Economics of Development Chapter 2 PDF

26 2 THE DEVELOPMENT GAP AND THE MEASUREMENT OF POVERTY Introduction The development gap and income distribution in the world economy Measures of inequality and historical trends International inequality (unweighted an...

26 2 THE DEVELOPMENT GAP AND THE MEASUREMENT OF POVERTY Introduction The development gap and income distribution in the world economy Measures of inequality and historical trends International inequality (unweighted and population weighted) Global (or world) inequality The measurement and comparability of per capita income Purchasing power parity (PPP) Per capita income as an index of development Measuring poverty Meeting the Sustainable Development Goal of poverty reduction Tackling poverty from the ‘grass roots’ Randomized control trials (RCTs) Human Development Index (HDI) Multidimensional Poverty Index (MPI) Can the poor countries ever catch up? Summary Discussion questions Notes Websites on poverty and income distribution  2 THE DEVELOPMENT GAP AND THE MEASUREMENT OF POVERTY 27 Introduction This chapter focuses on three major topics: first, the development gap in the world economy and the measurement of the world distribution of income; second, the measurement of global poverty and the problems associated with the use of per capita income as a measure of develop- ment; and third, the construction of alternative measures of economic and social development, including the Human Development Index (HDI) and the Multidimensional Poverty Index (MPI) developed by the Oxford Poverty and Human Development Initiative, and both published in the Human Development Report of the United Nations Development Programme (UNDP). To measure the development gap, and the degree of income inequality across countries of the world, we consider: the absolute gap between the richest and poorest country (the range) the relative gap between the richest and poorest countries the dispersion of income per capita around the average level of per capita income for all countries the Gini ratio, which is derived from the Lorenz curve of the income distribution. Using the Gini ratio, we distinguish between international inequality, which takes each country’s per capita income as just one observation (regardless of the distribution of income within coun- tries) and global inequality, which takes account not only of the distribution of income between countries but also within countries (using household survey data). We show that international and global inequality have both risen since the beginning of the nineteenth century, but the major cause of global inequality today is inequality between nations, not inequality within nations. It is important to recognize, however, that when measuring income inequality and poverty, the measures of income per head in US$ at official exchange rates are not necessarily a good measure of the purchasing power of local currencies (or what is called purchasing power parity, PPP), because official exchange rates do not take account of the much cheaper price of non-traded goods in poor countries relative to richer countries. We show how PPP is measured, and discuss in general the comparability of per capita incomes between countries, and the use of per capita income as an index of development. Turning to the measurement of poverty, we discuss the World Bank’s criterion of absolute poverty, which is $1.90 a day at PPP, and give the headcount index of the numbers living below this level of income across different regions of the world. The concept of the poverty gap is also discussed, because the headcount index does not take account of how far below the poverty line people live. Another method of measuring poverty is the food energy method, which measures the income necessary to buy a certain nutritional intake in different countries. The World Bank claims that it puts poverty reduction at the heart of all the work that it does, and in its World Development Report 2000/2001 (World Bank, 2000a), it proposed a three-pronged strategy for poverty reduction: promoting opportunity, facilitating empowerment and enhancing security. The meaning of these concepts is discussed in this chapter. Growth is, of course, central to poverty reduction, but how fast the poverty rate falls with growth (to meet the Sustainable Development Goal, for example) depends on the elasticity of the poverty rate with respect to growth. To overcome the limitations of taking a single measure of per capita income as an index of development, the UNDP annually constructs a Human Development Index (HDI) and also pub- lishes a Multidimensional Poverty Index (MPI). The HDI is based on three variables: life expectancy at birth, educational attainment (measured as the geometric mean of the average and expected years of schooling), and the standard of living measured at PPP. The MPI is also based on three dimensions of poverty: education, health, and standard of living, but with several sub-dimensions 28 DEVELOPMENT AND UNDERDEVELOPMENT of poverty as well. The rankings of countries by their level of per capita income are not the same as the rankings of countries by the HDI or MPI indices, because some countries devote more resources to social expenditure than others, particularly on education and health. Finally, we consider the question of whether poor countries are ever likely to catch up with the rich, and we reach the pessimistic conclusion that it will take the average poor country at least 100 years to achieve the current living standards in developed countries and probably 300 years for poor countries to equalize living standards with developed countries if they manage to grow, say, 1% faster than the rich countries. But this, of course, cannot be taken for granted. The development gap and income distribution in the world economy By any measure one cares to take, the evidence is unequivocal that the world’s income is dis- tributed extremely unequally between nations and people. There are many ways of classifying these divisions in the world economy. First, at a very basic level, there is the division between rich industrialized countries, mainly concentrated in the northern hemisphere, and poorer non- industrialized (or semi-industrialized) countries in the southern hemisphere – often referred to in the development literature as the ‘North–South divide’. Second, there is the division between continents: between the developed continents of Europe and North America on the one hand, and the continents of Asia, Africa and Latin America on the other. But the countries of Asia, Africa and Latin America are by no means homogeneous. They have many characteristics, and obstacles to development, in common, but there is also much that divides them, not least their economic performance since the 1960s, with Southeast Asia, China and India forging ahead, Africa left behind, and Latin America in the middle (often prone to financial crises). Third, the World Bank, which was established by the Bretton Woods Agreement in 1944 as a development agency to lend to poor countries, classifies countries in its annual World Development Report into three broad categories: low-income, middle-income and high-income countries, with the middle-income countries split into lower middle-income and upper middle-income. Table 2.1 lists the low-income and lower middle-income countries, and these are the countries normally thought of as developing countries. There are 36 low-income countries mainly concen- trated in Africa, and 48 lower middle-income countries spread across the different continents. Table 2.2 gives the population and average level of per capita income in current dollars measured at the official exchange rate, and at PPP for 2014. We see that for low-income countries with a popula- tion of 622 million, the average level of income per head is only $629 at current dollars and $1,570 at PPP (or roughly $5 a day). For lower middle-income countries, with a population of nearly 3 bil- lion, the average level of per capita income is $2,012 at current dollars and $6,000 at PPP (or roughly $16 a day). Compare these figures with an average income per head in high-income countries of roughly $40,000 (or just over $100 day). The discrepancies are huge. The richest country in the world is currently Norway, with an income per head of $66,330 at PPP, and the poorest countries are Democratic Republic of the Congo, with an income per head of $650, and Burundi with $770. Measures of inequality and historical trends 1. One measure of dispersion is the range or absolute income gap between the richest and poor- est countries. This gap is almost bound to grow over time if both the rich and poorest countries  2 THE DEVELOPMENT GAP AND THE MEASUREMENT OF POVERTY 29 Table 2.1 List of low-income and lower middle-income countries Low-income countries Afghanistan Gambia, The Myanmar Bangladesh Guinea Nepal Benin Guinea-Bissau Niger Burkina Faso Haiti Rwanda Burundi Kenya Sierra Leone Cambodia Korea, Dem. Rep. Somalia Central African Republic Kyrgyz Republic South Sudan Chad Liberia Tajikistan Comoros Madagascar Tanzania Congo, Dem. Rep. Malawi Togo Eritrea Mali Uganda Ethiopia Mozambique Zimbabwe Lower middle-income countries Armenia India Samoa Bhutan Kiribati Sāo Tomé and Príncipe Bolivia Kosovo Senegal Cameroon Lao PDR Solomon Islands Cape Verde Lesotho Sri Lanka Congo, Rep. Mauritania Sudan Cote d’Ivoire Micronesia, Fed. Sts. Swaziland Djibouti Moldova Syrian Arab Republic Egypt, Arab Rep. Mongolia Timor-Leste El Salvador Morocco Ukraine Georgia Nigeria Vanuatu Guatemala Pakistan Vietnam Guyana Papua New Guinea West Bank and Gaza Honduras Paraguay Yemen, Rep. Indonesia Philippines Zambia Source: World Bank, website. experience positive growth. For example, if the richest country, Norway, grows at 1%, this adds roughly $700 to the level of per capita income in Norway, while the same growth rate adds only $6–$7 dollars to the per capita incomes of the Democratic Republic of the Congo or Burundi. These countries would have to grow at 700% for the absolute gap between them- selves and Norway to be narrowed. But even the gap between the richest and poorest country is an understatement of the degree of income inequality in the world economy because it compares only the average income for poor and rich countries. If the income per head of the poorest people in poor countries is compared with the income per head of the richest people in rich countries, the absolute gap is even wider. 2. A second measure of dispersion, or division in the world economy, is the relative income gap, which is the ratio of the richest country (or group of countries) to the poorest country (or 30 DEVELOPMENT AND UNDERDEVELOPMENT Table 2.2 Population and income per capita, 2014 Population, millions Gross national income PPP gross national per capita, $ income per capita, $ World 7,261 10,787 14,931 Low income 622 629 1,570 Middle income 5,240 4,666 9,673 Lower middle income 2,879 2,012 6,000 Upper middle income 2,361 7,901 14,179 Low and middle income 5,862 4,238 8,811 East Asia and Pacific 2,021 6,156 11,872 Europe and Central Asia 264 6,892 13,580 Latin America and 525 8,990 14,053 Caribbean Middle East and North 357 4.222 11.834 Africa South Asia 1,721 1,496 5,299 Sub-Saharan Africa 973 1,638 3,382 High income 1,399 38,274 40,749 Source: World Bank, 2015. group of countries). At present, the ratio of income per head of the richest to the poorest country is approximately 700:1, and the ratio of income per head in the low-income countries to the high-income countries is approximately 60:1 at current exchange rates. This relative income gap is unprecedented historically. A necessary condition for the relative income gap to narrow is that the poorest countries grow faster than the richest. 3. A well-known statistical measure of dispersion is the standard deviation (SD), or the square root of the variance, which measures the average sum of the squared deviations of each coun- try’s per capita income from the average (or mean) income for all countries. Formally, the SD is measured as: n (Yi − Y)2 i= 1 SD = n where Yi is the per capita income of country i; Y– is the average level of per capita income of the whole sample, and n is the number of countries. In the growth and development literature, movements in this ratio, up or down, are referred to as ‘sigma (s) divergence or convergence’, respectively. 4. There is the coefficient of variation, which is the SD divided by the mean of the sample (Y–). This normalizes the SD because there is a positive correlation between the mean and the SD. 5. But the most widely used measure of income inequality is the so-called Gini ratio, derived from the Lorenz curve, which in this field of enquiry relates to the distribution of income in relation to the distribution of population across countries or groups of countries. Three  2 THE DEVELOPMENT GAP AND THE MEASUREMENT OF POVERTY 31 measures or concepts of inequality need to be clearly distinguished in using the Gini ratio. First, there is international inequality, with each country treated as a single unit and given equal weight in the measure. Second, there is weighted international inequality, with each country treated as a single unit but weighted by its size of population. Third, there is world or global inequality, which takes the individual person (or household), not the country, as the unit of measurement, and therefore takes into account not only differences in income between countries, but also between people within countries. Each measure has its own purpose, and there is no theoretical reason why the measures should move together (although, in practice, they tend to, taking a long historical perspective). Before describing how the Lorenz curve is constructed and the Gini ratio is measured, however, it needs to be said in advance that a single statistic does not say what is happening within the distribution, and, in particular, what is happening at the extremes of the distribu- tion. Ratios of extremes, such as the income of the poorest 10% of the world’s population compared with the richest 10% (often called the Kuznets ratio), or income of the poorest countries compared with the richest, can say as much, if not more, about income inequal- ity and social justice than any integral measure. In fact, the Gini ratio may indicate con- vergence or less inequality, while the ratio of extremes is increasing. For discussion of the conceptual problems relating to analysis of the world distribution of income, see Atkinson and Brandolini (2010). Consider now Figure 2.1. On the vertical axis is measured the percentage of income, and on the horizontal axis is measured the percentage of population. To draw the distribution of income (the Lorenz curve) first rank each country, groups of countries, or groups of individuals in ascending order according to the ratio of the percentage of income they receive and the percentage of popu- lation they represent; then cumulate the observations, and plot them on the diagram. To give a Figure 2.1 Lorenz curve diagram 100 90 80 70 60 % of income 50 40 30 20 10 0 10 20 30 40 50 60 70 80 90 100 % of population 25 2 9 81 32 DEVELOPMENT AND UNDERDEVELOPMENT simple example, suppose we take the World Bank’s division of countries into low, middle and high income, and that low-income countries contain 9% of the world’s population and receive only 2% of world income, middle-income countries contain 72% of the world’s population and receive 23% of world income, and rich countries contain 19% of the world’s population and receive 75% of world income. The cumulative distribution of income in relation to population would then be 2/9, then 25/81 (when middle-income country figures are added), and finally 100/100 when the rich countries are added. These points are plotted in Figure 2.1 and the curve joining them is the Lorenz curve. The diag- onal 450 line on the diagram shows an equal distribution of income. The position of the Lorenz curve in relation to the 450 line therefore gives a visual impression of the degree of inequality. The closer the Lorenz curve, the more equal the distribution, and the more ‘bowed’ the curve, the more unequal the distribution. The Gini ratio is calculated as the area between the Lorenz curve and the 458 line divided by the area of the triangle it lies within. If the Lorenz curve is coincident with the 458 line, the Gini coefficient would be zero – complete equality. If one person received all the world’s income, the Lorenz curve would follow the horizontal and vertical axes and the Gini coefficient would be one. When we examine international and global inequality, what we find is that through time, at least since the early 1900s, the Lorenz curve has been shifting outwards, and the Gini ratio has been rising, although according to some investigators, it may recently have levelled off, albeit at a high level. A central estimate for the current level of international inequality would be a Gini ratio of 0.54, and for global inequality a Gini ratio of 0.67. But, as we shall come to see, estimates vary depending on factors such as the sample of countries taken, how income is measured – whether by per capita income or household income – and whether income is measured at official exchange rates or at PPP rates. There have been many recent studies measuring and summarizing what has been happen- ing to international and global inequality historically (e.g. Milanovic, 2005, 2016; Bourguignon and Morrisson, 2002), and particularly since the 1950s (e.g. Norwegian Institute of Economic Affairs, 2000; Sala-í-Martin, 2002; Maddison, 2003; Ghose, 2004; Wade, 2004; Sutcliffe, 2004; Svedberg, 2004; Milanovic, 2005, 2016). What does the evidence show? We distinguish between international inequality (unweighted and weighted by population) and global (or world) inequality. International inequality (unweighted and population weighted) The unweighted Gini ratio of international inequality takes each country as one unit, regard- less of population size, and assumes that each person within the country has the same average income. The distribution of income within the country is not considered. It is countries that are the focus, not people. The ratio is basically, therefore, a measure of whether or not coun- tries are converging with each other, not whether the distribution of income across individuals in the world is becoming more or less equal. What does the evidence show? Using the best historical data available (Maddison, 2001; Bourguignon and Morrisson, 2002) for 26 coun- tries covering nearly 80% of the world’s population, the Gini ratio in 1820 was approximately 0.2. This is very low by current standards. Two hundred years ago, international differences in income per head were not great. Maddison (2003) and Easterlin (2000) show that the ratio of  2 THE DEVELOPMENT GAP AND THE MEASUREMENT OF POVERTY 33 per capita income of the richest to the poorest country in 1820 was only 3:1, compared with nearly 700:1 today. Table 2.3 shows the evolution of the unweighted Gini ratio through time, rising consist- ently to 0.54 in 2013 – a more than doubling of income inequality in the space of nearly 200 years. Some of the increase may be spurious due to the larger sample of countries used to cal- culate the ratio, but Milanovic (2005, 2016) shows that for the same 26 countries as used for the 1820 calculations, the Gini ratio still rises to just over 0.5 in 2000. For the period since the Second World War, the Gini ratio shows an increase for a consistent set of over 100 countries from 0.45 in 1952 to a peak of 0.58 in 2002. Since then, it has been falling slightly due to the fast growth of China and India, and some African countries because of the rise in commodity prices (Bourguignon, 2015). Turning now to the population-weighted measure of international inequality, Table 2.3 tells a slightly different story. It shows the Gini ratio peaking at 0.6 in 1988 and falling consistently to 0.47 in 2013. This implies that poor countries with large populations must have been growing faster on average than richer countries with smaller populations. In the 1950s and 1960s, this was due to the fast growth of some of the big Latin American countries, such as Brazil and Mexico, and Japan and South Korea, all using trade protection of one form or another. In the 1990s and 2000s, the decline in the weighted Gini ratio has been largely due to the rapid growth of a small number of poor populous Asian countries, especially India and China. Ghose (2004), in his study of 96 countries over the period 1981 to 1997, finds that the weighted Gini ratio fell by 0.7% per annum, Table 2.3 A comparison of Gini ratios International inequality Global (or world) inequality Year Unweighted Population Milanovic (2005, Bourguignon and Sala-í-Martin weighted 2016) Morrisson (2002) (2002) 1820 0.20 0.12 0.50 1870 0.29 0.26 0.56 1890 0.31 0.30 0.59 1913 0.37 0.37 0.61 1929 0.35 0.40 0.62 1938 0.35 0.40 1952 0.45 0.57 0.64 1960 0.46 0.55 0.64 1978 0.47 0.54 0.66 0.66 (1970) 1988 0.53 0.60 0.68 0.65 1993 0.56 0.59 0.70 0.66 (1992) 0.64 1998 0.56 0.57 0.69 0.63 2002 0.58 0.56 0.71 0.63 (2000) 2005 0.57 0.54 0.70 2008 0.55 0.51 0.69 2011 0.54 0.48 0.67 2013 0.54 0.47 Sources: Adapted from Milanovic, 2005, Table 11.1; Milanovic, 2016. 34 DEVELOPMENT AND UNDERDEVELOPMENT but only 17 of the 76 developing countries in the sample converged on the per capita income of the 20 developed countries. The majority of poor developing countries diverged. Despite the fall in the ratio since 1988, it is still high and much higher than the estimate of 0.12 in 1820. In other words, for a large part of the past two centuries, the world’s poorest and most populous countries have fared badly compared with the smaller, rich countries of the world. China and India are now reversing the trend, but for how long remains to be seen. Global (or world) inequality The Gini ratio of global (or world) inequality takes into account not only differences in average per capita income between countries, but also differences in income per capita within countries. Because internal income distributions are never equal, the measure of global inequality is bound to be higher than the unweighted measure of international inequality. It also means that changes in the global distribution of income are an amalgam of forces, including what is happening to the distribu- tion of income between countries, what is happening to population growth in rich and poor coun- tries, and what is happening to the distribution of income within countries. What does the evidence show? As far as the historical record is concerned, Bourguignon and Morrisson (2002) have tried to measure inequality among world citizens back to 1820 using a sample of 33 countries (or groups of countries) and measuring domestic income inequality by taking decile income shares (with the top decile of income earners divided into two). The results are shown in Table 2.3 above. It can be seen that the global Gini ratio in 1820 was already 0.5, more than double the level of international inequality, implying that domestic inequality was as great, if not greater, than inter­national inequal- ity. Through time, global inequality has increased, but because of rising inter­national ­inequality, not because of even greater inequality within countries. On the contrary, domestic income inequality, at least until recently, has shown a decrease historically, particularly in the richer countries of the world. Bourguignon and Morrisson calculate that within-country inequality accounted for 80% of global inequality in the first half of the nineteenth century, when most countries were more or less at the same income level, but by 1950, within-country inequality accounted for 40% of global inequality because of the increase in inequality between countries. Today, the contribution is about 20%. The Gini ratio of global inequality seems to have peaked in the early 2000s at 0.71 and has since declined (Bourguignon, 2015). But conflicting forces are at work. Population-weighted international inequality is falling because of the fast growth of India and China, but income distribution within some countries is widening, such as between the urban and rural sectors of China and India and in some developed countries such as the USA and countries of Europe. If the income distribution within countries was falling, the global Gini ratio would evidence a greater fall. Sala-í-Martin (2002) covers the period 1970–98, taking the income distribution of 125 coun- tries and aggregating them. Despite the much larger sample, the calculated global Gini ratios are remarkably similar to those of Bourguignon and Morrisson. The estimate for 1970 is 0.66, grad­ ually falling to 0.63 in 2000 (see Table 2.3). The explanation for the slight fall is that the lower ‘tail’ of the aggregated income distribution has shifted rightwards more dramatically than the upper ‘tail’, largely due to developments in China and India. Fast growth in these two countries has lifted millions of people above the poverty line, and has reduced the relative income gap with richer countries, and this has been enough to just offset the worsening income distribution within China and India, as mentioned earlier. Still, we have evidence again of global inequality on a vast scale. Milanovic (2005, 2016) has also undertaken the Herculean task of bringing together 300 household sample surveys of income and expenditure for over 100 countries for selected years from 1988 to 2011, covering 80–90% of the world’s population. The calculated global Gini ratio  2 THE DEVELOPMENT GAP AND THE MEASUREMENT OF POVERTY 35 using household income or expenditure measured at PPP is 0.68 in 1988 and 0.67 in 2011 (see Table 2.3). These estimates are remarkably similar to those of Sala-í-Martin, despite the difference in samples and the measure of income. Bourguignon (2015) gives a figure of 0.691 for 2010. These results are confirmed by Edward (2006) using national consumption distributions and collating them into a global distribution measured at PPP in US$ for 1993 and 2001. The global Gini ratio is estimated at 0.610 for 1993 and 0.614 for 2001. Just over 80% of this inequality is the result of between-country differences. All the evidence and studies show a massive degree of inequality in the world distribution of income, which shows some recent improvement but may not last if growth in China and India slows and the income distribution within countries gets worse. This development gap naturally extends into other aspects of human welfare such as health, nutrition, life expectancy, education, employment opportunities and so on, as we shall come to see later in this chapter and in Chapters 3 and 7. The UNDP (1997) has described the world as ‘gargantuan in its excesses and grotesque in its human and economic inequalities’. This is, of course, a normative statement, but economists should not be afraid of making normative statements, as Basu (2006) does when he argues that: the hiatus between the richest and the poorest people is too large, and the extent of poverty on earth is unacceptable. I like to believe that there will come a time when, looking back at today’s world, human beings will wonder how primitive we were that we tolerated this. Many other statistics can be given to illustrate the grotesque inequalities that exist. The richest 1% of people in the world receive as much income as the bottom 60%. Or, to put it another way, the 60 million richest people receive as much income as 2.7 billion poor. The total income of the richest 25 million Americans is equal to the total income of 2 billion of the world’s poorest people. The assets of the world’s 400 billionaires (mostly in rich countries) exceed the total amount of income of nearly one-half of the world’s total population. The most evocative graph comes from Wade (2001), who divides the world’s population up into equal 20% shares (quintiles) from poorest to richest, and then shows the percentage of income that each share receives. Interestingly, and ironically, the picture resembles a champagne glass with a very narrow stem in the hands of the poor and a wide open bowl (containing the champagne) in the hands of the rich (Figure 2.2). Below we discuss some technical problems concerning the measurement and comparability of per capita income across countries, and the measurement of poverty itself. Figure 2.2 D  istribution of world income (percentage of total, with quintiles of population ranked by income) SECOND RICHEST 20% 82.7% 20% 11.7% THIRD 20% 2.3% POOREST FOURTH 20% 1.9% Each horizontal band 1.4% 20% represents an equal fifth of the world’s people Source: Wade, 2001, based on Figure 3.2 in UNDP, 1992. 36 DEVELOPMENT AND UNDERDEVELOPMENT The measurement and comparability of per capita income When using per capita income (PCY) figures to measure poverty, to classify countries into rich and poor, and to compare the rate of development in different countries over time, the difficulties of measuring real per capita income and real living standards between countries must be continu- ally borne in mind. There are two issues to discuss. The first concerns the problems associated with national income accounting, particularly in developing countries. The second is the need to convert each country’s per capita income in its own domestic currency into a common unit of account (e.g. the US$) so as to be able to make meaningful international comparisons of living standards. This leads to the topic of purchasing power parity (PPP) estimates of PCY. Turning first to national income accounting, the first point to bear in mind is that only goods that are produced and sold at a price in the market are included in the value of national income, measured by either the output or the expenditure method. Much output in developing countries never reaches the market, particularly in the rural sector where production is for subsistence pur- poses. If no allowance is made for the subsistence sector, this will bias downwards the calculation of national income, and therefore PCY. This point also implies that growth rates will tend to have an upward bias as a result of the extension of the money economy and the shift of economic activities from the household and subsistence sector to the marketplace. Part of the growth in developing countries may be a statistical illusion arising from the changing balance between the informal subsistence sector and the modern exchange sector. Second, there is the sheer practical difficulty of measuring money national income in a rural economy where communications are bad, illiteracy is rife, and accounting procedures are rudi- mentary. Differences in the extent of the subsistence economy between developing countries, and differences in the ease and difficulty of collecting data, may markedly influence estimates of national income, and therefore of per capita income differences, between these countries and the rest of the world. Attempts are made in developing countries to make some allowance for pro- duction that never reaches the marketplace, but the estimates are likely to be subject to a wide margin of error. Some testimony to the role that the subsistence sector must play in the economies of most developing countries is provided by the inconceivability that 10% of the world’s population could remain alive on less than $1,000 per annum. But this is not the whole story. Purchasing power parity (PPP) The other part of the story, and probably the major part, concerns the understatement of living standards in developing countries when their national incomes measured in local currencies are converted into US$ (as the common unit of account) at the official rate of exchange. If the US$ is used as the unit of account, the national per capita income of country X in US$ is given by: GNPX 4 Exchange rate Population For example, if the GNP of country X is 100 billion rupees, its population is 5 million, and there are 10 rupees to the dollar, then the per capita income of country X in dollars is: 100 billion 4 10 5 $2,000 5 million  
t
5 ) &
% & 7 & - 0 1. & / 5
( " 1
" / %
5 ) &
. & " 4 6 3 &. & / 5
0 '
1 0 7 & 3 5 :

37 #VU
JG
UIF
MJWJOH
TUBOEBSET
PG
UIF
UXP
DPVOUSJFT
BSF
UP
CF
DPNQBSFE
CZ
UIJT
NFUIPE
JU
NVTU
CF
BTTVNFE
UIBU

SVQFFT
JO
DPVOUSZ
X
CVZT
UIF
TBNF
MJWJOH
TUBOEBSE
BT

JO
UIF
64"
*U
JT
XFMM
LOPXO
IPXFWFS
UIBU
PċDJBM
FYDIBOHF
SBUFT
CFUXFFO
UXP
DPVOUSJFT
DVSSFODJFT
BSF
OPU
HPPE
NFBTVSFT
PG
UIF
111
CFUXFFO
DPVOUSJFT
FTQFDJBMMZ
CFUXFFO
DPVOUSJFT
BU
EJĊFSFOU
MFWFMT
PG
EFWFM- PQNFOU
ʾF
SFBTPO
JT
UIBU
FYDIBOHF
SBUFT
BSF
MBSHFMZ
EFUFSNJOFE
CZ
UIF
TVQQMZ
PG
BOE
EFNBOE
GPS
DVSSFODJFT
CBTFE
PO
HPPET
BOE
BTTFUT
UIBU
BSF
USBEFE
UIF
QSJDFT
PG
XIJDI
UFOE
UP
CF
FRVBMJ[FE
JOUFSOBUJPOBMMZ
CVU
MJWJOH
TUBOEBSET
EFQFOE
BMTP
PO
UIF
QSJDFT
PG
non-traded goods, which are MBSHFMZ
EFUFSNJOFE
CZ
VOJU
MBCPVS
DPTUT
BOE
UIFTF
UFOE
UP
CF
MPXFS
UIF
QPPSFS
UIF
DPVOUSZ
"T
B
general rule, it can be said that the lower the level of development and the poorer the country, the MPXFS
UIF
QSJDF
PG
OPOUSBEFE
HPPET
SFMBUJWF
UP
USBEFE
HPPET
BOE
UIF
NPSF
UIF
VTF
PG
UIF
PċDJBM
FYDIBOHF
SBUF
XJMM
understate
UIF
MJWJOH
TUBOEBSET
PG
UIF
EFWFMPQJOH
DPVOUSZ
NFBTVSFE
JO
64 -FU
VT
HJWF
B
TJNQMF
FYBNQMF
ʾF
NPUPS
DBS
JT
BO
JOUFSOBUJPOBMMZ
USBEFE
HPPE
4VQQPTF
UIBU
UIF
EPMMBS
QSJDF
PG
B
QBSUJDVMBS
NPEFM
PG
DBS
JT
 
BOE
UIFSF
BSF

SVQFFT
UP
UIF
EPMMBS
*HOPSJOH
USBOTQPSU
DPTUT
UBSJĊT
BOE
TP
PO
UIF
QSJDF
PG
UIF
DBS
JO
*OEJB
XJMM
CF
 
3

5
 
rupees, otherwise a profit will be made by dealers buying in the cheapest market and selling in the NPTU
FYQFOTJWF
ʾF
GPSDFT
PG
EFNBOE
BOE
TVQQMZ
BOE
BSCJUSBHF
XJMM
FRVBMJ[F
UIF
QSJDF
PG
USBEFE
HPPET
#VU
MFU
VT
OPX
DPOTJEFS
B
OPOUSBEFE
HPPE
TVDI
BT
B
IBJSDVU
4VQQPTF
B
IBJSDVU
JO
UIF
64"
DPTUT

"U
UIF
PċDJBM
FYDIBOHF
SBUF
PG

SVQFFT
UP
UIF
EPMMBS
B
IBJSDVU
JO
*OEJB
TIPVME
CF

SVQFFT
#VU
TVQQPTF
UIBU
JO
GBDU
JU
JT
POMZ

SVQFFT
ʾJT
XPVME
NFBO
UIBU
BT
GBS
BT
IBJSDVUT
BSF
DPODFSOFE
UIF
WBMVF
PG
UIF
SVQFF
JT
VOEFSFTUJNBUFE
CZ
B
GBDUPS
PG
GPVS
ʾF
111
SBUF
PG
FYDIBOHF
GPS
IBJSDVUT
BMPOF
JT

4

SVQFFT
PS

5

SVQFFT
*G
UIF
OBUJPOBM
JODPNF
PG
DPVOUSZ
X NFBTVSFE
JO
SVQFFT
XBT
EJWJEFE
CZ

JOTUFBE
PG

UIF
OBUJPOBM
JODPNF
PG
DPVOUSZ
X in dollars, BOE
UIFSFGPSF
1$:
JO
EPMMBST
XPVME
OPX
CF
GPVS
UJNFT
IJHIFS
 
QFS
IFBE
JOTUFBE
PG
 
QFS
IFBE
BT
JO
UIF
FYBNQMF
BCPWF "T
EFWFMPQNFOU
QSPDFFET
UIF
SBUJP
PG
UIF
QSJDF
PG
OPOUSBEFE
HPPET
UP
USBEFE
HPPET
UFOET
UP
SJTF
BT
XBHF
MFWFMT
JO
UIF
OPOUSBEFE
HPPET
TFDUPS
SJTF
CVU
QSPEVDUJWJUZ
HSPXUI
JT
TMPX
o
TMPXFS
UIBO
JO
UIF
USBEFE
HPPET
TFDUPS
To make meaningful international comparisons of income and living standards, therefore, what is required is a measure of PPP, or a real exchange rate, between countries. ʾFSF
BSF
TFWFSBM
NFUIPET
PG
DPOTUSVDUJOH
111
SBUJPT
JO
PSEFS
UP
NBLF
CJOBSZ
DPNQBSJTPOT
POF
DPVOUSZ
XJUI
BOPUIFS
PS
ANVMUJMBUFSBM
DPNQBSJTPOT
JO
XIJDI
UIF
DVSSFODZ
PG
BOZ
POF
PG
B
group of countries can act as the unit of account without altering the ratios of living standards CFUXFFO
DPVOUSJFT ʾF
NPTU
DPNNPO
XBZ
PG
DPOTUSVDUJOH
B
111
SBUJP
CFUXFFO
UXP
DPVOUSJFT
TBZ
*OEJB
BOE
UIF
64"
JT
UP
UBLF
B
SFQSFTFOUBUJWF
DPNQBSBCMF
CBTLFU
PG
HPPET
BOE
TFSWJDFT
JO
CPUI
DPVOUSJFT
BOE
UIFO
UBLF
UIF
XFJHIUFE
BWFSBHF
PG
QSJDFT
XIFSF
UIF
XFJHIUT
wi
SFnFDU
UIF
QSPQPSUJPO
PG
FYQFOEJUVSF
PO
FBDI
HPPE
JO
UPUBM
FYQFOEJUVSF
ʾF
111
SBUF
PG
FYDIBOHF
CFUXFFO
*OEJB
BOE
UIF
64"
JT
UIFSFGPSF wiIPiI PPP 5 , wiUSPiUS where PiI is the price of the good in India and PiUS
JT
UIF
QSJDF
PG
UIF
HPPE
JO
UIF
64" ʾF
EJĊFSFODF
CFUXFFO
FTUJNBUFT
PG
1$:
BU
UIF
PċDJBM
FYDIBOHF
SBUF
BOE
111
FTUJNBUFT
PG
1$:
GPS
TBZ
UIF
MPXJODPNF
DPVOUSJFT
DBO
CF
TFFO
JO
5BCMF

Q
 
ʾF
EJĊFSFODF
JT
RVJUF
MBSHF
"U
UIF
PċDJBM
FYDIBOHF
SBUF
UIF
MFWFM
PG
1$:
JT
POMZ

QFS
BOOVN
CVU
BU
111
JU
JT
 
PS
NPSF
UIBO
EPVCMF
*O
PUIFS
XPSET
JO
MPXJODPNF
DPVOUSJFT
B
QFSTPOT
JODPNF
XPVME
CF
BCMF
38 DEVELOPMENT AND UNDERDEVELOPMENT to buy more than twice the goods and services that the official exchange rate would suggest. Notice that in high-income countries, there is hardly any difference between the two estimates because wage costs per unit of output in the non-traded goods sector match those of the traded goods sector. Per capita income as an index of development Now let us turn to the question of the use of per capita income figures as an index of development and for making a distinction between developed and developing countries, as well as between rich and poor. While there may be an association between poverty and underdevelopment and riches and development, there are a number of reasons why some care must be taken when using per capita income figures alone as a measure or indicator of development (unless underdevelopment is defined as poverty and development as riches). Apart from the difficulty of measuring income in many countries and the difficulty of making intercountry comparisons, using a single per capita income figure to separate developed from developing countries is inevitably somewhat arbitrary, because it ignores such factors as the distribution of income within countries, differences in devel- opment potential and other physical indicators of the quality of life. It is not so much a question of whether or not low-income countries should be labelled ‘underdeveloped’ or ‘developing’, but what income level should be used as the criterion for separating the developed from the develop- ing countries, and whether all high-income countries should necessarily be labelled ‘developed’. In many ways, it should be the nature and characteristics of the countries that determine which income level should be used as the dividing line. It also makes sense to categorize separately the oil-rich countries, which have high per capita incomes but cannot be regarded as developed by the criteria discussed in Chapter 1. Acronyms abound to describe the different stages of development. Perhaps the most amusing set is attributable to the Brazilian economist Roberto Campos, who distinguishes five categories of countries: the HICs, PICs, NICs, MICs and DICs. These stand for hardly industrialized countries, partly industrialized countries, newly industrialized countries, mature industrialized countries and decadent industrialized countries. The HICs and the PICs would certainly cover all the low- income countries and at least the lower half of the middle-income countries. The NICs cover most of the upper half of the middle-income countries – Brazil, Mexico, Hong Kong and Singapore being prime examples. The MICs and DICs cover most of those countries classified as ‘industrial market economies’, with the exception of New Zealand and Australia, which have become rich through agriculture. But bearing in mind the arbitrariness of per capita income, it is still very convenient to have a readily available and easily understandable criterion for classifying countries, and perhaps per capita income is the best single index we have. It also has one positive advantage, namely that it focuses on the raison d’être of development: raising living standards and eradicating poverty. And, in the last resort, per capita income is not a bad proxy for the social and economic structure of most societies. If developing countries are defined on the basis of a per capita income level so as to include most of the countries of Asia, Africa and Latin America, striking similarities are found between the characteristics and development obstacles of many of the countries in these contin­ ents. These include: A high proportion of the labour force engaged in agriculture with low productivity. A high proportion of domestic expenditure on food and necessities.  2 THE DEVELOPMENT GAP AND THE MEASUREMENT OF POVERTY 39 An export trade dominated by primary products and an import trade dominated by manufac- tured goods. A low level of technology and poor human capital. A high birth rate coupled with a falling death rate. Savings undertaken by a small percentage of the population. There are, of course, some countries that on a per capita income basis are classified as developed and possess many of the above-mentioned characteristics (e.g. some oil-producing countries), but the exceptions are few, and the reverse of this situation would be unusual. Also, these coun- tries have many social problems in common, such as growing unemployment in urban areas, inegalitarian income distributions, and poor health and standards of education – about which we shall say more later. In general, therefore, it can be said that per capita income may be used as a starting point for classifying levels of development, and can certainly be used to identify the need for development. The only major reservation that we shall have to consider later concerns the case of geographic­ ally dual economies, where an aggregate per capita income figure can disguise as great a need for the development of a sizeable region within the country as the need for the development of the country itself. There is a difference, however, between using per capita income as a guideline for classifying countries into developed or underdeveloped at a point in time and using the growth of per capita income as an index of development over time. The difficulty of using per capita income for the latter purpose is the obvious one that if, in a particular period, per capita income did not grow because population growth matched the growth of a country’s total income, one would be forced into the odd position of denying that a country had developed even though its national product had increased. This is an inherent weakness of linking the concept of development to a measure of living standards. This leads on to the distinction between growth and development. Development without growth is hardly conceivable, but growth is possible without development. The upswing of the trade cycle is the most obvious example of the possibility of growth without development; and examples of abortive ‘take-offs’ are not hard to find where countries have grown rapidly for a short time and then reverted to relative stagnation. Historically, Argentina is a case in point. On the other hand, development is hardly possible without growth; but development is possible, as we have suggested, without a rise in per capita income. It would be a strange, rather purposeless type of development, however, that left per capita income unchanged, unless the stationary per capita income was only temporary and a strong foundation was being laid for progress in the future. The ultimate rationale of development must be to improve living standards and welfare, and while an increase in measured per capita income may not be a sufficient condition for an increase in individual welfare, it is a necessary condition in the absence of a radical redistribution of income and the provision of basic needs to the poor. Measuring poverty The World Bank defines poverty as the inability of people to attain a minimum standard of living. The World Bank’s 1990 and 2000/2001 World Development Reports were devoted to a consider­ ation of the measurement, magnitude and nature of poverty in developing countries, and how to tackle it. This definition gives rise to three questions. How do we measure the standard of living? 40 DEVELOPMENT AND UNDERDEVELOPMENT What is meant by a minimum standard of living? How can we express the overall extent of poverty in a single measure? The most obvious measure of living standards is an individual’s (or household’s) real income or expenditure (with an allowance made for output produced for own consumption). The same level of real income and expenditure in different countries, however, may be associated with dif- ferent levels of nutrition, life expectancy, infant mortality, schooling and so on, which must be considered as an integral part of ‘the standard of living’. Measures of living standards based on per capita income, therefore, may need to be supplemented by further measures that include these other variables. Later in this chapter, we discuss the attempt by the UNDP to construct a Human Development Index, and the Oxford Poverty and Human Development Initiative to construct a Multidimensional Poverty Index, which take some of these factors into account. To separate the poor from the not so poor, an arbitrary per capita income figure has to be taken that is sufficient to provide a minimum acceptable level of consumption. There are two main ways of setting a consumption poverty line in order to measure poverty and make comparison across countries: the PPP method and the food energy method. As we have seen above, a country’s PPP is defined as the number of units of the country’s currency required to buy the same amount of goods and services in the domestic market as a dollar in the USA. The World Bank publishes the PPP levels of per capita income for all countries and regions of the world (see Table 2.2 on p. 30). For the measurement of poverty, to give an example, the PPP poverty line could be set at, say, $60 per month or $720 per annum. By definition, people on this PPP poverty line in any country have the purchasing power to obtain the same level of consumption of any person on the poverty line in any other country. But the composition of the consumption bundle is very likely to differ. The PPP poverty line is not explicitly linked to nutritional intakes derived from different consumption bundles, so there are likely to be intercountry differences in nutrition on the PPP poverty line. The food energy method of setting a consumption poverty line is one way of dealing with this problem by defining a minimum internationally agreed calorie intake line, and converting con- sumption bundles into calorie intakes using the nutritional values of consumption goods (with non-food goods having a zero value). The problem here, however, is that consumers in different countries may choose different combinations of food and other goods, which then require different incomes to meet nutritional requirements. Indeed, the nature of the society and the stage of devel- opment reached may require different combinations. What are regarded as optional extras in some countries may be necessities in others. The UN’s Food and Agriculture Organization defines under- nourishment as ‘food intake that is continuously insufficient to meet dietary energy requirements’. A consumption-based poverty line can therefore be thought of as comprising two elements: an objective measure of the expenditure necessary to buy a minimum level of nutrition; and a subjective additional amount that varies from country to country, reflecting the cost to individ­ uals of participating in the everyday life of society. All this is in theory. In practice, to measure the extent of extreme poverty in the world, the World Bank takes $1.90 a day at PPP at 2011 prices as the cut-off point. This used to be $1.25 a day at 1995 prices (and before that at approximately $1 a day), but the revised poverty line was introduced in 2014 to reflect the rising cost of basic food, clothing and shelter. The real value of $1.90 a day at 2011 prices is the same as $1.25 a day at 2005 prices. Given the poverty line, the simplest way to measure the amount of poverty is by the headcount index, which simply adds up the number of people who fall below the poverty line, which can also be expressed as a proportion of the total population, giving the poverty rate. In 2012, there were just under 900 million, or nearly 13% of humanity, living in extreme poverty. The numbers in poverty and the poverty rates for different areas of the world are shown in Table 2.4. Sub-Saharan  2 THE DEVELOPMENT GAP AND THE MEASUREMENT OF POVERTY 41 Africa has by far the highest incidence of extreme poverty, with 43% of the population living on less than $1.90 a day – a staggering figure. Students reading this book might like to try the experi- ment of living on such a meagre sum, and see how they fare. Paul Collier (2007) once called those living on less than $1 a day the bottom billion. On the new estimates of absolute poverty, the ‘bottom billion’ have become the ‘bottom 900 million’ caught in the four ‘poverty traps’ he identifies: the conflict trap (civil wars); the natural resources trap (the curse of natural resources – see Chapter 3); the trap of being landlocked with bad neigh- bours; and the trap of bad governance in a small country. One weakness of the headcount index, however, is that it ignores the extent to which the poor fall below the poverty line, so that comparisons between countries, or over time, using only the headcount index or the poverty rate, do not tell the full story. To overcome this weakness, the concept of the poverty gap is used and measured. This measures the proportionate gap between the average level of income below the poverty line and the poverty line itself. For example, if the poverty line is $1.90 a day and the average income for the poor below the poverty line is $1.50 a day, then the poverty gap is ($1.90 2 $1.50)/$1.90 5 0.21 or 21%. The poverty gap for different regions of the world is shown in Table 2.5. It can be seen that the poverty gap is by far the highest in sub-Saharan Africa at 16.47%. Table 2.4 Absolute poverty and poverty rates, 1990 and 2012 Global and regional poverty at the poverty line of $1.90 per day (at 2011 PPP) Region Number of poor in millions Poverty rate, % of population 1990 2012 1990 2012 East Asia and Pacific 996 147 60.6 7.2 Europe and Central Asia 9 10 1.9 2.1 Latin America and the 78 34 17.8 5.6 Caribbean Middle East and North 14 6.0 Africa South Asia 575 309 50.6 18.8 Sub-Saharan Africa 288 389 56.8 42.7 World 1,959 897 37.1 12.7 Source: World Bank, 2015. Table 2.5 Poverty gap at $1.90 a day (2011 PPP) (%) 2012 East Asia and Pacific (developing only) 1.47 Europe and Central Asia (developing only) 0.58 Latin America and Caribbean (developing only) 2.64 Low income 18.6 Lower middle income 4,69 Low and middle income 4.35 Middle East and North Africa (developing) Sub-Saharan Africa (developing only) 16.47 Source: World Bank, 2015. 42 DEVELOPMENT AND UNDERDEVELOPMENT The difference for every individual could also be summed and expressed as a proportion of total GDP. This would give the proportion of total income that would have to be redistributed to raise everyone above the poverty line. The focus of the World Bank is now very much on poverty eradication. When Robert McNamara was president of the World Bank in the 1970s, he defined absolute poverty as ‘a con- dition of life so degraded by disease, illiteracy, and malnutrition and squalor, as to deny its victims basic human necessities – [a condition] so limited as to prevent the realisation of the potential of the genes with which one was born’. In May 1992, Lewis Preston, the then president of the World Bank, declared that poverty reduction will be ‘the benchmark by which our performance as a development institution will be measured’. And in the World Development Report 2000/2001, James Wolfensohn, the then president, wrote: ‘Poverty amidst plenty is the world’s greatest chal- lenge. We at the Bank have made it our mission to fight poverty with passion and professionalism, putting it at the centre of all the work that we do’ (World Bank, 2000a). Jim Yong Kim, the current president, has said: ‘I want to eradicate poverty. I think there is a tremendous passion for that in the World Bank.’ As Collier (2007) writes: ‘an impoverished ghetto of 1 billion people [is] increas- ingly impossible for a comfortable world to tolerate’. Meeting the Sustainable Development Goal of poverty reduction To meet the Sustainable Development Goal of reducing extreme poverty to 1% of the world’s population by 2030 will require high sustained growth in the poorest countries. To calculate the growth required to go from the 2015 poverty rate to 1%, the elasticity of the poverty rate with respect to the growth of per capita income needs to be determined. As a rough guide, it is esti- mated by the World Bank (see Ravallion, 2013) that every 1% increase in per capita income leads to a 1.7% reduction in the poverty rate. This calculation, however, also depends on the distribution of income within countries. More equal countries cut poverty further and faster than unequal ones. In the most unequal countries, a 1% increase in income cuts the poverty rate by less than 1%, while in more equal countries a 1% increase in income reduces the poverty rate by as much as 4%. Ravallion (2013) calculates that if developing countries maintain their post-2000 growth performance, the number of extremely poor people will fall from the current level of 900 million to 200 million by 2030, or 3% of the world’s population. This would be a remarkable achievement. To reach the 1% poverty target, however, would require an increase in household consumption of 7.6% per annum – an unrealistically high level. Tackling poverty from the ‘grass roots’ Poverty not only means low income and consumption, and low levels of human development in terms of education and healthcare, but also feelings of powerlessness, vulnerability and fear, because poor people are not free, and are exposed to greater risk, living on the margin of subsistence. What it means to be poor is well illustrated in the World Bank’s study The Voices of the Poor (World Bank, 2000b), which asked 60,000 poor people in 60 countries to articulate their feelings about their physical and mental state. Some of the answers are contained in Case example 2.1, which are both moving and revealing. Feelings of helplessness, humiliation and lack of self-esteem are paramount.  2 THE DEVELOPMENT GAP AND THE MEASUREMENT OF POVERTY 43 Case example 2.1 The voices of the poor Poor people in 60 countries were asked to analyse and share their ideas of well-being (a good experience of life) and ‘ill-being’ (a bad experience of life). Well-being was various- ly described as happiness, harmony, peace, freedom from anxiety, and peace of mind. In Russia, people say ‘well-being is a life free from daily worries about lack of money’. In Bangladesh, ‘to have a life free from anxiety’. In Brazil, ‘not having to go through so many rough spots’. People describe ill-being as lack of material things, bad experiences, and bad feelings about oneself. A group of young men in Jamaica ranks lack of self-con- fidence as the second biggest impact of poverty: ‘Poverty means we don’t believe in self, we hardly travel out of the community – so frustrated, just locked up in a house all day.’ Although the nature of ill-being and poverty varies among locations and people – something that policy responses must take into account – there is a striking common- ality across countries. Not surprisingly, material well-being turns out to be very import­ ant. Lack of food, shelter and clothing is mentioned everywhere as critical. In Kenya, a man says: ‘Don’t ask me what poverty is because you have met it outside my house. Look at the house and count the number of holes. Look at my utensils and the clothes I am wearing. Look at everything and write what you see. What you see is poverty.’ Alongside the material, physical well-being features prominently in the character­ izations of poverty. And the two meld together when lack of food leads to ill health – or when ill health leads to an inability to earn income. People speak about the import­ ance of looking well fed. In Ethiopia, poor people say, ‘We are skinny’, ‘We are deprived and pale’, and speak of a life that ‘makes you older than your age’. Security of income is also closely tied to health. But insecurity extends beyond ill health. Crime and violence are often mentioned by poor people. In Ethiopia, women say, ‘We live hour to hour’, worrying about whether it will rain. An Argentine man says, ‘You have work, and you are fine. If not, you starve. That’s how it is.’ Two social aspects of ill-being and poverty also emerged. For many poor people, well- being means freedom of choice and action and the power to control one’s life. A young woman in Jamaica says that poverty is ‘like living in jail, living in bondage, waiting to be free’. Linked to these feelings are definitions of well-being as social well-being and com- ments on the stigma of poverty. As an old woman in Bulgaria says: ‘To be well means to see your grandchildren happy and well dressed and to know that your children have settled down; to be able to give them food and money whenever they come to see you, and not to ask them for help and money.’ A Somali proverb captures the other side: ‘Prolonged sickness and persistent poverty cause people to hate you.’ The following quotations are an illustration of what living in poverty means: Certainly our farming is little; all the products, things bought from stores, are ex- pensive; it is hard to live, we work and earn little money, buy few things or products; products are scarce, there is no money and we feel poor. (from a discussion group of poor men and women, Ecuador) We face a calamity when my husband falls ill. Our life comes to a halt until he recov- ers and goes back to work. (poor woman, Zawyet Sultan, Egypt) Poverty is humiliation, the sense of being dependent on them, and of being forced to accept rudeness, insults, and indifference when we seek help. (poor woman, Latvia) Source: World Bank, 2000b. 44 DEVELOPMENT AND UNDERDEVELOPMENT The World Bank proposes a three-pronged strategy for poverty reduction: promoting oppor- tunity, facilitating empowerment and enhancing security. Promoting opportunity is partly about expanding economic opportunities for poor people through the process of economic growth, and partly about expanding the asset base of poor peo- ple and increasing the return on those assets. The major causes of individual poverty can be linked to a lack of assets and/or a low return on assets. Important assets to enable people to grow out of poverty include natural assets, such as land; human assets, such as education and health; finan- cial assets, including access to credit, and social assets, such as networks of contacts. The return on assets once acquired depends on the institutional framework of a country, the performance of the economy, and what is happening in the world economy. The state has a role to play in expanding poor people’s assets because markets do not work well for poor people owing to lack of access, power and collateral. The state can help in three major ways: first, by using its power to redistribute resources; second, through institutional reforms to deliver services more effectively, particularly in the fields of health and education; and third, by facilitating the engagement of poor people in programmes that help them to acquire assets, such as land and credit. A growing economy is absolutely crucial for poverty reduction as emphasized by the World Bank’s Commission on Growth and Development headed by the Nobel Prize-winning economist Michael Spence (World Bank, 2008). Poverty cannot be reduced in a stagnant economy. The Commission finds a strong negative association across countries between the average growth of income and consumption and the share of people living on less than $1 per day. A 1 percentage point growth of income below the average is associated with a 2 percentage point increase in the share of people living in poverty. On the other hand, similar rates of growth of countries are associated with different rates of poverty reduction. This is the result of existing inequalities in the distribution of income, assets and access to opportunities. Growth is much more effective in reducing poverty where the income distribution is more equal than where there are big inequalities. The World Bank estimates that when inequality is low, growth reduces poverty by nearly twice as much as when inequality is high. Facilitating empowerment is a new departure in the thinking of the World Bank compared with its 1990 Report. Empowering poor people means strengthening the participation of poor people in decision-making, eliminating various forms of discrimination – ethnic, religious, sexual – and making state institutions more accountable and responsive to poor people. The great chal- lenge here is to tackle the institutional structures of poor countries that continue to marginalize, discrimin­ate against and disenfranchise vulnerable sections of society. The law, the Church, bureau- crats and local elites, and customs and traditions all play a part. The state has a role to play in help- ing to empower people by: 1. Curbing corruption and harassment, and using the power of the state to redistribute resources for actions benefiting the poor. 2. Ensuring that the legal system is fair and accessible to the poor. 3. Making sure that the delivery of local services is not captured by local elites. 4. Encouraging the participation of poor people in the political process. 5. Galvanizing political support for public action against poverty. Enhancing security means reducing poor people’s vulnerability to the various forms of inse- curity that affect people’s lives, such as economic shocks, natural disasters, crop failures, ill health, violence, wars and so on, and helping people to cope with these adverse shocks when they occur. The wide range of risks that poor people are exposed to is highlighted in Case example 2.2. This vulnerability to risk requires a range of insurance mechanisms for managing risk, such as health and  2 THE DEVELOPMENT GAP AND THE MEASUREMENT OF POVERTY 45 Case example 2.2 Poor people’s exposure to risk Poor people are exposed to a wide range of risks. Illness and injury Poor people often live and work in environments that expose them to greater risk of ill- ness or injury, and they have less access to healthcare. Their health risks are strongly con- nected to the availability of food, which is affected by almost all the risks the poor face (natural disasters, wars, harvest failures and food price fluctuations). Communicable diseases are concentrated among the poor, with respiratory infections the leading cause of death. A recent study of poverty in India found that the poor are 4.5 times as likely to contract tuberculosis as the rich and twice as likely to lose a child before the age of two. Illness and injury in the household have direct costs (for prevention, care and cure) and opportunity costs (lost income or schooling while ill). The timing, duration and frequency of illness also affect its impact. A study of South India found that households can compensate for an illness during the slack agricultural season, but illness during the peak season leads to a heavy loss of income, especially on small farms, usually ne- cessitating costly informal borrowing. Old age Many risks are associated with ageing: illness, social isolation, inability to continue working, and uncertainty about whether transfers will provide an adequate living. The incidence of poverty among the elderly varies significantly. In most Latin American countries, the proportion of people in poverty is lower for the elderly than for the population at large. In contrast, in many countries of the former Soviet Union, the inci- dence of poverty is above average among the elderly, particularly among people 75 and older. Women, because of their longer life expectancy, constitute the majority of the elderly, and they tend to be more prone to poverty in old age than men. The number of elderly people in the developing world will increase significantly in coming decades with the rapid demographic transition. Consultations with poor people show that income security is a prime concern of the elderly, followed closely by access to health services, suitable housing and the qual- ity of family and community life. Isolation, loneliness and fear all too often mark old people’s lives. Crime and domestic violence Crime and domestic violence reduce earnings and make it harder to escape poverty. While the rich can hire private security guards and fortify their homes, the poor have few means to protect themselves against crime. In 1992 in São Paulo, Brazil, the murder rate for adolescent males in poor neighbourhoods was 11 times that in wealthier ones. Poor people frequently voice their fear of violence and the resulting powerlessness: ‘I do not know whom to trust, the police or the criminals.’ Crime also hurts poor people indirectly. Children exposed to violence may perform worse in school. A study of urban communities in Ecuador, Hungary, the Philippines and Zambia showed that difficult economic conditions lead to destruction of social capital as involvement in community organizations declines, informal ties among resi- dents weaken, and gang violence, vandalism and crime increase. Violence and crime may thus deprive poor people of two of their best means of reducing vulnerability: human and social capital. continued overleaf 46 DEVELOPMENT AND UNDERDEVELOPMENT Case example 2.2 Poor people’s exposure to risk – continued Unemployment and other labour market risks Labour market risks include unemployment, falling wages, and having to take up pre- carious and low-quality jobs in the informal sector as a result of macroeconomic crises or policy reform. The first workers to be laid off during cutbacks in public sector jobs are usually those with low skills, who then join the ranks of the urban poor; a pattern observed in Africa and Latin America during the structural adjustment reforms of the 1980s and early 1990s. The East Asian crisis also had pronounced effects on labour mar- kets, with real wages and non-agricultural employment falling in all affected countries. As state enterprises in Eastern Europe and the countries of the former Soviet Union were privatized, poverty increased among displaced workers with low education and obsolete skills, not qualified to work in emerging industries. Fluctuations in demand for labour often disproportionately affect women and young workers. Most public sector retrenchment programmes have affected women’s employment more than men’s, and women are more likely than men to work for small firms, which tend to be more sensitive to demand fluctuations. As incomes fall, poor households try to increase their labour market participation, especially for women and children. Harvest failure and food price fluctuations Weather-related uncertainties (mainly rainfall), plant disease and pests create har- vest risk for all farmers, but technologies for reducing such risks (irrigation, pesticides, ­disease-resistant varieties) are less available in poor areas. Between 1994 and 1996, less than 20% of all cropland was irrigated in low- and middle-income countries (only 4% of such land was irrigated in sub-Saharan Africa). Fluctuations in food prices are a related risk. Since poor households spend a large part of their income on food, even small price increases can severely affect food intake. Households that meet their food needs through subsistence agriculture are less vulner- able than households that have to buy all their food. Liberalization of markets often boosts the price of staples – a benefit to small farm- ers if they are net sellers of food. Those hurt are the urban poor and the landless rural poor, as net food buyers, and farmers who engage in seasonal switching, selling food after the harvest when food is plentiful and cheap and buying it when it is scarce and expensive. Where transport facilities are good, traders can step in and equalize prices over the year through arbitrage, but such infrastructure is lacking in many areas. Source: World Bank, 2000a. old-age insurance, unemployment insurance and workforce programmes, social funds and cash transfers, microfinance programmes, insurance against crop failures and price instability and so on. The World Bank points out, however, that promoting opportunities, facilitating empower- ment and enhancing security are necessary conditions for tackling poverty, but not sufficient con- ditions in an interdependent, global economy. International action is also required to help poor people in at least five ways: Promoting global financial stability and reducing the risks of economic crisis. Opening up markets (particularly in developed countries) to the goods of poor countries. Encouraging the production of international public goods that benefit poor people; for exam- ple the control of disease, agricultural research and the dissemination of knowledge.  
t
5 ) &
% & 7 & - 0 1. & / 5
( " 1
" / %
5 ) &
. & " 4 6 3 &. & / 5
0 '
1 0 7 & 3 5 :

 t.PSF
GPSFJHO
BJE
BOE
EFCU
SFMJFG t (JWJOH
B
HSFBUFS
WPJDF
UP
QPPS
DPVOUSJFT
BOE
QFPQMFT
JO
UIF
HMPCBM
GPSVNT
BOE
NVMUJMBUFSBM
JOTUJ- UVUJPOT
PG
UIF
XPSME
TVDI
BT
UIF
8PSME
#BOL
UIF
*.'
BOE
UIF
850 3BOEPNJ[FE
DPOUSPM
USJBMT
3$5T 0OF
PG
UIF
NPTU
JOUFSFTUJOH
BOE
VTFGVM
XBZT
UP
mOE
PVU
XIJDI
JT
UIF
NPTU
FĊFDUJWF
XBZ
UP
mHIU
poverty at the micro-level is to conduct randomized control trials (RCTs), which is the approach UBLFO
CZ
UIF
"CEVM
-BUJG
1PWFSUZ
"DUJPO
-BCPSBUPSZ
TFU
VQ
CZ
"CIJKJU
#BOFSKFF
BOE
&TUIFS
%VnP
BU
UIF
.BTTBDIVTFUUT
*OTUJUVUF
PG
5FDIOPMPHZ
JO

BOE
EFTDSJCFE
JO
UIFJS
CFTUTFMMJOH
CPPL
Poor Economics: A Radical Rethinking of the Way to Fight Global Poverty
 
3$5T
FOBCMF
BOUJ QPWFSUZ
QPMJDJFT
UP
CF
CBTFE
PO
TDJFOUJmD
FWJEFODF
SBUIFS
UIBO
IVODI
PS
DVTUPN
"T
UIF
BVUIPST
BSHVF
AJEFPMPHZ
JHOPSBODF
BOE
JOFSUJB
PGUFO
FYQMBJO
XIZ
HPPEJOUFOUJPOFE

QPMJDJFT
GBJM
3FTFBSDI
PċDFST
MPDBUFE
JO
mWF
PċDFT
BSPVOE
UIF
XPSME
IBWF
DPNQMFUFE
PS
BSF
FOHBHFE
JO
PWFS

FYQFSJNFOUT
JO
NPSF
UIBO
GPSUZ
DPVOUSJFT ʾF
FTTFODF
PG
BO
3$5
JT
UP
UBLF
B
SBOEPN
HSPVQ
PG
QFPQMF
PS
GBNJMJFT
UP
DPOEVDU
BO
FYQFSJ- NFOU
BOE
UIFO
UP
DPNQBSF
UIF
SFTVMUT
PG
UIF
FYQFSJNFOU
PO
UIF
HSPVQ
XJUI
BOPUIFS
SBOEPN
TBN- QMF
PS
DPOUSPM
HSPVQ
OPU
TVCKFDU
UP
UIF
FYQFSJNFOU
BOE
UP
TFF
XIFUIFS
UIF
FYQFSJNFOU
MFBET
UP
B
TJHOJmDBOU
DIBOHF
JO
CFIBWJPVS
PS
PVUDPNF
'PS
FYBNQMF
UIFSF
XBT
B
NJDSPDSFEJU
FYQFSJNFOU
JO
)ZEFSBCBE
*OEJB
PO
XIFUIFS
NJDSPDSFEJU
XPSLT
UP
IFMQ
QPPS
QFPQMF
'JGUZUXP
OFJHICPVSIPPET
XFSF
DIPTFO
BU
SBOEPN
UP
SFDFJWF
NJDSPDSFEJU
IFMQ
BOE

PUIFS
OFJHICPVSIPPET
XFSF
UBLFO
BT
B
DPOUSPM
HSPVQ
"GUFS
FJHIUFFO
NPOUIT
UIFSF
XBT
DMFBS
FWJEFODF
PG
NJDSPDSFEJU
XPSLJOH *O
UIF
mFMET
PG
FEVDBUJPO
BOE
IFBMUI
3$5T
TIPX
UIBU
mOBODJBM
JODFOUJWFT
XPSL
*NNVOJ[BUJPO
DBNQT
BHBJOTU
QSFWFOUBCMF
EJTFBTFT
BSF
WFSZ
TVDDFTTGVM
XIFSF
QFPQMF
BSF
SFXBSEFE
GPS
BUUFOEJOH
'SFF
DIMPSJOF
EJTQFOTFST
OFYU
UP
XBUFS
TPVSDFT
GSFF
XPSNJOH
QJMMT
BOE
GSFF
OVUSJUJPOBM
TVQQMF- NFOUT
BMTP
AXPSL
$POEJUJPOBM
DBTI
USBOTGFST
UP
GBNJMJFT
UIBU
TFOE
DIJMESFO
UP
TDIPPM
IBWF
QSPWFE
FĊFDUJWF
JO
#SB[JM
UIF
#PMTB
'BNJMJB
QSPHSBNNF
BOE
.FYJDP
0QPSUVOJEBEFT 
*O
UIF
mFME
PG
population control, trials show that the provision of family planning services makes very little EJĊFSFODF
UP
GFSUJMJUZ
VOMFTT
UIFSF
JT
B
EFNBOE
GPS
UIFN
BOE
XPNFO
BSF
TFFO
BMPOF
GPS
BEWJDF
5FFOBHF
QSFHOBODJFT
IBWF
CFFO
SFEVDFE
XIFSF
HJSMT
IBWF
CFFO
QSPWJEFE
XJUI
GSFF
TDIPPM
VOJGPSNT
UP
TUBZ
JO
TDIPPM
SBUIFS
UIBO
SPBN
UIF
TUSFFUT 1PPS
QFPQMF
MJWF
TVSSPVOEFE
CZ
IVHF
SJTLT
UP
UIFJS
MJWFMJIPPET
TFF
$BTF
FYBNQMF

BOE
UIF
XBZ
UIFZ
DPQF
CZ
EJWFSTJGZJOH
BDUJWJUJFT
JT
WFSZ
JOFċDJFOU
ʾF
DIBMMFOHF
IFSF
JT
GPS
HPWFSONFOUT
UP
BTTJTU
UIF
EFWFMPQNFOU
PG
JOTVSBODF
NBSLFUT
UISPVHI
FEVDBUJPO
BOE
TVCTJEJ[JOH
JOTVSBODF
QSFNJVNT 0OF
XBZ
GPS
QPPS
QFPQMF
UP
FYQBOE
UIFJS
BTTFU
CBTF
JT
UP
CPSSPX
CVU
GSPN
NPOFZMFOEFST
JU
JT
FYQFOTJWF
BOE
UIF
GPSNBM
CBOLJOH
TZTUFN
JT
OPU
JOUFSFTUFE
JO
MFOEJOH
UP
UIF
QPPS
XJUI
OP
DPM- MBUFSBM
ʾJT
JT
UIF
DIBMMFOHF
PG
UIF
NJDSPDSFEJU
NPWFNFOU
TFF
$IBQUFS
 
ʾF
BMUFSOBUJWF
UP
CPSSPXJOH
GPS
JOWFTUNFOU
JT
QSJPS
TBWJOH
BOE
3$5T
TIPX
UIF
JNQPSUBOU
SPMF
UIBU
NPCJMF
QIPOF
banking can play in encouraging small saving and dealing with small accounts. ʾFTF
BSF
KVTU
TPNF
PG
UIF
mFMET
JO
XIJDI
3$5T
IBWF
UBLFO
QMBDF
8IJMF
UIF
SFTVMUT
DBO
CF
JOUFS- FTUJOH
UIFSF
BSF
MJNJUBUJPOT
PG
TVDI
USJBMT
JO
QBSUJDVMBS
JU
NBZ
CF
EJċDVMU
UP
HFOFSBMJ[F
UIF
SFTVMUT
PG
3$5T
CFDBVTF
UIFZ
BSF
DPOUFYU
TQFDJmD
&WBMVBUJPO
NBZ
CF
DPOEVDUFE
PO
POMZ
POF
TQFDJmD
sample, the trial may be implemented in such a way that it cannot be replicated, and if a specific QSPHSBNNF
JT
JNQMFNFOUFE
B
TMJHIUMZ
EJĊFSFOU
QSPHSBNNF
NBZ
OPU
IBWF
UIF
TBNF
SFTVMUT
"MTP
the evaluation of a trial itself may cause both treatment and comparison groups to alter their CFIBWJPVS
GPS
UIF
QFSJPE
PG
UIF
FYQFSJNFOU
MFBEJOH
UP
GBMTF
JOGFSFODFT 48 DEVELOPMENT AND UNDERDEVELOPMENT Notwithstanding these limitations, and that the trials are generally small scale, RCTs are begin- ning to make a major contribution to our understanding of the causes of poverty, the solution to poverty (what works and what does not), and what types of incentives poor people need to improve their health, value education and escape from the poverty trap. Human Development Index (HDI) To overcome the limitation of taking a single measure of PCY as an index of development and the problems of using PCY as a measure of living standards, the UNDP has developed the Human Development Index (HDI). This index gives a measure of the economic well-being of nations that does not necessarily accord with the usual measure: the level of per capita income. As the UNDP says in its Human Development Report (2014): ‘although GNP growth is absolutely necessary to meet all essential human objectives, countries differ in the way that they translate growth into human development’. The UNDP defines human development as ‘a process of enlarging peoples’ choices’. This depends not only on income but also on other social indicators such as life expect­ ancy, health, education and literacy. Originally, the HDI was based on three variables: life expectancy at birth; educational attain- ment measured by a combination of adult literacy (two-thirds weight) and combined primary, secondary and tertiary school enrolment ratios (one-third); and the standard of living measured by real PCY at PPP. An arithmetic average was taken of the three indices calculated. In 2010, the construction of the index changed. First, some of the variables changed, and second, the method of aggregation changed from an arithmetic average of the indexes to a geometric mean. The vari- ables for the construction of the HDI are: Life expectancy at birth Educational attainment measured as the arithmetic mean of the average years of schooling, and expected years of schooling Gross national income per head at PPP. These four variables are shown in columns 2, 3, 4 and 5 of Table 2.6 for 10 countries with very high human development, 10 countries with high human development, 10 countries with medium human development, and 10 countries with the very lowest human development, and also for regions of the world. To construct the index, fixed minimum and maximum values are taken for each of the variables. For life expectancy at birth, the range is 20–85 years. For average years of schooling, the range is from 0–15 years. For expected years of schooling, the range is 0–18 years. For per capita income at PPP, the range is from $100–$75,000 (taking logs). For any component of the HDI, the individual indexes can be computed according to the general formula: Actual value 2 Minimum value Index 5  (2.1) Maximum value 2 Minimum value Each index thus ranges from zero to 1. If the actual value of the variable is the minimum, the index is zero. If the actual value of the variable is equal to the maximum value, the index is one. Case example 2.3 shows how the index is calculated for Costa Rica. There is not always a close correspondence between the ranking of countries by their HDI and their ranking by per capita income. For example, many of the oil-producing countries, such as Qatar, Kuwait, Angola and Equatorial Guinea have much lower HDI rankings than per capita income rankings because they don’t use their riches for education and health, while countries  2 THE DEVELOPMENT GAP AND THE MEASUREMENT OF POVERTY 49 Table 2.6 Human Development Index and its components, 2013 HDI rank Country Human Life Mean years Expected Gross Development expectancy at of schooling years of national Index 2013 birth (years) (years) schooling income per (years) capita, PPP$ Very high human development 1 Norway 0.944 81.5 12.6 17.6 63,909 2 Australia 0.933 82.5 12.8 19.9 41,524 3 Switzerland 0.917 82.6 12.2 15.7 53,762 4 Netherlands 0.915 81.0 11.9 17.9 42,397 5 United States 0.914 78.9 12.9 16.5 52,308 6 Germany 0.911 80.7 12.9 16.3 43,049 7 New Zealand 0.910 81.1 12.5 19.4 32,569 8 Canada 0.902 81.5 12.3 15.9 41,887 9 Singapore 0.901 82.3 10.2 15.4 72,371 10 Denmark 0.900 79.4 12.1 16.9 42,880 High human development 50 Uruguay 0.790 77.2 8.5 15.5 18,108 51 Bahamas 0.789 75.2 10.9 12.6 21,414 51 Montenegro 0.789 74.8 10.50 15.2 14,710 53 Belarus 0.786 69.9 11.50 15.7 16,403 54 Romania 0.785 73.8 10.7 14.1 17,433 55 Libya 0.784 75.3 7.5 16.1 21,666 56 Oman 0.783 76.6 6.8 13.6 42,191 57 Russian 0.778 68.0 11.7 14.0 22,617 Federation 58 Bulgaria 0.777 73.5 10.6 14.3 15,402 59 Barbados 0.776 75.4 9.4 15.4 13,604 60 Palau 0.775 72.4 12.2 13.7 12.823 Medium human development 103 Maldives 0.698 77.9 5.8 b 12.7 10,074 103 Mongolia 0.698 67.5 8.3 15.0 8,466 103 Turkmenistan 0.698 65.5 9.9 s 12.6 p 11,533 106 Samoa 0.694 73.2 10.3 12.9 t 4,708 107 Palestine, State 0.686 73.2 8.90 13.2 5,168 of 108 Indonesia 0.684 70.8 7.5 12.7 8,970 109 Botswana 0.683 64.4 8.8 11.7 14,792 110 Egypt 0.682 71.2 6.4 13.0 10,400 111 Paraguay 0.676 72.3 7.7 11.9 7,580 112 Gabon 0.674 63.5 7.4 12.3 16,977 continued overleaf  50 DEVELOPMENT AND UNDERDEVELOPMENT Table 2.6 Human Development Index and its components, 2013 – continued HDI rank Country Human Life Mean years Expected Gross Development expectancy at of schooling years of national Index 2013 birth (years) (years) schooling income per (years) capita, PPP$ Lowest human development 178 Mozambique 0.393 50.3 3.2 9.5 1,011 179 Guinea 0.392 56.1 1.6 8.7 1,142 180 Burundi 0.389 54.1 2.7 10.1 749 181 Burkina Faso 0.388 56.3 1.3 7.5 1,602 182 Eritrea 0.381 62.9 3.4 4.1 1,147 183 Sierra Leone 0.374 45.6 2.9 7.5 1,815 184 Chad 0.372 51.2 1.5 7.4 1,622 185 Central African 0.341 50.2 3.5 7.2 588 Republic 186 Congo, 0.338 50.0 3.1 9.7 444 Democratic Republic of the 187 Niger 0.337 58.4 1.4 5.4 873 Regions Arab States 0.682 70.2 6.3 11.8 15,817 East Asia and the 0.703 74.0 74 12.5 10,499 Pacific Europe and 0.738 71.3 9.6 13.6 12,415 Central Asia Latin America and 0.740 74.9 7.9 13.7 13,767 the Caribbean South Asia 0.588 67.2 4.7 11.2 5,195 Sub-Saharan 0.502 56.8 4.8 9.7 3,152 Africa Least developed 0487 61.5 3.9 94 2,126 countries Small island 0.665 70.0 7.5 11.0 9,471 developing states World 0.702 70.8 7.7 12.2 13,723 Source: UNDP, 2014. such as Bangladesh, the Philippines, Ghana, Peru, Cuba and some of the Pacific Islands have high HDI rankings compared with per capita income because they invest in human development– education and health. Multidimensional Poverty Index (MPI) Poverty is not only about lack of income and low levels of education and health. It has many other dimensions. This is recognized by the Multidimensional Poverty Index (MPI), developed by Alkire and Santos (2010) with the Oxford Poverty and Human Development Initiative, and  2 THE DEVELOPMENT GAP AND THE MEASUREMENT OF POVERTY 51 Case example 2.3 Calculating the HDI for Costa Rica Life expectancy at birth in Costa Rica is 79.93 years Mean years of schooling is 8.37 Expected years of schooling is 13.50 Gross national income per capita at PPP is $13,011.7 Now plug these values into the general formula, equation 2.1 above, to derive an index (I) for each of the variables. 79.93 2 20 Health index 5 5 0.922 85 2 20 8.37 2 0 Mean year of schooling index 5 5 0.558 15 2 0 13.50 Expected years of schooling Index 5 5 0.750 18 0.558 1 0.750 Education index 5 5 0.654 2 ln (13,011.7) 2 ln (100) Income index 5 5 0.735 ln (75,000) 2 ln (100) The HDI is the geometric mean of each of the three indexes: HDI 5 (Ihealth 3 Ieducation 3 Iincome)1/3 where the education index is the mean of the two education variables. The HDI for Costa Rica is therefore: