Branko Milanovic: How we can strengthen the world’s fragile middle class

Globalisation has been a deeply contradictory process whose features are insufficiently recognised by both those who unquestionably support it and its detractors.

Between 1980 and 2000, countries’ gross domestic product per capita diverged, meaning poor countries grew more slowly than rich ones. This was because of the lost decade (or two) in Latin America and post-communist countries, and disasters and wars in Africa.


But after 2000, things changed: these three areas, and of course Asia, grew faster than the rich world. On the negative side, there was a steady and almost universal increase in income inequalities within nations, with the notable exception of Latin America.

When we look at the global population rather than at countries, however, there is a positive side. The unprecedented growth of China and, from the early 1990s, of India, as well as much of the rest of Asia has lifted millions out of poverty. For the first time since the industrial revolution, income inequality among world citizens has fallen.

So, oddly, the world faces today greater than ever concerns with income inequalities within nations – even as global inequality inches down and global poverty has been halved. But the fall in global poverty and the increase of what is described as the “fragile middle” class are two sides of the same coin. People who moved out of absolute poverty (defined as $1.25 per person per day) have swelled the ranks of those marginally better-off – and even of those who may be, with some effort, classified as a new “global middle class”. This term, however, may suggest a level of comfort not supported by the numbers. The global median income in 2011, calculated from the preliminary household survey data covering about 80 per cent of world population, is $5 per day (compared with a figure of slightly over $4 obtained in 2008). Taking a standard approach, used at the national level, the middle class consists of people with incomes from 25 per cent below the median (in this case, less than $4) to 25 per cent above ($6.5). By this measure, only about 13 per cent of world population qualifies. This is a lower share than in Honduras and Guatemala, not countries that one normally associates with a thriving middle class. So the importance of the global middle class is somewhat exaggerated.

It is also a fragile middle class because it is still relatively poor. Even when we include among the global middle the groups whose real incomes increased (in percentage terms) the most between 1988 and 2008 – that is people with incomes from $3 to $16 per day – it is only those at the upper end of the range who overlap with what is considered the lower middle class in rich countries. All others are poorer.


The global middle class is fragile in the sense that negative shocks (reduction in the number of hours of work, illness in the family, increase in housing prices) can easily push them back below the poverty threshold. A recent World Bank study of the Latin American middle class found that only if income goes over $10 per day does the likelihood of slippage back into poverty become small – and only then may a family be said to have escaped the scourge of poverty.

The existence of the global middle class, however fragile, is a great success of globalisation. But will this success deepen (by making slippages less likely) and broaden (by bringing more people into the global middle class)? This depends foremost on the continuation of high rates of growth in large, still relatively poor, countries. While the attention of the world was focused on Asia, the growing population in Africa means the “battlefield” on which both the reduction in poverty and expansion of the global middle class will be fought is gradually moving there. Overall Asia and Africa have survived remarkably well the worst years of the Great Recession. If their growth rates do not decelerate, we can look fairly confidently to the growth of the global middle class.

But many dangers loom. What is the role of national inequalities? On a purely arithmetic level, if real growth is given, greater inequality slows poverty reduction and probably the expansion of the middle class. But those who believe in trickle-down economics argue that without greater inequality there would not be strong growth. While this might have been true for China in the past 20 years, it is doubtful that further growth in inequality there will be so benign. China’s Gini – a measure of inequality – at about 44 is already greater than America’s. Can it rise further, deepening regional and urban-rural divides, without slowing the expansion of the middle? India’s inequality, long thought to be in the mid-30s Gini range, may if assessed in terms of income rather than consumption already be as high as 50, practically at the Latin American level.

It is therefore growth with redistribution (a familiar development formula from the 1970s) that should be our objective in the years to come, if we want both global poverty and global inequality to continue their downward trend.



Branko Milanovic is visiting presidential professor at the Graduate Center City University of New York and senior scholar at Luxembourg Income Study