Why Angus Deaton is (mostly) wrong to attack aid for undermining politics and accountability

April 15, 2014

     By Duncan Green     

Continuing aid week here on FP2P, here’s my response to Angus Deaton’s recent broadside against aid, and his claim that I agree with him. Tomorrow, Angus responds. Nervous, moi?

I’m both flattered and alarmed that Angus Deaton has been citing From Poverty to Power (the book, not this blog) in defence of his attack on aid in his

and aid can help

and aid can help

book The Great Escape (previously reviewed here). Flattered because Deaton is a development superstar. Alarmed because he comes down strongly against spending money on aid, and his critique is being picked up by aid critics, a position which I most definitely do not share. So I’ve gone back and reread Deaton’s book, and some of the other research on the impact of aid. Let’s try and make sense of it.

In his recent LSE lecture (from minute 47 onwards here) and in an excellent cross-examination by Owen Barder on Development Drums, Deaton’s most substantive criticism was on the link between aid and domestic politics:

‘Economic development cannot take place without some sort of contract between those who govern and those who are governed…. It is the need to raise taxes, and the difficulty of doing so without the participation of those who are taxed, that places constraints on the government and to some extent protects the interests of taxpayers…. One of the strongest arguments against large aid flows is that they undermine these constraints, removing the need to raise money with consent.’

In other words, bunging aid money to governments means they no longer have to listen to their citizens and opens the door to all kinds of bad practices.

What do other researchers say? When I did a quick trawl of the literature, I had a familiar sensation – how is it that with all those clever people researching away, we never seem to be able to answer really important questions (unless you count ‘needs more research’ as an answer?). The aid discussion follows the classic pattern of

1)      a blizzard of cross country regressions reaching contradictory findings, eg one IMF study finds ‘aid improves revenue performance significantly’, while another asserts with equal certainty ‘tax revenue declines by 9 cents for each grant dollar.’

2)      Baffled researchers then revert to ‘priors’ and political theory, to argue that aid does indeed undermine institutions. Eg see this 2006 CGD paper.

Is that the best they can do? What about some serious country case studies (preferably post Cold War – aid has rather moved on since Mobutu) investigating whether aid has indeed undermined the social contract, as claimed? There are plenty of examples of donors using conditionality to try and over-rule sovereign governments, but none of this more subtle erosion of the social contract alleged by Deaton.

Loading the C 130 planeAs it happens, my own priors mean that I suspect that some level of erosion does indeed take place (Deaton’s right about me on that), so I was surprised by the lack of evidence. What did I miss? Any links welcome.

Deaton accepts that ‘These harms of aid need to be balanced against the good that aid does, whether educating children or saving lives’ and is notably more positive on his area of genuine global expertise – health. But he argues that ‘Those who advocate more aid need to explain how it can be given in a way that deals with the political constraints’. I’m not convinced. In the balance of argument, we have lots of concrete, tangible benefits in the shape of nurses, teachers, vaccinations, avoided deaths etc, lined up against a rather vague and not very well evidenced claim of long-term institutional damage. Before taking an axe to the aid budget, I would say the burden of proof should be on the aid critics, and they haven’t done very well so far.

The best recent summary of the state of the evidence I could find was from veteran aid wonk Roger Riddell, who last month published an updated paper on the impact of aid. Riddell responds directly to Deaton’s attack:

‘In my view, this assessment goes too far. The conclusion drawn is based on what probably is a too partial and selective reading of the evidence. Stating that countries will always be better off without aid remains what it has always been – an unproven assertion. Donors may well have failed, at times undermined recipient-country ownership and (unintentionally) held back long-term development prospects, but they are now more aware of their shortcomings, as well as more knowledgeable about how they need to change current practices to give greater priority to transformational aid to make aid work better overall.’

Riddell believes that ‘beyond this rather sterile exchange, a far more important debate has begun to take place about the systemic effects of providing aid. Until recently, it was predominantly aid’s harshest critics who suggested that any short-term benefits that aid might achieve would be eclipsed by the indirect harm it could bring in its wake. Today, the main aid donors are ready to acknowledge that… a series of systemic problems have developed that are now seriously undermining aid’s potential impact. The gap between aid’s strongest critics and an important cluster of studies analysing aid’s wider impact has narrowed considerably.’

That recognition informs the Busan Partnership for Effective Development Cooperation (sorry for the descent into aid-speak  – see Nicola McIvor’s post yesterday for more), with its commitment to A Malawian activist takes part in a protest against corruption‘country ownership’ and ‘inclusive development partnerships’.

So what might the emerging common ground look like? Deaton is not against all aid, and suggests health, humanitarian emergencies, and countries where aid remains a low percentage of overall government revenues. He also advocates aid for basic research, technical assistance to governments that want it, and support for poor countries in the innumerable international negotiations. He advocates a good ‘put your own house in order’ list of sanctions on odious regimes, transparency in corporate payments to governments (e.g. on minerals extraction) and tackling northern restrictions on trade and migration, and perverse biofuel subsidies. He misses tax havens and illicit flows, but I imagine he would support their inclusion.

To that I would add the whole ‘thinking and working politically’ approach, including taking seriously bad aid’s potential for institutional damage. That means acknowledging and tackling any negative impacts/ strengthening the positives (eg by supporting better tax collection and other institutional strengthening, where it is demand-led; channelling a percentage of government-to-government aid to citizen or parliamentary watchdogs to keep them honest; maybe cash on delivery type contracts).

So yes, lots needs to be done to ensure aid strengthens poor people’s ‘freedoms to do and to be’. But no (sorry Angus) we shouldn’t be taking a hatchet to aid budgets.

Got a feeling Angus isn’t going to agree with this. Come back tomorrow to find out.

April 15, 2014
Duncan Green