Download Economic Development: Overview and more Slides Development Economics in PDF only on Docsity!
CHAPTER 2
Economic Development: Overview
By the problem of economic development I mean simply the problem of accounting for the observed pattern, across countries and across time, in levels and rates of growth of per capita income. This may seem too narrow a definition, and perhaps it is, but thinking about income patterns will necessarily involve us in thinking about many other aspects of societies too, so I would suggest that we withhold judgement on the scope of this definition until we have a clearer idea of where it leads us. —R. E. Lucas [1988]
[W]e should never lose sight of the ultimate purpose of the exercise, to treat men and women as ends, to improve the human condition, to enlarge people’s choices. [A] unity of interests would exist if there were rigid links between economic production (as measured by income per head) and human development (reflected by human indicators such as life expectancy or literacy, or achievements such as self-respect, not easily measured). But these two sets of indicators are not very closely related. —P. P. Streeten [1994]
2.1 Introduction
Economic development is the primary objective of the majority of the world’s nations. This truth is accepted without controversy, or so it would appear in public discourse at least. Raising the well-being and socioeconomic capabilities of peoples everywhere is easily the most crucial
social task facing us today. Every year, aid is disbursed, investments are undertaken, policies are framed, and elaborate plans hatched to achieve this goal, or at least to get closer to it. How do we identify and track the results of these e↵orts? What criteria do we use to evaluate the extent of “development” a country has undergone or how “developed” or “underdeveloped” a country is at any point in time? How do we measure development?
The issue isn’t easy to resolve. We all have intuitive notions of ”devel- opment.” Presumably, when we speak of a developed society, we have in mind a world in which people are well fed and well clothed, have access to a variety of goods and services, possess the luxury of leisure and entertainment, and live in a healthy environment. We think of a society free of violent discrimination, with tolerable levels of equality, where the sick receive proper medical care and people do not have to sleep on the sidewalks. In short, most of us would insist that a minimal requirement for a “developed” nation is that its physical quality of life be high, uniformly so rather than restricted to an incongruously a✏uent minority.
Of course, the notion of a good society goes further. We might stress political rights and freedoms, intellectual and cultural development, stability of the family, a low crime rate, social civility and so on. However, a high and widely accessible level of material well-being is probably a prerequisite for most other kinds of advancement, quite apart from being a worthy goal in itself.^1 Economists and policy makers therefore do well (and have enough to do!) by concentrating on this aspect alone.
It is, of course, tempting to suggest that the state of material well-being of a nation is captured quite accurately by its per capita gross national income (GNI): the per person value of income earned by the people of a country over a given year. (Or one might invoke its close cousin, gross domestic product, GDP, which restricts itself to domestically produced income, and ignores net income received from other countries, such as dividends, interest or repatriated profits.) Indeed, since economic development at the national level was adopted as a conscious goal, 2 there have been long phases during which development performance was judged exclusively by the yardstick of per capita income growth. In the last few decades, this practice increasingly has come under fire from various quarters. The debate goes on, as the quotations at the beginning of this chapter suggest.
(^1) This is not to suggest at all that it is su cient for every kind of social advancement. (^2) For most poor countries, this starting point was the period immediately following World
War II, when many such countries, previously under colonial rule, gained independence and formed national governments.
little. It may be that per capita income does not capture all aspects of development, but a weighty assertion that no small set of variables ever captures the complex nature of the development process and that there are always other considerations is not very helpful. In this sense, the view that economic development is ultimately fueled by per capita income may be taking things too far, but at least it has the virtue of attempting to reduce a larger set of issues to a smaller set, hopefully in a way that is supported by sound reasoning and empirical evidence.
This book implicitly contains a reduction as well, although not all the way to per capita income alone. In part, sheer considerations of space demand such a reduction. Moreover, we have to begin somewhere, so we concentrate implicitly on understanding two sets of connections throughout this book. One is how average levels of economic attainment influence development. To be sure, this must include an analysis of the forces that, in turn, cause average levels of income and other indicators to grow. The other connection is how the distribution of economic attainment, across the citizens of a nation or a region and across the nations of the world, influences development. The task of understanding these two broad interrelationships takes us on a long journey. In some chapters the relationships may be hidden in the details, but they are always there: levels and distribution as twin beacons to guide our inquiry.^4
This is not to say that the basic features of development will be ignored. Studying them is our primary goal, but our approach to them lies through the two routes described in the previous paragraph.
We begin with a summary of the historical experience of developing countries over the past few decades. We pay attention to per capita income, then to income distribution, as well as other indicators of development. We describe the structural characteristics of developing countries: the occupational distribution of the population, the share of di↵erent sectors (such as agriculture and services) in national income, the composition of imports and exports, and so on.
(^4) Even the double emphasis on levels and distribution might not be enough. For instance,
the Human Development Report (United Nations Development Programme [1995]) informs us that “the purpose of development is to enlarge all human choices, not just income. The concept of human development is much broader than the conventional theories of economic development.” More specifically, Sen [1983] writes: “Supplementing data on GNP per capita by income distributional information is quite inadequate to meet the challenge of development analysis.” There is much truth in these warnings, which are to be put side by side with Streeten and certainly contrasted with Lucas, but I hope to convince you that an understanding of our “narrower” issues will take us quite far.
2.2 Income and growth
2.2.1 Measurement issues. Low per capita incomes are an important feature of economic underdevelopment — perhaps the most important feature — and there is little doubt that the distribution of income across the world’s nations is extraordinarily skewed. Per capita incomes are, of course, expressed in takas, pesos, escudos, remimbi, and in the many other currencies of the world. To facilitate comparison, each country’s income (from all sources, but presumably largely in local currency) is converted into a common currency, typically U.S. dollars, and divided by that country’s population to arrive at a measure of per capita income. This conversion scheme is called the exchange rate method, because it uses the rates of exchange between the local and the common currencies to express incomes in a common unit.
The World Development Indicators (see, e.g., World Bank [2011]) contains such estimates for all countries. Using the yardstick of gross national income, the world produced approximately $59.2 trillion of it in 2009. A bit under $9.3 trillion of this came from all low-income and lower middle-income developing countries^5 Simply put, close to 70% of the world’s population have access to under 16% of world income. Norway, with per-capita income close to $85,000 under this system of measurement, would therefore be over 500 times as rich than the Democratic Republic of Congo, and close to 150 times as rich as Bangladesh.
Figure 2.1 contrasts per capita gross national income in di↵erent countries with the populations of these countries. The figure speaks for itself.
This book is not written with my heart on my sleeve, but the implied disparities are staggering. No amount of fine-tuning in measurement methods can get rid of the stark inequalities that we live with. Nevertheless, both for a better understanding of the degree of international variation that we are talking about and for the sake of more reliable analysis of these figures, it is best to recognize at the outset that these measures provide biased estimates of what is actually out there.
(^5) The 2011 World Development Indicators uses 2009 per capita income to define various
thresholds. Low income countries are those with per capita income under $995. Many African countries fall under this category, as do countries such as Bangladesh, Haiti, Myanmar and Nepal. Low middle-income countries are those that lie between per capita incomes of $996 and $3945; members of this group include China, India, Nicaragua, Nigeria, and Thailand. Upper middle-income countries include several of the richer Latin American economies, such as Argentina and Brazil, and countries such as Lebanon, South Africa and Turkey; they span the range between $3946 and $12195. The upper-income countries make up the rest.
(2) A more serious discrepancy arises from the fact that prices for many goods in all countries are not appropriately reflected in exchange rates. This is only natural for goods and services that are not internationally traded. Exchange rates are just prices, and the levels of these prices depends only on commodities (including capital) that cross international borders. The prices of nontraded goods, such as infrastructure and many services, do not a↵ect exchange rates. What is interesting is that there is a systematic way in which these nontraded prices are related to the level of development. Because poor countries are poor, you would expect them to have relatively low prices for nontraded goods: their lower real incomes do not su ce to pull these prices up to international levels. However, this same logic suggests that a conversion of all incomes to U.S. dollars using exchange rates underestimates the real incomes of poorer countries. This can be corrected to some extent, and indeed in some data sets it has been. The most widely used of these is the Heston–Summers data set (see box). Recently, the World Bank started to publish income data in this revised format.
PPP: The International Comparison Program. According to GDP esti- mates calculated on an exchange-rate basis, Asia’s weight in world output fell from 7.9% in 1985 to 7.2% in 1990—and yet Asia was by far the fastest growing region during this period^7. This same period also witnessed a sharp decline in some Asian countries’ exchange rates against the dollar. Now does that tell us something about the shortcomings of GDP exchange- rate estimates?
Actually, the trouble with market exchange rates for income calculations is not so much that they fluctuate, but that they do not fluctuate around the “right” average price, if “right” is to be measured by purchasing power. Even if exchange rates equalize the prices of internationally traded goods over time, substantial di↵erences remain in the prices of nontraded goods and services such as housing and domestic transportation. There is a simple reason for this: because developing countries have relatively low incomes, you would expect non-traded goods to be cheaper. By assigning international prices to a basket of goods and then estimating incomes relative to those prices, we’re carrying out a comparison that maintains “purchasing power parity”; hence the term “PPP incomes”.
The International Comparisons Program (ICP) began as a research project in 1968. Initially funded by the World Bank and the Ford Foundation,
(^7) See The Economist, May 15, 1993.
the ICP carried out comparison programs for 1970, 1973 and 1975 before its absorption as a regular component into the United Nations Statistical Division (UNSD). The UNSD coordinated regional comparisons carried out by regional commissions. Research on the conceptual basis of such com- parisons was initiated and continued by two economists at the University of Pennsylvania, Alan Heston and Robert Summers, both connected with the ICP since its inception. Their e↵orts resulted in the Penn World Tables (PWT; also called the Heston–Summers data set). It consists of a set of national accounts for a large set of countries dating from 1950. Entries are denominated in a set of ”international” prices in a common currency, which drew on the ICP. Hence, international comparisons of income can be made both between countries and over time.^8
Likewise, the ICP continues to provide price and expenditure comparison for an ever-expanding set of countries, further rounds of comparisons being completed in 1980, 1985, 1993 and 2005. The latest round (at the time of writing) is 2011, coordinated by the World Bank for 180 countries.
The first step of the ICP is simple: to compare prices for a well-defined commodity in (say) Bangladesh relative to the US. If it costs $2 to buy a kilo of potatoes in the United States but 100 taka in Bangladesh, then the “potato exchange rate” between the taka and the dollar is 50. Indeed, life at the ICP would be blissful if all we consumed were the lowly potato, but of course, there are thousands of items, all with their own prices. Some of these are traded, and the overall exchange rate as we know it comes from these prices. Many items are not traded. The di culties lie in aggregating from the lowest commodity heading (e.g., potato) all the way back up to overall income.
In particular, the ICP must gather detailed data on prices for thousands of items. These items are then classified numerous expenditure categories (broadly under the consumption, investment and government expenditure headings). By an averaging procedure following a method suggested by statistician R. C. Geary, the average price for each category is obtained relative to the price for that category in the U.S. In this way, numerous relative prices (or “price parities”) are made available for each country and each category. Finally, categories must be aggregated up to national income.
(^8) The PWT are housed at the Center for International Comparisons at the University of
Pennsylvania; see http://pwt.econ.upenn.edu/. Apart from income data, the PWT also o↵ers data on selected countries’ capital stocks and demographic statistics. In the revised GDP calculations based on PPP, Asia’s share in world output in 1990 jumped from 7 to 18%.
say, in dollars) for each such good and service over all di↵erent countries. National income for a country is then estimated by valuing its outputs at these international prices. In this way, what is maintained, in some average sense, is parity in the purchasing power among di↵erent countries. Thus we call such estimates PPP estimates, where PPP stands for “purchasing power parity.”
PPP estimates of per capita income go some way toward reducing the astonishing disparities in the world distribution of income, but certainly not all the way. For an account of how the PPP estimates alter the distribution of world income, consult Figure ?? , which compares PPP and exchange-rate calculations of per-capita GNI. The setting is just the same as in Figure 2.1, with countries ranked in ascending order of exchange-rate GNI per-capita. The new, darker line shows the PPP computations, which I’ve normalized for you by setting the world average PPP income (weighted by population) equal to its exchange rate counterpart. As expected, the PPP numbers are relatively higher for poorer countries. The main reason for this is that their nontraded prices are lower, as we’ve discussed.
!"
#!"
$!"
%!"
&!"
'!"
!"
#!"
$!"
%!"
&!"
'!"
#" &" (" #!" #%" #)" #" $$" $'" $+" %#" %&" %(" &!" &%" &)" &" '$" ''" '+" )#" )&" )(" (!" (%" ()" (*" +$" +'" ++" *#" &" (" #!!" #!%" #!)" #!" ##$" ##'" ##+" #$#" #$&" #$(" #%!" #%%" #%)" #%" #&$"
""
,-."/01230"24/56-7"$!!*"8-9/:04;-".03-"04<",,,=""
?@"1-."/01230"8,,,=" ?@"1-."/01230"8AB="
Figure 2.2. PPP versus exchange-rate GDP per capita, 2009.
(3) There are other subtle problems of measurement. Income measure- ment, even when it accounts for the exchange-rate problem, uses market prices to compare apples and oranges; that is, to convert highly disparate goods into a common currency. The theoretical justification for this is that market prices truly reflect preferences as well as relative scarcities. There are several objections to this argument. Not all markets are perfectly competitive; neither are all prices fully flexible. We have monopolies, oligopolistic competition, and public sector companies^10 that sell at dictated prices. There is expenditure by the government on bureaucracy, on the military, or on space research, whose monetary value may not reflect the true value of these services to the citizens. Moreover, conventional measures of GNP ignore costs that arise from externalities — the cost of associated pollution, environmental damage, resource depletion, human su↵ering due to displacement caused by “development projects” such as dams and railways, and so forth. In all of these cases, the going prices do not capture the true marginal social value or cost of a good or a service.
All these problems can be mended, in principle, and sophisticated measures of national income do so to a large extent. Distortions in prices can be corrected for by imputing and using appropriate “shadow prices” that capture true marginal values and costs. There is a vast literature, both theoretical and empirical, that deals with the concepts and techniques needed to calculate shadow prices for commodities. An estimated “cost of pollution” is often deducted in some of the measures of net national income, at least in industrialized economies. Nevertheless, it is important to be aware of these additional problems.
With this said, let us turn to a brief account of recent historical experience.
2.2.2 Historical Experience. Over the period 1980–2010, the richest 10% of the world’s nations averaged a per capita income (PPP) that was a bit over 4 times the world average income, while the poorest 10% had about 6–10% of the world average income, this latter ratio showing a small but significant tendency to deteriorate over the 1990s; see Table 1, which shows the same trend for both per-capita GDP and GNI. This broad constancy of extreme inequality among the very richest and poorest countries has been maintained since 1960, with some deterioration at the close of the 20th century. As Parente and Prescott (2000) quite correctly observe, interstate disparities within the United States do not even come close to these international figures. In 2010, the richest state in the United States
(^10) In many countries all over the Third World, sectors that are important or require bulk
investment, such as iron and steel, cement, railways, and petroleum, are often in the hands of public sector enterprises.
and in many cases there was no growth at all. Over 1980–1990, during the so called “lost decade” for Latin America, per-capita income for the region declined by an average of over 0.7% year over year, leading to an overall decline of around 10%.^13 Down went Argentina (-2.9% annualized over 1980– 1990), Brazil (-0.5%), Mexico (-0.3%), Peru (-3.0%), Uruguay (-0.7%) and their neighbors. Only Chile (2.1%) and Colombia (1.4%) had a significantly higher per capita income in 1990 than they did in 1980. It is certainly true that such figures should be treated cautiously, given the extreme problems of accurate GNP measurement in high-inflation countries, but they illustrate the situation well enough. With some notable exceptions (such as Chile, 4.7%, and Argentina, 3.6%), growth in incomes continued to be slow in the 1990s; around the world average at a bit less than 1.6%. We see a broader recovery over 2000–2010, with average growth rates well in excess of 2% annually; take note of Argentina (3.3%), Brazil (2.4%), Chile (2.6%), Peru (4.3%) and Uruguay (3.0%), though one of Latin America’s largest economies, Mexico, has not fared quite so well (0.8%).
Similarly, much of Africa stagnated or declined over the 1980s. Sub-Saharan Africa as a whole declined at an annual per-capita rate of over 1%, and things were not any better in the 1990s (-0.4%), though the first decade of this century has seen a relative improvement (2.2%). Countries such as Nigeria (-1.6%) and Tanzania (-2.0% 14 ) experienced substantial declines of per capita income through the 1980s, and essentially stagnated through the 1990s, before pulling back to a more robust recovery over 2000–2010 (3.9% annual for Nigeria, 4.0% for Tanzania). Kenya barely grew in per capita terms in the 1980s, and continued to decline in the 1990s before recovering to some extent in 2000–2010; overall conditions (0.2%) over this thirty-year period are near-stagnant. Uganda stagnated over the 1980s (-0.1%) before picking up pace and making substantial progress over 1990–2010, growing at over 3.5% annually. Another notable turnaround is Rwanda, crippled by negative growth in the 1980s (-1.2%) and 1990s (-0.7%) before a remarkable recovery over 2000–2010 (4.8%). Yet Burundi’s negative growth rate of 3.2% in the 1990s is barely compensated for by its near-stagnation over 2000– (0.4%). One of the largest countries in Africa, the Democratic Republic of the Congo, went into veritable freefall over the 1980s (-2.2%) and the 1990s (-8.5% annually) before its less surreal advance of around 1.8% annually over 2000–2010. And what of Zimbabwe, a country that stagnated in the 1980s (0.7%) and 1990s (-0.3%) before entering a freefall of its own (-4.8%) over 2000–2010?
(^13) See Morley (1995) on the fortunes of Latin America over this period. (^14) See Maliyamkono and Bagachwa (1990).
We began this account with the world rate of growth (roughly 1.5% annual over 1970–2010), and we end with another benchmark — the growth experience of country-members of the Organization for Economic Cooperation and Development (OECD). The twenty original members and the fourteen additions since contain all the developed countries, though a few middle-income countries are also members. Over 1970–1990, OECD growth was a bit over 2.4% annual, before falling to a more sedate 1.8% over the 1990s (but still a bit more than the world average over this period) and then under the world average at 0.8% during 2000–2010, the slowdown essentially an outcome of the “great recession” of 2009. Finally, the United States mirrors the OECD reasonably well, growing at over 2.2% over 1970– 1990, falling slightly to a bit under 2.2% in 1990–2000, before slowing to 0.7% in 2000–2010. Its overall growth over 1970–2010 is uncannily close to the world average; see below.
This discussion isn’t a comprehensive accounting of growth, but you can see that there’s been a lot of churning in the international distribution of incomes. Such diversity demands explanation, but this demand is ambitious. No single story can account for the variety of historical experience. We know that in Latin America, the sovereign debt crisis triggered enormous economic hardship in the 1980s. In sub-Saharan Africa, low or negative growth is due in large measure to unstable government, civil strife and consequent infrastructural breakdown, as well as to high rates of population increase. The heady successes of East Asia are not fully understood, but a conjunction of farsighted government intervention, a relatively equal domestic income distribution and a vigorous entry into international markets played an important role. We will take up these topics, and many others, in the chapters to come.
Growth experiences such as these can change the face of the world in a couple of decades. One easy way to see this is to look at the ”doubling time” implicit in a given rate of growth; that is, the number of years it takes for income to double if it is growing at some given rate. The calculation in the footnote^15 reveals that a good approximation to the doubling time is seventy divided by the annual rate of growth expressed in percentage terms. Thus an East Asian country growing at 5% per year will double its per capita income every fourteen years! In contrast, a country growing at 1% per year will require seventy years. Percentage growth figures look like small numbers, but over time, they add up very fast indeed.
(^15) A dollar invested at r% per year will grow to two dollars in T years, where T solves the
equation [1 + (r/100)]T^ = 2. This means that T ln (^) e [1 + (r/100)] = lne 2. However, ln (^) e 2 is approximately 0.7, whereas for small values of x, ln (^) e (1 + x) is approximately x. Using this in the equation gets you the result.
Obs Cat ¿ ¡ ¬ √ ƒ 32 ¿ 84 13 3 0 0 21 ¡ 43 43 14 0 0 26 ¬ 0 27 50 23 0 20 √ 0 0 20 70 10 29 ƒ 0 0 0 3 97
Figure 2.4. The Income Mobility of 128 Countries, 1982–2009.
rough symmetry between changes upward and changes downward, which partly accounts for the fact that you don’t see much movement in the world distribution taken as a whole. (As the previous edition of my book shows, the same remarks are also true of 1960–1985.) This observation is cause for much hope and some trepidation: the former, because it tells us that there are probably no traps to ultimate economic success, and the latter, because it seems all too easy to slip and fall in the process. Economic development is probably more like a treacherous road, than a divided highway where only the privileged minority is destined to ever drive the fast lane.
This last statement must be taken with some caution. Although there appears to be no evidence that very poor countries are doomed to eternal poverty, there is some indication that low incomes are very sticky. Even though we will have more to say about the hypothesis of ultimate conver- gence of all countries to a common standard of living (see Chapters 3–5), an illustration may be useful at this stage. Following Quah (1993), we can use per capita income data to construct “mobility matrices” for countries. First convert all per capita incomes to fractions of the world’s per capita income. Thus, if a country has per capita income of $1,000 and the world average is $3,000, assign the number 1/3. Now create categories that we will put each country into. Quah used the following categories (you can certainly construct others if you like): income less than a quarter of the world average (1), income between a quarter and half the world average (2), income between half the world average and the world average (3), income between the world average and twice the world average (4), and income exceeding twice the world average (5).
Now imagine doing this exercise for two points in time, with a view to finding out if a country transited from one category to another during this period. You will generate what we might call a mobility matrix. Figure ?? illustrates such a matrix for the period 1982–2009, using per-capita GDP data (a very similar observation holds for GNI). The numbered rows house countries in the corresponding categories in 1982; the italicized numbers
tell you how many in each; e.g., in 1982, 32 countries in the 128-country set had per-capita GDP less than a quarter of the world average. The columns house the corresponding categories in 2009. Thus a cell of this matrix defines a pair of categories. Each entry in the cell records the percentage of countries that made the transition from the row category (1982) to the column category (2009). For instance, 13% of the countries in category 1 made it to category 2 over the twenty-seven year period. A matrix constructed in this way gives you a fairly good sense of how much mobility there is across nations. A matrix with very high numbers on the main diagonal, consisting of those special cells with the same row and column categories, indicates low mobility. According to such a matrix, countries that start o↵ in a particular category have a high probability of staying right there. Conversely, a matrix that has the same numbers in every entry (which must be 20 in our 5 ⇥ 5 case, given that the numbers must sum to 100 along each row) shows an extraordinarily high rate of mobility. Regardless of the starting point in 1982, such a matrix will give you equal odds of being in any of the categories in 2009.
With these observations in mind, continue to stare at Figure ??. Notice that middle-income countries have far greater mobility than either the poorest or the richest countries. For instance, only half of all the countries in category 3 (between half the world average and the world average) remained where they were in 1962. In stark contrast to this, fully 84% of the poorest countries (category 1) remained where they were in 1962, and none of them made it over the world average by 2009. Likewise, 97% of the richest countries
in 1982 stayed right where they were in 2009.^17 This is interesting because it suggests that although everything is possible (in principle), a history of underdevelopment or extreme poverty puts countries at a tremendous disadvantage.
This finding may seem trite. Poverty should feed on itself and so should wealth, but on reflection you will see that this is really not so. There are certainly many reasons to think that historically low levels of income may be advantageous to rapid growth. New technologies are available from the more developed countries. The capital stock is low relative to labor in poor countries, so the marginal product of capital could well be high. One has, to some extent, the benefit of hindsight: it is possible to study the success stories and avoid policies that led to failures in the past. This account is not meant to suggest that the preceding empirical finding is inexplicable: it’s
(^17) Of course, our categories are quite coarse and this is not meant to suggest that there were
no relative changes at all among these countries. The immobility being described is of a very broad kind, to be sure.
!"#
$%#
$"#
%%#
%"#
&%#
&"#
'%#
'"#
(%#
)# ')))# !))))# !')))# $))))# $')))# %))))# %')))# &))))# &')))#
!"#$"%&'(")%$+,")-.'#"/)*
01!)2"#)$'23&')'#+4%5)6777)
*+,-.#/01#$)2# *+,-.#30405#&)2#
(a) The full range of per-capita GDP
!"
#!"
$!"
%!"
&!"
'!"
(" $(((" &(((" !(((" )(((" #((((" #$(((" #&((("
!"#$%&'()%+&$,-%./(#%0***
12!3%#$(34'((#,5&67888***
*+,-."/01"$(2" *+,-."30405"&(2"
(b) Per-capita GDP below $15,
Figure 2.5. Income shares of poorest 40% and richest 20% in various countries. Source: The World Bank; http://databank.worldbank.org.
observation that history matters in maintaining persistent di↵erences needs more of a justification than might be obvious at first glance.
2.3 Income distribution in developing countries
The international disparity of national income is only one indication that something is fundamentally askew with global development. Add to this the astonishing inequalities observable within each of the vast majority of developing countries. It is commonplace to see enormous wealth coexisting with great poverty, and nowhere is this more evident than on the streets of Bombay, Rio de Janeiro, Manila, Mexico City, and the other great urban conglomerates of the developing world. It isn’t that such inequalities do not exist in the developed world—they certainly do—but coupled with the low average income of developing countries, these disparities result in an outcome of visible poverty and destitution.
We will have much more to say on the topic of income distribution later in this book. As an overview, however, it is useful to get a feel for the magnitude of the problem by looking at some data.^18 Figure 2.6 summarizes recent information on inequality for over 90 countries, arranged in order
(^18) One can imagine that the statistical problems here are even more severe than those involved in measuring per capita income. The goal is to measure the incomes earned by di↵erent groups in the same country and compare them, so all the measurement di culties are compounded (except for the problem of international price comparability), because no
of increasing per-capita GDP. The bulk of the data comes from 2000, but in the interests of coverage I have added several more observations from 1998–2002. The figure records the income share of the poorest 40% of the population as well as the income share of the richest 20% of the population.
Look first at Panel A of Figure 2.6. Observe that both the share of the poorest 40% and the richest 20% exhibit a distinct transition around the $15, mark. Before this threshold, the poorest 40% of the population earn, on average, around 15%—perhaps less—of overall income, whereas the richest 20% earn well over half of total income. Then (for the richer countries) the former share rises to around 20%, while the latter falls to about 40%. These — especially the shares below the $15,000 threshold — are sizable inequalities. When compounded with the inter-country di↵erences that we’ve already discussed, it is no surprise that the poor in the developing world are twice cursed.
One would be tempted to conclude from Panel A that inequality unam- biguously falls in the course of development. But such a conclusion would be premature for several reasons. To begin with (as you will be warned over and over again), variation in some variable across countries is not necessarily to be equated with a change in that variable as a particular country develops over time. In addition, there are arguments that suggest more complex variations in inequality over the course of development, including the famous “inverted-U hypothesis” of Simon Kuznets: inequality first rises and then falls over the course of development.
Here is a quick summary of the Kuznets argument. At very low average levels of living, it is di cult to squeeze the income share of the poorest 40% below a certain minimum. For such countries the income share of the rich, although high, is not close to the extraordinarily high ratios observed in middle-income countries. Flipping this argument around, it is possible that as economic growth occurs, it initially benefits the richest groups in society more than proportionately. This situation is reflected in a rise in the income share of the upper segments of the population. The share of the poorest groups tends to fall at the same time, although this does not necessarily mean that their income goes down in absolute terms. At higher levels of per capita income, economic gains tend to be distributed more equally—the poorest groups now catch up in income share. The overall picture follows, then, an “inverted U”.
As both Table 2 and Panel B of Figure 2.6 appear to suggest, there is some truth to the story. In Panel B, we’ve truncated the data of Panel A at
system of overall, national accounts can be used to estimate the incomes of any one subgroup of the population.
