raeo

63 MulEdimensional poverty tells a different story than income poverty The results showed that knowing income poverty is not enough if our concern is with muledimensional poverty MulEdimensionally, Bangladesh is substaneally less poor - but Pakistan substaneally poorer - than would be predicted by income poverty Ethiopia is far more muledimensionally poor, and Tanzania much less so, than predicted by income poverty Most LaEn American countries e.g. Brazil rank worse on muledimensional poverty than on income poverty; but Colombia s income and MPI poverty ranks are about same

64 OPHI: Why Multidimensional Poverty? It is different from monetary poverty And different policies reduce it. MPI Poor $1.25 a day 100% 90% 80% 70% 60% 50% 40% 30% Percentage of the Population 20% 10% Slovenia Slovakia Belarus Serbia Kazakhstan Armenia Bosnia and Herzegovina United Arab Emirates Macedonia, The former Yugoslav Republic of Georgia Tunisia Russian Federation Albania Occupied Palestinian Territory Montenegro Latvia Thailand Uruguay Moldova, Republic of Ukraine Ecuador Uzbekistan Jordan Brazil Mexico Argentina Czech Republic Viet Nam Croatia Hungary Dominican Republic Belize Kyrgyzstan Maldives Azerbaijan Sri Lanka Colombia Syrian Arab Republic Trinidad and Tobago Suriname Egypt Turkey Estonia Guyana Peru Morocco South Africa Iraq China Tajikistan Paraguay Philippines Indonesia Mongolia Honduras Nicaragua Gabon Swaziland Bolivia, Plurinational State of Guatemala Bhutan Djibouti Vanuatu Ghana Lao People's Democratic Republic Sao Tome and Principe Lesotho Zimbabwe Namibia Congo, Republic of Nigeria Pakistan Nepal Cambodia Cameroon Kenya Haiti Togo Bangladesh Yemen India Cote d'ivoire Gambia Mauritania Chad Zambia Tanzania, United Republic of Afghanistan Malawi Madagascar Timor-Leste Rwanda Mozambique Uganda Benin Sierra Leone Congo, Democratic Republic of the Senegal Guinea-Bissau Central African Republic Burundi Somalia Guinea Liberia Burkina Faso Mali Ethiopia Niger 0%

65 2016 Update from OPHI: Why MulEdimensional Poverty?

66 Incidence and Intensity by Country; OPHI: Here we see what intensity adds Average Intensity of Poverty (A) 75% 70% 65% 60% 55% 50% 45% 40% 35% Brazil China Indonesia Tajikistan Argentina Lao Guatemala Nepal Namibia Ghana Nigeria India Cambodia Zimbabwe Poorest Countries, Highest MPI Guinea-Bissau Afghanistan Mozambique Benin Gambia Bangladesh Somalia Ethiopia Liberia Burundi Congo DR Tanzania Rwanda Malawi Niger The size of the bubbles is a proportional representation of the total number of MPI poor in each country 30% 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% Percentage of People Considered Poor (H)

67 OPHI: MPI also varies greatly across subnaeonal regions within a country e.g. Cameroon 75% 70% 65% Incidence: 6.5% to 86.7% Intensity: 36.4% to 62.3% Extrême-Nord Average Intensity of Poverty (A) 60% 55% 50% 45% 40% Yaoundé Douala Littoral (Excluding Douala) Sud-Ouest Sud Cameroon Adamaoua Est Nord-Ouest Centre (Excluding Ouest Yaoundé) Nord 35% 30% 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% Percentage of People Considered Poor (H)

68 Table 5.6 MulEdimensional Poverty Index, Data for

69 Country Survey Year MIP Percent Poor Thousands Poor Poverty Intensity Bangladesh , Brazil , Burundi , Bolivia, PlurinaEonal State of , Burkina Faso , Cambodia , Colombia , Congo, Republic of 2011/ , Cote d'ivoire 2011/ , Dominican Republic Egypt , Ethiopia , Ghana , Guinea , HaiE , Honduras 2011/ , India 2005/ , Indonesia , Kenya , Lao People's DemocraEc Republic 2011/ , Liberia , Mali 2012/ , Mexico , Madagascar 2008/ , Malawi 2013/ , Mozambique , Nepal , Nigeria , Pakistan 2012/ , Peru , Philippines , Rwanda 2014/ , Senegal , Sierra Leone , South Africa , Tanzania, United Republic of , Timor-Leste 2009/ Uganda , Table 5.6 Update: MulEdimensional Poverty Index, Data for

70 Poverty reduction may cause growth Inclusive growth can cause poverty reduction, directly and indirectly Less appreciated: Poverty reduction can cause growth and development, while failure to address poverty can constrain prospects for development because: Although small, poverty income is part of total income Programs that raise productivity of poor directly even at small scale has direct contribution to growth; but of course tiny increments because incomes are so low; of course has an opportunity costs Poor health, nutrition, and education lowers economic productivity of people in poverty, leading directly and indirectly to slower growth Often, the poor lack access to credit, which constrains growth, e.g.: Lost opportunities for entrepreneurship which may benefit society Leaves them unable to finance their children s education, also limiting the skilled labor force needed for development Incentives for high fertility as a source of old-age financial security Higher income for the poor raises demand for locally produced goods Social exclusion/injustice, which is associated with poverty, also likely causes economic stagnation:

71 Social exclusion/injustice, associated with poverty, likely causes economic stagnation Social exclusion/injustice, which is associated with poverty, also likely causes economic stagnation in the long run Political and social reform needed to overcome constraints to access to land, water, basic resource based livelihood opportunities Elite control of natural resources translates to social and political power to protect elite interests that may be inconsistent with modernization Inadequate voice for the poor who know their public goods needs Other features of a broader social justice agenda which as many have pointed out is also a foundation of economic efficiency The poor may be susceptible or coercible to participating in civil conflict Among its other benefits improved social justice may positively contribute to growth

72 5.5 Economic CharacterisEcs of High- Poverty Groups Rural poverty Women and poverty Ethnic minoriees, indigenous populaeons, and poverty

73 Table 5.7 Poverty: Rural versus Urban

74 Update: Poverty, Rural vs Urban - Percentage below poverty line Region and Country Survey Year Rural Population (%) Sub-Saharan Africa Urban Population (%) National Population (%) Benin Burkina Faso Cameroon Malawi Tanzania Uganda Zambia Poverty: Rural vs Urban Asia Bangladesh India Indonesia Uzbekistan Vietnam Latin America Bolivia Brazil Dominican Republic Guatemala Honduras

75 Table 5.8 Indigenous Poverty in LaEn America

76 Workfare vs Welfare? Basic cost effeceveness consideraeons Workfare, such as a Food for Work Program, are more likely to represent a bezer policy than welfare on a current program efficiency basis when these criteria are met: 1. It is harder to screen the poor without a workfare requirement 2. Poor workers have lower opportunity cost of Eme (so the economy loses lizle output when they work in the program) 3. Non-poor workers have higher opportunity cost of Eme (so they are unlikely to parecipate to get the benefits) 4. The fraceon of the populaeon living in poverty is smaller (so the extra costs of a universal welfare scheme would be high) Note: Each of the above are factors in the efficiency tradeoff: It s important to keep in mind that all these factors must be accounted for together in order to determine whether welfare or workfare is more comparaevely efficient on these criteria.

77 Workfare vs Welfare: Human Capital Factors Another significant factor in a workfare vs welfare choice is that the program does not reduce inceneves for the poor to acquire human capital Note: This factor is largely addressed to incenevize school age children to not see future workfare parecipaeon as an alternaeve to pu{ng high effort into school and conenuing their educaeon; a program design response might be very high work hour requirements for parecipaeon as a deterrent but cannot be so high that those who later need the program do not parecipate. This is also a social efficiency design consideraeon, that can save expenditures when viewed across Eme An addieonal consideraeon could be the inceneve for adults to build skills this might require a different approach to program design A requirement might be parecipaeon in work aceviees that also build relevant skills, or devoeng part of work Eme to training Final note: Another strategy is to add requirements for school azendance if used over Eme; if necessary, consider providing schooling on workfare site

78 Workfare vs Welfare: Other Factors Other significant factors in a workfare vs welfare choice There are greater net benefits of the program s work output There is less social segma of visible workfare parecipaeon, so the poor do not suffer humiliaeon or be deterred from needed work (otherwise, a discreet welfare transfer may be preferable) AlternaEves available for disabled or others prevented from taking part in the program Which can bezer encourage farmers and microentrepreneurs to take favorable business bets (a funceoning safety net), other factors considered?

79 5.6 Policy OpEons on Income Inequality and Poverty: Some Basic ConsideraEons Areas of IntervenEon: Altering the funceonal distribueon MiEgaEng the size distribueon ModeraEng (reducing) the size distribueon at upper levels ModeraEng (increasing) the size distribueon at lower levels

80 5.6 Policy OpEons on Income Inequality and Poverty: Some Basic ConsideraEons Policy opeons Changing relaeve factor prices Progressive redistribueon of asset ownership Progressive taxaeon Transfer payments and public provision of goods and services

81 5.7 Summary and Conclusions: The Need for a Package of Policies Policies to correct factor price distoreons Policies to change the distribueon of assets, power, and access to educaeon and associated employment opportuniees Policies of progressive taxaeon and directed transfer payments Policies designed to build capabiliees and human and social capital of the poor Some specific programs covered in later chapters include: condieonal cash transfers (Chapter 8); agricultural extension (Chapter 9); and micro-finance (in Chapter 15, and Chapter 11 cases)

82 Concepts for Review Absolute poverty Asset ownership Character of economic growth Decile Disposable income Factor share distribueon of income Factors of produceon Foster-Greer-Thorbecke (FGT) index FuncEonal distribueon of income Gini coefficient Headcount index Income inequality Indirect taxes Kuznets curve Land reform

83 Concepts for Review (conenued) Lorenz curve MulEdimensional poverty index (MPI) Personal distribueon of income Progressive income tax Public consumpeon QuinEles RedistribuEon policies Regressive tax Size distribueon of income Subsidy Total poverty gap (TPG) Workfare programs

84 Appendix 5.1: Appropriate Technology and Employment GeneraEon: The Price IncenEve Model Choice of techniques Factor Price distoreons and appropriate technology PossibiliEes of Labor-Capital subsetueon

85 Figure A5.1.1 Choice of Techniques: The Price IncenEve Model

86 Appendix 5.2: The Ahluwalia-Chenery Welfare Index ConstrucEng poverty-weighted index of social welfare

87 Table A5.2.1 Income DistribuEon and Growth in 12 Selected Countries