By: Patrick W. Zimmerman

It looks as if the narratives about how moderate House Republicans got punished by the voters in the midterms because of their ideology or because of their votes on healthcare and tax cuts is a little hasty.

Why? Because if you look at the ideological positioning and voting record of ousted vs. returning Republicans (and Democrats, for that matter), in the House of Representatives, other than party affiliation, things look pretty random.  While one can make the argument that moderates got wiped out on both sides in the Senate, the House is all over the place.

The DW nominate score difference among the outgoing republicans is basically indistinguishable from those who won their reelections. Staying R: +0.494.  Departing R: +0.456.  Staying D: -0.384. Departing D: -0.394.  On the Republican side, if you squint real, real hard and are ok with a p value of 0.11, then you can talk yourself into thinking that centrism was loosely kinda sorta correlated with people leaving.

Well, maybe it’s a correlation with how people voted on the ACHA?  Or Tax cuts?  Did voters punish them for their votes on those two signature pieces of legislation during the 115^th Congress (note, this only applies to Republicans, since there was no variation at all among the Democrats; the universally rejected both in the House).  Nope, not that we can see.  Sure, of the small number of Republicans who flip-flopped (15 voted no on the ACHA and yes on tax cuts, 7 did the reverse, and 5 gave a nay on both), a number of them lost (11 of the 27), but the flipping seems a lot more like an act of desperation from people who knew that they were in trouble more than something that caused that trouble.  And those only represent 11 of the 72 Republicans who exited the 115^th Congress in 2018.¹

So what gives?

The short-short version

It looks a lot like Republicans who got ousted or retired before they were going to didn’t lose primarily because of their personal policy choices or ideology.  Assuming candidate quality on an individual basis is somewhat random for incumbents, this sure looks a lot like voters rejected the Republican party as a bloc, rather than individual candidates because of what they did or did not do vis-à-vis­ Trump or the GOP’s policy agenda.  It sure looks like suggestions that being a “maverick” or rebelling against one’s own party doesn’t earn the same kind of local cachet that it used to.

The results

To help measure things here, we’ve made an interactive chart of each member of Congress that was involuntarily removed, and then charted their ideological position, their position relative to their own party, and their votes on healthcare and tax cuts.  We are not considering representatives who voluntarily gave up their seats to run for another office (whether or not they won) or due to another administrative posting, as those reasons aren’t reflective of either rejection by the voters or the fear thereof.

How do the outgoing reps fit on the scale of ideology and legislative effectiveness?  Well, from both sides of the aisle, it’s a bit random.   There are some from both ends of the spectrum who either retired or were defeated, and a whole bunch in the middle of their party’s ideological blocs.

Mouseover for details.

To more closely explore the question of whether or not moderate Republicans got particularly hammered in the midterms, we calculated each member’s DW-nominate score relative to their own party, which we’re calling “position” for now, which is calculated using the z-score of their DW-nominate rating compared to the party average.

The average position of an ousted Republican?  -0.183σ.  That is to say, less than 2 tenths of a standard deviation from the mean.  New Jersey moderates Tom MacArthur (NJ-2) and Frank LoBiondo (NJ-3) both lost with positions more than 1.7σ towards the political center.   So did Steve Russell (OK-5, position of +0.03σ) and David Brat (VA-7, position of +2.13σ).

The average position of an ousted Democrat?  Even more towards the party average: -0.095σ.

How did the outgoing candidates vote on the ACHA, and did that seem to matter?

Mouseover for details.

In short: nope.  Not really.  Of the candidates that bucked the Republican party on healthcare (and there were only 20 in total), exactly 50% of them left office, but that doesn’t really square with the narrative that voters were mad at the GOP for trying to take away their healthcare and, thus, punished reps who voted that way.  They certainly didn’t reward Republicans who didn’t vote in favor of the AHCA.

How about tax cuts, then?

Mouseover for details.

Again, no real pattern here; just under 50% of the Republicans who rebelled against their party leadership (all but one from the high state-tax states of NY, CA, and NJ) managed to survive November 2018.  Sure, Republicans who voted against the tax cuts didn’t do great, but that doesn’t really mean that voters were punishing those votes in general.  They sure didn’t hate on the Democrats who voted that way.

So what gives?

It looks like a couple of things are at work here.

• It was a blue year. Democrats won the popular vote by 8 points, which means that sentiment was going to hurt Republicans anyways.
• People seem to be voting less and less for individual candidates than for party identity. Split-ticket voting has been on the decline for decades, and 2018 continued the trend. It would make sense that in such an environment voters would pay less attention to how the candidate on the ticket in front of them voted than what the party they represent enacted and, in a symbolic or culture-war sense, what identity that party stands for.

Notes:
¹  the GOP lost a net 40 or 41 seats in the 2018 midterm elections 35 or 36 of those 72 outgoing members of the House were replaced by other Republicans, and 3 were balanced out by districts that flipped in in the other direction (the uncertainty is due to the ongoing election fraud investigation in NC-9.  If Mark Harris eventually wins the seat, he will be replacing outgoing GOP rep Robert Pittenger, who he beat in the primary).^