Global Income Inequality

Article excerpt

Branko Milanovic looks at changing patterns of global inequality, and different ways of measuring it.

What do we mean by global inequality? How large is it? Is it increasing? Why does it matter? What can be done about it? These are some questions addressed here.

The three ways to look at global inequality

There are three main concepts of global inequality that need to be clearly distinguished, but are often confounded: inequality among countries' per capita incomes - either unweighted (Concept 1) or weighted by countries' populations (Concept 2) - and inequality between world individuals (Concept 3).

Concept 1 deals with convergence and divergence of countries' average incomes, but not with income inequality among all people in the world. Economic growth in a small country will not have the same effect on global welfare as growth in a populous country. Concept 2 takes this into account by weighing each country by its population. However, this implicitly assumes that each individual within a country receives the per capita income of that country. Thus, it does not take into account within-country inequalities. Only Concept 3 includes inequalities both between and within countries.

The reason why Concepts 1 and 2 (the latter often used as a proxy for 'true' inequality between world individuals) have been calculated more frequently is simply that the data requirements are much less onerous. To calculate Concepts 1 or 2 we need to have data only for per capita incomes (often GDP per capita) and population. Thus annual calculations of Concept 1 and Concept 2 inequality across most countries in the world are relatively simple matters. For Concept 3, however, we need to have information on national income distributions. We also need, of course, to convert all the incomes expressed in local currencies into a common yardstick generally represented by international (PPP) dollars. National income distribution figures are obtainable only from household surveys, and surveys often differ in their design, definition, coverage, etc - between years for the same country, even more so between countries - and are not available for all countries annually. In addition, national income distribution data for China, USSR, and most of Africa were unavailable until the mid-1980s. This is the reason why the first detailed and relatively comprehensive (representing more than 90 per cent of the world population) Concept 3 calculations could be done only for the period after around 1985. This is also the reason why such a calculation cannot be executed annually, but only for more or less evenly spaced four or five year intervals.

But once we have these methodological issues straightened out, we enter into what is popularly called 'the tyranny of numbers'. As Figure 1 shows, the evolutions of Concept 1 and Concept 2 inequality have, particularly over the last quarter a century, taken different paths. (All three concepts are measured using the Gini coefficient, a standard measure of inequality that ranges from the theoretical value of 0 - everybody has the same income - to the equally theoretical value of 100 - all income is appropriated by one country or one person.) Since the late 1970s, Concept 1 inequality has been on the rise, and this has finally slowed down only in the last couple of years. The reasons are clear: stagnation and decline of middle-income countries in Latin America, Eastern Europe and the former USSR, and almost relentless declines of mean incomes in most of Africa. At the same time, Western countries have notched up respectable growth rates, just over 2 per cent per capita annually. Consequently, while middle and bottom countries have deteriorated in relative, and often absolute, terms, the top has grown richer. But exactly the opposite evolution is charted by Concept 2 inequality. China and India, since around 1978 and the mid-1990s respectively, have dramatically increased their growth rates; but whereas in Concept 1 calculations they do not count for more than any other two countries, in Concept 2 calculations they account for almost 40 per cent of world population. …