18
Towards A Better Understanding of the
Global Distribution of Income
This paper describes a new methodological approach and some innovative graphical techniques for presenting consolidated information on income distributions. These techniques are seen to be especially useful in describing regional and global distributional patterns and for assessing the number of people who fall below a predefined income or expenditure threshold, in particular in relation to poverty. This work is still in its early stage and thus it has not been possible in the time available to compile detailed country level data on income distributions over comparable periods to provide a comprehensive statistical perspective. The resulting measures are thus referred to as “quasi-exact.” Work is in progress to obtain a better all-embracing measure of income and to define this on an international comparable basis using purchasing power parities (PPPs) and PPP adjusted GNP per capita data.
1. Introduction
There is a substantial and growing body of statistical and circumstantial evidence to show that the global distribution of income is getting worse. Yet, in some sense paradoxically, the latest information available from the World Bank's data base[1], and its Global Economic Projections[2], as well as from the IMF World Economic Outlook[3] (WEO) indicate that for a number of years now, the real net output of the global economy (world GNP) has been growing faster than the annual increase in the world's population. In addition, the developing countries of the World have been growing, on average, almost twice as fast as the rich industrially developed countries. On the surface and in principle, therefore, the basic conditions appear to have been established for a potential reduction in total poverty and in the number of people who are poor. In theory, too, if the tenets of market capitalism are believed to hold, then disparities in income should gradually disappear as economic forces move towards greater equilibrium. According to the dictates of this theory, current differences in well-being should narrow as assets and employment get switched to lower cost centers and resources flow to people with low incomes. Unfortunately, despite this evident capacity to resolve the world's most pressing development agenda priority, progress in this area of poverty reduction appears to move backwards rather than forward. The explanation for this lies partly in institutional features, partly in the data and partly in the economic behavior of those who control the World's productive capital.
Income, as a combination of gross operating surpluses and labor remuneration, is the outward manifestation of the unequal distribution of wealth and the holding of income generating assets. These characteristics are heavily concentrated and unequally distributed around the globe. Income inequality thus tends to expand rather than contract, because the processes of accumulation are associated with attempts to gain increased market control and long-term income protection. These forces thus tend to be self-reinforcing, concentrating economic power rather than making it more re-distributive. Appealing to (any) assumed fundamental spontaneous market forces to bring about the changes desired in the pattern of distribution, may thus prove to be of little avail. More powerful moral imperatives and fairer globally negotiated taxation agreements backed up by stronger international institutional arrangements will be required to alter existing conditions.
In an attempt to provide a more substantive information basis for understanding the problem, this paper proposes a new approach to measuring global inequality. Using partial micro data relating to the incomes and expenditures of households and individuals, drawn from different countries, the paper proposes a simple method to generate a ‘quasi-exact’ perspective as to how the global GNP of all nations taken together is shared between the world’s population. The results, perhaps, do not reveal much that would be surprising to many observers of the development scene. But they have important implications for the overall strategies of the major development institutions. The results are discussed against the background of other more traditional perspectives.
On a conventional national per capita basis, the evidence appears ambiguous. The overall position, however, has probably been getting even worse because the net population growth in most richer countries has been either stable or falling, while that in the poorest regions has still been rising fast. Also the natural rate of growth of people with above average incomes in low income countries has generally been much lower than that of the majority of people who receive income below this level.
2. The Importance of Global Distribution
There has clearly been, for some time, a need to re-examine the overall relationship between growth and inequality as they actually impact on individuals incomes. Development strategists wish to have a general overview of the wider impact of doctrines of “free market capitalism” and their incidence on well-being. The need to analyze whether these approaches give rise, in their various guises, to more economic freedom and capability or lead to the greater entrapment and impoverization of poor people is important. Evidence suggests that economic progress is good primarily for the already well-established and those endowed with wealth and productive assets. Those with very little resources invariably remain mostly untouched by the higher incomes earned by this empowered minority who have already secured a way of life whereby each is able to enjoy an acceptable level of well-being. An approach that generates a “within” as well as” between” dimension to distribution measurement is thus a prerequisite.
3. Global Income
For the purposes of this exercise, global income is taken to be sum of the reported as well as estimated and imputed GNP values of all countries as calculated by the World Bank. Ideally, GNP values in local currencies should be first converted into ‘international’ dollars using 1993 base reference purchasing power parities (PPPs), which are then updated to 1997 by applying actual or implied prices and growth rates. World Bank group aggregation and gap filling procedures are used where information is missing. These techniques are detailed in the reference notes to the 1999 World Development Indicators and are already fairly well known and recognized. The imputation procedures apply, for the most part, to only a few and generally speaking economically small countries. For comparison purposes only, other numeraires and conversion factors—such as the current average annual US dollar exchange rate—and aggregation methods are used. The primary objective is to envision how the ‘global GNP’ is distributed across groups by their income levels, and whether this changes markedly as a result of applying different assumptions.
With this approach, it is also possible to derive various regional (geographical) sub-aggregates of the global distribution and to compare these (a) with other regions; and (b) with the global position. In principle, this can also be done for specific economic ‘blocs’ such as "low income developing countries," "middle income countries," "industrial countries," trading countries, or any other grouping.
No adjustment is made to reported GNP measures for the under-recording of informal and ‘shadow economy’ transactions although, at the country level, these can be significant where the authorities exercise only a weak control over policy management and taxation. In some countries, however, the GNP numbers will usually incorporate some official estimates to account for missing values.
4. The World Population
The World Population as estimated by the United Nations for 1997 and reported in the Bank gives the equivalent ‘global’ population. Because of the nature of its country operations policy, the national total is usually the preferred World Bank number. In this type of analysis, however, the underlying demographics are also important because these have a direct bearing on any given country’s income distribution. This is not only related to the total number of women and children in the population and aged dependents, i.e., differences in direct economic contribution. The assumption is that the greater the number of people there are falling within the normal occupied labor force age group (15-65), the more likely it is the individual income distribution will be less unequal, i.e., it will be affected by the population momentum as well.
5. The Distribution of Global Income
The global income distribution is obtained by mapping the global GNP to the World’s population by way of explicit representative country level income distributions. These are measured initially across households or persons and transformed to an individual basis. No adjustment is made at this stage for "adult equivalence." Income distributions for countries where there are no income data are estimated on the basis of expenditures. Where there is no data income at all, either from administrative (tax) files or household surveys, the distributions are proxied using the available information for ‘near’ countries or groups of countries. Hence, the use of the term ‘quasi-exact’ to depict the shape of the overall distribution where it incorporates countries for which there is little information other than average per capita income in national currency terms or in PPP adjusted values.
One major drawback to this information is that the distributional data often relate to consumption rather than income. In particular, when drawn from household survey results, the outlay data usually refer to actual expenditures rather than consumption. Whichever variable is chosen, this will have a further impact on the basic shape of the curve. The use of reported expenditure figures, for example, tends to condense the format of the actual income distribution curve and compress the range of inequality (see graph 3).
6. Traditional Perspectives of Global Income Distribution
It is informative to trace how, over time, analysts have looked at the distribution of global incomes and why it has been necessary to move to more relevant and refined measurement approaches.
6.1 ‘The Aggregated National’ Approach
Historically, early studies of international inequality were just that; simple comparisons between countries across the globe of national per capita income levels. At first, these were calculated using GNP, population totals and average annual foreign exchange rates (the ‘rf’ principle exchange rate line in the IMF’s International Financial Statistics) to convert GNPs in nominal domestic currencies into a common base currency numeraire such as the US dollar. The World Bank, although applying its own “Atlas” conversion method (see appendix) to even out (over a 3-year period) volatile foreign exchange fluctuations from year to year, still basically relies on this method to distinguish the “poor” from the “middle income” and ‘rich (industrial)” countries. The GNP per capita of a country is used as a prime basis for establishing the Bank’s “Operational Guidelines” for lending and policy support. The 1999 WDI and 1999 World Bank Atlas give details of these rankings. More recently, the Bank has turned to PPP adjusted GNP and per capita income estimates to ensure a more conceptually sound international comparability and to facilitate a consistent aggregation procedure across countries and component GNP (expenditure) categories.
UNDP in both its annual “Human Development Report” (HDR) and in the derivation of its Human Development Index (HDI) now also resort to PPPs, although the text of the HDR makes extensive reference to exchange rate based values. PPP estimates, in equalizing price level differences between countries, provide a quite different perspective on income levels—compressing the range in incomes between rich and poor countries. The assumption of “one price” for each outlay item may conceal the fact, however, that the poor pay more for many of the goods and services they buy (and especially when such prices are adjusted for quality factors). These comparisons suffer from the fact that the “global” view is still dependent on a comparison of the adjusted mean incomes which are specific to each country, even when these are appropriately modified to take account of differences in population size. As is readily recognized, mean income (or even the median income as interpolated from household income studies to a “national”) level[4] fails to provide any information about the potentially wide variations in incomes within any given country. Indeed, in many other respects, GNP or GDP per capita is a very imperfect “locator” of representative average individual income level. Both include imputations for implied incomes that are fixed and generally non-realizable or fungible in any other more liquid form. GNP also includes gross operating surpluses that are usually attributable—but not necessarily disbursable or distributed—to a very limited number of people. Furthermore, GNP incorporates a “depreciation” or “capital consumption” provision that is tied to a specific element that is not disbursed generally.
Nevertheless, an important instance where this approach to global assessments has achieved a high level profile is the latest 1999 UNDP Human Development Report on “Globalization with a Human Face.” Page 2 in the “Overview” section of this report provides, in the very first chart, a picture of the “Stark disparities between rich and poor in global opportunity.” Using exchange rate based numbers, the HDR starts with ‘shares of world GNP’ as measured by the control of global GNP produced by the richest 20% (86%), middle 60% (13%) and poorest 20% (1%) of countries. The Report goes on to add that more than 80 countries (the UN now recognizes 188 member countries and 174 are scaled in the HDI) still have per capita incomes lower than they were a decade or more ago. It emphasizes the fact that, as between countries, inequality has increased. The income gap between the fifth of the world’s people living in the richest countries and the fifth in the poorest was 74 to 1 in 1997, up from 60 to 1 in 1990 and 30 to 1 in 1960. In just over thirty years, the average income gap doubled. The HDR points out that the nineteenth century, too, inequality grew rapidly. During the last three decades of that century, in an era of rapid global integration, the income gap between the top and bottom countries increased from 3 to 1 in 1820 to 7 to 1 in 1870 and 11 to 1 in 1913. These comparisons are confused, however, by exchange rates which have varied considerably across countries over time, and by the Gerschenkron effect that distorts income measures over that period.
The 1999 HDR argues, convincingly, that taking just the past decade alone, there has been an increasing concentration of income, resources and wealth among people, corporations and countries -- rather than any growing convergence as had been widely predicted as a consequence of increasing globalization and freer trade. Instead of improving general well-being, globalization has lead to financial volatility and economic insecurity and may thus pose a greater threat to those who have in fact no responsibility or control for over the state of affairs in which they find themselves. These groups, furthermore, have fewer means to withstand any exposure and vulnerability to such external risks.