More on the New York Times Hypothesis:

And for those running a mental regression on the New York Times Hypothesis, its not clear how to code this outcome:

The Supreme Court reached opposite results yesterday in a pair of cases involving the Ten Commandments, upholding a display of the commandments on the grounds of the Texas Capitol, while striking down displays in Kentucky courthouses. The rulings will be criticized from all sides; we would have preferred the Texas case to have come out the other way. But taken together, they are an important reaffirmation of the nation's commitment to separation of church and state.

And just to clarify--the precise hypothesis, as I understand it, is that the best predictor of the outcomes in Supreme Court cases is the preference of intellectual and societal "elites." Because the preferences of this group is difficult to measure, if one were to run a regression, the operational variable that one could use to proxy for this variable would be the views of the New York Times Editorial Board.

So the hypothesis is that the prevailing views of the country's elites is the best predictor of outcomes in Supreme Court cases, as opposed to other variables (such as political ideology, general public preferences, etc.). The 10 Commandments case would be hard to code on that basis. (The Washington Post is similarly equivocal, "The court's approach may not be philosophically satisfying, but in practical terms, it isn't a bad way to evaluate public religious monuments.").