Thinking, fast and slow and the transparency agenda in development

I was reading “Thinking, fast and slow” by Daniel Kahnemann. It is a very good book. It challenges conventional wisdom and is so full of meaning that it asks for a regular reread to discover more hidden treasure.

He explains how humans actually think, and not just how we think we think. He doesn’t really believe these insights will change anything: he considers it mostly an enrichment for water cooler discussions. As our illogical ways are hard-wired, even when we know we are illogical, we can not help it but to proceed on the beaten path. The Homo sapiens and Homo economicus are 2 entirely different species.

However, this will not stop me from trying to apply some of his insights.

One theme from the book of direct relevance for development work is our relationship with data. I will touch on 2 aspects: how to predict success of an intervention and how to convince people an intervention is a success.

Apparently, a conviction is formed when the story behind the conviction is convincing. Now apparently a story is convincing when it is in the first place coherent. Real life stories however (what some people call the reality) are never very coherent: a lot of things happen that blur the story. People have lots of reasons, not just one, and the one they tell you might not be the one that is relevant. So to be convincing, only the coherent data should be presented. Otherwise, the conviction will be weaker.

The prime example of this effect is of course the diplomatic cable: a coherent and short analysis is explicitly required. Clarity and conciseness are cardinal virtues. A political decision taken on the basis of this kind of analysis will of course be convincing. Meanwhile the simplifications in the analysis yielding to the demands for a coherent story can lead to important errors. Only the elements mentioned in the cable will be taken into account. Kahnemann says: “what we see is all there is”. This might be why sometimes bad choices are made in foreign affairs.

Closer to the development world is the International Aid Transparency Initiative.

The International Aid Transparency Initiative (IATI) aims to make information about aid spending easier to find, use and compare.

Those involved in aid programmes will be able to better track what aid is being used for and what it is achieving. This stretches from taxpayers in donor countries, to those in developing countries who benefit from aid.

Improving transparency also helps governments in developing countries manage aid more effectively. This means that each dollar will go as far as possible towards fighting poverty.

From what was explained above, we understand that the transparency towards the taxpayers will not lead to more trust in the expenditures and trust in the whole venture of development aid. To the contrary: the full exposure to all the data will probably lead rather to more distrust, because the simple story we need to convince of the good aid does, will get complicated. For every straight success story, there will be more lots of maybes, and too many unfortunate failures. Full disclosure is perhaps necessary for moral reasons and to keep the practitioners honest, but will not lead to more trust by the public.

More to the core of our work is how we predict the success of an intervention. Experts, it seems, can be very good at analysing complex situations, but most of them seem unable to predict what will happen next. His prime example concerns newborn babies. Before, it was the gynecologist who would decide how to care for the newborn. When they started to base the decision on simple indicators that can be gathered by every nurse, infant survival started to improve. Apparently, when the feedback loop is not immediate, statistics and simple indicators are more accurate in guiding what to do than experts. The anaesthetics get direct feedback from their work, and develop very good gut reactions, less so the gynaecologists when deciding on what to do with the newborn.

This seems to be very relevant for the approval process of projects and other interventions. The approval process is normally a mixed bag of expert analysis and some indicators. The indicators are not really chosen because they predict success, but because they measure political priorities, such as the mainstreaming of women’s issues and environmental awareness.

Experts who analyse projects for approval are seldom around when the results are obtained, and evaluations happen even later, too late to inform the next phase. Moreover, as political priorities shift, chances are that, even when the results are good, attention shifted a long time ago and we won’t continue the project.

This is a typical situation where, according to Kahnemann, expert advice is next to worthless. An alternative should be to use simple indicators that are known from the statistical analysis to predict success.

We know about some of the main elements that can predict success in development: proven interventions (de-worming etc, the whole CRT stuff) interventions done by trustworthy partners, and interventions tackling in a serious way the main issues at stake. Most of these elements are quite straightforward, and could form the basis for a simple analysis based on indicators. But not a lot of elements are available for analysing the substance of the project.

And here the IATI comes in. We just don’ t have the statistics on interventions to go beyond the most simple results predictions. IATI should strive to offer these statistics asap.

Perhaps humanitarian assistance, with its short feedback loop, the urgency to get the results right and existing standards, would be a good place to start.

This entry was posted in development, humanitairan. Bookmark the permalink.

Leave a Reply

Your email address will not be published. Required fields are marked *