On the power of ethnography in public policy research

I was going to write this blog post a long time ago, every since Ryan Briggs (Virginia Tech) alerted me to these posts by Tom Pepinsky (Cornell University), Ken Opalo (Stanford University) and Chris Blattman (Columbia University), but then the “worm wars” debate happened on Twitter, I got pulled into it (inadvertently and unwillingly) and well, here I am writing about a topic I wanted to write well after I wanted to write about it, but much later than I wanted to. Anyhow…

Ryan discusses Chris‘, Tom‘s and Ken’s posts which more or less call for more ethnographic research and institutional analysis to help improve aid effectiveness on the ground (I hope I didn’t misread and misinterpret anyone, all their posts are linked so do read them). I actually think it was Tom who suggested that there is a bigger role for ethnography and more institutional analysis, to help understand which interventions work best, as Chris asked for, instead of blindly implementing skills-training programs.

Ken furthers Tom’s point and emphasizes the importance of understanding local elites in how these programs are implemented. Local elites may be factors to push forward or bring into a stalemate any intervention you want to do on the ground. And Ryan calls for iterative, repetition-style trials where simple interventions are tried out, tested and analyzed.

All four posts really struck me as highlighting the need for more cross-disciplinary, multi-method research. I don’t work in international development or foreign aid, but I am a specialist in comparative public policy, and it seems to me as though the interventions that Chris, Tom, Ken and Ryan discuss are in fact policy interventions, and as such, we need to examine them closely. I am a big fan, believer, proponent and practitioner of ethnographic work. I study wastewater governance, sanitation and solid waste management. I visit and spend extended periods of time in landfills, follow scavengers through their rummaging activities, and blend into communities where access to toilets, sewerage, drainage and clean drinking water are in short supply. For me, ethnographic work is fundamental to what I do, and one of the most powerful tools in my research methods arsenal.

So, what I seem hear from Chris, Ryan, Tom and Ken is a call for more institutional ethnography. While the more traditional version of institutional ethnography was developed by Canadian sociologist Dorothy Smith, and you’ll find it defined as the study of textually-mediated relations, the work I do, and I follow more closely is that of Elinor Ostrom, who studied how institutions emerge through the ethnographic observation, on the ground, of routines and actions of actors. Traditional institutional ethnography focuses a lot on texts, whereas the model that Ostrom implemented places more emphasis on observation of routines, rules, actions and norm emergence. For example, in my own work doing institutional analysis of water governance organizations, I have studied how rules and norms emerge in river basin councils, by actually participating in these roundtables, following closely how each actor interacts with each other, studying their routines, rules (both written and unwritten) and their practices. This is institutional analysis merged with ethnography (and quite obviously, I simplify for purposes of this blog).

If I were to translate this to the development field, I think we (all five of us) agree that we need on-the-ground, in-depth analyses of simple interventions, repeated through time, focusing on all the actors (especially elites with unbridled power). This is definitely something ethnographic work, and especially institutional ethnography are particularly well-suited to do. And one of the challenges I will have in future years with my undergraduate students in Public Policy is to teach them how to do robust institutional analysis and how to implement ethnography in policy settings. It will be fun!

You can share this blog post on the following social networks by clicking on their icon.