Roy Moore is not a traditional Republican politician. His claim to fame is defying court orders to remove a statue of the Ten Commandments and to allow same-sex marriage in the state as the elected chief of the Alabama Supreme Court. Each incident, years apart, cost him that position. This year, he decided to run for the Senate seat vacated by Attorney General Jeff Sessions, against the obvious desire of the establishment in the state.

On the campaign trail, Moore has embraced a robust indifference to the nuances of policy. Asked about the Deferred Action for Childhood Arrivals program in a radio interview, Moore wasn’t familiar with the issue — an issue that, at the time, was dominating the national political conversation. He was not familiar with the idea of “right-to-work” states, an important issue for most conservative politicians.

In other words, Moore seems an awful lot like another stick-to-his-guns, to-hell-with-the-establishment, policy-nuance-doesn’t-matter politician: Donald Trump.


Yet President Trump supports Luther Strange, the opponent who will face Moore in next week’s runoff election. Strange is the choice of much of the GOP establishment, including Senate Majority Leader Mitch McConnell (R-Ky.). Trump will travel to Alabama this weekend to rally with Strange.

The race, then, sets up a weird dynamic. Can Trump’s appeal to voters in Alabama — an appeal that earned him a 27.7-point margin of victory in the state — help push Strange to a win after coming in second in the first round of voting? Or will voters prefer to again vote for an outsider such as Moore, as they did last year in the Republican primary and the general election?

In the first round of voting, it seems the latter was the case.

If we compare how Strange and Moore did in each county with how they did statewide and then compare that with how Trump did in the 2016 primary in each county versus statewide, the result looks like this.

The dots represent each county, and the lines represent the trend. The red Moore line goes up and to the right, meaning that the better Trump did relative to his statewide percentage last year, the better Moore did relative to his overall support. The green Strange line is almost flat, suggesting that more Trump-friendly counties didn’t show him much more support than more Trump-unfriendly places. The good news for Strange is that the places in which he did better tend to have more voters.

If we compare the results of the Alabama Senate primary this year with last year’s general election, the lines head in the opposite direction. The better Trump did relative to his state percentage, the better Moore did in that county in the primary. The worse Trump did, the better Strange did.

These correlations are not terribly strong, but they’re there. There are other factors that might be at play, including Strange’s inability to consolidate the anti-Moore vote in the primary, meaning that others in the race pulled votes he’d earn in a runoff. But that’s why we chose this metric: Strange does worse relative to himself in places that were more strongly pro-Trump in the general last year. Moore does better relative to himself in places where Trump did better in the 2016 Republican primary.

Again, this suggests that what people liked about Trump is what they like about Moore, and they’re more interested in that than whom Trump thinks they should vote for. If that holds true, it’s a warning sign for other Republican candidates up for reelection in 2018: Trump is the establishment now, and the endorsement of the establishment still can’t necessarily carry the day against an insurgent.