About this Blog
About The Author
Email Me

RealClearPolitics HorseRaceBlog

By Jay Cost

« Obama's Polarized America | HorseRaceBlog Home Page | Should Obama Be Faulted for the Lack of Bipartisanship? »

Should Republicans Support Toomey's Challenge to Specter?

Stuart Rothenberg's column today argues that Chris Dodd, rather than Jim Bunning, is the most vulnerable senator up for reelection in 2010. Rothenberg has a good point, and there is little doubt that both Dodd and Bunning are in trouble. But so also is Arlen Specter, who is headed toward a tough primary battle with former representative Pat Toomey.

Specter is feeling the heat, so much so that he has already released an ad against Toomey, which he subsequently had to walk back. He has good reasons to be nervous. Toomey mounted a robust challenge in 2004, and Specter squeaked out a narrow victory.

Specter's chief problem is that Pennsylvania's primaries are closed, meaning only Republicans can vote. This could make the difference because Republican registration has been falling off. At primary time in 2004, Democrats held a 500k registration edge. Last November, it was 1.3 million, thanks to new registrants and party switchers. Presumably, the voters who have drifted to the Democratic Party are more moderate - and thus more amenable to Specter. So, the remaining Republicans are presumably now more conservative, and more amenable to Toomey.

Pennsylvania is a difficult state to represent because it is so diverse. It's a bit rural, a bit urban, a bit industrial, a bit post-industrial. And then of course there is Philadelphia. Arlen Specter has dealt with this problem by racking up a studiously moderate voting record. His lifetime ideological score is 0.06 (where -1 is entirely liberal, 1 is entirely conservative). This is identical to the score of the late John Heinz, but much more moderate than Rick Santorum, whose 0.349 score made him a darling of conservatives, but a fish in a barrel in 2006.

Specter's persistent political problem is the fact that a not insignificant minority of the state's population is conservative, especially in the central and western portions of the state. This presents an opportunity for an ambitious candidate like Pat Toomey.

However, is it good for the party for Toomey to challenge? Obviously, it is good for Toomey - and many conservatives have become frustrated with Specter over the years. So, they'd like to see him go. But frustration is more an emotional response than a rational one. Specter's lifetime voting record has been moderate, but he can win the state - and he has never failed to side with the GOP on the all-important question of organizing the chamber. Toomey's record would be more conservative, but his chances of victory are much lower. Additionally, a tough, negative primary battle might damage both of them.

I want to put some hard numbers to this - or more specifically, allow you to do that. I have configured the following spreadsheet. It calculates the expected ideological score of the next U.S. Senator from Pennsylvania, whether it be Arlen Specter, Pat Toomey or somebody else.

To do that, it weighs seven relevant factors against one another. The first three we hold constant:

(a) The expected ideological score of Specter. We'll hold this constant at 0.06, his lifetime score.

(b) The expected ideological score of Toomey. He had a very conservative 0.694 score when he was in the House. However, he'd have to moderate in the Senate. Let's assume that he would be as conservative as Santorum, which would put him at 0.349.

(c) The expected ideological score of the the Democratic challenger. Let's assume he would be as liberal as Bob Casey, Jr., which would put him at -.304.

The final four statistics are yours to manipulate, though I have put some baseline numbers in to get you started:

(d) Specter's chance of defeating Toomey in the primary.

(e) Specter's chance of winning the general if Toomey challenges Specter.

(f) Specter's chance of winning the general if Toomey does not challenge Specter.

(g) Toomey's chance of winning the general.

[Disclaimer: the baseline figures are not my actual estimates. They're just there to get you started.]

As mentioned above, the ideological scores here go from -1 (perfectly liberal) to 1 (perfectly conservative). They're based on the DW-Nominate methodology that is a mainstay of political science research.

The goal is to find reasonable numbers so that (a) Toomey challenges Specter and (b) the Senate is made more conservative as a consequence. I tried my hand at this for half an hour or so, and the only reasonable situations I could find where the Senate shifts to the right are where Specter's chances of defeating Toomey increase.

Final word. As you noodle with this, remember that in 2004 Toomey and Specter spent a total of $20 million between the primary and the general election. That number presumably will be higher next year - so whatever movement to the right you can generate is purchased at a very high price, with dollars that could go to help other Republican candidates.

-Jay Cost