We’re more than 11 months out from Election Day, and there are too many moving parts (changes in national environment, primary elections, possible retirements, fundraising, strategic decisions, and more) to know anything for certain.
But it’s not too early to get an initial feel for key races—that is, to figure out exactly what it would take for Republicans to hold an important seat or for Democrats to take one over. Tennessee might become one of those key races.
For Democrats to retake the Senate, they’d have to successfully defend all of their current seats (some of which are in Trump-y states such as Montana, West Virginia, and Missouri), and flip two Republican seats. The obvious targets are Arizona and Nevada—Nevada voted for Hillary Clinton in 2016 and Arizona is, partially because of Trump’s unpopularity, well within reach for Democrats. But if Democrats fail to take one of those seats or have trouble holding one of their red-state seats, they might try to make up the shortfall in Tennessee.
Tennessee is probably the third most vulnerable Republican-held Senate seat, after Arizona and Nevada. Democrats have recruited Phil Bredesen, a former two-term governor who won every county in the state in his 2006 bid, to run for the seat that incumbent Republican senator Bob Corker is leaving. And it hasn’t been that long since there was a close race in Tennessee—Corker only beat Harold Ford Jr. by three points during the Democratic wave of 2006.
So how vulnerable is Tennessee? And what would Bredesen have to do to take the seat? The first step in tackling these questions is to turn the clock back about a decade.
Tennessee Has Changed a Lot Since 2006
Casual election watchers probably haven’t paid much attention to Tennessee. The Volunteer State has voted for the Republican presidential candidate in each of the last five presidential elections, and it’s given the GOP double-digit victories in the last four. But these topline numbers hide an important trend. Since the mid-2000s, the state has reddened significantly.
To see this, just look at these three maps.
The first map shows the partisan index of each county—that is, how each county voted relative to the national vote—in the 2004 election. This correlates pretty strongly with the 2006 senate election result where Corker beat Ford.
You’ll notice a pattern here: The red counties were more Republican than the country, and blue counties were more Democratic. So Republicans won by enormous margins in the eastern part of the state—a mountainous section whose Republican loyalty goes back more than a century—while Democrats fared better in the more historically Democratic central part of the state (which also contains the Nashville metro area), Memphis (in the southwest corner), and areas with a larger share of black voters.
This isn’t a picture of a winning Democratic coalition in Tennessee—remember Corker beat Ford—but it contrasts with what we see in the next map.
The second map shows the partisan index of each county in 2016. And the third map shows the difference between Bush’s 2004 vote share and Trump’s vote share in each county. Between 2016 and 2004, Democrats lost significant support outside major metro areas, while gaining some ground in the central cities (the dark blue county in the southwest contains Memphis, the bluest central county contains Nashville). This shift favored Republicans: Bush won Tennessee by 14 points in 2004; Trump won the state by 26 points.
White voters are likely responsible for much of this movement. According to the New York Times Upshot’s estimates, white voters in Tennessee moved about 15 percentage points to the right between 2004 and 2012. County-level data suggests that between 2004 and 2016, areas with fewer college graduates, fewer black voters, and more residents who were born in-state were more likely to shift towards Republicans. The same county-level data shows that Trump tended to improve on Romney’s two-party vote share in areas where whites were less educated.
These patterns should sound familiar. The Republican party has been gaining strength with white voters (especially in the South) for a long time, and non-college educated whites who voted for Obama and Trump played a huge role in 2016. The long-term trends have pulled the state right since Bredesen was last re-elected, which could create problems for him in 2018.
So Can Bredesen win?
It’s easy to see how these changes might be a problem for Bredesen. According to the Census, 74.2 percent of Tennesseans are non-Hispanic white and 17.1 percent are African-American. In order to win, Bredesen will probably need to find a way to win a greater share of white voters than other recent Democratic candidates. There are also some potential warning signs in the polls: In a December Gravis poll, Bredesen narrowly led both Republican Stephen Fincher and Rep. Marsha Blackburn—but only polled at 42 percent.
Moreover, as Governing has documented, not every red-state Democrat who performed well in the past has been able to replicate their performance in the new political environment. In 2016, Evan Bayh, a former Democratic senator and governor from Indiana, came out of political retirement to run for his old Senate seat. He earned more votes than Hillary Clinton did in Indiana, but still lost by ten points. Bob Kerrey, a former Nebraska Democratic Governor and Senator who left the upper chamber in 2001, tried to run again in 2012. He got a larger share of the vote than Obama, but also lost by double digits. Ted Strickland of Ohio is a slightly different case because he lost his 2010 gubernatorial re-election bid before failing to unseat Republican Rob Portman in 2016—but it’s still not a great precedent for Democrats.
Yet sometimes it does work. Joe Manchin, a popular Democratic governor from West Virginia, was able to win a Senate seat in quickly-reddening West Virginia in 2010.
The most important factor for Bredesen is that he will likely be running in a much more Democratic national environment than Bayh and Strickland—or even Kerrey in 2012—did. It’s possible to imagine Bredesen outperforming a generic Democrat at a time when the national conditions favor the blue team. And if he performs well in the state’s major metro areas (something other Democrats appear to have done in 2017) while clawing back some ground in the rest of the state, it’s possible to imagine him winning.
Handicappers disagree on exactly how to think about this race. Sabato’s Crystal Ball rates it as “Likely Republican” while the Cook Political Report rates the race as “Toss-up.” These are both reasonable perspectives, but, in light of all of this data, I would advocate for splitting the difference and thinking of the race as “Leans Republican” until we get further into election season.
This post originally appeared on Weekly Standard