No Room for Moderates in Trump's Party

By Colin McAuliffe (@colinjmcauliffe)

The night before the election National Review columnist and Trump critic David French wrote that Republican voters should not let Trump alone decide their vote in this years midterms. French implored his readers to judge each candidate on their own individual merits, saying “The best form of accountability is candidate-by-candidate, not party-by-party.” This sounds like a reasonable standard, but the truth is that nearly every Republican legislator is almost identical in their voting behavior, meaning that it is extremely difficult to judge Republicans on a candidate-by candidate basis. Every elected Republican is simply a foot soldier for Trump and his agenda, and therefore any vote cast for a Republican is a vote for Trump.

A few weeks ago, we compared how frequently Senators from the two parties voted with Trump to Trump’s vote shares in those states (these data were collected by FiveThirtyEight). This showed that Democrats are responsive to their general electorates, meaning that Democrats in redder states vote in line with the Trump administration’s position more frequently than Democrats in bluer states. The reverse was not true for Republicans, who vote with Trump nearly all the time regardless of whether they are in a swing state or a deep red state. The pattern holds for the House as well, as the chart below shows. Democrats are responsive to their general electorates, Republicans uniformly vote with Trump regardless of what kind of district they represent.


Conventional understandings of congressional voting behavior would predict representatives in swing districts to have more moderate voting records, but since Republican legislators are completely homogenous, that means that the supposedly moderate Republicans in swing districts are the most likely to be out of line with their voters. It’s also the case that these same Republicans were the most likely to get ousted last Tuesday.

While several outgoing Republicans may have recognized that performatively fussing about Trump’s personal conduct could give them some leeway while they quietly vote for Trump’s agenda, Tuesday’s election hardly changed the ideological composition of house Republicans In practical terms. Some commentators have argued that it is likely that following these losses, the Republican party will become more wedded to Trump, but the actual voting behavior of these members indicates that any daylight between moderate Republicans and Trump was mostly fiction. There is no downside to losing representatives who refused to use their leverage to reform their party from within, and with Democrats in their place, Trump’s legislative agenda is effectively halted.

We can certainly speculate about whether Republicans could have minimized their losses or even held the house if a few Republicans in swing districts had actually chosen to move to the center or stand up to Trump. However, since so few Republicans appears to have actually tried anything resembling moderation, it’s nearly impossible to test this hypothesis. It may simply be the case that any Republican in a district that was Trump won by less that 6 points or so had their fate written in stone. We’ll never know if moderation will improve Republicans electoral prospects until we see Republicans actually start moderating their voting behavior.

French’s standard sounds nice and it would be great if we could judge individual candidates based on their individual merits. You can certainly do this for Democrats, for whom there is clear variation in voting behavior that tracks with district competitiveness. However Republican legislators don’t seem to have individual merits. They vote as a package. If you vote for a Republican, you are voting for the whole Trump agenda. That’s exactly the point for the overwhelming majority of Republican voters, but #NeverTrump conservatives are going to have to accept that the only way to repudiate the president is to elect Democrats.

Colin McAuliffe (@colinjmcauliffe) is a co-founder of Data for Progress.

Ethan Winter