Christian Grose alerts me to this piece (pdf), co-authored with Antoine Yoshinaka:
What are the electoral consequences of switching parties for incumbent members of Congress? Do incumbents who switch fare better or worse after their switch?…We find that the primary and general election vote shares for party switchers are not as high after the switch as before. Additionally, we learn that party switching causes the primaries in the switcher’s party and in the the opposing party (the switcher’s “old” party) to become more competitive in the short run. Over the long run, however, primaries in the switcher’s new party are less competitive than those in the old party before the switch.
Via email, he writes:
Does this apply to the current case of Specter vs. (possibly) Sestak? In some ways yes. Even though Specter v. Toomey was a primary Specter didn’t want this time and had trouble with 4 years ago, his choice to run as a Democrat does not guarantee an easy path to reelection via a Democratic primary. Our results suggest that Specter should worry about angry Democratic primary voters who may prefer Sestak to Specter in the Dem. primary.
On the other hand, our results might not apply to the Specter case, as Specter has time to shift his positions in order to cater to the Democratic primary electorate (and most other switchers didn’t have the president as strongly behind their switch). In many of the cases we studied, the switchers were lower-profile House members that switched midway through their 2-year terms, thus allowing only a few months, in some cases, to make nice with their new primary electorates.
I agree. Either scenario seems possible. If I had to guess, I’d say #2 is more likely.
All in all, I am quite impressed by how much political science research speaks to Specter’s switch, and how well it helps us understand his decision and what may result from it.