Sotomayor decisions "defy depictions of her record as falling neatly into either a liberal or conservative category"

Yesterday's Odd Washington Post Story. By Roger Clegg
Bench Memos/NRO, Jun 09, 2009

The Washington Post yesterday had a long story arguing that Judge Sotomayor's decisions on "race, discrimination and the law" somehow "defy depictions of her record as falling neatly into either a liberal or conservative category." The article picks the eight cases that involved divided rulings. But, accepting arguendo the Post's methodology, I don't see how the article shows that Sotomayor is not a predictable liberal.

In six of the eight cases, it seems to me that the Post acknowledges that Judge Sotomayor came to a liberal result. Of the remaining two, one involved a policeman who was fired for mailing out racist and anti-Semitic fliers. Judge Sotomayor, in dissent, wanted to rule against the police department — just as the ACLU's New York affiliate had urged the court to do. Sure, her position favored a bigoted policeman, but she also wanted to use an aggressive interpretation of the First Amendment to tie the hands of the police department.

That leaves Hankins v. Lyght, involving a minister who sought to invoke the Age Discrimination in Employment Act when his church forced him to retire at age 70. The district court dismissed the case, but the Second Circuit remanded for the judge to consider whether applying the Religious Freedom Restoration Act (RFRA) would change the outcome; Judge Sotomayor dissented. Here's a link to the appellate opinions. Judge Sotomayor seemed especially unhappy that the panel went out of its way, in her view, to uphold the constitutionality of the RFRA; but that's an odd sort of "violat[ion] . . . of judicial restraint" (her words), and since RFRA was enthusiastically supported by the religious right, I'm not sure that her position is necessarily the un-liberal one. In any event, the appeal did not involve race and was more about procedure than substance, and it's hard to argue that this one ambiguous case out of eight proves the Post's thesis.