Skip links

The Senate: How 2018 sets up 2020


— The 2018 Senate races are important not just for determining the majority next January, but also for setting up both parties for the next election.

— Republicans are defending almost double the seats that Democrats are defending in 2020, something of a reversal from this cycle, where Democrats are playing a lot of defense.

— However, the most vulnerable 2020 seat (Alabama) is currently held by a Democrat. But Democrats have some decent targets of their own.

— — If Republicans can net a seat or two this year, it would make their position much stronger in 2020, much like their strong showing in 2014 protected them in 2016. Likewise, Democrats holding the line in 2018 would be a victory in and of itself, given both this year’s difficult map and the potential for a limited gain on what for Democrats is a better map in 2020.

The Senate: A first look at 2020

Throughout the 2014 cycle, the Republicans appeared favored to win control of the U.S. Senate, but there was some doubt about whether they would maximize their gains. They needed six net seats to take control, but given the map, it seemed possible that they could win nine or 10. As it ended up, their eventual gains came out on the higher end, as Republicans won narrow races in Alaska, Colorado, and North Carolina (by about two points apiece, give or take) along with six others by significantly larger spreads to score a nine-seat net gain, going from 45 to 54 seats and the majority. Republicans also came close to winning New Hampshire and Virginia, losing by just about three and one points, respectively.

In an assessment published right after the election, we argued that the GOP’s nine-seat gain in 2014 provided “the buffer the GOP needs to be better able to preserve its Senate majority in 2016.” Sure enough, had the Republicans fallen short in their three closest wins in 2014, they would have won the Senate in 2014 but lost it in 2016, when Democrats netted two Senate seats. As it was, Republicans won a 52-48 Senate majority in 2016 that has since been reduced to 51-49 thanks to the Democratic victory in Alabama’s special Senate election late last year.

We bring this up just to note that when assessing the Senate, keep an eye on the future. This year’s Senate races aren’t just about who wins the chamber for the 2019-2020 Congress, but also about which party has better set themselves up for the 2020 election.

There is an imbalance in the three Senate classes right now. As we’ve noted many times this cycle, Democrats are disadvantaged this year by the fact that they hold 26 of the 35 seats being contested in November (including the two independents who caucus with Democrats). In order to net the two additional seats they need to win the Senate, Democrats need to win 28 of 35 seats (80% of all the Senate races this year). That is an achievement a party has only accomplished twice in the history of Senate popular elections (since 1913): 1932 and 1964, when Democrats won 80% or more of the Senate races those years amidst presidential landslides for Franklin Roosevelt and Lyndon Johnson, respectively.

Given that they are defending several red state Senate seats in 2018, Democrats would be very fortunate to win Senate control this November — and, in truth, just holding the line at 51-49 would be a victory, too. That could set Democrats up to win control in 2020, when Republicans are defending 22 of 34 seats being contested. That gives Republicans incentive to make the most of the opportunity they have this year to add to their majority — even an extra seat or two could give them the insulation they need to weather what 2020 could have in store.

The good thing for Republicans is that while they are playing defense in many more states than Democrats are next cycle, a significant share of the GOP-held seats coming up in 2020 are in states where Republicans should be fine. Just take a look at Map 1, which shows the current party control of the 2020 Senate seats up for election.

Map 1: 2020 Senate races, shaded by current party control

Note: *Signifies special election.

One big factor on next cycle’s map is the lack of “crossover” Senate seats, meaning states that voted differently for Senate in 2014 and president in 2016. There are only 15 senators (of 100) that represent states that voted for the other party for president in 2016, and 11 of them are on the ballot in 2018. That leaves only four crossover senators, all of which are on the ballot in 2020 (on the 2022 Senate map, last contested in 2016, every state voted for the same party for both president and Senate, something that had never happened before in the Senate popular election era). On the 2020 map, the crossover senators are Sens. Doug Jones (D-AL) and Gary Peters (D-MI), Democrats in Donald Trump-won states, and Sens. Cory Gardner (R-CO) and Susan Collins (R-ME), Republicans in Hillary Clinton-won states.

Naturally, a couple of the crossover senators are the first ones worth looking at when trying to find exposure for either party in 2020.

The most obviously vulnerable senator on either side isn’t that big of a surprise given the circumstances of his election: Sen. Doug Jones (D-AL). Jones won a hotly-contested special election in December 2017 over former Alabama state Supreme Court Chief Justice Roy Moore (R), whose already weak and controversial candidacy imploded over credible accusations of lurid behavior toward underage women. As it was, Jones needed all the breaks to win as a Democrat in ruby red Alabama: He beat Moore by just a 50.0%-48.3% margin. Assuming the GOP nominates a strong challenger — not always a given — Jones probably will have a hard time winning, particularly because one would assume the GOP presidential candidate (be it Donald Trump or someone else) will carry Alabama by at least 20 percentage points. Alabama likely starts as a Toss-up, or even with Jones as an underdog.

The other most vulnerable senator on the ballot in 2020 is Sen. Cory Gardner (R-CO), who won an impressive but close victory over then-Sen. Mark Udall (D-CO) in 2014. Colorado remains a competitive state, but its purple has acquired a slightly bluish tint in recent years: Democratic presidential candidates ran a little bit better in Colorado than they did nationally in each of the last three elections; prior to that, Republicans usually did better in Colorado than they did nationally. Looking ahead to 2020, one would probably expect the Democratic presidential nominee to have at least slightly better odds of carrying Colorado, at least at the start of the campaign. So Gardner might need some split tickets to win; he is capable of that, but his race would start as a Toss-up as Democrats figure out who might challenge him.

The third race that probably would start as a Toss-up is a new addition to the 2020 Senate map: Arizona. After the passing of Sen. John McCain (R-AZ), Gov. Doug Ducey (R-AZ) appointed former Sen. Jon Kyl (R-AZ) to fill the seat temporarily. However, Kyl says he does not intend to run for a full term, so the GOP nominee in 2020 will be someone else (it’s also possible that Kyl will not serve until the 2020 election, so Ducey potentially could give someone else an appointment to create a weak form of incumbency). In any event, given how close and competitive this cycle’s open seat Senate race between Reps. Kyrsten Sinema (D, AZ-9) and Martha McSally (R, AZ-2) appears to be, another open-seat (or quasi-open seat, if an appointee is running) Senate race in 2020 should also be competitive in a year where Arizona might be hotly contested in the presidential race (although Trump would start as a favorite to carry the state given its typical right-of-center presidential positioning). It seems possible that the loser of the McSally-Sinema race could turn around and run for the 2020 special; also, remember that whoever wins in 2020 would have to run again in 2022 if that person wanted to capture a full term.

To us, these are the three most clearly competitive 2020 Senate seats. One is held by a Democrat, and two are held by Republicans, but the Democratic seat (Alabama) is the most obviously endangered.

Of the remaining current Democratic seats up in 2020, Sens. Gary Peters (D-MI) and Jeanne Shaheen (D-NH) would probably start as modest favorites in the two closest states in the 2016 presidential election: Trump won Michigan by just over 0.2 points, while Hillary Clinton won New Hampshire by just under 0.4 points. The presidential race might help determine how close those Senate contests would be. As noted above, Sen. Mark Warner (D-VA) had a very close call in 2014, but presidential-level turnout probably will help insulate him in a state that, like Colorado, also seems to be trending Democratic. Assuming that she wins this fall and wants to run again for a full term, appointed Sen. Tina Smith (D-MN) would be on the ballot again in 2020. Trump did almost win Minnesota in 2016, so if he competed there again that could potentially hurt Smith.

Of the remaining Republican seats, there are a few that stand out as decent, or at least potentially decent, Democratic targets.

One big question mark is in Maine, a state that Clinton carried (though only by a relatively weak three points) but where Sen. Susan Collins (R-ME) has been an extremely strong incumbent. One wonders what effect the ongoing Brett Kavanaugh saga might have on Collins, a Senate swing vote, although whatever happens with Kavanaugh’s embattled nomination to the Supreme Court might be ancient history by the time 2020 rolls around. The key question would be whether Collins is on the ballot: If she is, she’d start as a clear favorite; if she retired, the race would be a Toss-up.

North Carolina, often home to competitive Senate contests, likely will have another one in 2020 as Sen. Thom Tillis (R-NC) presumably seeks a second term. That one might go the way of the presidential race, and the Tar Heel State’s persistent GOP lean helps Tillis. In Iowa, Sen. Joni Ernst (R-IA) probably would get help from the presidential race, too, although Iowa bounces around a lot and the state might not replicate the big margin (more than nine points) it gave Trump in 2016. Georgia, another right-of-center state that could be getting more competitive, might also be a Democratic target, although Sen. David Perdue (R-GA) would also start as a favorite. And while it remains difficult to imagine him losing, Senate Majority Leader Mitch McConnell (R-KY) probably will attract a decent challenger and will have some vulnerability thanks to his weak favorability numbers (a penalty he pays for serving as majority leader).

Put it all together, and you get a competitive landscape to start the next cycle that looks something like this:

Toss-ups to start:

Democratic seats: Alabama

Republican seats: Arizona and Colorado

Probably or potentially competitive:

Democratic seats: Michigan, Minnesota, New Hampshire, and Virginia

Republican seats: Georgia, Iowa, Maine, Kentucky, and North Carolina

Probably not competitive, at least to start

Democratic seats: Delaware, Illinois, Massachusetts, New Jersey, New Mexico, Oregon, and Rhode Island

Republican seats: Alaska, Arkansas, Idaho, Kansas, Louisiana, Mississippi, Montana, Nebraska, Oklahoma, South Carolina, South Dakota, Tennessee, Texas, West Virginia, and Wyoming

A note on the last category: One or more of these races could come into play based on candidate selection and other factors. For instance, what if Gov. Steve Bullock (D-MT), term-limited in 2020, decided to challenge Sen. Steve Daines (R-MT)? That might turn into a Toss-up race, but it’s also possible that Daines wouldn’t be pushed that hard if someone like Bullock took a pass on running (he may be one of the dozens of Democrats who could run for president). Additionally, besides the Minnesota special, another one of these 2020 races is on the ballot this year. That is the special election in Mississippi that is likely to go to a runoff, probably between appointed Sen. Cindy Hyde-Smith (R) and former Clinton administration Agriculture Secretary Mike Espy (D). The winner — we rate the special election as Likely Republican — likely will be on the ballot again in 2020 for a full term. But if Espy pulled off the upset this year, he’d be in the same position as Doug Jones will be: trying to hold a Senate seat in two years as a Democrat in a state that seems certain to vote substantially in favor of the GOP presidential nominee in 2020.


Overall, the 2020 picture looks like this: A bit more exposure for Republicans than Democrats, but Democrats hold the most vulnerable seat on the board (Alabama).

We’ve consistently thought that, in this cycle, there’s probably not going to be a lot of net change either way in the Senate, and that the GOP is favored to hold control (though that’s not a certainty as the calendar turns to October). And, as we look ahead to 2020, the seats in play do not point to a big advantage for one side or the other. Yes, Republicans are defending almost double the seats, but many of the seats Republicans are defending start off as basically safe for them, which reduces their exposure.

That’s all a long way of saying this: If Republicans can eke out a net gain this year — even if it’s just a seat or two — it could go a long way toward helping them protect the majority next cycle, when they probably have more exposure than the Democrats, but just a little bit more (at least to start). Meanwhile, if Democrats hold the line or even pick up a seat or two (the latter gain would give them the majority), they might be well-positioned to take, or hold, the majority next time.