Update 5/25/2016: This post originally linked to a draft of my forthcoming CGD book Results not Receipts. As planned, we’re taking the draft offline today. I’ll spend the next couple of months thinking about and reacting to the very helpful comments below, some from Duncan Green and his readers, others that were emailed, and more from a CGD review meeting. It has been a very valuable process and I’m hugely grateful to all who took the time to read and comment. I’ll be making substantive changes to tone and content as a result. Look for the book in about six months.  

Here [was] a draft of my (hopefully) forthcoming book for CGD Results Not Receipts: Counting the Right Things in Aid and Corruption [working title] I’d love comments and reactions from the vast, knowledgeable network of development thinkers and doers that read Views From the Center –they’ll surely make the final version far better.  The elevator pitch: donors frequently suggest corruption is the biggest obstacle to development and aid effectiveness, and that they can accurately measure corruption risk while protecting their projects from it at a reasonable cost.  It isn’t and they can’t.  Below, a brief synopsis.

Chapter One discusses the two problems of corruption and poor governance.  There is the well-known problem that corruption can and does derail aid projects and (at times) broad based-development.  But there is a second ‘problem of corruption’ –and that is one of donor perception and response.  A narrative that suggests corruption is the major problem of development and an intractable one justifies aid fatigue (its broke, and we can’t fix it).  And concern with the risk of malfeasance leads to aid programs that involve heavy oversight or even direct selection, design and management of projects from distant donor capitals. 

Chapter Two suggests the narrative that corruption is the major problem of development and an intractable one might be wrong: we have seen significant development progress despite stagnation in perceptions of governance.  The chapter also points to the limited relationship between common measures of corruption and development outcomes. 

Chapter Three provides one potential reason for the scale of the disconnect: weaknesses in the measures of corruption.  The three measures of expert perceptions indices, surveyed evidence of bribe payments, and popular concern with corruption vary considerably.  Expert perceptions appear to be biased and inaccurate measures of either bribery levels or popular concern about corruption and it is very unclear what they actually measure.  Even surveyed estimates of bribery appear open to considerable uncertainty, but they do suggest that any one measure of national corruption is likely to hide considerable within-country heterogeneity.

Chapter Four turns to country-level responses to the corruption threat –focusing on competition, results measurement and transparency as key tools to improve outcomes.  And Chapters Five and Six discuss the weak empirical basis for current donor approaches to anti-corruption and call for a focus on transparency and outcomes.  The conclusion suggests the inconvenient truths for supporters of the current approach to corruption that the book has outlined emphasizes the need to focus on results not receipts.

Be warned: it’s a draft.  There are rough edges.  The pictures and references need work.  The text needs a good edit.   And I’m sure there are weak arguments and missing points a-plenty.  That’s why I’m hoping for feedback, but also why I’d be grateful if you didn’t cite or quote it (yet…).  And as the plan is to properly publish the book in the end, the draft text will only be available for a month, but I hope that’s ample time for people to download a copy and take a look.  I’ll be revising the text over the next couple of months.

I’m very grateful for any and all feedback, on chapters or the whole text, by email or in the comments, if possible by the end of May.  And thanks in advance!