Read on Twitter
Why should you do qualitative research (interviews, focus groups, observation of practices, ethnography) as an economist? A 
https://twitter.com/D_Kuehn/status/1360762409938526210
https://twitter.com/D_Kuehn/status/1360762409938526210
1. The Institutionalists were right that economic behavior is shaped and mediated by institutions, norms, practices, law, & relationships. This doesn’t mean supply & demand don’t matter, but it does mean they play out in complicated and contextualized ways, and power matters too.
2. Welfare economics as an enterprise is extremely weak and the subjectivists are also right. This is a point I’ve grown to appreciate in both Buchanan and Nutter. When assessing impacts on welfare it’s helpful to talk to people about what they value and don’t value.
3. The potential outcomes framework requires an intimate knowledge of the flow of people into a treatment: how they’re hired, recruited, assigned, tracked, guided, and informed of opportunities. These processes are complicated and contextual so usually you have to ask in detail.
4. Sometimes empirical results align nicely with theory and that’s nice but sometimes there are puzzles. You’d be amazed at how often if you ask someone why you found that X has beta effect on Y, they have a perfectly sensible explanation that has nothing to do with price theory.
5. Often we do work to inform decisions. Sometimes it’s the decisions of policy makers & sometimes others. The people we study have a perspective on what they want & need. If we have the ear of decision makers, we have an opportunity to communicate those wants & needs if we ask.
6. I mentioned potential outcomes above but assignment to treatment isn’t the end of your causal concerns. What exactly is happening to your counterfactual? You learn that by asking “if a participant didn’t have this opportunity what other programs could they access in the area?”
7. Sticking with causal inference, you also identify things like collider bias, and establish exclusion restrictions by a detailed understanding of the phenomenon you’re studying. Sometimes you get this by being well read but you get a much better understanding in the field.
8. The lamppost problem. When we stick with quantitative methods we study what is easily quantified or expend great effort trying to quantify what isn’t quantified. Qualitative research opens up the questions we can ask to economic phenomenon that can’t be quantified.
9. Some categories that can be quantified are best not. Two theories I find promising are identity economics (Kranton & Akerlof) & the task approach to labor (Autor), but I find their formal treatment unwieldy and not so insightful. Qualitative work on both is rich & meaningful.
10. Power matters in economics. A lot of people appreciate this and as work progresses on monopsony, inequality, and political economy this will be increasingly unavoidable. Power relations are also contextual and depend on institutional levers and personal circumstances.
OK that’s a nice round number. Maybe I should write these up?
Also, it’s fun and fascinating to go out and talk to people.
Also, it’s fun and fascinating to go out and talk to people.
