I just wanted to add a couple of thoughts to yesterday's class discussion. First, the criteria from Baker v. Carr are certainly relevant to determining whether a case presents a non-justiciable political question, but they are dramatically overinclusive, making them relatively useless as a predictive guide as to what a court will do. The first two criteria--whether there is a textual commitment of the matter to another branch, and whether the issue is not amenable to judicially manageable standards--seem to be the most important.
Second, the courts seem to use the political questions doctrine in practice as a sort of safety valve, a means to staying out of a controversy (or type of controversy) where the judiciary's involvement is apt to cause it some long-term institutional damage. This might be because courts would have real difficulty deciding cases according to rules that look at all judicial in character rather than legislative; that is, the judiciary's involvement might make the courts look especially political (or results-oriented), thus damaging its reputation. Or it might be (as Brian mentioned in class yesterday with respect to Ashwander) because the Court realizes its decision might not be obeyed, and thus would expose it as powerless.
Third, and relatedly, it makes some sense to think of the political questions doctrine--though it is a constitutional rule, derived from the "case" or "controversy" requirement of Article III--as working in practice as a sort of constitutional avoidance principle, much like those listed in Ashwander. It is a dramatic avoidance principle, as it entails the judiciary's complete withdrawal from the field, leaving resolution of the constitutional question entirely to the other branches. But it operates as means of avoidance all the same, and is invoked by the courts for largely the same reasons.