Senator Lee has the distinction of being the most ideologically extreme senator in the 113th Congress.
At first glance, it may seem as though Utah’s Tea Party Senator, Mike Lee, and Massachusetts’ liberal firebrand, Elizabeth Warren, may not have much in common. In terms of policy, they disagree on almost everything. However, there is one area in which the two are quite similar. Lee and Warren share the distinction of being the most ideologically extreme senators in their respective parties.
How do we measure the ideological positions of senators? Political scientists often use a measure of legislators’ ideology called NOMINATE (Nominal, Three-Step, Estimation) scores. These scores are based on the roll-call votes cast by legislators and condense the thousands of votes cast in the House and Senate into a single score representing a legislator’s typical ideology. The scores range from -1 to 1 with lower values indicating more liberal voting while positive scores indicate more conservative voting records. Using this measure, Lee scored a .991 in the 113th Congress while Warren scored -.622. There were no senators with scores larger than Lee’s or smaller than Warren’s. The chart below shows a histogram of senators’ ideology scores and notes Lee and Warren’s scores at the ideological extremes.
Another way of thinking about a legislator’s ideology is to consider how far she is from the typical legislator in her party. By this measure, Lee was .56 from the median Republican (the median Republican senator’s score in the 113th Congress was .43) while Warren was .29 from the median Democrat (the median Democratic senator’s score in the 113th Congress was -.33). Using this measure, Senator Lee has the distinction of being the most ideologically extreme senator in the 113th Congress.
Given that this blog focuses on Utah politics, lets take some time to consider Senator Lee’s ideology score in greater detail. One reason that Lee could appear so far from the typical Republican senator is that he is voting in line with his constituents (Utah voters), who are also more conservative than the typical American voter. To investigate this possibility, I plot the average ideological position of each state’s voters as measured in a large nationwide survey in 2010 (the year Lee was first elected to office) and each Republican senator’s ideology score. The vertical axis shows Republican senators’ ideology scores from the most moderate Republicans (Collins from Maine) to the most conservative (Lee). The horizontal axis shows the average ideology of each state’s electorate from moderate (South Dakota) to conservative (Louisiana). The two arrows point to Senators Lee and Hatch.
The figure shows that Lee’s conservative voting record is not necessarily because of Utah voters’ policy preferences. Orrin Hatch, who represents the exact same voters as Lee, has a much more moderate voting record (he is lower on the vertical axis). Furthermore, states like Oklahoma, Louisiana, and Alabama have more conservative voters (larger values on the horizontal axis) but their senators have acquired voting records that are more moderate than Lee’s.
Let’s also consider Lee’s voting record in the context of other Utah politicians. The figure below shows the ideological locations of 7 of the last 8 people to serve in the House and Senate from Utah. Mia Love is not shown because she has yet to acquire a voting record. Lee is by far the most conservative of the bunch.
How does Lee’s voting record compare to other legislators from Utah? For a more historical picture, the histogram below shows the ideological scores for all 49 representatives and senators who have served from Utah since Utah became a state in 1896. Again, Lee stands out among his Utah colleagues as the most conservative member of the Utah delegation ever.
Does any of this matter for the upcoming 2016 election? Because of Utah’s conservative electorate, it is unlikely that Lee would lose to a Democrat in November. However, given the recent changes to Utah’s caucus and primary system, it is possible that he could face a difficult and expensive primary challenge in 2016. Political science research suggests that legislators who are “out of step” with the ideology of their voters are more likely to face challenges in the primary and general elections.
To investigate this possibility, I first use a statistical model to predict the relationship between a Republican senator’s primary electorate and her ideology score. I then measure the degree to which the actual values from the data deviate from the model’s prediction. The distance between the data and the model prediction gives us a rough estimate of how “out of step” a legislator is from her primary electorate. Positive deviations indicate legislators who are more conservative than their primary electorate. Negative deviations indicate legislators who are more liberal than their primary electorate.
The figure below shows that Lee has the largest deviation score of any Republican senator in the Senate today. Furthermore, the deviation score is positive, indicating that Lee is more conservative than we would expect from the composition of the Republican primary electorate in Utah. This is unusual since we expect candidates to strategically position themselves between an ideologically extreme primary electorate and a more moderate general electorate. Consequently, Lee is vulnerable to a moderate challenger in a primary election.
What this tells us, contrary to the popular belief of “extremes on both sides”, is that the most extreme on the right, and objectively extreme. And that the most extreme on the left, Warren, is…pretty much the center left. In other words, while there IS a “far right”, “far left” is a myth, at least when it comes to the senate.
Actually you are misreading the data. Nominate scores are bases on voting data data and sense the majority determines what gets voted on nominate scores are always skewed toward the majority.
Are you saying that since the nominate scores are skewed toward the majority, and the recent majority has been liberal, that all these results are skewed somewhat in the liberal direction?
(previous comment intended to be addressed to smo…)
Nice work Mike! (both Mikes!) I love the firestorm of discussion you have sparked but I’m disappointed that few have bothered to actually read you post and think about what the data means and where it comes from.
Tony R. above: I don’t think you can draw that conclusion as the scoring of votes is inherently relative to other votes instead of absolute against a non moving standard. In other words the base line can move and there is no way to substantively evaluate ideology. This is merely a measure of consistency in voting along partisan lines. Perhaps the better conclusion is that Mike Lee is more likely to vote along party lines than any other Senator and that includes more than Warren. But that doesn’t speak to the qualitative analysis of the particular items they are voting on and whether those policy items are far right or far left.