Hiring networks in law and political science

Daniel Katz, Josh Gubler, Jon Zelner, Eric Provins and Eitan Ingall have a new paper suggesting that hiring networks in legal academics and political science work quite differently. Here’s their depiction of hiring and placement among legal academics (click on image to see larger graph).

law_network.jpg

As Katz et al. note, this is quite different from the pattern observed in political science.

unlike departments of political science studied by Fowler, Grofman & Masuoka (2007), the aggregation of the micro-level decisions by law hiring committees converges not upon a cluster of institutions but rather upon two institutions – Harvard and Yale.

While the modes of representation aren’t directly comparable, a rough eyeballing of this graph from Fowler, Grofman and Masuoka demonstrates Katz et al.’s point.

ps_network.jpg

One plausible interpretation of this striking difference might be differing kinds of specialization in schools (such as law schools) that both train future academics and practitioners. Nationally oriented law schools can strive for two kinds of reputational excellence – excellence in turning out future scholars, and excellence in turning out practicing lawyers who land good jobs in top 20 firms etc. Given limited resources, it is likely that only a very few schools can strive for excellence in both, and if you have to choose one of the two as a law school dean, you are likely to choose excellence in turning out practicing lawyers, since the immediate rewards are greater. Thus, only a couple of schools at the very top of the pile are really going to have sufficient resources to concentrate on building up a program aimed at turning out future academics. (In all of this, I bracket questions such as what ‘excellence’ means in substantive terms, self-fulfilling expectations etc – which isn’t to say that these aren’t important). In contrast, nationally oriented political science departments don’t usually have the same direct material incentives to turn out practitioners, and don’t get much prestige for turning out, say, brilliant undergraduates who go on to do great things. Thus, you might plausibly expect that top political science departments will focus more on turning out future academics.

The other interesting phenomenon that Katz et al. allude to is the phenomenon of self-looping – that is, of departments hiring their own. They suggest that this is much more prevalent among legal academics than among political scientists, creating some methodological concerns. There is a very strong norm in political science against hiring your own (at least before they’ve gone out into the marketplace, worked elsewhere for several years etc, and even then it is uncommon). But there appears to be no equivalent norm in the legal academy, as best as I can see. Here, I don’t have a plausible explanation (if readers do, they are invited to put them in comments).

5 Responses to Hiring networks in law and political science

  1. Daniel Nexon March 19, 2009 at 11:11 am #

    The important thing is that neither of these patterns look like the ones you would expect with markets. They both appear to reflect status-based segmentation, but with differences in the size of the core clique.

  2. Daniel Nexon March 19, 2009 at 11:14 am #

    Indeed, you should be using the first Figure 2 in the paper, not the second, to compare with the Political Science one.

  3. Patrick Rogers March 19, 2009 at 1:47 pm #

    It actually doesn’t look too tremendously different to me. I wonder if the shape of the graph is just an artifact of the visualization software. Different algorithms will present the same data differently. It also looks like there are more law schools in the Katz et al data set, leading to a higher density of third-tier schools in the network.

  4. Daniel Nexon March 19, 2009 at 2:57 pm #

    The paper has such a figure, and it looks a lot like the Political Science one in terms of basic structure. That’s why I took Henry to task for choosing which “Figure 2” (they’re both labeled that in the draft) he used.

  5. Henry March 19, 2009 at 11:50 pm #

    Dan – I originally was going to post the two graphs that were most comparable, but then decided that the information content of the first was greater, even if it made the two more difficult to compare. The distributions do look pretty different to me at the top end – for law schools, there is a sharp drop off after the first two institutions, while for political science, there is a much more graduated drop. But this is all based on eyeballing the graphics, so take it for what it is worth …