The high share of aid provided to relatively better-off middle income countries is the biggest challenge to the aid record of the European Commission - and the biggest obstacle to maintaining political support, in the UK at least, for spending nearly a fifth of our aid through Brussels.

The bias to MICs is significant. The precise numbers depend on which source and year is used. However, the UK’s annual statistical report, Statistics in Development, has a comparative table (17) which gives the share of oda to low income countries from different donors. It shows that in 2009, only 46% of aid disbursed through European institutions was to low income countries, compared with a figure for all donors of 62%, and for the UK of 72%.

A crude extrapolation from these numbers suggests that every time the UK transfers £1 from the bilateral aid programme to Brussels, 26p is taken from low income countries and given to middle or upper income countries. Given that the UK spends about £1300m through Brussels each year, the transfer from LICs to MICs amounts to as much as £340m a year. For the EC as a whole, spending close to €11 bn euros on official development assistance in 2010, the equivalent figure is close to €3 bn per annum ‘lost’ to low income countries.

Who are all these MICs receiving aid from the Commission? There is one category of countries which have recently graduated to middle income status, and where aid programmes are relatively easily defensible: Laos, Pakistan and Vietnam fall into this category. There is another category which offers more of a challenge, or requires a different kind of argument. Europe's top ten aid recipients include Turkey, Serbia, Kosovo, and Morocco. Turkey and Serbia, at least, are at the upper end of the middle income bracket. Other recipients (in 2010) include Algeria, China, Colombia, Croatia, Trinidad and Tobago, and Tunisia: upper middle income countries all. Actually, Trinidad and Tobago is classified as a high income country, and does not appear on the latest DAC list. Full details of disbursements can be found in the annual report from the European Commission on development and external spending, the latest of which covers 2010 and was published in July 2011.

What can be said about this?

First, it is important to be clear that there is no sleight of hand. All the countries listed (with the exception of Trinidad and Tobago) qualify as eligible for official development assistance ( oda) under DAC rules. A new list was approved in October 2011 and is reproduced for ease of reference here. It shows that countries eligible for oda must be least-developed, ‘other low income’, lower middle income, or upper middle income, using World Bank estimates of GNI per capita, though NB in nominal terms and not adjusted for purchasing power parity. It is notable that the middle income country category spans quite a range, from $US 1000 pc to $US 12,275 pc. It is also worth pointing out that quite a few least developed countries, a UN classification which tends to lag income status, and is also broader in scope, would otherwise qualify as middle income. In fact, 18 of the 48 least developed countries fall into the middle income country category. Anyway, the point here is that the DAC rules are applied correctly. We will come back later to the question of whether the DAC rules are correct.

Second, it is obvious from the list of middle income countries receiving aid from the EC that many are neighbourhood countries in Eastern Europe or the Mediterranean, funded not by development cooperation instruments, including the European Development Fund, but by the Neighbourhood Instrument or the Pre-Accession Instrument. This is the case for Turkey, for example. In that sense, the high share of MICs in development cooperation is an accident of DAC labelling.

Third, though, that does not exactly obviate the problem. British ministers present and defend the aid programme as being designed to tackle the worst of poverty in the poorest of countries. They are legally bound by the 2002 International Development Act to ensure that all development spending is designed to reduce poverty, which presumably applies to spending even in Neighbourhood and Pre-Accession countries. I would like to see that assumption tested! In addition, ministers and their many stakeholders in international development must feel exposed by the scale of spending in countries which hardly qualify as developing countries, and which could arguably afford to tackle poverty from their domestic budgets. We need to come back to that argument also.

Fourth, the neighbourhood apart, there is also a surprisingly large number of MIC aid recipients funded from aid budgets proper, both the Development Cooperation Instrument and the European Development Fund. It is not easy to arrive at a breakdown of funding by instrument and income group, but the annual report shows that 41 upper middle income countries received aid from the EC in 2010, including many ACP countries funded from the EDF and many Asian or Latin American countries funded from the Development Cooperation Instrument. Antigua, Barbados, Botswana and Mauritius fall into the first group; Argentina, Brazil and Malaysia into the second.

Now, Andris Piebalgs, the Development Commissioner, has recognised the problem and pledged to tackle it. His new policy paper, Agenda for Change, introduces the concept of ‘differentiated development partnerships’. It states specifically that ‘the EU must seek to target its resources where they are needed most to address poverty reduction and where they could have greatest impact.  Grant-based aid should not feature in geographic cooperation with more advanced developing countries already on sustained growth paths and/or able to generate enough own resources’.

In December 2011, the Commission published its proposals for the new financial regulations which will govern EU external spending for the period 2014-2020, including both the DCI and the EDF. There is no commitment to differentiation in the EDF proposal, but the DCI proposal states that ‘in principle, high income, upper middle income and other large middle income countries, which are on a sustainable development path and/or have access to large domestic and external resources to finance their own development strategies, would graduate out of bilateral aid programmes’. It goes on to say that ‘partner countries representing more than 1% of the world's GDP and/or upper middle income countries according to the list of recipients of Official Development Aid ( ODA) of the OECD/ DAC are in principle excluded; however, additional criteria relating to their need and capacity is used, such as Human Development Index, the Economic Vulnerability Index and aid dependency, as well as economic growth and foreign direct investment.’

Practically, this means that 19 previous recipients of aid are excluded from so-called ‘bilateral’ programmes, though they may continue to receive aid from the DCI for regional or thematic programmes. The 19 are listed in Figure 1 below. Together, they received some €350m in 2020, including for regional and thematic programmes, and humanitarian aid, so this is a modest step in the direction of redistribution, just 12% of the potential transfer.

Figure 2; List of countries proposed to be excluded from bilateral EC aid programmes in 2014-20

How serious a problem is this? The answer depends partly on whether aid to middle income countries can be justified, a much-debated topic, which Jonathan Glennie explored in an ODI Working Paper. Some people feel that aid should be provided to poor people in so-called developing countries, pretty well regardless of their level of income. Certainly, they would argue for support to anti-poverty programmes in lower middle income countries. They would also argue for support to successful MICs because of the spillover effects to poorer countries in their regions.

There is indeed a case for being sensible about the arbitrary nature of dividing lines, and also recognising that middle income countries sometimes slip back below the poverty line. However, most observers would accept that as countries become richer, grant aid should be phased out and loans and other forms of aid phased in. ‘Blending’ is the current buzz-word, a topic on which Mikaela Gavas led work at ODI. This approach is consistent with EU thinking, and was also an implicit theme of the aid effectiveness meeting in Busan.

At the same time, a carefully calibrated aid approach to countries at different levels of development does not recognise the political imperative behind EU funding in the neighbourhood. Does anyone really think that EC pre-accession and neighbourhood budgets are read off from World Bank calculations of GNI per head and the DAC list of aid recipients? Does anyone think they should be? I thought not.

For British ministers, it is an inconvenience that aid is ‘diverted’ from poor countries in the developing world to less poor countries in the neighbourhood. For some others, the opposite may be the case. For the EU-12, for example, the new Member States, it is an advantage of aid spending that it helps to support countries in their neighbourhood, like Moldova or Belarus. The same is true for many Member States, both old and new, with respect to the Mediterranean. The EU has made ambitious aid pledges, including to reach 0.7% of GNI by 2015. It might well be that this pledge would not survive if neighbourhood funding had to be provided in addition to funding developing countries. There might even be an effect on the UK Government’s willingness to provide aid to traditional recipients, if it were asked to contribute to additional budgets, as it would surely want to do, to support stabilisation in former Yugoslavia or the Arab Spring.

That makes one obvious recommendation look shaky: to define the problem out of existence, by eliminating all Upper Middle Income Countries from the list of DAC recipients. There is probably a good case for that, actually, but it would have to be approached with great care, and in tandem with cast-iron commitments not to cut aid overall. Otherwise, it would be self-defeating. I’d be interested in views on that subject.

The alternative strategy is for donors that really want to concentrate on poor countries to vote, somehow, with their feet – in other words, to allocate resources so as to maximise the impact on poverty. Such an approach was implicit in the UK’sMultilateral Aid Review, completed in 2011, which ranked the EDF and the EC’s humanitarian work as ‘strong’, but the budget as ‘weak’, largely because of its lower focus on poor countries.

What position should the UK take in the current negotiation over the Multi-Annual Financial Framework 2014-20? Some would probably argue that the answer is to minimise the flow of money through Brussels, and spend the aid budget instead either bilaterally or through the UN or the World Bank. That option would ignore the good work done by the development and humanitarian instruments of the European Commission. It would also reduce diversity and contestability in the overall aid architecture.

An alternative is to minimise the amount of money going to less desirable parts of the EC programme and maximise parts going to the more desirable. From a UK perspective, that would mean arguing for smaller neighbourhood and pre-accession instruments, and transferring money instead to development instruments, including the EDF. Such a strategy would be feasible, but, taken to extremes, (a) would probably lose friends among countries concerned for the neighbourhood, and/or (b) open the UK to demands for additional funding for the neighbourhood.

Perhaps the answer is not to take the strategy to extremes. It would be possible, at the margin, to rebalance the EC external budget, favouring the EDF and the DCI, and encouraging Andris Piebalgs to move faster in the direction of differentiation, across the programme (including vis a vis the ACP).

In this connection, it is notable that the Commission proposes to increase the DCI by 19% in real terms, but the EDF by only 13%. That seems the wrong way round. The EDF is outside the budget, and therefore perhaps easier to increase, given fiscal stringency. Achieving differentiation within the EDF would be a challenge, given the contractual nature of the Cotonou Agreement, but should at least be proposed.

A final alternative is to revert to an idea I proposed back in 2006, which is to create a new, supplementary and voluntaryMDG Fund for the EU, devoted to low and low and middle income countries, funded from aid programmes, poverty-focused, and subject to regular replenishment, a bit like the IDA. Would that not solve the problem?

In the meantime, it will not be surprising if the UK’s International Development Select Committee takes a hard look at the allocation of EC development funds, and at the differentiation agenda.