Paul L. Caron

Wednesday, March 2, 2016

Chilton & Posner:  Is Legal Scholarship Politically Biased?

Adam S. Chilton (Chicago) & Eric A. Posner (Chicago), An Empirical Study of Political Bias in Legal Scholarship, 44 J. Legal Stud. 277 (2015):

Law professors routinely accuse each other of making politically biased arguments in their scholarship. They have also helped produce a large empirical literature on judicial behavior that has found that judicial opinions sometimes reflect the ideological biases of the judges who join them. Yet no one has used statistical methods to test the parallel hypothesis that legal scholarship reflects the political biases of law professors. This paper provides the results of such a test. We find that, at a statistically significant level, law 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. These findings raise questions about standards of objectivity in legal scholarship.

Figure 3

Professors who are Democrats (adjusted)—shown in the left panel—have an average article ideology of -2.67 with a 90% confidence interval of -3.13 to -2.21. Using a t-test, we can say that this is statistically different from zero (p-value < 0.00). Professors who are Republicans (adjusted)—shown in the right panel—have an average article ideology of 0.17 with a 90% confidence interval of -0.72 to 1.10. For these professors, we cannot reject the possibility that the true net ideology of their articles is zero (p-value = 0.72). 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. ...

[I]f 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 leftleaning 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 nondonors 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.

Table 4

The data presented in Table 4 suggest that constitutional rights scholars are less ideologically diverse than other legal scholars. Among constitutional rights scholars, 77% are net Democratic donors, and 4% are net Republican donors. In the rest of the sample, 40% are net Democratic donors, and 20% are net Republican donors. It also shows that constitutional rights scholars are more likely to produce biased research (mean of -3.85 conservative articles) than Republican and Democratic scholars in other fields (mean of -1.35 conservative articles).

Legal Education, Scholarship | Permalink


The authors conducted the study by finding law profs who had donated to Democrats and looked at their publications. Then they did the same thing for law profs who had donated to Republicans. They had enough trouble finding the latter that they changed their selection criteria to get enough (R) data points.
If they had done this study with physical scientists, they would perhaps have found that donators to Democrats are more likely to write pro-Climate Change articles. So one interpretation of their results is that, just as reality has a liberal bias, so law has a liberal bias.

The coding of the publications is, to my mind, somewhat peculiar. In criminal cases, if the defendant wins, it is “liberal.” Likewise in civil cases where a woman or a minority wins, it is considered liberal. Where the coding system had omissions, the coders were to use their “common political intuitions.” To no surprise, the study authors said that they disagreed with some of the coders decisions. The authors did not change the decisions, nor, apparently, did they otherwise attempt to correct for such disagreements (by using multiple coders per publication, for example).

Given these problems, and others that the authors themselves bring up, I think their conclusions, which include “Law faculties may need ‘balance’” is not justified.

Posted by: Drew | Mar 5, 2016 1:20:53 PM

The p-value can't be less than 0.

Posted by: Lisa | Mar 3, 2016 7:44:41 AM

Is there a mistake in the initial summary? It says Dem donors write liberal scholarship, GOP donors write conservative. Figure 3 shows that is true for Dems but is NOT true for GOP.

Posted by: motionview | Mar 2, 2016 3:47:14 PM

This might be the most meaningless article I've read in years. Exactly how was bias determined? We have no idea from this post. That two politically conservative writers determine that liberals write biased articles is somewhat less than stunning.

Posted by: joseph | Mar 2, 2016 3:08:43 PM

And in other breaking news, the sky is blue.

It's nice to have numbers, but does this surprise anyone, even the slightest bit? Academia is a hive mind and progressives are as tolerant of dissenting opinions as the Politburo was.

Posted by: Todd | Mar 2, 2016 7:30:40 AM