Rumors of the Republican Party’s Demise is Greatly Exaggerated

The Grand Old Party isn’t going away anytime soon.

Peter Licari, PhD
6 min readSep 29, 2018
Perhaps a tad premature.

Recently, I happened across an interesting piece from FiveThirtyEight talking about what the gender gap might look like in the 2018 midterms. For those who didn’t know, over the last few decades, women have been more likely to support Democratic party candidates over Republicans. Political scientists have been calling this difference in support the “gender gap” — because scientists in general tend to be very literal in their naming choices. There are a multitude of reasons why (Democrats tend to run more women, their platform tends to be more pro-choice, etc) but the piece argued that the gender gap may actually be shrinking in 2018. Not because women are more likely to vote Republican this time around but because some evidence suggests that men are becoming less likely.

This erosion of support among a key GOP constituency (men are much more likely to identify and/or lean Republican then women are according to recent Pew Research data) is commensurate with a number of articles coming out over the last few months claiming that the base, writ large, is shrinking. And many of these write-ups are strongly implying that President Trump is responsible. “Trump owns a shrinking Republican party” claims the Brookings Institute. “As Trump Governs For His Base, The Republican Party is Shrinking” writes Rantt. Others drop the pretense and directly insinuate that it’s the President’s doing like this Washington Post Op-Ed writing “Evidence Trump is Shrinking the GOP.”

The idea certainly makes a lot of intuitive sense. The President can’t seem to go a day without lying — multiple times. His braggadocious remarks managed to get many in the UN to openly laugh at him. His landmark legislative achievement to date (the 2017 tax reform bill) is pulling lukewarm support at best from his base. Even members of his own administration admit to trying to thwart his baser impulses and have apparently considered invoking a constitutional amendment that would expel him from office. If ever there was a President that could single-handedly repel people from their own party, it would be this one, right?

Not right. Rumors of the GOP’s demise (and Trump’s finger on the trigger) are definitely over-dramatized. Or downright wrong. Take your pick.

“It would be just as accurate to say that ‘President Trump presides over a shrinking Democratic party.’ And not even Fox News is audacious enough to run that headline.”

First and foremost, one should be pretty suspect whenever people take polling data and assume that people are changing their minds/party affiliations/attitudes/air filters (friendly reminder for you to do that, by the way). The vast majority of these polls don’t track individuals over time. They take different samples of the population and track the differences between them. The people they asked last time could very well be continuing on unabated in whatever was measured. We don’t know. We haven’t asked them. You can really only say that the population’s characteristics appear to be changing over time — and even then, it’s rarely statistically significant.

But with that standard warning in place, it’s pretty easy to make it appear that there is an exodus from the Republican camp. You may have seen charts tracking Republican party identification since 2016 or reading someone’s summary of Republican party ID over that time frame. If you limit it to just that time period, it definitely seems to be on the decline.

The problem, however, is that this sort of myopic analysis completely disregards longer trends in the nation’s partisan makeup. What happens if you extend the analysis out longer? Using data from Gallup[1], I did exactly this: I analyzed the change in Republican party identification from January of 2008 all the way up through September of 2018. This is what that looks like.

On first glance it may still seem like President Trump is causing a downturn but pay attention to the difference in the severity of the two slopes. They are not drastically dissimilar — something that is evident when you compare it to a long-term trend that ignores the 2016 election. Not only is there not a severe difference between all the slopes, a statistical analysis suggests that the differences are not significant [2].

Concluding that Republican Party ID is trending down over the last decade would be a far more accurate picture. But it still isn’t the whole truth. Party ID has been trending down over the last decade regardless of party and regardless of who sits in the Oval Office. This has been met with a concomitant increase in the number of self-identified “Independents.” (Or, more accurately, people who lean towards one party or another but, for whatever reason, do not prefer to overtly identify with any one in particular [3]). Those headlines insinuating that President Trump has been at the helm of a degrading Republican ship? It would be just as accurate to say that “President Trump presides over a shrinking Democratic party.” And not even Fox News is audacious enough to run that headline.

So why do we have so many articles insisting that Trump is really, truly the final nail in the GOP’s coffin? My hunch is that it’s because there’s this fairly widespread expectation for it to be true. Well, expectation for some; hope for others.

I was recently having a conversation with another UF Political Science PhD, James Fahey, on a topic similar enough that my butchered paraphrasing of his words ring true here:

I think with Trump people want to believe that there’s this universal backlash against him and his actions. He’s been openly bigoted, bragged about sexual assault, he’s transgressing against presidential norms left and right — all of these things that are bad for people like academics and others who know why those norms are important. We just expect people to feel shame by default because, in our minds, how couldn’t they? But that’s just us transposing our feelings on a group onto everyone else.

And I think James is right. (Or, my memory of what James said at least). There are certainly a number of people who study American politics who are totally flabbergasted by what they are witnessing. And, for the record, that includes a notable number of conservative observers as well. But they falsely expect people to a) know as much as they do; b) care as much as they do; and c) come into the conversation with the same broad ideological commitments as they do. The fact is that not all that many people meet those criteria — and they don’t work for more reasons than I can actually get into here.

The Republican party will survive the 2018 midterms — although they will almost certainly lose ground in the House if not the majority. And although some may be tempted to believe that he’s somehow eroding the base, it will undoubtedly survive President Trump too.

Peter R. Licari is a PhD student in Political Science at the University of Florida specializing in American Politics, Political Behavior, and Political Methodology. The opinions expressed are his own. He can also be found on YouTube and on Twitter(@prlitics13). What little spare time remains is dedicated to long-distance running, video games with his ever-patient wife, and to oddly productive one-sided conversations with his cat, Asia.

1. The data and Stata replication file for this analysis can be found on my Github.

2. For the curious, I used a regression discontinuity analysis assuming a linear trend both prior and after the 2016 election. The results were insignificant at any traditionally accepted level for social science research (p=.216)

3. If you follow up and ask people which way they tend to lean, party ID between both parties has largely been pretty stable over the last couple decades.

--

--

Peter Licari, PhD

I’m a data scientist and social scientist specializing in political behavior. I’m also a runner, writer, gamer, YouTuber, and dinosaur enthusiast.