Aims: ACADEMIC YEAR 2009/2010 ECO1.2 MEASURING POVERTY, INEQUALITY AND HUMAN DEVELOPMENT Enrica Chiappero Martinetti 12 hours - To briefly discuss main conceptual and methodological issues in poverty and inequality analysis - To outline the standard uni-dimensional poverty and inequality measurement -Beyond economic poverty: Amartya Sen s capability approach and multidimensional wellbeing analysis Contents: - Some conceptual issues: welfarist (utility, income) and non-welfarist approaches (basic needs, human development and capability approach) - Measurement issues in the welfarist approach (indicators, unit of analysis, etc.) - Inequality measurement: Lorenz curves and Gini index - Poverty measurement: identification (poverty lines) and aggregation (poverty indexes) - The economic growth, poverty and inequality triangle - Measurement issues in non-welfarist approach: multidimensional poverty analysis and empirical applications of the capability approach 1
Useful readings: Coudouel, Hentschel e Wodon, Poverty Measurement and Analysis (Ch.1, Poverty Manual, World Bank) (available online) Robeyns I., "The Capability Approach: A Theoretical Survey", Journal of Human Development, 2005, 6(1), pp. 93-114. (available online) F. Bourguignon, The poverty-growth-inequality Triangle, WB, 2004. (available online) Ferreira F., Ravaillon M. (2008), Global Poverty and Inequality: a Review of the Evidence, World Bank, Policy Research Working Paper n. 4623, May 2008 (available online) Alkire S., Deneulin S. (2009), The Human Development and Capability Approach, in Deneulin S. and Shahani L., An introduction to the Human Development and Capability Approach. Freedom and Agency, Earthscan, UK, ch. 2 (available online) Further readings: Sen A.K. Development as freedom, Oxford University Press, 1999 Sen A.K., The idea of Justice, Allen Lane, Penguin Book, London, 2009 Sen A.K., Capability: reach and limits, in Chiappero-Martinetti E. (ed.) Debating Global Society. Reach and Limits of the Capability Approach, Feltrinelli Foundation, 2009 Fukuda-Parr and Shiva Kumar, Readings in Human Development. Concepts, measures and policies for a development paradigm, Oxford University Press, 2003 Deneulin S. and Shahani L., An introduction to the Human Development and Capability Approach. Freedom and Agency, Earthscan, UK 2
Some conceptual issues Why equality? Equality of what? (A.Sen, 1982, in Choice, Welfare and Measurement, Blackwell, Oxford and MIT Press A. Sen, 1992, Inequality Re-examined, ch. 1, Oxford University Press ) Two main issues: 1 st issue: every normative theory of social arrangement asks for equality of something equality of what? a) Equal liberty and equal distribution of primary goods (John Rawls, Theory of Justice, 1973) b) Equal treatment of individuals; equality of welfare or equality of resources? Dworkin argues that the latter is the right way of defining distributive justice (Ronald Dworkin, What is equality? Part I: equality of welfare and What is equality: Part II: equality of resources, 1981) c) Economic equality (Thomas Nagel, 1979, 1986) d) Equal weights on everyone's utility gains in the utilitarian philosophy (max of the sum of the utilities of all people taken together); i.e. Bentham s principle: the main social goal is to achieve the greatest happiness for the greatest number of people e) But also antiegalitarian positions must define a space for equality: equality of libertarian right. A distribution is just if and only if everyone is entitled to his holdings(robert Nozick, Anarchy, State and Utopia. 1974); no one has any more right to liberty than anyone else; equal legal and political treatment (James Buchanan, The limits of liberty, 1975). 2 nd issue: Human diversity and equality. "All human beings are equal". Is it really true? If all people were identical, equality in one space (e.g income) would be congruent with equalities in others (e.g, health, well-being, happiness, etc.). However, human being differ from each other in many different ways: personal characteristics (age, sex, physical and mental abilities, etc.), endowments of wealth and liabilities, natural and social environments. Moreover, societies and communities to which we belong offer very different opportunities, etc. Thus, equality/inequality in one space may take us in a very different direction from equality/inequality in the space of another variable. Finally, various types of inequality are interrelated and reinforce each other. 3
Two main approaches for well-being, poverty and inequality analysis 1) welfarist approach - basic assumptions (microeconomics theory): representative agent (homo oeconomicus) and unidimensional space (welfare, utility or happiness) for determining economic wellbeing individuals are rational and best judges of the activities which maximize their welfare, utility or happiness; given their initial endowments (time, land, physical, financial and human K) individuals make production and consumption choices according to their sets of preferences over bundles of consumption and production activities (given technology and prices in competitive markets, perfect information and no externalities). Pure welfarist approach: subjective well-being (utility or happiness) and comparison across individuals. Usually, imperfect but observable proxies for utility such as income or consumption 2) non welfarist approaches - basic assumptions: focused on multidimensional spaces, they pay more attention to human diversity and heterogeneity of contexts 2.1. basic needs approach (Paul Streeten, First things first: meeting basic needs in developing countries, 1981; Frances Stewart, Basic needs in developing countries, 1985): need to attain some basic multidimensional outcomes. Physical inputs, minimum specified quantities of food, shelter, water, sanitation necessary to prevent ill health, undernourishment etc. 2.2. capability approach (Sen, Nussbaum and others): the capability to function represents the various combinations of functionings (beings and doings) that the person can achieve. Main features: - well-being instead of welfare: what a person is actually able to do or to be; living is a set of interrelated functionings and capabilities (constitutive elements of w-b) - relationships between good/resources and utility/well-being 4
- interpersonal differences and plurality of (socio, economic, natural, political, cultural etc.) contexts capability = a vector of functionings that a person can achieve (i.e. it reflects the person's freedom to choose what kind of life to live; real opportunities for a person to have wellbeing, including the freedom to have alternatives other than the chosen combination) functionings= constitutive elements of a person's w-b; set of beings and doings (i.e. to be nourished, to avoid escapable morbidity and premature mortality, having self-respect, take part in the community life, etc.) standard of living = a set of achieved functionings difference between positive (freedom from) and negative freedom (freedom to): doing x is different from choosing to do x and doing it Functionings e capabilities. A functioning is an achievement whereas a capability is the ability to achieve. Functionings are, in a sense, more directly related to living conditions since they are different aspects of living conditions. Capabilities, in contrast, are notions of freedom in the positive sense: what real opportunities you have regarding the life you may lead (Sen, The standard of living, 1987) A person s capability to achieve functionings that he or she has reason to value provides a general approach to the evaluation of social arrangements and this yields a particular way of viewing the assessment of equality and inequality (Sen, 1992: 5) Functionings: the various things a person may value doing or being (Sen, 1999:75): being educated, having a good job, being able to participate effectively in social and political life, having the social bases of self-respect and non-humiliation; being able to be treated with dignity and without discrimination (Nussbaum) Capability: freedom to enjoy various functionings. [ ] Various combination of beings and doings a person can achieve : set of vectors of functionings reflecting the person s freedom to choose from possible livings (Sen, 1992:40): the real freedom to achieve a given level of education or to aspire to a certain job, to take part to the political life, to have an own family Agency: ability to pursue goals that one values and has reason to value. An agent is someone who acts and brings about change (Sen, 1999:19) ; close to the concepts of self-determination, empowerment, autonomy etc NB: Two aspects of freedom: a) capability refers to the opportunity aspect of freedom b) agency refers to the process of freedom 5
Conversion factors: Personal (or internal): age, gender, race, ability/inability External: socio-economical, natural, cultural, institutional environment Conversion rates: ability to convert through a conversion function f i (.) inputs (commodities, private and public resources) into outputs (achievement). Interaction and combination of conversion factors NB: special attention to potential marginalisation factors: gender, disability, migration, socio-economic background Individual Capability Endowment set Goods and resources Personal and social conversion factors Functioning Choice Achieved Functionings To sum up: the CA is a way of thinking, not a theory of distributive justice nor an alternative economics paradigm to the neoclassical theory comparing with other multidimensional approaches of well-being, the CA is not simply a way to enlarge the evaluative well-being to variables other than income, but it is a radically different way to conceive the meaning of well-being. 6
it considers a plurality of well-being dimensions, the relationships among them are investigated and through which poverty, deprivation and inequality assume a new and clear meaning There are, however, some relevant "open questions". Among others: 1. Capabilities or functionings? Does a "given" list of functionings exist? How can we chose what include and what exclude in this list? All the functionings are equally important or some priority among them does exist? - Related issues: a) relativism vs. universalism (culture, history, values): focus on what is "common" (essential) for all human beings or focus on the differences (defining "norms" for each context)? b) freedom of choice: is it really a value for all? Someone outlines that universalism do not recognize the right of each individual to chose his own individual project of life deciding in an autonomous way what is important. 2. How the Sen's approach could be empirically implemented? Methodological issues related to complexity and inadequacy of the traditional tools of analysis (see next section). Given the rich array of functionings that Sen takes to be relevant, given the extent of disagreement among reasonable people about the nature of good life, and given the unresolved problem of how to values sets, it is natural to ask how far Sen s framework is operational. Is it a realistic alternative to the methods on which economists typically rely? (Sudgen, 1993) 3. Policy implications? - Related issues: a) Common critiques: more redistributive policies, paternalism, "intrusion". Is it government able to decide what is a good or valuable life? b) The role of individual responsibilities: equality in terms of f/c doesn t imply equality in the realization of life plans but only equalization of what s needed (inputs) to produce wellbeing (output): i.e equality of opportunity not equality of outcomes c) The role of the Institutions and democracy 7
From the capability approach to the human development approach The meaning of human development (UNDP Report n. 1/1990): The basic objective of human development is to enlarge the range of people's choices. These choices are not fixed for ever. They change over time as circumstances and aspirations change. But at all levels of development, the three essential capabilities for human development are for people to lead long and healthy lives, to be knowledgeable and to have a decent standard of living. If these basic capabilities are not achieved many choices are simply not available and many If these basic capabilities are not achieved many choices are simply not available and many opportunities remain inaccessible.. CA as conceptual foundations A holistic concept (Mahbub ul Haq, 1995) a) Development must put people at the centre of its concerns b) Purpose of development is to enlarge all human choices, not just income c) Four essential pillars: a. Equality b. Sustainability c. Productivity d. empowerment Relationship between growth and development: "Human development is the end, economic growth a means Economic growth human development Human development economic growth 8
Measurement issues in the welfarist approach Some basic and common problem in inequality and poverty measurement a) economic well-being: what kind of indicator is more adequate to describe it? income: command over resources; only flows or also wealth stock? Monetary or in-kind income? (+) (comparing with C) limited number of sources of income makes it easier to collect information; more neutral (it measures the degree of command over resources and is not affected from preferences) (-) underreporting, affected by short-term fluctuations (i.e seasonality), how to collect income from informal labor activity and home agricultural production; consumption: command over commodities in general or over specific types of consumption (i.e. food or housing) (+) C is smoother than I (life-cycle profile); it shows current standard of living; it could be easier to recall expenditure and consumption, including self-consumption (-) it reflects not only opportunities but also preferences so that consumption choices might be misleading (i.e. a rich household with low consumption), some expenses are not made regularly (durable goods), how take into account differences in prices, rationing, home production Generally speaking: consumption based measures are preferred especially in low or middle income countries. The use of both would be the optimal solution. b) Unit time : week? Month? Year? The whole life cycle? Are agents perfect foresights in their C and saving decisions? Perfect credit markets? Panel data can be useful for studying the dynamic of inequality, poverty and social mobility but are scarcely available. Usually, year for income and month for consumption c) Individuals or households? - economic theories and measures of inequality/poverty: individual (representative agent) - however, available statistical data are often collected to household level (always, when refer to consumption); as a matter of fact, most economics decisions refer to household (C, S, labor supply, etc.) and needs are usually satisfied within the family. NB Less inequality at the household level (compensation mechanism) 9
- distinction between household, family etc. Two main issues: 1) From a theoretical point of view: can we assume that well-being is equally distributed within the family? Inequality in terms of opportunities, chances and standard of livings between men and women, adults and children. We need a better understanding of the decision process (dictatorial, democratic, cooperative?), the "sharing rule", the distributive mechanism and so on; 2) From a technical point of view: comparison among household with different size/structure/characteristics equivalence scales. - Positive correlation between size and needs but economies of scale, household public goods (i.e. heating, housing) and different needs for different household members - Total income: each household is consider as a single unit - Per capita income Y/N: no economies of scale and each individual has the same weight - Equivalence scales : estimated by expenditure data, mainly in food consumption; i.e. Italy: n. of components 1 2 3 4 5 6 7+ coefficient 1 1.67 2.22 2.72 3.17 3.60 4 arbitrary defined, useful for international comparisons; i.e. - standard OECD e.s.: first adult = 1; other adults = 0.7 (reflecting economies of scale); children < 15 = 0.5 (reflecting their presumably lower needs) - modified OECD: coefficient s are equal to 1, 0.5 and 0.3 Nb some authors (Szekeley, Hilgert, 1999 et al.) using different equivalence scales show that poverty rate varies between 13% and 66% in 19 Latin American countries: others reach similar results for transition economies. - Subgroups of population? Need to pay attention not only to vertical inequality (between individuals or households) but also horizontal inequality between culturally defined groups (see F. Stewart on Horizonal inequality as neglected dimension of development): geographical or socio-economic criteria Emphasis on differences between groups more than within groups 10
d) Some other technical issues in comparisons analysis: - time series data: nominal and real indicators - cross data: purchasing parity power (PPP) - sample structure and dimension: significance of estimations, statistical inference, etc. Inequality measurement Lorenz Curve: order n individuals from poorer to richer (x 1 x 2 x 3... x n ) and compute the share of the cake that is the proportion of the total income that each equal share of population received (for ex. the bottom 10% of population receives 5% of total income, etc.) indicate in horizontal axis the cumulative proportion of population and in vertical axis the corresponding cumulative proportion of income; if income is equal distributed among the population, the Lorenz curve corresponds to the diagonal line - the line of perfect equality - (the first 5% of population get 5% of the cake, the first 10% obtain 10% etc.); otherwise the Lorenz curve is a convex curve and the degree of convexity is higher when inequality is higher. Gini index (or concentration index) xi xj G = 2 n 2 µ i j G=0 perfect equality; G=1 max inequality Advantage: we can consider the Gini index also in term of Lorenz curves A/(A+B) Warning: an identical value of G could be associated to two different Lorenz curves. 11
% of income 1 A B 0 1 % of population Ex: Lorenz curves for world income distribution for 1988 and 1993 10 0 8 0 % of income 6 0 4 0 1988 1993 2 0 0 0 2 0 4 0 6 0 % of population 8 0 10 0 12
A potential problem: Lorenz curves crossing 100 90 80 70 Income share 60 50 40 Coutnry B Country A 30 20 10 0 0 10 20 30 40 50 60 70 80 90 100 Population share Other common inequality indexes 1. quantile ratio (decile or quintile ratio): it compares income share of different share of population, i.e. the income share of the bottom 10% (i.e. the poorest) with the income share of the top 10% 2. variance: 1 σ = n n i= 1 3. coefficient of variation: ( xi µ ) 2 C = σ µ Measuring international and global inequality (Milanovic B., 2005, Worlds Apart: ) Three concepts of world inequality 13
Concept 1: unweighted international inequality country as unit of analysis, income or GDP percapita; disregards the population and compares representative individuals from all the countries Concept 2: population-weighted international inequality compare mean incomes among countries but now weighted by the population of each country [nb: concept 1 and 2 are based on the assumption that within-country distribution is perfectly equal] Concept 3: true world income distribution inequality is calculated across all the individuals in the world lining up all individuals from the poorest to the richest regardless of the country [Pen s parade] 14
15
Poverty analysis: who are the poor (identification phase) 1) Absolute poverty lines (you are poor if you have not enough) directly defined in terms of basic needs (food, clothing, shelter) or indirectly in terms of income (minimum amount of money for basic necessities); arbitrarily chosen (1 or 2 $ a day in Word Bank approach) (-) only basic physiological needs; what is fundamental for the human life is arbitrarily chosen; is an adequate framework just for developing countries? 2) Relative poverty lines: (you are poor if you have less than others) it refers to the actual standard of living of the whole population (income or consumption); a given % of the mean/median income or expenditure (e.g. ISPL=50% of the mean income; EU-wide pl = 60% of European equivalised median income); the lowest decile/quintile of population (-) do not allow for comparisons across countries or over the time (to the same poverty rate can correspond a substantial different welfare level in absolute terms) (-) the choice of the cutoff point is arbitrary; (-) poverty measures are insensitive to overall well-being: because P depends on the distribution of income or consumption not on their level. If poor are getting poorer, all things equal for the others, mean income decreases, z decreases, the poverty measure decreases despite the drop in absolute living standard of the poor (idem if rich people are getting richer) Poverty analysis: how many poor there are (aggregation phase) Incidence of poverty: Head-count index Simply the share of population with income/consumption below the PL [1] H = q/n q = poor people (i.e. population living below the PL) N = whole population (+) simple to construct; easy to understand 16
(-) it does not take into account the intensity of poverty (it assumes that all poor are equally poor); the index does not change if individuals below PL become poorer or richer as long as they remain below PL; policies based on H might be sub-optimal; in fact, the easiest way to reduce H is to target benefits to people just below PL. But what about the poorest? See the example below: country A 100 100 150 150 p.l = 125 H= 50% country B 124 124 150 150 p.l = 125 H= 50% Depth of poverty: Poverty gap (PG) and poverty gap index (PGI) 1 q PG = ( z x i ) N i = 1 How far off households are from the poverty line. PG is the average, over all people, of the gaps between the income of the poor (x i ) and the PL (z). It indicates the average extent to which individuals fall below the poverty line. 1 q ( z PGI = xi ) z N i = 1 PGI expresses the poverty gap as a percentage of the PL (+) they can be interpreted as the average shortfall of poor people; they show how much would have to be transferred to the poor to bring their income up to the PL. Put in other words: they are the minimum cost for eliminating poverty (without targeting costs or distortion effects); PG does not imply that there is a discontinuity, a jump at the poverty line (-) they do not capture differences in the severity of poverty amongst the poor and ignore inequality among the poor ; they are insensitive to transfers among the poor. See ex. below country A 100 100 150 150 p.l = 125 PG=12.5 PGI=10% country B 80 120 150 150 p.l = 125 PG=12.5 PGI=10% 17
Severity of poverty: Squared poverty gap index (PGI 2 ) PGI 2 = 1 q z x i N i = 1 z 2 Is the average of the square relative PG of the poor: it is a weighted sum of the poverty gaps as a proportion of the PL. The weights are the proportionate poverty gaps themselves (+) it takes into account inequality among the poor: a transfer from a poor to an even poorer would reduce the index; a transfer from a very poor to a less poor would increase the index (-) it is relatively difficult to read and interpret Poverty analysis: some policy implications PG, PGI and SPG are important complement of the incidence of poverty: some groups of population can have a high incidence but low poverty gap and viceversa; important for policy evaluation 18
Group A: just below the PL Group B: average poor Group C: very poor According to the above figure: - PGI is more sensitive to the well-being of the poorest than H and SPGI more than PGI - H treats all the poor similarly : people in group A, B and C are given an equal value (=1) - PG gives a value relative to the distance to the PL (those in group B have on average twice the value of those in A) - SPG gives a value of the squared distance (those in group B have on average a value 4 times that of those in group A) - Using H policies would focus on group A (with 1 $ one household in group A goes out of poverty) - Using PGI focus would be on A, B and C (with 1 $ the poverty gap will be reduced by 1 for any household in group A, B or C) - Using SPG focus would be on group C (with 1 $ one household goes from group C to group B and reduces SPG by 3 2-2 2 =5: from B to A SPG will be reduced only by 1) 19
Economic growth, poverty and inequality a) The inverted U hyphotesis of Kuznets : relationship between GNP per-capita and income inequality - inequality index used: quantile ratio (i.e the richest 20% /the lowest 60% ) - countries: India, Sri Lanka, Puerto Rico; UK and USA - results: India 1.96 Sri Lanka 1.67 Puerto Rico 2.33; UK 1.25 USA 1.29 - conclusions: inequality grows in the initial phase of development but decreases in the following phases when GNP increases nb: A large number of studies have attempted to test this hypothesis; controversial results. There is some evidence that an inverted-u relationship sometime is present but there are reasons to be skeptical of these findings. b) Effect of growth and inequality changes on poverty - Fig. 5a: impact of a uniform growth (all individuals get an increase in income by 30) without any change in terms of inequality. Distribution is shifted to the right: if poverty is measured in absolute term and pl doesn t change the % of poor people will be lower - Fig. 5b: impact of a decrease in inequality with constant mean (i.e. without growth). Both distributions have equal mean but the lower has a lower dispersion. Poverty is lower. See the table below on the relationship between P, I and G 20
21
Human Development Indexes An overview of the 5 HDI used in the HDR (see graph) 22
1) Human Development Index (HDI) is a summary measure of human development and measures the average achievements in a country in 3 basic dimensions of HD: a long and healthy life, measured by life expectancy at birth knowledge, measured by the adult literacy rate (with two-thirds weight) and the combined primary, secondary and tertiary gross enrolment ratio (with one-third weight) a decent standard of living, measured by GDP per capita (PPP US$) for calculating these dimension indices (life expectancy, education and GDP) minimum and maximum values (goalposts) are chosen : INDICATOR MAXIMUM VALUE MINIMUM VALUE Life expectancy at birth (years) 85 25 Adult literacy rate (%) 100 0 Combined gross enrollment ratio (%) 100 0 GDP per capita (PPP US$) 40000 100 performance in each dimension is expressed as a value between 0 and 1 by: Dimension index = (actual value minimum value)/ (maximum value minimum value) Finally, HDI is calculated as a simple average of dimension indice (see example) 2) Human Poverty Index (HPI) a) for developing countries (HPI-1) it measures deprivations in 3 basic dimensions of HD captured in the HDI: i. a long and healthy life, measured by the probability at birth of not surviving to age 40 ii. knowledge, measured by the adult illiteracy rate iii. a decent standard of living, measured by: 23
1. % of population not using improved water sources 2. % of children underweight for age (an underweight average between 1. and 2. is calculated to determine iii) the indicators are already normalized between 0 and 100 because they are expressed as % HPI-1 is calculated as follows: HPI-1 = [(P 1 3 + P 1 3 + P 3 3 ) / 3] 1/3 b) for selected OECD countries (HPI-2) HPI-2 refers to the same dimensions as HPI-1 plus social exclusion: i. a long and healthy life, measured by the probability at birth of not surviving to age 60 ii. knowledge, measured by the % of adults (16-65) lacking functional literacy skills (according to the literacy scale of the International Adult Literacy Survey) iii. a decent standard of living, measured by the % of population living below the income PL (50% of the median disposable household income) iv. social exclusion, measured by the rate of long term unemployment (12 months or more) o the indicators are already normalized between 0 and 100 because they are expressed as % o HPI-2 is calculated as follows: HPI-2 = [(P 3 3 1 + P 1 + P 3 3 ) / 3] 1/3 (see examples) Why a parameter α = 3 has been chosen: if α = 1, HPI would be the arithmetic average of these dimension. As α rises, greater weight is given to the dimension in which there is most deprivation. The choice of a value = 3 is used to give additional but not overwhelming weight to areas of more acute deprivation (see technical note of HDR 1997) 24
3) Gender Related Development Index (GDI) GDI adjusts the average achievement to reflect the inequalities between men and women in the same dimensions of HDI: 1) Three steps: 1. F and M indices in each dimension are calculated according to the general formula: Dimension index = (actual value minimum value)/ (maximum value minimum value) INDICATOR MAXIMUM VALUE MINIMUM VALUE Female Life expectancy at birth (years) 87.5 27.5 Male Life expectancy at birth (years) 82.5 22.5 Adult literacy rate (%) 100 0 Combined gross enrollment ratio (%) 100 0 GDP per capita (PPP US$) 40000 100 2. F and M indices are combined in a way that penalizes differences in achievement between men and women: equally distributed index = {[female population share (female index 1-ε )] + [male population share (male index 1-ε )]} 1/1-ε ε measures the aversion to inequality and in the GDI ε = 2 (moderate penalty on gender inequality in achievement) so that the above formula becomes: {[F population share (F index -1 )] +[M population share (M index -1 )]} -1 3. GDI is calculated by combining the three equally distributed indices in an unweighted average. (see example) 4) Gender Empowerment Measure (GEM) GEM captures gender inequality in 3 key areas: 1. Political participation and decision making power as measured by women s and men s % shares of parliamentary seats 2. Economic participation and decision-making power as measured by 2 indicators: 25
a. M and F % shares of positions as legislators, senior ifficials and managers b. M and F% shares of professional and technical positions 3. Power over economic resources as measured by F and M estimated earned income (PPP US$) For each dimensions an equally distributed equivalent percentage (EDEP) is calculated (see the above EDE); for economic and political participation and decision making the EDEP is indexed by dividing it by 50 (in an ideal society with equal empowerment the GEM variables would equal 50%) Finally the GEM is calculated as a simple average of the three indexed EDEPs Links between inequality, poverty and human development Inequality broader than poverty defined over the whole distribution and not only on part of it (below a given poverty line) but narrower than HD and CA relative and not absolute income-based measure (how the cake is distributed, not how big it is neither what are the other ingredients ) Degrees of complexity/disaggregating level of empirical analyses Unit of analysis: which relates to the way in which human diversity is accounted for Evaluative space(s) or object of analysis: which refers to the number and the nature of evaluative spaces involved Context of analysis, which refers to the degree to which structural elements, spatial factors, and environmental diversities are taken into account. Standard Approach: most basic and common conceptualization of welfare (based on mainstream neoclassical economics) avoids any kind of complexity. Nothing other than income distinguishes individuals, and the amount of income of an individual is all that is 26
required to identify deprivation, evaluate inequalities, or depict quality of life and human well-being in line with the neoclassical perspective. Capability Approach: Plurality of units of analysis and human diversity: individual, family/household, specific sub-groups of population (i.e. women, children, elderly, by caste, by ethnic or religious group), whole community Plurality of evaluative spaces and domains: agency-empowerment, standard of living - well-being, capability, achieved functionings and multiplicity of variables for each domain (qual/quant, obj/subj) Plurality of contexts: socio-economics, geographical, cultural Human Development Indexes: Refer to people: conceptually recognizes human diversity but concretely pay attention only to gender Multidimensional evaluative spaces (from 3 to 4 in HPI-2) and various variables (a dozen) Context: conceptually pay attention to external factors that can affect HD achievements but only partially include differences in context (i.e. HPI-1 and HPI- 2) Standard Human Capability Approach Development Approach Unit of Representative People Human diversity analysis Agent (gender) (gender, age, etc.) Evalutative Unidimension. Multidimens. Multidimension. space(s) Economic Opportunities Capability, functioning, resources (I or C) (I, health, educ.) Freedom, empowerment Context, External factors No Partially Social, economic, political, cultural 27
PROBLEM SET N. 1 INEQUALITY MEASUREMENT 1. Compare the following distributions: X=(1,5,7,2,14) Y=(2,3,2,8,2) a) Is income distributed more equally in X or Y? b) Would you prefer to live in society X or Y? c) Graph and compare the corresponding Lorenz Curves d) Calculate and compare Gini coefficients PROBLEM SET N. 2 INEQUALITY MEASUREMENT 1. Compare the following distributions: M=(2,4,8,6) S=(2,5,8,5) e) Graph and compare the corresponding Lorenz Curves f) Calculate and compare range, variance, CV and Gini coefficients PROBLEM SET N. 3 POVERTY MEASUREMENT Given the distribution X = (2,3,4,11) and the poverty line z=10 a) calculate FGT indexes b) If the distribution becomes X =(2,3,3,11) which measures will increase and why? c) If the distribution becomes X =(3,3,3,11) which measures will decrease and why? d) Suppose you have money to allocate for reducing poverty. How would you do if the poverty measure were: 1) the head-count, 2) the poverty gap, 3) the squared poverty gap. 28
PROBLEM SET N. 4 HUMAN DEVELOPMENT INDEXES a) Calculate the HDI for China on the basis of the following data: Life expectancy: 72,5 years Adult literacy rate: 90,9% Combined gross enrolment ratio: 69,1% GDP per capita: 6757 b) Calculate the HPI-1 for United Arab Emirates on the basis of the following data: Probability at birth of not surviving to age 40: 2,1% Adult illiteracy rate: 11,3% Population not using improved water source: 0% Children under weight under age 5: 14% c) Calculate the HPI-2 for USA on the basis of the following data: Probability at birth of not surviving to age 60: 11,6% People lacking functional literacy skills: 20,0% Long term unemployment rate: 0,5% Population below income poverty line: 17% d) Calculate the GDI for Uruguay on the basis of the following data: Life expectancy: F= 79,4 years; M=72,2 years Adult literacy rate: F=97,3%; M=96,2% Combined gross enrolment ratio: F=95,0%; M=83% Estimated earned income: F= 7203; M=12890 Population share: F=0,505; M=0,495 e) Calculate the GEM for Italy on the basis of the following data: Seats in Parliament held by women: 16,1% Female legislators, senior officials and managers: 32,0% Female professional and technical workers: 46.0% Estimated earned income: F=18501; M=39163 Female population share =51,5% 29