The correlation between district partisanship and legislator ideology is not perfect, but it is nevertheless strongly positive.
This analysis was performed by Robert Richards, a student research fellow at BYU’s Center for the Study of Elections and Democracy, in collaboration with me and other faculty. The writing is mostly his. Inquiries about this research should come to me (Adam Brown).
Utah legislators voted on 585 separate bills during the 2011 session. Citizens generally hope their legislators are voting the way they themselves would if given the chance, but even the most attentive citizen doesn’t keep track of every vote his/her legislator casts. This means that legislators can conceivably vote contrary to the wishes of their districts (knowingly or unknowingly) without attracting any negative attention.
We can actually test this: We can get a rough estimate of each legislative district’s ideology by looking at its gubernatorial vote in 2010. We can also give an ideology score to each legislator’s voting record using the NOMINATE scoring procedure we have written about in the past. Using those data sources, then, we can ask the question: Do legislators from more conservative districts vote in a more conservative way? Do legislators from more liberal districts vote in a more liberal way?
The figure below depicts these data. Each “D” represents a Democrat in the 2011 Utah House of Representatives, and each “R” represents a Republican. Legislators in more Republican districts appear further to the right and legislators with more conservative voting records appear higher up in the figure. For example, Rep. Jim Nielson’s (R-Bountiful) district gave 72% of its vote to the Republican ticket in 2010. Rep. Nielson’s ideology score for 2011 was 17, placing him on the conservative end of the legislature and near the ideological middle of House Republicans. On the other hand, Rep. Larry Wiley (D- West Valley) has an ideology score of -72, and his district voted only 37% Republican in 2010.
Using a statistical technique called ordinary least squares regression, I found that legislator voting records do have a strong relationship with district ideology.1 In all of the years I examined (2007-2011), the average legislator’s ideology score was markedly more conservative in more Republican districts. In 2011, for example, if a House district shifted from being 50% Republican to being 60% Republican, I would expect the district’s representative to move 26 points to the right on the ideology score. This is illustrated by the red line in the figure above. The results are similar for the Senate.
There are some important caveats to this analysis. As the graph shows, there are a lot of legislators who do not exactly fit the trend I’ve described. Some lie very far from the line, meaning their ideology is not what we’d predict based on their district partisanship. However, these legislators aren’t necessarily bad representatives of their district’s views. Partisanship and ideology are only two aspects of representation, and they’re not exactly the same thing. This means that the “perfect representative” may not always have an ideology score that lines up with his district’s partisanship in the way we’d expect based on this model.
We can see, then, that this analysis describes only a general trend, not a hard-and-fast rule, but the trend is optimistic: Conservative districts generally have conservative legislators, and liberal districts generally have liberal legislators. The correlation between district partisanship and legislator ideology is not perfect, but it is nevertheless strongly positive.2