This is a guest post from Andy Sumner, Co-Director of the King’s International Development Institute at King’s College London.

GLOBAL consumption grew by $10 trillion between the end of the Cold War and 2010. The $10 trillion questions are who benefited and by how much? Some of the answers may come as a surprise. A global middle class may be emerging but it might not be breaking down the old distinctions between rich and poor as much as some people think.

Start with the basic facts about inequality. In a new study published by the Centre for Global Development, a think tank in Washington DC, we show that income inequality rose slightly from the late 1980s to 2005 and was relatively flat after that. This is for inequality within countries.

If you look at inequality between countries — which measures whether the poorest ones are catching up — you find this fell slightly up to 2005 and then fell more quickly afterwards.

However, these trends are largely attributable to rising prosperity in China. Take out China and the picture looks rather different.

Throughout the entire period, inequality within countries in the rest of the world has been constant overall because as some countries have become less equal, others have become more so. And inequality between countries has risen, pushing up global inequality overall.

However, all of this is based on a dominant measure of inequality – the Gini index – which is increasingly being questioned (see here). The problem with the Gini is that drawing conclusions from the commonly used aggregate inequality measures makes it difficult to take a nuanced view of how global growth interacts with changing national and international inequality.

In the light of these reservations we propose an alternative approach based on four consumption layers. They are the ‘global absolute poor’ ($0-$2 per person per day); the ‘global insecure’ ($2-$10 a day); the ‘global secure’ ($10-$50 a day); and the ‘global prosperous’ (above $50 a day).

Identifying such benchmark thresholds is inevitably a rough and ready exercise. But we can derive them in relation to patterns of actual consumption and the use of the average poverty line for developing countries of $2/day. A $10/day line makes for a rough approximation for ‘security from poverty’ (see here).

We can then consider how each layer of global society has fared since the end of the Cold War. About 15% of the growth in global consumption from 1990 to 2010 went to the richest 1% of the world’s population (those living on more than $75/day). At the other end of the scale, the bottom half of the world population who were the $2 poor in 1990 got less than 12% of global growth. Chart 3 shows that in terms of the number of people in the world, the largest number is bunched around the poverty line, that is, between $1.25 a day and $2 a day.

But chart 4 paints a different picture. It looks at who consumes the most globally; here, the largest number is based around the $50 a day a mark (ie, the relatively few people who earn $50 a day consume far more than the $2 a day group, even though there are far more of the latter.

Over time, the number of poor people escaping poverty is growing (this is shown in chart 3; the 2010 line is to the left of the 1990 line). But the rich are also doing better. Chart 4 shows that the amount of global consumption attributable to those who are richer is also growing; the 2010 line is also to the right (and steeper) than the 1990 line.

This shows a rather different pattern from the one described in the recent Dollar, Kleineberg and Kraay study, which found that growth (still) is good for the poor in a dataset covering 118 developing countries (this paper revisited their earlier studies on the subject). While it is true that growth (still) is good for the poor in a general sense, patterns of growth can differ substantially between countries. And the poor are likely to do better from growth when inequality is low or falling rather than high or rising. It is also worth considering the picture with and without China, which makes a huge difference to the pattern of growth observed and who benefits and how much.

What is most sobering is that, yes, growth is good for the poor. But we find a surprisingly modest amount of redistribution could have ended $2 poverty already. Our calculations show that it would take 0.2% of world GDP to end global poverty of $1.25 a day; ending global poverty of $2 a day would take 1% of world GDP. Ending $10-a-day poverty, though, would be hugely expensive, taking 20% of world GDP. The proposed UN goal of eradicating poverty only aims at ending $1.25 poverty by 2030.

Importantly not all growth is the same. In fact it can be quite different. We find five stylised patterns of growth that can be compared to the global average.

There is growth that helps the poorer end of the distribution most (as in Ethiopia, South Africa, Malawi and Mali); growth that helps the middle most (eg, Brazil); growth that – startlingly – hurts the poor and/or the middle (Nigeria and Zambia) and finally, growth that is equitable across the income spectrum (such as Vietnam, Nepal and the Philippines).

At a national level, therefore, the question is one of the governance of growth – meaning different countries can manage or govern growth very differently and this, along with other factors, plays a big role in who benefits from growth and how much.

So what about the new global middle class, also called the ‘buoyant billions’? We wonder if they are quite as ‘middle class’ as their name suggests given, first, that many of them are actually found in the layer of the global insecure, and second, given the doubts about whether many of them exist beyond a few large countries, notably China.

We identify a persistent global structure of two relatively homogeneous clusters, the poor or insecure and the prosperous or secure – a dumb-bell shape. True, we also detect the emergence of a rapidly changing and heterogeneous new global middle class. So does that mean the old dumb-bell world of two clusters is breaking down? We are not convinced that it is. It turns out that most of the change in the distribution of global consumption is confined to upper middle-income countries, those with an income per head of between approximately $4,000 and $12,000 a year; the category includes China and Brazil for example and soon Indonesia may join, too.

This leads us to suspect that the emerging global middle may not represent a move away from the old two-cluster world but rather a transitional phase, in which some groups in richer emerging markets are shifting out of the poor/insecure cluster into the secure/prosperous one.