Two things you need to know about trends in US party affiliation

A new Gallup poll raises some old questions about sampling error and non-response

Here are a few paragraphs from a recent CNN article covering a new Gallup poll that shows the Democratic advantage in party affiliation is the largest since the party’s 2012 lead:

This is a *major* warning sign for Republicans….

Gallup polling for the first three months of 2021 shows that 49% of the public identify as Democrats or Democratic-leaners, while just 40% call themselves Republicans or say they lean toward the GOP.

That's the largest gap between Democrats and Republicans in Gallup's quarterly study of party identification in nearly a decade. The last time Democrats had a larger lead on party ID was early 2009…

But it's not just the party ID gap that stands out in Gallup's first-quarter polling. It's this: Just 25% of the public calls themselves Republicans -- close to the lowest (22%) that Gallup has ever measured since it started doing telephone-based polling. (Another 15% say they lean to Republicans.)

When you combine those two data points, you get this: Not many people want to be a Republican at the moment. The party's brand is quite clearly damaged after four years of Donald Trump seeking to break every political norm possible.

The Point: Smart GOP strategists look at these numbers and know the best strategy is to immediately start charting a course away from Trump. The problem? The party base still loves the former President -- and has no plans on abandoning him.

The Gallup data come with this graph:

OK, so, first off, while the current Democratic margin is technically the largest since 2012, it’s not by much. Several gaps between 2013 and 2020 look nearly as large, including in the third quarter of 2018 and the first quarter of 2017. I just feel like it’s worth getting the misleading headline out of the way before we go on.

Sure, I can come up with a lot of theories why this trend might be real. Maybe Republicans were put off by the many members of their party who stormed the US Capitol on January 6th. Or, on the other hand, maybe some Republicans who were particularly supportive of Trump are deciding to call themselves “Independents” instead of GOPers since many prominent leaders — Mitch McConnell and Mitt Romney among them — disavowed Trump after the attack.

Or maybe the trend is just noise. Consider two possibilities.

First, the Democrats’ current advantage could be caused by sampling error — the random chance that a sampled population doesn’t match its population because an unrepresentative set of people were chosen to respond. The potential for such error is guaranteed in polls even of large sample sizes, such as Gallup’s — indeed it is evident in their data here, as seen by the frequent small jumps in the share of the population identifying as Republican, Democratic, or Independent over the last few months. Observing a phantom swing across the last several quarters wouldn’t be surprising at all given how much the underlying share of self-proclaimed “Independents” in the electorate has jumped around within the last quarter.

Second, it’s possible that news events turned Republicans off from taking phone polls, even from a reputable source like Gallup. This would decrease their share of party identifiers in the polls, making it appear that the GOP share had decreased in the population even when it hadn’t. Such patterns of differential partisan non-response are common after events that cause news coverage to be unusually good or bad for one of the parties. Here is a paragraph from Mark Blumenthal, who ran the Huffington Post’s polling blog from 2010 to 2015:

during the 2018 midterms, following the nationally televised congressional hearing in which Christine Blasey Ford leveled accusations of sexual assault against Brett Kavanaugh, who was then a Supreme Court nominee, we saw something similar in the polling I helped conduct at SurveyMonkey.

The survey completion rate jumped 3 percentage points among those who expressed approval of Trump while remaining essentially flat among those who expressed disapproval. The changes in completion rate among Trump approvers vs. disapprovers increased the share of self-identified Republicans and conservatives, non-college-educated women and respondents in rural ZIP codes in our unweighted data.

Put more simply, the composition of our samples changed, and even after demographic weighting, the respondent pool included more of the kinds of people who tend to support Trump. As a result, we also saw what appeared to be an increase in Trump’s approval rating among all U.S. adults. Yet that trend was mostly an artifact of the response patterns; it didn’t reflect a meaningful real-world change in how Americans viewed Trump.

Indeed, by the end of October, the higher completion rates among Trump approvers, the change in the demographic and geographic composition of our unweighted data, and the corresponding bump in Trump’s approval rating had all faded, reverting back to where they had been through much of the summer.

…

There are two possibilities with Gallup’s data. First, that the share of Republicans hasn’t changed that much, if at all — a pattern consistent with the trend from 2016 to 2020, between which points Donald Trump’s vote share increased — constituting “a *major* warning sign,” or that this is the result of various well-documented statistical issues with public opinion polling.

What are you putting your money on?