Research we conducted this week provides further evidence for the core argument of Election Reformers Network: that our rules themselves are a source of political dysfunction and partisanship.
Political scientists measure partisanship with a tool call the NOMINATE score, which ranks Representatives and Senators based on their roll call votes (with +1 being extremely conservative, and -1 extremely liberal). Whether a given district’s Congressperson has an extreme or moderate score is of course determined by political factors such as the demographics of the district. But we believe the mechanics of our elections also play a role, in particular the mechanics of primary elections and whether primaries are won with a majority, or with a small plurality in a crowded field. This theory directly relates to the election reform we’ve committed to first, Ranked Choice Voting, which is well designed to identify the majority-supported candidate in a crowded election.
To test our idea we compared US Representatives on the basis of their NOMINATE scores and the primary election result the year the representatives first won a seat in Congress (before the power of incumbency weighs in). The table below shows initial support for our hypothesis, that low plurality primary winners tend to be more ideologically Right or Left.
|Average NOMINATE Score (absolute value)|
|Reps with majority primary victories||0.410|
|Reps w < 35% in primary victories*||0.524|
To address the possibility that deeper red and blue districts inherently generate more crowded primaries, we used recent presidential election results to calculate an “expected nominate score” for each Congressional district and then measured Representatives on the difference between the actual and expected scores. This approach shows an even clearer pattern: Reps who entered Congress with a low plurality primary victory were significantly more partisan relative to their district than those who won majority support
|Average difference between real and expected NOMINATE scores|
|Reps with majority primary victories||0.233|
|Reps w < 35% in primary victories*||0.322|
(*Not including reps elected under “Top two” primaries in CA & WA)
We’ll bring on support from true political scientists to really test this analysis, assess if we have something we can publish, etc, but at a minimum this is a strong indication that candidates who win split primaries with a low plurality tend to be more ideologically extreme, which further strengthens the argument for Ranked Choice Voting.
RCV allows voters to support multiple candidates in order of preference and nearly always results in a winning candidate who has support of the majority. We are moving to a world of much more crowded elections, as grassroots organizations challenge central party control and as support for independents strengthens, and this context reinforces the need for RCV.
For more information on NOMINATE see https://voteview.com/data