I’m thinking about a couple of loose ends from this weekend’s post about the graduate program in economics. I suggested a set of first year courses to try to separate economic ideas and their context, from the modern economic toolbox:
2 x history of economic thought
2 x working with data and coding
1 x mathematics concepts
1 x probability and statistics concepts
1 x experimental methodology (field, lab, natural)
1 x theory modeling methodology
The decline in the relative share of “theory” papers that I mentioned yesterday is no doubt due in part to computing power, new data, new tools, and new methodologies. But another aspect is that the barriers to entry for “publishable” theory are relatively high.
Decent economic theory is totally reasonable for any student in an economics grad program. Thinking through economic problems—how stuff works, trying out what-ifs, and constructing mental frameworks—is exactly what we have done through the undergraduate curriculum and the coursework stage of the grad program.
But how reasonable is it for any student to produce Economic Theory? The field is a particular facet of the thing, and the difference is mathematical formalism. Small-t theory is a thing, and Theory is a field. The current setup makes for fantastic research that I certainly don’t want to lose, but the fact that Economic Theory, the field, is arcane makes for an unhealthy household in the profession. Sniffing that students should just “know more math” is not going to solve this problem. Everyone has enough on their respective plates.
I discussed a couple of weeks ago, in the context of McCloskey and the recent Offer and Söderberg book, the bias towards studying fields rather than studying things, times or places in economics. The first year curriculum is a fine place to start looking for the reason why. Binding topics to tools in micro, macro, and econometrics courses creates silos that impede both our ability to do holistic research on real times, places, or things and our ability to talk freely across fields. The second and third years are elective field courses: IO, growth, mechanism design, labor. Holistic-toolbox courses happen only by accident or the grace of a particular instructor.
So that’s why I think we could do better to separate ideas from tools in our pedagogy, at least in the first year. In the end, I don’t think this changes much about the way economics researchers—post Ph.D.—see themselves. I don’t know many economists who are blinkered to other fields. Most economists do describe their research interests in terms of what they work on rather than how they work on it. Most economists are willing and able to bring a diverse set of tools to bear in their work.
Really my point then is just to let the graduate curriculum catch up to these changes (or constants, maybe) on the research side. I have heard it said a lot that the top rookie job candidates are those that demonstrate command of “theory and empirics” in their job market paper. Yes, that, and more! But then I’ve also heard it said that it’s disappointing to see “theory and empirics” mean just a bolted-on model to an empirical paper, or some cursory data analysis in a theory paper. Well, what do we expect? We aren’t giving our grad students the best chance to do what we’re asking them for. We need to build the conditions for success.
Ideas / tools in the first year would help, I think. So would field courses based on things, not fields, maybe co-taught if that helps get over the hump. I appreciate that through the lens of the “fields” status quo there is a potential risk to depth and rigor. But, as economist well know, there are always constraints and there are always trade-offs. Field-based training might be great at producing students who can turn out good papers for a dissertation almost by formula, but I’m not sure it does enough to encourage the kind of papers we seem to value later in the broader research arena.