Sunday, April 5, 2020

On inter-university online seminars

Marginal revolution linked to this tweetstorm (*) regarding the pandemic-induced appearance of a number of on-line seminars in economics not tied to particular universities. I had it on my list to do a post about that very topic, so I can now frame my post in part as a response to the tweetstorm.

I am aware of at least five such seminars: one on the pandemic and related topics run by the NBER, an econometrics seminar run out of Harvard, and seminars on crime, education, and experimental / behavioral economics started by individual researchers or small groups of researchers.

Some remarks:

Most regular seminars in economics are 80 or 90 minutes, though departmental seminars in which students present work in progress are sometimes 50 or 60. My sense is that most online inter-university seminars are 60 minutes. Perhaps there is a view that attention spans are shorter when not in person or perhaps the start of a "new" seminar allows a change that would have (or should have) happened to in-person seminars but for the stickiness of the old equilibrium.

One thing the tweetstorm omits is the relatively unique character of economics seminars that arises because they typically involve substantial back-and-forth between the speaker and the audience during the seminar, not just at the end as is common in other disciplines. This feature has important implications for the extent to which economics seminars can scale up in the same way that a lecture does. In my view, the in-person seminars that the NBER runs in labor economics and education (especially at the summer insitute where the audience for the labor talks sometimes rises to over 100) already well exceed the optimal scale given the interactions.

There are a number of solutions to the interaction / scale issue. One is to generate and maintain strong social norms about who gets to ask questions, so that what appears to be a seminar with 100 participants is actually a seminar with 25 participants and 75 non-speaking onlookers. That's how NBER meetings work. Another solution is to limit the size of the seminar. At least one of the online seminars that I am familiar with asked invitees not to broadcast the existence of the seminar (which is why I am being coy about it here) for precisely this reason. There is, of course, an aspect of this to the NBER meetings as well - you have to be invited and not everyone gets invited.(**) A different solution dials down the interaction, limiting questions during the presentation and then having a formal discussant or a formal period for questions and answers (or both) after the presentation. A couple of the online seminars I am familiar with have gone down this road.

A related point is that the online seminars I have participated in - including the informal student public economics seminar here at Wisconsin, which has moved on-line (as have some of the student seminars at Michigan) - is that they make use of the virtual hand-raising facility built into the meeting software rather than allowing audience members to simply shout out their comments, as often happens in person. I like the change to raising hands and hope it carries over to in-person seminars when they return.

The tweetstorm thinks these inter-university seminars will persist and worries about who should run them. I agree with the tweetstorm author's prediction and also with his focus on the importance of the gatekeeper role. I predict that either organizations like the NBER, or its European semi-analogues like CEPR, IZA and CESifo, will end up managing the on-going on-line seminars, with the subject area leaders within those organizations playing the gatekeeper role, just as they do now with group membership and in choosing who presents at the reguular subject area conferences / seminars. It will be interesting to see how this plays out, and we all know the track record of economists' predictions.

One last point: on-line seminars are only a partial substitute for in-person departmental seminar visits, which I do not think will go away. Much business gets done at seminar dinners, and much is learned during in-person office visits with the seminar speaker on the day of the seminar. One of the highlights of seminar visits for me is often meeting with the local gradual students and hearing about their work. Online seminars omit all of these features.

(*) A "tweetstorm" is a blog post made harder to read by being shared on twitter in little bits.

(**) I do not mean to pick on the NBER here; I am using it as an example that I think most economics readers will have some familiarity with, even if from a distance. These same issues arise with, e.g. the Institute for Research on Poverty Summer Research Workshop of which I am a co-organizer. It is a real problem. No matter how inclusive and open and non-hierarchical and encouraging of junior scholars you want to be, there are real fiscal and technological constraints: budgets are finite and seminars / workshops that get too big are not as intellectually useful.

No comments: