Liberal bias at the American Bar Association?

Kobi pointed me to this news article that discusses this research article by Richard Vining, Amy Steigerwalt, and Susan Navarro Smeicer which claims that the American Bar Association has a liberal bias in its evaluation of Supreme Court nominees. They write:

We [Vining et al.] find both that, all else being equal, Democatic/liberal nominees are more likely to receive the ABA’s highest rating of “Well Qualified” than their Republican counterparts, but also that the ABA relies on more traditional measures of professional qualifications, such as prior experience as a judge or Circuit Court clerk, when rating nominees to the federal appellate courts. Our results lead us to conclude that the ABA should take affirmative steps to ensure liberal candidates are not being unconsciously favored and rated.

I have mixed feelings about this research. On one hand, I can only assume that the leaders of the American Bar Association are mostly liberal Democrats and so it makes sense that they would view liberal Democratic judges more positively, much in the same way that I’d expect the Chamber of Commerce to generally have conservative Republican views. I also have mixed feelings about the quasi-governmental role played by an organization such as the ABA: on one hand, it seems vaguely inappropriate; on the other hand, I presume these people do take their job seriously, and there are a lot of judicial nominees to look at.

Getting back to the Vining et al. article: as noted above, I’m predisposed to accept their basic claim of liberal or Democratic bias, but I don’t really buy their data analysis. Their measure of partisanship is fine–they compare Clinton’s nominees to those of Reagan, George H. W. Bush, and George W. Bush–but, as a measure of ideology, they use the ideology of the same-party senators from the home state of the judicial nominee. I don’t see this as making much sense, but it’s not a fatal flaw, as they replicate all their analyses using president’s party and get the same basic results. That said, I think that at the very least their claims are really about partisanship, not ideology.

Given the basic finding–that Clinton’s nominees were more likely than Reagan/Bush/Bush’s to get high ABA ratings, after controlling for some background variables–I wonder if it can be explained by differences among the judges being nominated. I just don’t know enough about the courts to even try to say anything definitive on this, but surely Rehnquist, Bork, Scalia, and Thomas are more partisan than Breyer and Ginsburg, no? This is not to make any statement about whether Bork et al. are more conservative than Clinton’s nominees are liberal, but just to suggest that there maybe are some systematic differences between Clinton’s nominees and those of his Republican predecessors and successor. If the ABA is responding negatively to partisanship and positively to moderation, this could show up in their model as a bias toward Democrats (i.e., Clinton’s nominees) without actually implying anything about an ideological tilt than needs to be corrected.

P.S. Just a quick comment on the graphs: more could be said here, but to start with, I’d recommend taking a graph such as Figure 2 and doing the following:
– Restrict the y-axis to [0,1]. Probabilities can’t be below 0 or above 1.
– Clean up the axes: label the x-axis at 0, 5, 10, and label the y-axis at 0, 25%, 50%, 75%, 100%.
– Display the lines cumulatively, i.e., Pr(0), Pr(0 or 1), Pr(0 or 1 or 2), etc. Then you can label the areas between the lines. See Figure 10 of this article for an example of what to do.
There’s lots more to be said about the displays; I’d recommend thinking carefully about what information you want to convey.

4 thoughts on “Liberal bias at the American Bar Association?

  1. I'm reminded of the Colbert quote:

    Reality has a well known liberal bias.

    Most likely the Republicans are just nominating less qualified people (Harriet Miers?).

  2. Another explanation is that there is a supply of judicial candidates with varying levels of quality. All presidents nominate judges from the highest quality to the lowest quality. Thus, as tenure of a party increases the quality of the marginal and average judicial candidate declines. This would yield a similar finding of "liberal bias" simply because of a mechanical association between tenure and quality.

  3. I was going to write what Martin wrote above about selection bias of Democratic vs. Republican nominees. I guess that one could empirically examine whether the quality of nominations goes down as a party makes more nominations (particularly consecutive nominations).

    Also, I think that there is a long-standing tradition in nominations (that was strained under G.W. Bush) that a the senators from the state are generally consulted before a nomination goes to the Senate, so it might make sense that the ideological/partisan position of the home senators is a good metric.

  4. you might do better comparing alito and roberts to rehnquist, bork, scalia and thomas. here, the levels of conservatism are not incredibly different, but there are different levels of partisanship.

    other factors: chances of the nominations being stopped. democrats didn't seem at all likely to "bork" either roberts or alito, they just didn't have the political power. thus an ABA unqualified rating might just seem like sour grapes. with bork and other reagan appointies, the democrats did have the power to at least put up a credible fight. i'm not sure where republican political will stood in relation to ginsberg and breyer.

    this relates to the potential for the time in which the nominations took place being a factor. maybe the ABA was more liberal during the reagan administration than it was during the bush 43 administration. also, if we limit this to SCOTUS appointments, we have a very small sample size.

    where the study might actually have something to say is in non-supreme-court appointments. here, there are enough judges to have meaningful sample sizes and really understand what is going on.

Comments are closed.