What Would Happen if Thirty House Members Resigned?

G. Elliott Morris

Dec. 07, 2017

Categories: 2018 Midterms Tags: house midterms




Republican Representative Trent Franks, of Arizona’s 8th Congressional District, just resigned amid rumors of inappropriate conduct. He will leave behind a House seat that voted for Donald Trump by a 22 point margin. There will likely be a special election in Frank’s district (and my updated forecast model suggests Democrats will have a 12% chance of winning the seat), but AZ-08 is not the big takeaway from this story.

Trent Franks will join at least two other current House members (TX-27’s Blake Farenthold and NV-04’s Ruben Kihuen) who face calls to resign amid allegations of misconduct. And today, of course, Minnesota Sen. Al Franken announced he’ll retire “in the coming weeks”. The first of the three (Farenthold) occupies what would be a Likely Republican seat if he resigns. But Kihuen and Franken both could leave possible openings for Republicans that are looking for crucial buffer seats to mitigate losses next year.

But, like a good TV infomercial, “wait, there’s more!”

Democratic Strategist Michael Trujillo tweeted that Washington Post and CNN reports are working on investigations into up to 30 other members of Congress for sexual misconduct. Given the torrent of allegations over the past week this is not all that surprising, though to be clear, the report has not yet been validated.

But what if Trujillo’s story is true? What if thirty Congressmen really do resign in the wake of their scandals? Let’s assume (1) the 30 members are from the House — because, let’s face it, it’s the better Chamber (I have James Madison on my side) — and (2) that neither party’s Representatives are more/less likely to engage in this behavior. Undeniably, thirty members exiting the lower chamber would cause a huge disruption in our expectations for 2018.

To start, there have already been quite a few analyses of the impacts of retirements on the House. They note a few things: First, as representatives retire their party loses a significant 5-6% “incumbency advantage” in our expectation of that district’s vote. And second, given that (1) the average representative is a Republican and (2) Republicans overperformed Hillary Clinton’s vote share in most districts, we should expect the average retirement to benefit Democrats. The benefit ends up being 0.5 to 1 percent on average.

What happens if thirty representatives retire? First, let’s take out the 84 women in Congress and the 29 male retiring members. That leaves 312 representatives for us to randomly choose retirements from.

Using the power of the forecast model I can predict what will happen in each congressional district if a representative retires or if they stay put. Randomly picking 30 of them, I can compare the average forecast now with the average forecast after retirements. However, this one test won’t be enough to get a good measure of the true aggregate increase in vote share after 30 retirements. So I repeat that process 100,000 times and average all the different estimates.

Using this method, the average difference in district vote share after 30 retirements is a 2.1% boost for Democrats. To be sure, that’s an average — whatever combination of retirements/resignations do come could provide anywhere from a 6% boost to Republicans to a 6% boost for Democrats.


At any rate, the intent of this post was to show yet another example of the insights we can glean from my forecast model. In this case, it seems like we can expect a torrent of retirments to help two groups:

If the dam does break for misbehaving Congressman, look here for some analysis of the implications.






comments powered by Disqus