Various outlets, including this blog and The Washington Post more broadly, have dissected record-high turnout in Republican primary contests this year, concluding that it is unlikely to boost Republicans in the general election and probably a result of the more-competitive GOP contest.
Yet we’re still faced with some striking and perplexing numbers. Republican turnout is up more than 50 percent from both 2008 and 2012, and the reason for the spike is not entirely clear. Is Donald Trump motivating new people to participate after sitting out past contests, motivating a dormant sector of Republican or independent voters who now think they have a voice, or has Trump’s controversial candidacy and the high-attention contest heightened the stakes for all Republicans, including Trump’s opponents?
Here's a good way to tell: If Trump’s supporters are fueling the rise, he should be performing best in places with the biggest increases in turnout. We divvied up the states that voted in both 2008 and 2016 (which are comparable because they included contested primaries for both parties) into three groups -- those with huge increases in vote totals (85 percent or more compared with 2008), those with moderate increases (33 percent to 84 percent) and those with increases below 32 percent or decreases.
The data show Trump’s support didn’t peak in high-turnout states. Among those states with the largest turnout increases compared with 2008 -- 85 percent or more -- Trump garnered 34.3 percent of the vote, lower than his average overall (34.9 percent). He performed slightly better in states with a moderate turnout increase, and worse in states where turnout did not rise as much. The comparison does not differ much when looking at 2012 turnout data; Trump won 32.9 percent in states that had the largest turnout increases, which is below his cross-state average.
A look at individual states shows the tenuous tie between Trump’s support and rising turnout. The trend line running through states tilts slightly upward -- meaning a positive correlation -- but it’s small and not statistically significant. (For data nerds, it’s a 0.12 Pearson's correlation coefficient.)
In Kansas, where the caucus turnout increased by a greater percentage over 2008 than any other state we examined, Trump won just 23 percent of the vote. Yet he performed very well (47 percent) in Mississippi’s primary, where the turnout increased by 182 percent compared to 2008.
State-level comparisons are fairly rough, so we asked Post voting data guru Ted Melnik to investigate county-level data for any evidence of Trump-driven surge. He found the opposite: In primary contests, counties that were more supportive of Trump had slightly smaller increases in turnout compared with 2008. The correlation was negative but fairly small, at -0.13.
What do all these numbers mean? Basically, Republicans’ turnout surge is not being caused by a hypodermic shot of Trump voters into the primary electorate. Non-Trump Republicans also have been inspired to vote at higher rates -- some probably in opposition to Trump and others simply because the contest is competitive.