Ian Ayres raises an interesting set of points over on the Freakonomics blog here. Seems we have both been running logit models this week of the failed House vote on the $700 bailout bill (make that the “economic rescue” bill). Ayres’ contribution is to isolate the impact of electoral vulnerability and membership in the Congressional Black Caucus on legislators’ votes. Both of these forces increased the likelihood that legislators would vote against the bill. Ayres concludes that because the more vulnerable members were more likely to vote against the bill, vote seekers for the next House vote should concentrate on those most vulnerable members. How? Ayres suggests a non-aggression pact between those incumbents and challengers to take the issue off the table.
Here at the Monkey Cage, we have a different perspective on the House members most likely to switch their votes. We also have a different perspective on the likelihood that challengers—particularly those in the most competitive races—would disarm themselves in the run up to the November elections. The most viable challengers tend to be more strategic than your run of the mill guy who wants to run for Congress (like the blind rabbi from New Jersey), and thus are unlikely to disarm.
The Ayres model is a good start. But we can go a step further by thinking more broadly about the range of forces likely to shape legislators’ votes on such a mammoth and unpopular government intervention into financial markets. As Mark Spindel and I noted in our post this morning, we find strong evidence that electoral vulnerability, ideological extremity (both strong liberals and strong conservatives), and impending retirement help to account for legislators’ votes. Like Ayres, we found that the extent of the subprime foreclosure problem in each state tends not to explain much of the variance in votes. That may just be an artifact of noisy data; high foreclosure rates in Oakland, CA, for example, are probably not well captured by the state foreclosure rate. We also find that members who represent districts with higher median household income tended to vote disproportionately in favor of the bill, even after controlling for their policy views and the competitiveness of their race. If we think about median income as a proxy for the degree of exposure to Wall Street, the finding makes a lot of sense. Who knows. Wall Street “Fat Cats” might even be a term of endearment, not scorn, in these more affluent communities.
But back to the implications of the model for this week on the Hill. Does this modeling exercise generate any useable knowledge? Off with the quotation marks around “science” in political science. It’s prediction time! (Yes, we’re going out on a limb that we will surely regret.)
We can use the model to generate the list of members erroneously predicted to vote in favor of the bill. Looking at Republicans (since we assume that Speaker Pelosi expects the GOP to cough up the lion’s share of vote switchers), these members tend to be moderate, electorally safe, with a strong finish in 2006 and no viable challenger in 2008. If we add in the two moderate GOP who are retiring (yet voted against the bill), this generates the following list of potential GOP vote switchers. Let’s see how close (or far) we get after Friday’s vote.
Whitefield (KY), Frelinghuysen (NJ), Gallegly (CA), Lobiondo (NJ), Ros-Lehtinen (FL), Hall (TX), Platts (PA), Miller (Michigan), Smith (NJ), Young (FL), Lucas (OK), Jones (NC), Aderholt (AL), Moran (KS), Ramstad (MN), Hunter (CA).
Think we’re wrong? Let us know, so long as you offer your own predictions.