I don't think I've blogged yet about the new paper by Seth Masket and Steve Greene showing the costs to individual Democratic Members of Congress for voting for health care reform during the 111th Congress. It's an excellent paper, and a useful corrective to the general conventional wisdom (among scholars of elections and Congress, at least) that voters are too inattentive to have their vote affected by actions taken by their representatives.
That said, the one caveat I have about this kind of research is that they only study two of the possible outcomes Members were faced with when casting their votes: legislation becomes law, and the Member votes either aye or nay. As I said way back early in the health care debate (can't find it right now, alas), it was likely that a lot of Democrats believed that their best course of action was to vote no while the bill passed. However, alas for the party, if too many did that, then they would wind up with no bill at all. Not only might they prefer having the bill pass even if it put some Members in electoral jeopardy, but it's also possible that a failed bill might have had severe electoral consequences of its own.
Now, we have no idea whether such an effect was possible, let alone whether that hypothetical effect was larger than the direct effect Seth and Steve estimate for voting for the bill. We also don't know whether passing the bill will ultimately help Barack Obama win reelection (which might incidentally help Democrats lower on the ballot in 2012 -- although it could then hurt the Dems in 2014! Hey, this stuff can get complicated!). So I'm not criticizing, at all, the Masket and Greene study. Just pointing out that it's also worth keeping in mind the larger context.