Will Trump Voters Vote Democratic In 2018?

Donald Trump has historically bad approval ratings for a first-year president, which makes sense for a man who also had record-low favorability ratings with the electorate, losing the national popular vote in his election.¹ What is remarkable is that despite upside down approval numbers, Trump remains incredibly popular with most of those who voted for him. According to a September Voter Study Group report, only 6% of Trump voters regret their 2016 vote.² But with Trump voters who had previously voted for Barack Obama in 2012 (Obama-Trump voters), that number jumps to 16%, making them the only subgroup of his voters to register a double-digit regret. Contrast that to the fact that no subgroup of Clinton, or third-party voters even come close to that 16% number. This presents an opportunity for Democrats to reclaim some of these voters—especially in 2018 contests in which Trump won’t be on the ballot.

Who Isn’t Gettable?

The most important place to start is by admitting that Democrats can’t win every vote. The Party has to keep to its role as the voice for the country’s center and left. That means that for many on the right, the Democratic Party won’t even be an option.

We took a look at 2016 voters using the progressive data firm Catalist’s national voter file database. Of the 138 million people who voted, Catalist lists around 50 million as likely to identify as Republicans. This matches up well with exit polls that found that 33% of voters self-identified as Republicans.³ Catalist also offers a model of the likelihood that voters will be willing to split their ticket—or vote for candidates from different parties for different offices. On this test, Catalist found that about 39 million of these likely Republicans (around 78%) are not likely to vote for a non-Republican at any level.

Democrats should not be talking about trying to flip these hardcore Republicans, who have limited to no history of voting Democratic down-ballot. To target them would be a waste of resources, and even worse, could cause the Democratic Party to abandon its values. But that also means that around 11 million Republican voters, or 22%, of these voters that Catalist has identified as likely to identify as Republican have a history that indicates that they are open to splitting tickets. To be clear, even in a good election, the majority of these voters will be voting Republican, but winning some of them could be a key component in winning a close election. Add that to the number of Independents who are open to Democrats, and you’ve got a lot of potential votes on the table.

The Obama-Trump Voters

Numbers from the Cooperative Congressional Election Study back up the Catalist data cited above. And not surprisingly, they point to potential ticket splitters as being mostly Obama-Trump voters rather than Romney-Trump voters. Romney-Trump voters stuck with Republicans in House races by an 86% to 6% margin. But Obama-Trump voters voted GOP in House races by a 60% to 28% margin—meaning 1-in-3 pulled the lever for Democrats even while voting for Trump. Democrats did even better with Obama-Trump voters in statewide elections, picking up 34% of their votes for Democratic Senate candidates and 35% for Democrats in Governors’ races.

These numbers offer a grounded hope that many of these Obama-Trump voters are gettable. Twelve House Democrats currently represent districts that voted for Donald Trump, Missouri Democrat Jason Kander outran Clinton by 16 percentage points in his Senate race, and Montana Democrat Steve Bullock outran her by 30 percentage points in his gubernatorial campaign. Clearly, a significant slice of these voters are open to appeals from the right kind of Democrat with the right message.

Obama-Trump voters won’t be a singular silver bullet for Democrats to take back Congress, but there is no path to a majority without them, thanks to the fact that there are six million of these voters nationwide, and they are geographically located in strategic areas for electoral wins (for more information about the size and spread of Obama-Trump voters, check out the memo cited here).⁴

Just like Obama-Trump voters, the Romney-Clinton voters are up for grabs in 2018, and they will also play an important role in any winning Democratic coalition. But some have argued that Democrats can win next year without getting a single Trump voter, simply by boosting turnout and engaging potential new Democratic voters like the former Romney voters who voted for Clinton. While Romney-Clinton voters may in fact be slightly more likely to vote Democratic in 2018 than Obama-Trump voters, the strategy of focusing solely on Clinton voters and ignoring Trump voters falls apart when one considers Romney-Clinton voters only went Democratic by the slimmest of margins—46% to 43% at the House level in 2016. That could help explain why all 15 Romney-Clinton districts are currently represented by a Republican in Congress. Clearly a vote against Trump didn’t translate directly into a vote against Republicans generally. 

As illustrated in the chart above, Romney-Clinton voters were more likely to vote Democratic for House in 2016 than Obama/Trump voters. And very few of them regret supporting Clinton, which indicates they might be willing to vote Democratic down-ballot in greater numbers next year. In fact, the Voter Study Group report found that around half said they were planning on voting Democratic for Congress in 2018 compared to around 30% for the Republican. That contrasts to Obama-Trump voters, among whom a plurality say they are undecided and less than 40% say they plan to vote Republican. Though much work is still needed as only about 20% say they currently plan on voting Democratic. These sets of numbers point to the fact that while neither group is lost, neither is a slam dunk for Congressional Democrats either—the Party will have to work for them. 

Third party voters won’t be a panacea either. Evan McMullin voters were more reliable down-ticket Republican voters than Obama-Trump ones, and a plurality of Gary Johnson voters chose a Republican for Congress. Jill Stein voters voted Democratic by a wide margin, but nearly half didn’t vote Democratic. And these groups are all fairly small. Johnson and Stein voters are probably worth targeting in certain areas, but McMullin voters are almost certainly not gettable on down-ballot 2018 races.

Also, about 10% of voters in 2016 were ‘new voters,’ or those that hadn’t voted previously. We don’t have data on how they voted for House by presidential preference, but Clinton won this group of voters by a 57% to 38% margin overall. Therefore, if they turnout in a midterm year like 2018, it would be a big boon to Democratic hopes as well. 

It is hard to imagine Democrats winning back the House or holding their own in the Senate by using the strategy of completely ignoring gettable Trump voters. Winning back a significant chunk of Obama-Trump voters next year is likely necessary to rebuilding a Democratic Congressional majority.

Implications of Giving Up Obama-Trump Voters 

The fact that many Democratic candidates were able to win a significant number of Trump voters even at the very moment Trump was winning the White House suggests that they can do so again, and better, in 2018. If we were to assume no Trump voter is gettable, that is essentially writing off the 23 states that Trump won with over 50% of the vote. That is 198 electoral votes, 46 Senate seats, and 205 House seats.

On the Senate side, six of those seats (Indiana, Missouri, Montana, North Dakota, Ohio, and West Virginia) are currently held by Democrats up for reelection in 2018. If Democrats pick up the swing state Senate seats of Arizona and Nevada, that would still be a net loss of four seats.  So giving up on Obama-Trump voters and other Republicans who voted for Democratic Senators six years ago would inevitably mean a smaller Democratic Senate caucus. 

The House wouldn’t be as disastrous on paper, but ignoring Obama-Trump voters would create a strategy that would be too narrow to realistically be able to pull off a majority. Trump won 205 seats with a majority and an additional 25 with a plurality. Fully 21 districts supported Obama in 2012 and Trump in 2016, and about half of ‘swing seats’ as defined by Cook Political Report were won by Trump at the top of the ticket.⁵ Democrats have to win in a lot of different types of places to win a majority, and in many of them, it will require winning back quite a few Obama-Trump voters.⁶

Conclusion

There is a reason these Obama-Trump voters voted Democratic at the top of the ticket just four years ago. These voters are focused on jobs and the economy, and while Trump is not delivering on his promise of being the jobs president, Trump’s economic message was rooted in what they cared about. Third Way recently finished some qualitative research with two distinct groups—Obama-Trump voters and Rising American Electorate (unmarried women, people of color, millennials) voters.⁷ They had very similar concerns and outlooks of how the Democratic Party could address their economic concerns. In both, we found that by returning to messages focused around how Democrats can make the government work for the betterment of the middle class and become the party of jobs again, Democrats just might win back enough of these voters to win in 2018.