In the 110 years from 1871 to 1981, Alabama never elected anyone to serve the state in the U.S. Senate who wasn’t a Democrat. For much of that period, that was true throughout the South, a Democratic bastion stronger than either party has seen in any other region over the course of U.S. history. After the civil rights movement, the region flipped, and the once solid-blue South became a deep hue of red.

It happened quickly. In Alabama, Sen. Richard C. Shelby switched parties to Republican in 1994; Jeff Sessions, now attorney general, won election in 1996. Since, both senators in the state have been Republican, and there was every reason to think that those two senators would continue to be Republicans into the indefinite future.

Until Thursday.

To be clear, it’s still very likely that the man who will win the upcoming special election in Alabama to replace Sessions will be the Republican, Roy Moore. But with a report by The Washington Post that Moore allegedly sought sexual contact with a 14-year-old girl in 1979, it seems worth reevaluating that assumption. Could Doug Jones be the first Democratic senator from Alabama in two decades?

On Twitter on Friday, Charles Franklin of the Marquette Law School Poll noted some interesting data. Moore has twice won election to statewide office in Alabama, being elected chief justice of the state Supreme Court in 2000 and 2012. In both cases, he left office after being censured for his behavior: In 2003, he was removed from office after refusing to take a Ten Commandments monument out of a state building, and in 2016, he was suspended for demanding that the state’s ban on same-sex marriages be maintained contrary to a U.S. Supreme Court decision.

What Franklin noted was a sharp downturn in Moore’s performance between his two races, suggesting that his having left in disgrace the first time tamped down his support the second time.

Here’s a visual version of what Franklin noted, with the bars scaled to the number of votes.

In 2000, Moore won about 55 percent of the vote in his race, a bit less than George W. Bush won in the presidential race in Alabama that year. Moore got fewer votes than Bush, by about 63,000 votes.

In 2012, the gap between Mitt Romney and Moore was much wider. Moore got about 52 percent of the vote, down about three points from 2000. Romney, though, outperformed Bush in 2000 by more than four points. The margin of Moore’s victory in 2000 was 5.5 points lower than Bush’s; in 2012, his margin of victory was almost 19 points worse than Romney’s. Moore won by only 3.5 points in 2012 and received 343,000 fewer votes than Romney.

Notice that Moore’s Democratic opponents did better than the Democratic presidential candidates, Al Gore and Barack Obama, both in terms of percentages and raw vote totals. The suggestion is that, perhaps, a national Democrat is perceived less positively than an Alabama Democrat. In 2016, Donald Trump won Alabama by a wider margin than even Romney, 27.7 points. At the same time, Democratic Senate candidate Ron Crumpton, taking on Shelby, lost by a wide margin, even though he got more votes than Hillary Clinton in the state. (He got 749,000; she got 730,000.)

There are two ways to interpret this.

One is that Moore is more vulnerable than one might have originally thought, particularly given:

  • The newly reported revelations, which led, among other things, to endorsements being reconsidered and party funding shut off.
  • The results of Tuesday’s elections, in which the Democratic candidate for governor in Virginia (a state Clinton won) ended up seeing a groundswell of Democratic support that helped him perform significantly better than polling in the race predicted.

A poll conducted in the wake of the new reports shows the Alabama race as essentially even — though there are good reasons to be skeptical of what it found. Moore’s lead in the polling average has been about six points, narrower than the lead he enjoyed in the runoff for his party’s nomination.

The other way to look at this is that the fundamental structure of the race hasn’t changed. This is an election in December of an off year, the sort of election that tends to reward the candidate supported by voters who turn out more often, which generally means voters who are older and wealthier. That generally means Republican voters. That’s the caveat to the comparisons to 2000 and 2012 that Franklin himself notes: A Democrat coming within four points of a Republican in statewide race in a high-turnout presidential election doesn’t mean another Democrat would come similarly close this December.

There will likely be Republican voters who are discouraged from heading to the polls by the idea of having to cast a ballot for either a scandal-clouded Republican or, worse, a Democrat. But how many will be so discouraged is hard to say. It’s also hard to say how many Democrats are planning on flooding the polls next month, as they did in Virginia. As a percentage of the state’s population, there just aren’t as many Democrats in Alabama to do that flooding.

Were the Alabama contest between Jones and any other Republican, it would be very hard to make the case that Jones might win. (Such a case would lean heavily on the second bullet point above.) A race against Moore, though?

Let’s just say that it’s not a gimme for the Republicans. This may be about as close to a non-solidly-red Deep South as we’re likely to see for a while.