The Great Escape

Read book. Took notes on book. Listened to Development Drums podcast. Read review. Went on holiday to Bologna. Got flu. Now finally ready to pass on a few thoughts about Angus Deaton’s The Great Escape.

First thing, just to highlight, is the amazing, bird’s eye view you get of progress in a) global health and b) income/economic growth over the last couple of hundred years. One general message is that, on a world scale, things have been getting better at an incredible rate over the last couple of hundred years and that a big part of the inequality we see is a result

of the some parts of the world becoming rich (which is undeniably positive as a fact in and of itself). And Deaton provides loads of lovely statistical snippets to show it. For example in the health ‘section’, for the last 160 years, for every four years of calendar time, the world’s highest life expectancy increased by a year.  Meanwhile, between 1820 and 1992, the average income of all the inhabitants of the world increased between seven and eight times and the fraction of the world’s population in extreme poverty fell from 84 to 24% (at the same time as the world population increased by billions). Deaton is clearly a master of the data and a giant in his field and it shows.

Then of course we get some broad ideas of the ‘why’ without there being an overarching grand theory, which is quite refreshing and probably bang on given the scope and scale of the changes we are talking about. Deaton also focuses on some other pertinent issues like the hazards of measurement and indicators, economic growth as an overwhelmingly positive force (or not) and the (topical) issue of income inequality within the USA and worldwide (I have a word doc to which I add links to contributions on the inequality debate: currently on about 15 articles/reports in the last month).

But the most polemical, and (for me) interesting stuff is found in the final section on aid and why it’s bad, which got me writing angry comments in the margins. To cut straight to the chase, the central argument is that “foreign aid makes governments less responsive to the needs of the poor, and thus does them harm”. Essentially if we accept that poverty (broadly defined) is a consequence of poor institutions, and that aid to governments undermines institutions and democracy, weakening the social contract, aid is bad. Which isn’t revolutionary stuff, very broad-brush, reminding me a bit of Easterly with the idea that aid mucks with incentives, combined with sort-of Duncan Green/Oxfam change coming from combination of active citizens and progressive states working together. Deaton therefore also criticises what he characterises as the ‘hydraulic approach’ to aid whereby monetary inputs are automatically assumed to lead to positive outputs, with development conceived as a technical issue – definitely agree here, as does pretty much everyone up to date with current development thinking. And regarding the main argument about aid and institutions, the theory sounds eminently plausible...

However. The first problem is, as Owen Barder points out, this is an empirical not a theoretical question, and there isn’t hard evidence supporting Deaton’s view that all to-government aid ends up wrecking the given country’s political institutions and gets siphoned off by corrupt governments rather than being used to help the country’s people. Anecdotally, Barder mentions Ethiopia’s human development progress (see Development Drums podcast for this and further references) and Rwanda struck me as an example after listening to a Guardian podcast listing the positively stunning gains made in health there. This reported that in the last ten years alone: deaths from AIDS and TB were down 78%. Maternal mortality down 60%. Under five mortality down 70%. Paul Farmer called these developments ‘the steepest declines in mortality ever recorded, at any time, and in any place’. In this light, and even despite serious concerns about political freedoms in Rwanda, Deaton’s caricatured claims about corrupt African governments just don’t square with reality. It might be possible, in fact, for aid to strengthen democracy and the social contract when its use is made accountable to the people through traditional institutions. Rwanda’s imihigo system, whereby if the local ‘mayor’ fails to deliver on developmental promises he is made a social pariah, is one example. Chris Blattman in his review makes this point even more strongly:

 “And, frankly, I personally find it hard to believe that levels of democracy in Africa would be as high as they are without aid. I think the most important forces driving democracy are probably internal to Africa, and the example and economic success of advanced democracies comes second. But aid and foreign meddling comes a close third”

On the general point, Owen Barder, with characteristic succinctness, crystallises the underlying criticism when he accuses Deaton from arguing from extremes. And I’d agree that the sentence “the director of one aid agency gave me a bloodcurdling account of how aid funds had gone to gangs of murderers” isn’t the most nuanced or helpful reflection on aid I have read recently.  Basically there is definitely something lazy if not worse about taking a homogenised image of the African other (based on Mobutu or Mugabe for example) as a key premise of an argument about aid.

And second, even conceding Deaton’s argument about aid’s effects on political institutions, to claim, in the absence of hard evidence, that this overrides the (widely accepted) results from worldwide health initiatives (just one example) is a little strong for me. When first reading the chapter I’d scribbled down a note to Owen Barder’s ‘pocket defence’ of aid based on the smallpox case as potential rebuttal, and I’ve got to admit I did a fist-pump when he unleashed it in the podcast. And after Deaton took this road, it was a bit of a slippery slope, where he seemed to edge towards the view that political rights and freedoms trumped everything. The USA should consider stopping trade with Saudi Arabia (Deaton agrees to), okay, but then should the rest of the world stop trading with the USA because of Guantanamo? Deaton’s refrain that dealing with countries like Ethiopia under repressive dictatorships was “like having blood on our hands” seemed unconvincing.  Surely not supplying healthcare when people will die without it is also ‘having blood on our hands’? The world isn't that black and white.

And this is where my angry notes in the margin come in. Because it seemed to me that Deaton was, in general, pretty sloppy in the final chapter. My quibbles include: his use of hyperbole (see ‘gangs of murderers’ example above); his reliance on anecdotes, for example about 'WaBenzi'/corruption in Kenya; his random pot-shots about loosely connected issues ranging from RCTs (which he doesn’t like) to aid apparently being used as a way for rich country governments to boost their popularity (erm not sure I buy this given recent findings from ODI “the public may be becoming less supportive of maintaining, let alone increasing, current levels of UK spending on aid”). But worst of all (what Owen Barder with wonderful tact referred to as the chapter being “less well-evidenced”) his disingenuousness with statistics – his bread-and-butter! So he pulls up graphs showing that there is no positive correlation between aid and growth, or between aid and democracy, and then leaves these facts 'floating' in his argument, flirting perilously with the causation-correlation fallacy that jumped out even at me as a non-economist/statistician.  For a social scientist renowned in his field, to toy with these arguments seemed pretty petty.

Basically ‘Deaton-on-aid’ got a bit rant-y (though I do appreciate the sentiment of frustration with the aid sector, as Owen Barder put it, “a cry of exasperation”) and obscured the more important take-home ideas that most people agree on now and which did also come out partly in the final chapter:

1. Aid isn’t the solution – we can do far more through migration, climate change, trade policies, 'aid for development' in promoting global public goods like scientific knowledge etc;
2.  Aid and development is political - I think we can all agree that the technical, depoliticised, 'problem-solving' conception of aid and development can be put to bed;

But alongside these two conclusions drawn by Deaton, Chris Blattman's fundamental objection to Deaton is worth keeping in mind: that 'The answer to “Does aid work?” is the same answer to every question in social science: “It depends” ’(from his own review). So Deaton is coming up with (probably) the wrong answer to (definitely) the wrong question. The Great Escape is a triumph in terms of tracing broad global trends in health and income with rigorous statistical evidence, but the final chapter is best taken with a large dose of salt. Time to move on from Sachs-vs-Easterly methinks.