Median Voter Theorem Revisited
Via Twitter, Garrett Jones challenged me to revisit out dispute from back in June about the explanatory power of the median voter theorem. He seems to feel that Olympia Snowe’s vote in favor of health reform offers strong support for his position.
To which I say . . . sure, Snowe’s vote is a reminder that public opinion has some bearing on legislator behavior. Nevertheless, I would just repeat my initial points. Richard Burr and Kay Hagan have the exact same constituents, but different views on health care. Same with Chuck Grassley and Tom Harkin. Look at every state represented by senators from different parties. Insofar as the median voter theorem has explanatory power, we would expect those senators to be pulled toward the same position on health care. But that’s happening in either zero of those states or else only in Nebraska. Or consider the public option dispute among Democrats. Webb and Warner are listed with different views, so are Dorgan & Conrad, and Dodd & Lieberman.
The point is, for a first cut at predicting legislator behavior you should look at their partisan affiliation, not the views of their constituents. Beyond party affiliation, there is evidence that constituent public opinion makes a different. But legislators have a fair amount of leeway.
9 Responses to “Median Voter Theorem Revisited”
Register
RSSLogin
Topic Cloud
ArchivesBlog RollAbout Matt YglesiasContact Me
Donate
October 15th, 2009 at 8:34 am
Ohio is another good example. Brown is vocally in favor of a strong public option. With Voinovich all signs are he would not consider voting for the Baucus bill which does not have a public option and is watered down. And remember, Voinovich doesn’t even need to worry about ever facing a primary challenge again.
October 15th, 2009 at 8:47 am
I would guess that there is some path dependency at play with each individual politician. Each politician has a different history of specific positions and past behavior through which they are viewed by their constituents. The public opinion of their constituents probably influences the positions they take but the idiosyncrasies of each politician might mean that even two similar politicians will have to respond differently to identical shifts in their constituents views. In other words, there is no reason to expect that changes in public opinion will be reflected one-to-one in a politician.
October 15th, 2009 at 8:56 am
Dude hits 21 homers in 82 games, he can say want he wants about the median voter theorem.
October 15th, 2009 at 8:58 am
I believe the the MVT doesn’t actually refer the the constituents’ preferences, it applies to the Member’s preferences. Constituents often don’t vote based on one issue, they vote for a Member based on many factors, which includes but is not limited to voting record. The extent to which a Member’s vote reflects the preferences of his/her constituents is constrained only by the degree to which they perceive that vote will affect their chances for re-election. Iowa voters like both Grassley and Harkin largely because they have seniority and influence, and keeping them in office increases seniority and influence.
Or, Members vote how their constituents want if: a) they agree, or b) voting differently would be detrimental to re-election. This doesn’t conflict with the MVT, but adds a dimension to the preferences by factoring in the Member’s preference to get re-elected.
October 15th, 2009 at 9:15 am
Does this relate at all to your thesis with regards to Republican party discipline? Do Democrats have more leeway? Do they use this leeway to express median voter preferences? Does this help them win elections?
October 15th, 2009 at 9:17 am
I think Jack Heckelman had a paper on this a few years ago. The MVT predicts something different than Yglesias thinks about senators. Since senators aren’t elected at the same time, if for any reason one happens to differ from the median voter, then the median voter wants the other senator to differ too, in the equal & opposite direction. So if I’m a median voter & a moderate Dem and one of my senators is a liberal Dem, then I want to give the other slot to a moderate Republican. There is no reason why I would push them towards each other at all.
October 15th, 2009 at 9:18 am
That’s not the US Capitol…
October 15th, 2009 at 10:01 am
MVT is interesting, but here’s a simpler possibility: Take a look at Snowe’s bio on Wikipedia.
There’s a LOT of illness/death in her family background (for instance, she was orphaned as a kid), which was by no means economically upperclass. I suspect she’s had a lot of firsthand experience w/ the medical & insurance probs faced by regular folks and this, in part at least, had something to do with her vote.
October 15th, 2009 at 2:22 pm
Don’t forget about Bob Bennett (working… with Wyden?) and Orrin Hatch (working… to create the least productive amendments possible)