Home / Publications / Interviews / Bob Sutcliffe on Global Inequality

February 23, 2005

---

Whether economic inequality is rising or falling globally is a matter of intense debate. Economist Bob Sutcliffe of the University of Basque Country in Bilbao has been analysing both the statistical details and the broader political-economic import of the debate and shared some of his insights in a recent interview with PERI.

PERI: If someone asked you whether global inequality has grown over the past 25 years, I assume you'd say, “It depends—on how inequality is defined, on what data is used, on how that data is analysed.”  Is that fair?

Bob Sutcliffe: Yes, it's fair but it's not enough. First, the most basic fact about world inequality is that it is monstrously large; that result is inescapable, whatever the method or definition. As to its direction of change in the last 25 years, to some extent there are different answers. But also there are different questions. Inequality is not a simple one-dimensional concept which can be reduced to a single number. Single overall measures of world inequality (where all incomes are in principle taken into account) give a different result from measures of the relation of the extremes of income (the richest compared with the poorest). Over the last 25 years, you find that the bottom half of world income earners seems to have gained something in relation to the top half (so, in this sense, there is less inequality) but the bottom 10% have lost seriously in comparison with the top 10% (thus, more inequality) and the bottom 1% have lost enormously in relation to the top 1% (much more inequality). None of these measures is a single true measure of inequality; they are all part of a complex structure of inequalities, some of which can lessen as part of the same overall process in which others increase.

We do have to be clear about one question of data because it has caused huge confusion. When looking at the distribution of income in the world you have to reduce incomes of different countries to one standard. Traditionally it has been done by using exchange rates; this makes inequality appear to change when exchange rates change, which is misleading. But now we have data based on “purchasing power parity” (the comparative buying power, or real equivalence, of currencies). Using PPP values achieves for comparisons over space what inflation-adjusted index numbers have achieved for comparisons over time. Although many problems remain with PPP values, they are the only way to make coherent comparisons of incomes between countries. But they produce estimates which are astonishingly different from exchange rate-based calculations. For instance, U.S. income per head is 34 times Chinese income per head using exchange rates, but only eight times as great using PPP values. (And, incidentally, on PPP estimates, the total size of the U.S. economy is now only 1.7 times that of China, and is likely to be overtaken by it by 2011.) So when you make this apparently technical choice between two methods of converting one currency to another you come up not only with different figures on income distribution but also with two totally different world economic, and thus political, perspectives.

PERI: So even if some consensus were reached on the choices of definition, data, and method, you're urging a complex, nuanced portrait of what is happening to global inequality rather than a yes or no answer.  Could you give a brief outline of what you think that portrait looks like?

BS: Most integral measures—integral meaning including the entire population rather than comparing the extremes—using PPP figures suggest that overall income distribution at the global level during the last 25 years has shown a slight decline in inequality, though there is some dissent on this. In any event this conclusion is tremendously affected by China, a country with a fifth of world population which has been growing economically at an unprecedented rate. Second, there seems to me little room for debate over the fact that the relative difference between the very rich and the very poor has gotten worse. And the smaller the extreme proportions which you compare, the greater the worsening. So the immensely rich have done especially well in the last 25 years while the extremely poor have done very badly. The top one-tenth of U.S. citizens now receive a total income equal to that of the poorest 2.2 billion citizens in the rest of the world.

There have also been clear trends within some countries. Some of the fastest growing countries have become considerably more unequal. China is an example, along with some other industrializing countries like Thailand. The most economically liberal of the developed countries have also become much more unequal—for instance, the U.S., the U.K., and Australia—and so have the post-communist countries. The most extreme figures for inequality are found in a group of poor countries including Namibia and Botswana in southern Africa and Paraguay and Panama in Latin America.

Finally, the overall index of world inequality (measured by the Gini coefficient) is about the same as that for two infamously unequal countries—South Africa and Brazil. And in the last few years it has shown no signs of improvement whatsoever.

PERI: Can you explain the difference between a unimodal and a bimodal international income distribution? Also, at one point you refer to a possible trimodal distribution—what does that refer to?

BS: The mode of a distribution is its commonest value. In many countries there is one level of income around which a large proportion of the population cluster; at higher or lower levels of income there are progressively fewer people; so the distribution curve rises to a peak and then falls off. That is a unimodal distribution. But in South Africa, for example, due to the continued existence of entrenched ethnic division related to economic inequality, the curve of distribution has two peaks—a low one, the commonest income received by black citizens, and another higher one, the commonest received by whites. This is a bimodal distribution because there are two values which are relatively more common than those above or below them. Because of its origins you could call it the “Apartheid distribution.” The world distribution is in many respects uncannily like that of South Africa. It could be becoming trimodal in the sense that the frequency distribution of income has three peaks—one including those in very poor countries which have not been growing economically (e.g., parts of Africa), one in those developing countries which really have been developing (e.g., in South and East Asia), and one in the high income industrialized countries, a kind of “apartheid plus” form of distribution.

PERI: In 2002, you wrote that many institutional players in this debate were not being exactly honest: for example, emphasizing results based on data or methods that they elsewhere acknowledged to be poor.  Has this changed over the past few years?  Has the quality of the debate over trends in global income inequality improved?

BS: The most egregious pieces of statistical opportunism have declined. But I think there is a strong tendency in general for institutions to seize on optimistic conclusions regarding distribution in order to placate critics of the present world order. This increasingly takes the form of putting too much weight on measures of welfare other than income: for instance, life expectancy, for which there has been more international convergence than in the case of income. Life expectancy often replaces income or is combined with it. But there has been very little discussion of the philosophical basis for this. If poor people live longer but in income terms remain as relatively poor as ever, has the world become less unequal?

The problem of statistical opportunism is not confined to international organizations defending the world economic order; it also exists on the left. So, on the question of inequality, there is a tendency to accept whatever numerical estimate shows greatest inequality on the false assumption that this confirms the wickedness of capitalism. But capitalist inequality is so great that the willful exaggeration of it is not needed as the basis of anti-capitalist propaganda. It is more important for the left to look at the best indicators of the changing state of capitalism, including indicators of inequality, in order to intervene more effectively in the story.

Finally, the quality of the debate, regardless of the intentions of the participants, is still greatly restricted by the shortage of available statistics about inequalities. That has improved somewhat in recent years, although there are many things about past and present inequalities that we shall probably never know.

PERI: Do you see any contexts in which it's more important to focus on absolute poverty levels and trends in those levels rather than on inequality?

BS: The short answer is no, I do not. Plans for minimum income guarantees or for reducing the number of people lacking basic necessities can be important. But poverty always possesses a relative as well as an absolute constituent. It is a major weakness of the Millenium Development Goals, for example, that they talk about halving the number of people in absolute extreme poverty without a single mention of inequality. And there is now a very active campaign on the part of anti-egalitarian, pro-capitalist ideologues in favour of the complete separation of the two. That is wrong not only because inequality is what partly defines poverty but more importantly because inequality and poverty reduction are inseparable. To separate them is to say that redistribution should not form part of the solution to poverty. Everyone is prepared in some sense to regard poverty as undesirable. But egalitarians see riches as equally pathological. The objective of reducing poverty is integrally linked to the objective of greater equality and social justice.

PERI: Can you explain the paradox that China's economic liberalization since the late 1970s has increased inequality within China and at the same time reduced global inequality?  Some researchers and policymakers interpret China's experience over this period as teaching us that it may be necessary for poor countries to sacrifice some equality in order to fight poverty.  Do you agree with this—if not, how would you respond?

BS: Global inequality is not the sum of inequalities in individual countries. In theory all individual countries could become more unequal and yet the world as a whole become more equal and vice versa. In China, a very poor country in 1980, average incomes have risen much faster than the world average and this has reduced world inequality. But different sections of the population have done much better than others so that inequality within China has grown. If and when China becomes on average a richer country than it is now, further unequal growth there may contribute to increasing rather than decreasing world inequality.

China's rapid growth has been of a very inegalitarian kind, but it has been very fast. And the proportion of the population in poverty seems to have been reduced. But it is possible to envisage a more egalitarian growth path which would have been slower in aggregate but which would have reduced the number of poor people at least as much if not more. So I do not think it is right to say that higher inequality is the cause of reduced poverty, though it may for a time be a feature of the rapid growth that, in turn, creates employment and reduces poverty.

This does not mean that all increases in inequality are necessarily pathological. The famous Kuznets Curve sees inequality first rising and then falling during economic growth as an initially poor population moves by stages from low-income, low productivity work into high-income, high productivity work, until at the end of the process 100% of the population is in the second group. If you measure inequality during such a process, it does indeed rise and then fall again to its original level—in this example at the start everyone is equally poor, at the end everyone is equally richer. That might be called transitional inequality; many growth processes may include an element of that. In that case equality is not really being “sacrificed” to reduce poverty—poverty is reduced by a process which increases inequality and then eliminates it again. But at the same time inequality may be growing for many other reasons which are not, like the Kuznets Effect, self-eliminating, but rather cumulative. When inequality grows, this malign variety tends to be more important than the self-eliminating variety. But many economists are far too ready to see growing inequality as the more benign, self-eliminating variety.

PERI: Where do you think the question of what is happening to global income inequality fits into the broader debate over neoliberalism and globalization?

BS: Many people say that some measures of inequality started to improve in about 1980 and that is when neoliberalism and globalization accelerated, so it is those processes which have produced greater equality. There are many things which are wrong with this, among them the fact that at least on some measure global inequality has grown since 1980. In any case, measures that show global inequality falling in this period are, as we have seen, very strongly influenced by China. China's extraordinary growth has, of course, in part been expressed in and permitted by greater globalization (its internationalization has grown faster than its production) and it is also clear that liberalization of economic policy has played a role, though China is hardly a neoliberal economy. But to permit is not to cause. The real cause is surely to be found not so much in economic policy as in a profound social movement in which a new and highly dynamic capitalist class (combined with a supportive authoritarian state) has once again become an agent of massive capitalist accumulation, as seen before in Japan, the U.S., Western Europe. So, an important part of what we are observing in figures that show declining world inequality is not any growth of egalitarianism, but the dynamic ascent of Chinese and other Asian capitalisms.