Skip to main content
Menu
Senate And Governor Races Are Diverging Sharply In Some States

More and more Americans vote for the same party up and down the ballot these days. But some people still split their tickets. Who? Where? Well, we can use our forecast data to try to answer that. There are 22 states holding multiple statewide races this year — meaning a Senate race and a governor’s race — and with no presidential race to muck things up, most states will vote for the same party in each. But there are some notable exceptions.

In three states — Massachusetts, Vermont and Maryland — it’s especially likely that voters will elect a Republican governor and a Democratic or independent senator. Other states with close contests for Senate or governor, such as Nevada or Wisconsin, could also see split outcomes. Even in states where the same party wins both the Senate and gubernatorial races, there will be at least some difference between the margin of victory in the two races. In other words, just because New York is likely to elect Democrats to both offices, it doesn’t mean all voters cast a straight-ticket ballot — there will be some difference in how big a margin Sen. Kirsten Gillibrand and Gov. Andrew Cuomo win by in November.

So, to see just how far apart voters might be in 2018, I looked at how much each candidate was expected to win by1 according to the “Classic” versions of our Senate and gubernatorial models, and then I calculated the difference between the margin of victory in the statewide races for each state.2 The difference gives us a sense of how common, relatively speaking, split tickets are in a given state. And as you can see in the table below, Vermont has the biggest difference between the forecasted margins in its races for Senate and governor — about 60 percentage points all told. What this means is roughly 30 percent of Vermonters may vote for both Sen. Bernie Sanders and Republican Gov. Phil Scott, since Scott’s expected share of the vote is about 30 points higher than the share the Republican Senate candidate is forecasted to get.3

Where voters might split their tickets on statewide races

Difference in the size of the forecasted margin of victory for the leading party’s candidate in 2018 Senate and governor races, as of Oct. 18

Senate Governor
State Incumbent Forecast Margin Incumbent Forecast Margin Diff
VT Sanders* D+46.2 Scott R+13.4 59.5
MA Warren D+29.3 Baker R+29.2 58.6
MD Cardin D+35.5 Hogan R+17.6 53.1
HI Hirono D+53.1 Ige D+29.6 23.4
WY Barrasso R+40.9 Open R+25.0 15.9
CT Murphy D+22.1 Open D+6.4 15.7
NM Heinrich D+22.4 Open D+7.5 14.9
ME King* D+25.9 Open D+11.4 14.5
OH Brown D+12.8 Open R+1.2 14.0
TX Cruz R+5.2 Abbott R+18.8 13.6
AZ Open D+1.8 Ducey R+11.4 13.3
MN‡ Klobuchar D+20.8 Open D+10.2 10.6
WI Baldwin D+12.7 Walker D+2.3 10.4
TN Open R+4.6 Open R+14.1 9.4
RI Whitehouse D+28.9 Raimondo D+19.6 9.3
NE Fischer R+14.1 Ricketts R+22.7 8.6
NY Gillibrand D+30.1 Cuomo D+21.9 8.2
MI Stabenow D+18.1 Open D+11.7 6.4
FL Nelson D+2.3 Open D+3.5 1.2
PA Casey Jr. D+14.4 Wolf D+15.6 1.2
MN‡ Smith† D+9.7 Open D+10.2 0.6
NV Heller R+1.0 Open R+0.8 0.3

“Open” indicates that the incumbent senator or governor is not running for re-election. Some numbers may not add due to rounding. The forecast margin is based on FiveThirtyEight’s “Classic” model as of 4:30 p.m. Eastern time on Oct. 18.

*Bernie Sanders and Angus King are independent senators who caucus with the Democrats. They are counted as Democrats in this analysis.

†Tina Smith is an appointed, rather than elected, incumbent. FiveThirtyEight’s model treats seats with no elected incumbent as open.

‡Minnesota is included twice because it has two Senate elections this year.

Massachusetts, where our forecast shows Democratic Sen. Elizabeth Warren and Republican Gov. Charlie Baker as sure bets to win, has the second-biggest difference — close to 59 points. And Warren and Baker are expected to win by nearly identical margins — just shy of 30 points in both cases, as of about 4:30 p.m. Thursday — which means a lot of people will likely be splitting their tickets in statewide races. In third with a 53-point difference is Maryland, where Republican Gov. Larry Hogan appears comfortably ahead and Democratic Sen. Ben Cardin looks to be headed toward an easy win. But these are all states where the incumbents are members of different parties. In states where the same party has incumbents seeking re-election for both Senate and governor, the differences tend to be smaller. This is best exemplified by Pennsylvania, where Democratic Gov. Tom Wolf and Democratic Sen. Bob Casey Jr. have nearly identical leads, so their margins of victory are expected to be separated by about 1 point. Similarly, states with open-seat races for both the Senate and governor, like Minnesota4 and Tennessee, also have smaller differences between their forecasted margins of victory; that’s partly due to the lack of elected incumbents, which shifts the states closer to their partisan defaults.5

So why are there variations between the margins by which a state’s Senate and governor candidates are expected to win? We have two contributing factors that might help explain it: incumbency and elasticity.

Incumbents tend to win re-election, even if their political party does not match the one their constituents generally prefer. For example, Govs. Baker and Hogan won their respective races in Massachusetts and Maryland in 2014, which was a strong Republican cycle. Neither faced an incumbent that year, but they only won by narrow margins in deeply Democratic states. Now they are both very popular — Baker and Hogan came in first and second, respectively, in Morning Consult’s third-quarter approval rankings — which helps them run against the Democratic lean of their states and the overall national environment, which is also pretty Democratic-leaning this year. On the other side of the party ledger, Democratic Sen. Sherrod Brown is up for re-election in Ohio, which is 7 points more Republican-leaning than the nation as a whole, according to our partisan lean metric.6 But despite being a Democrat in a pretty red state, Brown has a 29 in 30 chance of winning per the “Classic” version of our model as of Thursday afternoon, and he is forecasted to win by about 13 points. It’s a different story in the Ohio governor’s race, however. With no incumbent, that race is likely more influenced by the baseline partisan lean of the state and the national environment, which has helped make it a toss-up contest in our gubernatorial forecast model.

Another factor at play may be elasticity, or how likely a state’s voters are to be swingy. Highly elastic states are more prone to big shifts in voter preferences, while more inelastic states fluctuate less. More flexible electorates in states such as Massachusetts and Vermont have shown a greater willingness to vote in opposite directions for different seats,7 which we are seeing play out dramatically in 2018. But elasticity is an imperfect explanation in some cases. Whereas Massachusetts and Vermont rank among the states with the highest elasticity, Maryland is toward the bottom of the list,8 yet the difference between its Senate and governor races is similar to the difference we see in the more elastic states. That said, elasticity can help to partly explain some of the seemingly sizable gaps in states where the same party looks likely to win both seats. Take Hawaii, for instance. Although Democratic Sen. Mazie Hirono and Democratic Governor David Ige are both overwhelming favorites to win, Hirono’s forecast margin runs about 23 points ahead of Ige’s. While part of that discrepancy is caused by the presence of third-party candidates in just one of the races (the governor’s race), that would only amount to a small portion of the difference — less than 4 percent of the vote is going to third-party candidates, according to our forecast. It’s possible, therefore, that Hawaii’s relatively elastic electorate9 makes it more open to voting for Ige’s Republican opponent, state Rep. Andria Tupola.

So despite increased political polarization and high levels of straight-ticket voting, our look at the 22 midterm Senate and gubernatorial races this year shows that some voters will still likely vote for different parties on the same ballot. And on election night, we will see at least a few states where the Senate and governor races run very far apart.

Footnotes

  1. To do this, I first subtracted the Democratic vote share from the Republican vote share in each race in our forecasts (as of Wednesday, Oct. 18, at 4:30 p.m. Eastern time) to find the forecasted margin of victory for each race. (Negative values signify a Democratic lead and positive values signify a Republican lead.) I then calculated the difference between the margin of victory in a state’s Senate race and its governor race. So if the margin in the Senate race was -25 (so D+25) and the margin in the gubernatorial race was +5 (so R+5), the overall difference was -30 (meaning the Senate race margin was 30 points more Democratic). We’re showing the absolute value of the difference because we’re trying to show which state’s races had the largest difference in margin regardless of which party the difference leaned toward.

  2. I excluded California because there’s no Republican on the ballot in the Senate race. Minnesota is listed twice because it has two Senate races on the ballot this year. And I counted the Vermont and Maine Senate races as having a Democratic incumbent because Sens. Bernie Sanders and Angus King caucus with the Democrats even though they are independents.

  3. I say “roughly” not only because the forecast could change, but also because my calculation ignores third-party and write-in votes, which will factor into overall the vote shares. For example, there’ll be some Vermonters who vote for Sanders for Senate and a third-party candidate for governor, or some who vote for Scott and a third-party candidate in the Senate race.

  4. Because Democratic Sen. Tina Smith was appointed, not elected, she doesn’t accrue the incumbency advantage in FiveThirtyEight’s model, which means this race functions like an open-seat race.

  5. While the incumbency advantage has ebbed in recent years, incumbent senators, for example, outperformed their state’s partisan baseline by 8 points in 2016. By removing this factor from races with no incumbents, the national environment and state fundamentals are left to exert more influence, at least in theory.

  6. FiveThirtyEight’s partisan lean is based on how a district voted in the past two presidential elections and in state legislative elections. In our formula, 50 percent of the weight is given to the 2016 presidential elections, 25 percent to the 2012 presidential election and 25 percent to state legislative elections.

  7. Respectively, Massachusetts and Vermont ranked as the fourth- and sixth-most-elastic states in our latest calculation.

  8. Maryland ranked as the 11th-least-elastic state in our most recent calculation.

  9. It ranked as the 13th-most-elastic state in our latest calculations.

Geoffrey Skelley is an elections analyst at FiveThirtyEight.

Comments