The Future of Aid – Poverty and the Middle Income Countries

24 September 2013

Andy Sumner

A series of papers since late 2010 has discussed a shift in the location or 'geography' of global poverty. The shift is quite simple: the distribution of global poverty has shifted from countries officially classified by the World Bank as low-income countries towards countries classified as middle-income countries (MICs). This has led to the following stylised fact: three-quarters of the world's extreme poor (US$1.25) live in middle-income countries.

In short, there is a 'new bottom billion' of extreme poor. And these people live not in the absolute poorest countries by per capita income, nor in low-income countries (LICs), nor the least developed countries (LDCs), but in countries—middle-income countries—which many donors treat differently, and consider that a middle-income country classification itself a reason to be reducing or even ending aid.

Of course the world's poor have not 'moved'. What has happened is that the countries where many of the poor live experienced drastically rising average incomes and poverty did not fall as much as one might expect in absolute numbers—especially when China is excluded from estimates.

Of course nothing happens when a country crosses a (somewhat) arbitrary threshold in per capita income but it does matter to traditional donors because not only are those thresholds used in numerous and various ways, the crossing of that arbitrary line is viewed as cause enough for some donors to at least consider ending aid. And higher levels of average per capita income do imply substantially more domestic resources available for poverty reduction and greater access to private capital markets.

The number of aid-dependent countries has fallen to about 30 countries and 10 islands. Eighty developing countries are on a convergence path with OECD average incomes and 130 developing countries get less than 2 per cent of their gross national income from aid. Even among the world's poorest countries a third are 'convergers' with OECD incomes (but of course have a long way to go). By 2030 there could be as few as just 16 low-income countries.

Two competing perspectives on this changing pattern of global poverty are as follows: the first is that the thresholds used to classify countries by the World Bank and extensively used by aid agencies, albeit with other indicators and in various ways, are moribund—meaning they do not represent 'poor' or 'non-poor' countries in any meaningful sense any longer (if they ever did) from the point of view of aid donors.

The second, and by no means necessarily mutually exclusive, is that global poverty is gradually in the process of 'nationalising' at least in terms of resources, meaning the bulk of extreme poverty is in developing countries with rapidly rising average incomes and where resource constraints are less pressing. This is not only because of additional resources produced by economic growth, but also because private capital markets can be accessed, and thus official development assistance (ODA) is becoming of lesser importance over time as domestic resources allocation becomes an ever more significant variable.

On the first point, the country income categories have limited usefulness in identifying countries where most of the world's poor live, and thus for aid donors to focus on those categories runs the risk of divorcing aid from the bulk of world poverty. This is to assume aid ought to be linked to the world's poor. However, one could argue, as some have, that poverty in middle-income countries is no longer an international concern. Indeed, many of the world's extreme poor may already live in countries where the total cost of ending extreme poverty is not prohibitively high as a percentage of GDP and, by 2020, even with fairly conservative estimates, most of the world's poverty may be in countries that do have the domestic financial resources to end at least extreme poverty. However, constraints remain relating to the heterogeneity of new MICs and their economic growth patterns, as well as differing administrative state capacities and constraints of domestic political economy in terms of the taxation base and support for redistributive policies among the emerging but largely insecure new 'middle classes' (many of whom are barely out of poverty themselves).

All of the above would suggest a redefining of categories to emphasise countries with large poverty gaps relative to gross domestic product (GDP) amongst other indicators, and an emerging post-ODA relationship between OECD aid donors and countries with substantial domestic resources—whether they are labelled middle-income countries or categorised in some other way in order to address those points noted above related to the non-resource 'bottlenecks' to ending poverty.

Within a decade or so traditional donors could face a world where, for all but 15-20 countries, concessional lending and policy coherence is more in demand from the vast majority of developing countries than relatively small amounts of ODA (small relative to domestic GDP), where supporting inclusive growth could be increasingly important but a political tightrope, and where traditional donors could even consider co-financing of global or regional public goods with middle-income countries.

Surely, the poor will increasingly live in fragile states as the world moves towards 2030? Maybe. Maybe not. Here's the rub. A significant amount of world poverty could easily remain in stable middle-income countries. Further, poverty in fragile states is increasingly, not in low-income fragile states, but middle-income fragile ones—such as Pakistan and Nigeria. That suggests the cause of poverty in such countries is not solely a lack of resources; it also suggests, counter intuitively, that fragility is not necessarily a barrier to economic growth.

All of this may well imply some significant restructuring of aid ministries in OECD countries. The kind of administrative unit fit for pursuing engagement with MICs is unlikely to be large aid ministries with an existing portfolio of projects, programmes and spending. Rather, one might imagine smaller, cross-governmental administrative units with unequivocal mandates across government, technical capacity, and staffed by those with 'soft skills', meaning strong political sensitivity (in addition to the technical skills already required). Such units may well be more fit-for-purpose to facilitate the shift from spending money on projects and sectors in MICs to cultivating quite new collaborative relationships that require careful negotiation of objectives, co-financing arrangements, policy coherence agreements between parties and working sub-nationally in MICs. Thus 'soft skills' and a premium on political sensitivity and negotiation would be the core skills rather than 'old school' project and programme planning, management and evaluation which will likely only matter in an ever decreasing number of aid-dependent countries in the decades ahead.

Sumner, A. (2013) Global poverty, aid and middle-income countries: Are the country classifications moribund or is global poverty in the process of 'nationalising? WIDER Working Paper No. 62. UNU WIDER: Helsinki.

WIDERAngle newsletter
September 2013
ISSN 1238-9544