Observations on “An Empirical Study of Political Bias in Legal Scholarship”

August 13th, 2014

I read with great interest Adam Chilton and Eric Posner’s new article about political bias in legal scholarship. I encourage you to read the entire piece, as it is quite thought-provoking. Their general conclusion was “professors at elite law schools who make donations to Democratic political candidates write liberal scholarship, and law professors who make donations to Republican political candidates write conservative scholarship.”

I had a few observations, that I will point out in this post.

First, the authors had a tough time finding enough law professors at the top 14 schools who donated more money to Republicans.

In the initial sample of 140 professors, however, only 8 had donated more money to Republicans than Democrats (which is our principle measure of ideology).

The authors had to turn to an alternate methodology to try to find more conservatives at the top universities:

That said, we still believed that 8 Republicans was too small a sample for drawing reliable inferences, so we proceeded to oversample Republicans. To do so, we used the Database on Ideology, Money in Politics, and Elections (DIME) developed by the political scientist, Adam Bonica.30 Using the DIME database, we searched for employees of the 14 law schools in our sample who were net conservative donors, and who met our criteria of being tenured academic scholars. This resulted in the identification of an additional 16 Republican donors who were not already in our sample. We added these 16 professors to our initial sample of 140 professors, resulting in a sample of 156 total observations.

I’m not sure how this impacts the randomness of their sample, but in the end, there were 24 “Republican” law profs, and 132 “Democrat” law profs sampled at the top 14 law schools. This doesn’t surprise me, and is consistent with research by John McGinnis, James Lingdren, and others.

Second, the authors broke down these professors by subject matter they wrote in. My initial thought, was, how many of those 24 “Republican” Profs teach constitutional law. The authors note (p. 21) that virtually none of them teach ConLaw:

Third, the distributions may reflect the influence of constitutional law scholarship. Just by chance, none of the net Republican donors in our dataset are constitutional law scholars, and only one of the adjusted Republicans is. If constitutional law scholarship is more ideological than other forms of scholarship,50 the different distributions may show that Democrats are more likely to write constitutional law scholarship but not that they are otherwise more likely to write ideologically than Republicans.

Specifically, among the ConLaw professors sampled in the revised set, 20 were Democratic donors, 1 was a Republican donor, and 5 were non-donors. Again, this isn’t surprising.

As David Hyman notes in a new symposium issue of the Illinois Law Review on why the Law Professors Misunderestimated the Obamacare lawsuits, Akhil Amar carefully revised his discussion of professors who supported the challenge from top 20, to top 10:

In a public debate before the Supreme Court held oral argument, Professor Amar emphasized the fact that only one law professor at the top ten U.S News-ranked law schools agreed that the challenges to PPACA had merit.48 A year earlier, Professor Amar made a broader claim: that there was only one “constitutional scholar that I know at a top twenty law school (there are hundreds of them, they’re left, right and center) that thinks this is constitutionally problematic.”

I’m pretty sure he was referring to Richard Epstein in the former example–though this excluded Randy Barnett (Georgetown is not top 10). My contribution to the symposium, Obamacare and Man at Yale, makes similar observations.

Third, the authors confirm what seems to be the conventional wisdom–many closeted conservatives get ahead by stifling their ideology, or writing in field where ideology isn’t as important.

With those caveats in mind, if it is in fact the case that Republicans write less ideologically biased scholarship than Democrats do, then one would naturally ask why. The most plausible explanation is that if the dominant ethos in the top law schools is liberal or left-wing,51 then Republicans are likely to conceal their ideological views in their writings. Republican professors might fear that scholarship that appears conservative may be rejected by left- leaning law review editors, and disparaged or ignored by their colleagues, which will damage their chances for promotions, research money, and lateral appointments. This would explain why even non- donors tilt left. Republicans could suppress their ideological views by avoiding controversial topics, taking refuge in fields that have little ideological valence, focusing on empirical or analytical work, or simply writing things that they don’t believe.

This is an important point the authors make. When I first considered going on the market, I received many forms of conflicting advice. One of the bigger dividing points was whether I should pursue a position teaching constitutional law with my views. Some argued that I should pick something non-controversial, like Torts or Corporate Law, and maybe write about ConLaw when I got tenure. There are some academics who have taken this strategy. Others said this strategy was ridiculous, I would be stifling my own creativity, and punishing myself as an academic. I ultimately sided with the latter approach–in no small part because I had already written numerous ConLaw articles–and listed ConLaw as the top class I wanted to teach. I am thankful things turned out the way they did, but I’m sure others have taken the former approach.

Fourth, for the small number of “Republican” law professors, there sure were a lot of “Conservative” articles written. The article found:

Of the 780 articles in our dataset, 512 are liberal and 237 are conservative.

The authors don’t draw any conclusions here, but it would seem this small cadre of conservative law professors is quite productive. I don’t know quite how the numbers break down, as some liberals authored conservative papers, some conservatives authored liberal papers, and some neutral professors authored both. But, if I read the study correctly, “professors at elite law schools who make donations to Democratic political candidates write liberal scholarship, and law professors who make donations to Republican political candidates write conservative scholarship.” More likely than not, a liberal article was authored by one of the many liberal professors, and a conservative article was authored by one of the few conservative professors. Extrapolated, 24 conservatives law professors wrote (some large share of) 237 conservative articles–about 10 articles per professor. The remaining 132 liberal law professors wrote (some large share) 512 articles–about 4 articles per professor.

Update: After closer study, I see that the authors only considered a professor’s 5 most recent articles, so this wouldn’t really measure productivity in this way.

If so, this may confirm a piece of conventional wisdom I’ve heard many, many times–a conservative scholar must be able to outpublish any concerns from a tenure committee. In other words, if you publish enough quality scholarship, it can provide a rebuttal to any concerns from others.

Fifth, the authors find that  “net Democratic donors write highly ideological articles, whereas net Republican donors write articles that are distributed widely across the spectrum.”

According to our coding system, net Democratic donors write highly ideological articles, whereas net Republican donors write articles that are distributed widely across the spectrum. The average net Democratic donor writes 2.63 liberal articles on net, while the average article of a net Republican donor writes 0.17 conservative articles on net, which is even closer to 0 than the number of article written by non-donors, who on average write 1.44 liberal articles on net..

The authors state their position later:

In other words, our data suggest that Democrats in our sample do not write articles that are on balance neutral, but that Republicans in our sample may write articles that are on balance neutral. Forty percent of the articles written by Democrats (adjusted) could not be classified with high confidence, 51 percent of articles written by Republicans (adjusted) could not be classified with high confidence. If such articles are “neutral,” then Republicans wrote substantially more neutral articles.

So the few “Republican” authors, more likely, author “neutral” articles than the “Democrat” authors.

Fascinating article, with a lot of food for thought.

Update: Adam Chilton offered this reply, which he gave permission for me to post here:

In response to your first question, we unfortunately cannot say anything about the productivity of liberal or conservative professors. For each of the 156 professors in our sample, we coded their 5 most recent articles. So in our dataset, everyone is equally productive so to speak (although obviously some professors finished their 5 most recent articles in the last year and others took upwards of a decade to write the 5 articles we coded).
In response to your second question, of the 237 conservative articles in our dataset, the conservative professors wrote 91 of them, liberal professors wrote 109 of them, and 37 were written by professors whose ideology we were unable to code. So, liberal professors wrote a plurality of the total conservative articles, but there were many more liberals than conservatives in the sample. When we look at just the 110 articles that we were able to code with “high confidence”, this changes slight. Of those 110 articles, 51 were written by conservatives, 44 by liberals, and 15 by neutral professors.