How Democratic Incumbents React to Primary Challenges
Activists from both parties have long claimed that mounting a primary challenge to the incumbent, even if unsuccessful, will build up pressure to move the incumbent in that direction. A recent study showed that this was true during the campaign, and incumbents that lost a primary would immediately send out more moderate tweets. While this is helpful for campaign posturing, I wanted to examine the more long-term effects. This analysis, which focuses only on the Democratic party, will look at whether an incumbent adopts more conservative or liberal positions after a primary challenge from the right or the left. It will also look to see if the magnitude of any ideological change is impacted by how well the challenger did against them.
Methodology Discussion
Usually when I look at ideology, I prefer to Adam Bonica’s DIME measure, like I’ve used in this article: https://medium.com/@rudnicknoah/a-dimes-worth-of-moderation-84a1fa45d53b. I like to use this in my model because this measure ideology by seeing who has donated to a campaign, so you get a view of how liberal or conservative a candidate is perceived to be which matters in elections. But for this, I wanted to look at substantive changes to the ideology so I borrowed the 2018 ideology score based on cosponsoring and voting for legislation provided by GovTrack here: https://www.govtrack.us/congress/members/report-cards/2019/party-house-democrat/ideology.
All of the results were taken from the 2018 Ballotpedia results and whether a candidate was to the left or the right was taken from a collection of defunct candidate websites and an endless constellation of local news sources for which I will remain thankful for their coverage once again. Because this will later look at incumbent ideological shift, members of congress that were successfully primaried are not included. I know that will throw up the first section but in 2018 that was too small of a number to move the medians.
Who Gets a Challenger?
The first visual is a set of box plots that maps the ideology of every Democratic House incumbent for whether or not they received a left-wing challenger or a right-wing challenger.
This chart surprised me because it shows that representatives across the ideological spectrum got left or right-wing challengers in 2018 and overall, they weren’t concentrated. Left wing incumbents did not get right-wing challengers more often and vice versa.
The next two charts plot out the totals received by left wing or right-wing candidates against the ideology of the candidate on the x-axis. A reminder that the farther to the right on graph, the more conservative the incumbent.
Combined with the earlier chart, it seems that not only are ideological challengers not targeted to the opposite wing of the party, but they don’t perform better based on the ideology for the incumbent. There is no good trend fit like I had expected with an increasing slope line on the left-wing plot and a decreasing trend line on the right-wing chart.
Now that we have looked at who receives a challenger, I wanted to see if the strength of a challenge would move an incumbent in that direction. To mark the change, I took the 2020 report card value and subtracted that from the 2018 report card. A negative value means that the incumbent has moved to the left and a positive value means that they have become more conservative. This is the metric on the y-value of the axis for the final chart below. The x-axis is the net right and left candidate share in a primary sorted out into bins. For example, if an incumbent faced a left-wing challenger who received 25% of the vote and a right-wing challenger received 4% of the vote, the net would be 21% left. This would put the data point in the 20% to 30% Left bucket. This net measure where someone faced challengers from both sides only happened a few times but I believe is useful when accounting for cross-pressure and strategic maneuvering for the winning incumbent.
This chart shows a few key takeaways, with the first two being the most important in my opinion:
· A very narrow loss drove incumbents to become more liberal than the median representative
· A decisive but close win for the incumbent where the left-wing nets 30% to 40% moves the Democratic incumbent to become more moderate after the primary ends
· A no-hope challenger (0% to 10%) sent the incumbent moving sharply in the other direction. A single digit left wing challenger had the same median effect as a right-wing challenger taking double digits.
· The n is small but a decent performance for a right-wing candidate had a relatively larger effect driving an incumbent in that way than a decent performance from a left-wing candidate. A left-wing primary challenger that took between 10 to 30% still drive an incumbent to the left but not necessarily more than the median of leaving an incumbent completely uncontested.
From this data and cycle, it seems that if you wish to challenge an incumbent in the House for the sake of dragging them along the ideological scale, it is best to put up a real fight and not wind up in the single digits and that if the race looks close you should get it as close as possible or just win because if it falls noticeably short there will be a backlash. Whether that is because the left upsets them or that they need to rely on crossover votes in the primary will take more analysis but the effects here are clear in how much of an outlier that group is.
