The World’s New Middles - Implications for the Future of Development and Aid
Andy Sumner lays out why international aid must continually revisit the tensions between structural transformations and inclusive growth if it is to address the problems facing a global precariat.
The basic storyline of this blog is economic growth since the Cold War. It is outlined how growth has produced two new middles. The first is middle-income countries, and the second is ‘middle people’ who live above the global poverty line—but not too far above. Thinking about contemporary and future development, and development co-operation, this blog identifies five important patterns of growth, precarity, and structural change in developing countries.
The first new middle: Countries no longer stuck at the bottom
The first new middle of countries relates to the burgeoning number of countries no longer stuck at the bottom. Regardless of how one feels about the thresholds, there are now fewer countries ‘stuck’ at the bottom and not growing.
Figure 1 shows the decline in the number of LICs since about the year 2000. The number of LICs started falling drastically over the 2000s, to about 30 LICs today. The number of HICs has doubled from about 40 in 1990 to about 80 in 2013. At this point one could simply dismiss all of this as a set of arbitrary lines, as indeed one could do with the declines in global poverty. However, as in need of review as the LIC/MIC/HIC lines are, they do have symbolic meaning in terms of greater policy freedom in terms of access to non-aid finance in private capital markets (in contrast to donor conditionality). Indeed, they seem to most donors as a simple mechanism for identifying the ‘deserving’ poor and the ‘undeserving’ poor.
There is some logic in the groups though. The remaining LICs have weak recent growth history, suggesting weak growth prospects in the immediate future, and they also face the structural economic handicaps that characterize the least developed country classification—such as literacy rates and an export structure which is in general dominated by primary goods. Seemingly so, the fragility classifications have little diagnostic capacity to determine which countries will and won’t grow in contrast.
The second new middle: People living above extreme poverty but not too much so
The second new middle is that of people. Figure 2 shows the proportions and absolute numbers of the global population living in poverty according to various poverty lines. The proportion of those living under US$1.90 has fallen from 54.8% of the developing world population in 1981 to 14.8% in 2012 (taking ‘raw’ povcal data and not ‘filling’ or adjusting). And the proportion of those living between US$1.90 and US$5 has risen from 24.7% to 40.1% over the same period. The proportion of those living between US$5 and US$10 has increased substantially too, from 11.7% to 23% over the period.
What has happened in the developing world since the Cold War is a large expansion of people into what could be called a lower ‘precariat’ (US$1.90–US$5) in particular across the developing world, and an upper precariat (US$5–US$10) in China, as poverty rates have fallen in many countries with economic growth pushing people across the consumption lines discussed above. One could liken this to a transition in the sense of a transition from poverty to precarity to security. People do not jump out of poverty into secure lives with one big jump but by a slow ascent to higher consumption levels from destitution to poverty to precarity. US$5 is the average poverty line for all countries, and US$10 is a poverty line associated with security from falling back into poverty in the future.
Five patterns of growth, precarity, and structural change in developing countries of importance to thinking about contemporary and future development and development cooperation
1. Poverty and precarity
Different new MICs face different problems in sustaining poverty reduction. In new MICs where growth has not been driven by structural change the bulge of new population is in the lower precariat (US$1.90–US$5), and is thus more vulnerable to growth slowdowns as those people are not far above the poverty line. In contrast, in the countries where growth has been driven by structural change, the bulge of new population is closer too or in the upper precariat (US$5–US$10), and is thus much less vulnerable to growth slowdowns. However, countries attaining growth with structural change have also experienced notable rises in inequality (in capture of the richest of GNI) and this will weaken the responsiveness of poverty reduction to growth ahead at any given poverty line, and countries will need to grow faster to attain the same rates of poverty reduction as the income share of the poorest deciles is squeezed.
2. The seeming premature graduation of a set of poor MICs
Almost twenty countries have had very large increases (meaning doubling or more) in GDP PPP per capita between 1990–2012 and, where there are two data points, in GNI PPP per capita too (GNI PPP per capita is not available for five of the 19 in the earlier period). Henceforth, these could be called ‘genuine’ MICs for this large increase in average output and income. However, the other half of the sample of countries have had much smaller proportional increases in GDP PPP per capita and, where there are two data points, more limited expansion of GDP PPP or GNI PPP per capita (and in three of those countries GNI PPP per capita has actually shrunk). This second set of counties may even be ‘premature MICs’ for this reason. This could simply be (a) that many of the premature MICs are not far over the LIC/MIC threshold; (b) a PPP problem, in that PPP conversion for those countries is based on poor-quality price data; (c) related to the differences between GNI and GDP—output versus income but, more significantly, national versus domestic or, alarmingly, countries have achieved higher income per capita and crossed the line into MICs status by exchange rate conversion but this rise in incomes or output largely evaporates (seemingly) when US$PPP are used.
3. The drastic decline of aid dependency
The good news is that in the new MICs there has been a drastic decline in aid dependency, especially so in the ‘genuine’ MICs but also evident in the ‘premature’ new MICs too. Aid dependency, measured as net ODA/GNI has been falling across both groups. This is to the point of very low levels on average in the genuine new MICs (less than 2% of GNI on average). In contrast, the remaining LICs have a net ODA/GNI ratio of almost 12 per cent. Those new MICs with drastically rising GDP PPP per capita and structural change already have very low ODA/GNI ratios (e.g. Angola, China, India, Indonesia, Nigeria, Pakistan, and Sri Lanka). However, many of the premature new MICs, where GDP PPP per capita has not risen drastically, retain ODA/GNI ratios in the order of 2–10% of GNI, suggesting that, for the premature new MICs at least, moderate ODA may be important for some time. That said, even these ‘poor MICs’ have experienced drastic drops in aid dependency since the early 1990s.
4. Growth with and without structural change
In the genuine new MICs, a shift away from agriculture as a proportion of GDP is very much evident in almost all of the countries. The mean is a virtual halving of agriculture in GDP, value added. Other sectors have expanded, particularly industry, and in some countries that includes manufacturing as well. Interestingly, even the countries with a drastic increase in GDP PPP per capita and structural change retain surprisingly high proportions of the labour force in agriculture. On average, almost 40% of the labour force remains in agriculture as at 2012. In contrast, in most of the premature new MICs there has been less structural change away from agriculture. There is some shift evident in some of the premature new MICs, but on a smaller scale and with much less expansion of industry, including of manufacturing. In fact, on average, manufacturing as a proportion of GDP value added has actually declined.
5. Kuznets’ revenge
Finally, in the ‘genuine’ new MICs, where growth has been accompanied by structural change away from agriculture, the share of the richest decile has increased—call it Kuznets’ revenge—with the exception of Pakistan, where little structural change was noted. At the same time, the GNI share of the poorest 40 per cent has been squeezed and reduced over the period (again with the exception of Pakistan). In the premature new MICs, only four countries have two data points. In those, the share of the richest decile has risen in two countries and fallen in two others. And the share of the poorest has fallen where the share of the richest has risen and vice versa. In short, the Kuznets curve and underlying logic, although largely dismissed, has reappeared in the post-Cold War era in the countries that have experienced growth with structural change. Almost all of the countries experiencing rapid growth with structural change since 1990 have experienced rising inequality (rising GNI share to top decile and falling share to poorest four deciles) from relatively low levels of inequality in the early 1990s. So, while there may be no universal law in the cross-sectional plots, the experience of post-1990 MICs has been that growth and structural change have been accompanied by rising inequality in their time-series data.
So, what does this all mean for the future development and development cooperation?
First, the large numbers of people just above the global poverty line suggest there’s a longer haul for development co-operation beyond the lowest poverty lines. If many MICs can now fund poverty programmes, perhaps there is a role for development co-operation on the precariat in terms of supporting the expansion of insurance mechanisms? Second, the potential premature graduation of some MICs from exchange rate movements, perhaps due to commodity-led growth, may give pause for thought to those advocating the end of aid to only-just-MICs. Third, the growing importance of aid beyond money—meaning agreements on trade or tax havens—may be more important than financial transfers to many developing countries. That said, most MICs still borrow at 10% for ten-year treasury bonds so maybe there is a large role for long-run concessionary finance if, say, the EU borrows at 0.2% on ten-year treasury bonds. Fourth, the lack of more substantial structural change in many growing countries would make Arthur Lewis turn in his grave, and those countries where growth is structural change-driven, rising inequality points towards a need to look again at tensions between the twin goals for development of structural transformation and inclusive growth.
Andy Sumner is Co-Director of King’s International Development Institute and a Reader in International Development at King’s College, London. He is an editorial board member of the Journal of International Development; the European Journal of Development Research and of Global Policy. This post first appeared on UNU-WIDER website.
Photo credit: DonkeyHotey via Foter.com / CC BY