Grad school from below

Noahpinion shares what he learned in econ grad school. The post is enjoyable and worth reading because it speaks to an enduring divide between what faculty think graduate students are learning in grad school and what the lived experience of graduate students actually is. Moreover, Noah confirms a private notion of mine that Ph.D. students should run sections not just to "learn how to teach" but more to learn what, exactly, their doctoral coursework has been about.

As a current doctoral student, let me begin by saying that grad school has been both the most challenging and the most fun experience I've ever faced. But in my first year, I constantly felt as if I had no idea whatsoever what was going on, or why I should care about statistics, philosophy of science, the paradigm wars, or a good deal else that was being covered. Noah's summary of a first-semester macro course, with substitutions, could equally be a journal of my first semester in IR and then, later, in American Politics:

My first semester was on business cycle theory. (the second semester was all growth theory). We spent a day covering the basic history of the field - the neoclassicals, Keynes, Friedman, Lucas and the RBC people, and finally the neo-Keynesian movement. I recall reading the Summers vs. Prescott debate but not really getting what it was about. From then on it was all DSGE. We did the Ramsey model and learned about Friedman's Permanent Income Hypothesis. We spent a lot of time on RBC. We took a big break to learn value function iteration and how to numerically solve DSGE models by fixed-point convergence. Then we did Barro's model of Ricardian Equivalence, learned a basic labor search model, briefly sketched a couple of ideas about heterogeneity, touched on menu costs, and spent a good bit of time on Q-theory and investment costs. Finally, at the very end of the semester, we squeezed in a one-week whirlwind overview of Calvo Models and the New Keynesian Phillips Curve...but we weren't tested on it.

In the concatenation of models and the dearth of explanation and exploration, this seems familiar. In fact, the biggest difference between my first-semester experience and Noah's is that although he feels that he never learned how economists knew anything was "right"(HINT: they probably weren't!) I never felt as if, in IR, I was given a firm base to stand on. The joke in my comps study group later on was that if we were asked to name one thing that IR had "learned," we would answer "democratic peace" and throw away our other outlines full of citations contending that the democratic peace theory was bunkum.

In almost every way, this is a healthy approach. I'd rather have a humble field that attempts to redefine science not as a disciplinary cudgel but as a series of structured debates that at least allow us to rule out a lot of policymakers' opinions about causation as wrong than a shiny, quanted-up, scientific edifice that leads policymakers to cheerfully create the Great Recession (among other pathologies). If modern IR had the same track record at successful advising as modern mainstream econ, one wonders whether the world (or at least vast swathes of the U.S.S.R.) would not have been a charred cinder by the mid-1950s.

But let me be parochial for a second, and speak not about the virtues of this approach for humankind in general but the problems it poses for graduate students. (Such navel-gazing is common in the discipline; hence the fact that every quant IS scholar has wished, however briefly, there were more great-power wars, since that way they could have larger Ns to play with.) This experience of constant doubt and questioning makes political science graduate school, and IR Ph.D.s in particular, really difficult. Teaching, though, helped me to reconnect all of these abstract debates to the big questions I had begun graduate school in order to answer. In fact, teaching sections forced me to write down, for the first time, what I thought those big questions were.*

In reformulating my understanding of the field (and re-reading, and then reading more deeply) in order to answer my undergraduates' questions, however, I discovered what must be the graduate school version of teaching your grandmother to suck eggs--namely, that the grad syllabi I had found so mysterious the semester before had actually been meant as answers to exactly the same questions my students were now asking me.

Noah quotes Feynman, and so let me offer another Feynman vignette, this time from his second book of memoirs. When he found himself stuck in research, he found he could retreat to teaching, not only because at least the daily process of helping others learn offered some solace to someone who thought he had dried up but also because it eventually allowed him exactly the chance to return to the broader questions with fresher eyes.

So why were my undergrad classes so much more educational for me? The simplest explanation is just that this was the second time I'd gone through the material, and so the review made clearer connections that had been obscure. The more profound difference, though, is that undergrads are more incentivized to ask questions. Graduate students are vastly more risk-averse about asking dumb-sounding questions, not least because their professors will also be their colleagues and one simply doesn't want to make a bad impression. (The reverse calculation--that failing to learn something correctly will lead to catastrophically bad impressions down the road--almost never seems to be made, which I leave as an unsolved puzzle for rational-choice theorists.) Accordingly, a great many people in every seminar--I will wager 80 percent of students in 80 percent of seminars--are faking it, or, worse, wrongly confident in their abilities. And 100 percent of students fall into those categories at some point.

Returning to Noah's piece: More teched-up readers may enjoy the former physics major's sniffing at economics (although, recall that physicists have a low opinion of a lot of things). These same readers should also enjoy Donald Mackenzie's discussion of economic theory and assumptions about probability distributions).

*For the record, my first version was "Will the states system experience war or peace? Will it foster prosperity or poverty?" I've since revised that to be even simpler--"Will the states system experience conflict or cooperation?"--while adding the constant Leninist refrain "Who--whom?" Obviously, these are hardly original, but the point is that these questions are a little more fundamental, and admittedly simplistic, than the standard journal fare.)