Still no evidence of a decline in Republican party identification since 2016
But Democrats are at their highest pre-election level since 2007, and doing significantly better versus 2015
The takeaway: Surveys from Gallup show that a steady percentage of American adults identify as Republicans. These data counter some academic theories that Donald Trump has driven partisans out of the Grand Old Party. On the other side of the political aisle, Democratic party identification has indeed increased. (Given their performance in the 2018 mid-terms, perhaps that’s not all too surprising.)
Editor’s note:
Thanks for subscribing! Your membership adds up and makes all this newslettering possible (reminder: I do all this work independently). Please consider sharing online or with a friend; the more readers, the merrier. Remember that apart from getting special articles subscribers can also comment below each post and participate in exclusive threads.
As always, send me your tips about what you’d like to read about next. Or what you don’t want to read. Or your feedback otherwise. Also, cat pictures are nice, so please send them to me! I’m elliott@thecrosstab.com, or @gelliottmorris on Twitter.
Thanks all,
—Elliott
PS! I’m giving away free trials to this newsletter from now until November 1st. I’d love if you shared it with your friends. Click the button below or share this link: https://thecrosstab.substack.com/subfriends
Donald Trump has scored some the highest same-party approval ratings of any modern president. But academics and political analysts have speculated that the president’s high levels of support are not due to his popularity with Republicans writ large, but because Republicans who don’t agree with him are turning away from the party, thereby artificially inflating Trump’s in-party ratings with the pro-Trump Republicans who have remained.
Repeatedly throughout Trump’s presidency, I have not found much evidence for this claim.
Last Monday, Gallup released their third-quarter 2019 estimates of party identification in the US. According to 12,000 interviews of adults, 42% of Americans identify as Republicans or Independents who lean Republican and 47% identify as Democrats/lean Democrats. Here is their graph of how party affiliation has changed since the first quarter of 2015:

The trend clearly shows that Trump has not caused a large decrease in the share of Americans identifying as Republicans. In Gallup’s partisanship estimates for the third quarter of 2016, 43% of Americans affiliated with the GOP, only one percentage point higher than today’s numbers. Indeed, there has been no statistically significant decrease in GOP party identification at all; with a sample size of 12,000, the margin of error for the poll is just under 1% and overlaps with the confidence interval for the 2015 proportion; we cannot tell whether the proportions are statistically different.
In case you’re wondering, that math is: ( sqrt( ( 0.42 * (1 - 0.42) ) / 12000 ) * 1.96 ) = 0.0088 * 100 = 0.9%.
For what it’s worth, the changes remain insignificant when using the margin of error for a difference in proportions, which is a better way to compare results from political polling. From what I can tell, the 2015 survey interviewed 3,566 people, so the margin of error for the difference would be: sqrt( ( (0.42 * (1-0.42) ) / 12000 ) + ( (0.43 * (1-0.43) ) / 3566) ) * 1.96 = 0.0185 100 1.85%.
Though Republican identification has held steady, compared to the earliest point in Gallup’s trend, there is real evidence of an overall increase in Democratic Party identification. According to Gallup’s surveys, 44% of Americans identified as Democratic in the first quarter of 2015; in the first quarter of 2019, 49% did; and in the third quarter, 47% of adults said they were Democrats. That’s a difference of 3-4%. Using the same formula as above, we see that changes from 2015 to Q1/3 2019 are outside the margin of error.
sqrt( ( (0.49 * (1-0.49) ) / 12000 ) + ( (0.44 * (1-0.44) ) / 3566) ) * 1.96 * 1.96 = 0.0185 * 100 = 1.85%.
The 3-4% difference is greater than this 1.85% threshold for statistical significance, so we would say the increase is real and significantly greater than zero. However, given the extra sources of error (i.e. “non-sampling” error) in polling, I wouldn’t bet the farm on the precision of the difference, it could be closer to 2 or even 6, we just don’t have precise enough data.
Let’s take a look at the historical context for the Democrat’s margin in party identification. Today, Gallup registers the highest Democratic advantage since 2007/8. If you believe that party identification is a proxy for voter behavior—and you should!—then these data smile upon Democrats.

Now, to end my meandering and get back to the original point:
Given the steady percentage of Republican identifiers over time, the theory that Trump’s approval rating is being artificially inflated by a group of stalwart Republican loyalists has little staying power. Still, it could be true that Republicans who disapproved of President Trump no longer call themselves Republicans, but then the percentage of GOP identifiers has stayed steady because Trump-approving Independents or Democrats have come to call themselves Republicans.
But I don’t think that’s all that likely.