Learning Needed

More Excel tables needed.

I jested in my post on Saturday about failure being okay for aid workers as long as they could write a "lessons-learned" document about it.  Sometimes it feels like that, but--and I say this in all seriousness--development is a field where we still don't know squat.  We write LOTS of lessons-learned documents.  Every NGO, UN agency, regional development bank and the World Bank write their own.  I have my doubts about how many of them are read, even by people designing almost identical programs.  Whether these are read or not, I don't think we actually ever find out about the vast majority of our failures.

Many times, we judge the performance of programs on first-hand witnessing or talking to stakeholders in the programs.  The pressure is on for us to show good results to donors so they'll want to keep working with us, and recipients often feel like they have to show appreciation or be positive in order to be considered for future programs.

I've done some work on a water, sanitation and hygiene program in primary schools funded by the Gates Foundation.  The program is far from typical and suggests some ways forward.  While I don't know that the quality of the work the program is doing is really any higher than most development projects, I do know that we're really going to be accountable for our results.  The Center for Global Safe Water at Emory University is also on the project, and they essentially serve as an independent learner, measuring things NGOs wouldn't have the staff time or capacity to get at like levels of fecal coliform on kids hands or absenteeism levels in schools using a randomized and controlled study.  We'll know exactly how effective this kind of program is, and the Gates Foundation will know if it's worth investing in.

If we're really going to learn what works, development agencies need to invest far more in learning. Donors need to prioritize learning over feeling good about themselves.  There's plenty of time for that later when we really know what works.  The bigger and more professional aid organizations are already heading in this direction, but there are thousands of smaller NGOs getting money without such efforts.  Some first-class research organizations of note working on this problem already are the MIT Poverty Action Lab, the Center of Evaluation for Global Action at UC Berkeley and the Overseas Development Institute in the UK.