What was the single biggest determinant of whether a Democrat voted yea or nay on the health care bill yesterday? Actually, it isn't too hard to figure out:
This reflects the breakdown in the roll call based on the percentage of the vote that Barack Obama received in each congressional district in 2008. Members who are retiring from politics are excluded; those who are vacating their seats to run for statewide office are listed by the percentage of the vote that Obama received throughout the state, rather than their home district. Those technicalities aside, the trend is pretty clear. All 12 Democrats running in a place where Obama received less than 40 percent of the vote decided against the bill, whereas 195 of 202 running where he received a majority voted for it (including 64 of 64 in places where Obama had at least 70 percent of the vote). The swing votes tended to come from those districts in which Obama received less than 50, but greater than 40 percent of the vote two Novmebers ago. (Although it makes only the slightest bit of difference, the Obama vote share alone was a better predictor than PVI, which accounts for the results of the last two elections.)
By contrast, the competitiveness of a Democrats' 2010 race, as determined by a consensus of the four major forecasters, made much less difference:
Although there is a trend there as well, it isn't nearly as definitive. Of the 22 Democrats running in races that are classified as toss-ups (or lean Republican), 14 in fact voted for the bill. Moreover, the likelihood that a member's race is competitive is of course related to Obama's standing in the district; once both variables are accounted for simultaneous in a regression analysis, the competitiveness of a race basically turns out to be irrelevant. There are a handful of Democrats, like Dan Boren of Oklahoma, running in heavily anti-Obama districts but who nevertheless have only token opposition; they were almost uniformly unwilling to vote for the bill. On the other hand, members running in some very tough races in places where Obama did even modestly well were usually willing to back it. At the end of the day, the explicit appeal the White House made -- that Obama's presidency depended on the outcome of the vote -- may have been persuasive to many Democrats, except those for whom Obama is unambiguously a liability.
This is not to suggest, however, that Obama was the only variable that mattered. Ideology -- as measured by DW-Nominate scores, mattered some too:
Roughly the 110 most liberal Democrats all voted for the health care bill; Stephen Lynch was the most liberal member not to do so, and his track record is actually closer to the midpoint of the Democratic party than to its left flank. This ideology variable still qualifies as statistically significant even if Obama's vote share is accounted for along with it.
Even though Democrats picked up enough "Stupak bloc" voters to get them across the finish line, a Democrat's position on abortion also made some difference:
This variable holds up as being modestly statistically significant even if an overall ideology score is included as well, and you do even better with more sophisticated metrics on a member's abortion position like his Progressive Punch score. In other words, the concerns that pro-life Democrats had about the bill appear at least to have been reasonably sincere.
Some other interesting variables, however, did not matter very much. For instance, the percentage of uninsured in each district had no systematic impact:
And the magnitude of campaign contributions from insurance-industry lobbying groups mattered only at the margins:
Basically, each Democrat's vote determined by two things: a member's confidence that Obama could be an asset to them (he tended to get the benefit of the doubt on this -- but only up to an extent) and frankly their conscience -- as it regards both health care overall and the side issue of abortion. Each Democrat wrestled with these things in his own way and there were certainly a few exceptions and surprises -- like Lynch or Mike Arcuri voting against the bill, or Betsy Markey (who was a no-to-yes flip) or the two pro-life members from West Virginia voting for it. But in retrospect, the vote might not have come at such a high price for Democrats if Obama had invested more political capital in the bill earlier in the process and made more explicit moral appeals for it -- as these were the things that seem to have been most persuasive to Democrats at the end.