Bernie Sanders fans and #NeverTrumpers might not be ready to discuss a Hillary Clinton versus Donald Trump general election matchup, but the rest of the political world is—and the picture right now looks pretty grim for the Republicans.
It didn’t necessarily have to turn out this way. There’s good reason to think that, had the Republicans run a “generic” candidate, the GOP would have had an electoral edge in 2016. But Trump is anything but generic, with historic unfavorables that surpass even Clinton’s significant ones—and the math shows that Clinton, as unpopular as she is, could potentially be the first candidate since 1984 to win the two-party popular vote by more than 10 percentage points.
Story Continued Below
That said, we’re six months out from Election Day—plenty of time for more Trump surprises in an election year already riddled with them.
And with that, here are 5 points to keep in mind as we kick off the main campaign—from what to expect if Trump doesn’t change course, to what to consider if he does.
1. A generic Republican might have been a favorite for the White House.
Way back in early 2015, there were at least two major reasons to think the GOP would have a leg up in a presidential contest.
For one, fundamental factors such as the economy and the president’s approval rating were operating in the Republicans’ favor. The incumbent president’s party typically pays a price for economic troubles, and, while far better than just a few years earlier, in 2015 the economy was not exactly roaring. Average annual growth from 2013 to 2015 was only about 2 percent, compared to an average annual growth rate of more than 3 percent since World War II. On top of that, President Barack Obama’s approval rating was about 45 percent at the end of 2015, which, based on past elections, would give a slim edge to the GOP in a two-party presidential contest.
Second, parties almost never win three straight White House terms. Since Harry Truman’s time, only the Republicans have managed to accomplish this feat—and this just once, when George H.W. Bush won in 1988 after Ronald Reagan’s two terms in office.
The numbers are a little different in 2016: Obama’s approval rating sits around 50 percent, and the economy remains decent enough that Americans feel relatively good about it—similar to how they felt before the Great Recession hit. But neither Obama’s job performance rating nor economic growth are high enough to give a significant edge to the Democrats, especially when coupled with the fact that the nominees are competing for an open seat (incumbents usually have an advantage). In fact, political scientist John Sides and his colleagues built a probability model based on those three factors that suggests a generic edge for the Republican nominee in the fall.
With that said, this year’s Republican nominee is not a generic nominee—and that changes everything.
2. In a Clinton vs. Trump race, Clinton begins as the favorite.
One would not expect a candidate with a -12 net favorability rating to enter a general election campaign as the favored competitor. But Hillary Clinton will indeed begin the long march toward November as the favorite. What Clinton needed is an opponent who is even more disliked by the public than she is, and Donald Trump is just what her doctor ordered: Trump’s net favorability is currently -24 according to HuffPost Pollster’s polling average. We appear to be headed for a matchup between perhaps the two most loathed general election candidates in modern U.S. political history.
Yes, it’s true that Trump is an unprecedented political figure who has been consistently underestimated, only to remarkably end up in his current position as the presumptive GOP presidential nominee. But, in reaching that elevated standing Trump has also alienated large swaths of key constituencies, including many Republicans. His unprecedented unpopularity will likely have serious, negative consequences for his electoral chances.
The worst number for Trump may be his rating among women. At the start of April, Gallup found that 70 percent of women held an unfavorable view of the real estate mogul, compared to 58 percent of men. While more women vote Democratic than Republican—a partisan gender gap that has existed in every presidential election dating back to 1980—women will likely form a slight majority of the electorate in November, just as they have for decades, so they are still a constituency that Trump should worry about—a lot. And, considering Trump’s hits on Clinton for “playing the woman’s card”—which Clinton happily embraced in a fundraising appeal—his gendered language and attacks probably aren’t going away. While Trump’s campaign believes this will help him improve his support among white women, who have backed all GOP nominees since 1996, that strategy is a bit of a gamble, and could well backfire.
And then there are Hispanic voters, who appear to abhor Trump. The research firm Latino Decisions recently found Trump’s net favorability among Latinos to be -78 percent, while Hillary Clinton’s is +29 percent. To put Trump’s numbers into perspective within his party, Ted Cruz’s net favorability was -16 percent, and John Kasich’s was -10 percent. Although Hispanic voters will be heavily concentrated in uncompetitive California and Texas, they will be very important in at least three swing states: Colorado (where Hispanics made up 14 percent of the state’s 2012 electorate), Florida (17 percent) and Nevada (19 percent). In light of how Trump is viewed by this demographic group, it’s not difficult to imagine Clinton winning 80 percent of Latinos after Obama won 71 percent in 2012. And, most projections expect Latinos to make up more of the electorate than they did in 2012, when they comprised 10 percent of all voters. That assumption is based partly on the growing Latino population, but also on the fact that hatred of Trump may motivate more Hispanics to register to vote and turn out to the polls.
Lastly, party unity is likely to be a bigger problem for Trump than Clinton. There’s little question that #NeverTrump is a larger force within the GOP than the anti-Clinton contingent is within the Democratic Party. Take the April 26 Pennsylvania primary as an example. Based on the exit poll, 84 percent of Democrats said they would definitely or probably vote for Clinton if she won the Democratic nomination, and 11 percent said they would be “scared” if Clinton became president. Overall, 69 percent of Democrats felt the Clinton-Sanders contest had energized the party while 26 percent felt it had divided Democrats. Contrast those numbers to views of Republicans in the Keystone State: Only 39 percent felt the GOP campaign had energized the party while 58 percent felt it had divided Republicans. In total, 77 percent said they would definitely or probably vote for Trump in the general election, and 22 percent said they would be “scared” if he became president. Of course, there is ample time for Trump to bring anti-Trump Republicans back into the fold, and his favorability numbers among party members have improved in recent weeks. Still, via Gallup, his net favorable rating among Republicans was +29 as of May 5 versus +44 for Clinton among Democrats.
These factors, coupled with Clinton’s healthy lead over Trump in early horserace polling, led the Crystal Ball to make Clinton a large favorite in our first Clinton-Trump Electoral College map, in which we give Clinton a 347 to 191 edge in the electoral vote. (Many people, including some Republicans, have told us they believe this projection is actually too kind to Trump.)
3. Recent presidential elections have been relatively close in the popular vote. This one might not be.
Dating back to 1988, seven straight White House contests have been decided by less than 10 percentage points in the two-party popular vote. This competitive streak matches the country’s previous record, which occurred from 1876 to 1900.
We at the Crystal Ball have been fairly adamant that it will be hard for either party to win more than 55 percent of the two-party vote in 2016, in part because of election fundamentals—economic conditions and the incumbent president’s approval rating—and because of the political polarization that exists in the country today. Months of anti-Clinton ads and presidential campaigning are likely to push many recalcitrant Republicans toward backing their party’s nominee, improving his electoral chances.
However, it’s possible that Clinton could break this competitive streak and win by a little more than 10 points—particularly if Trump struggles to unite his party around him, continues to poll terribly with nonwhite voters, and remains weaker than Mitt Romney’s 2012 support level among whites, particularly white women. Here’s some basic demographic math: Take the
2012 exit poll as a starting point and calculate the two-party vote based off those voting percentages. That year, the electorate was 72 percent White, 13 percent African American, 10 percent Hispanic, 3 percent Asian and 2 percent something else. Now, add in
the assumption that the electorate will be slightly more nonwhite in 2016, with an uptick in Latino vote share to 12 percent and Asian to 4 percent, and with Blacks falling to 12 percent without Obama on the ticket. If Trump performs about three points worse among white voters than Romney did in the two-party vote, perhaps as a result of losing some highly educated suburbanites, while Clinton wins four out of every five Latino votes because of Trump’s anti-immigration rhetoric, that could produce the conditions for Clinton to win the two-party popular vote by just a little bit more than 10 points.
4. The down-ballot consequences may be grim for Republicans.
In 2014, Republican Govs. John Kasich of Ohio and Brian Sandoval of Nevada each won reelection and dramatically increased their share of the vote from their initial victories four years earlier. Kasich went from 49 percent in 2010 to 64 percent, and Sandoval jumped from 53 percent to 71 percent. Yet, while their vote share increased, their actual number of votes did not: Kasich added just about 56,000 votes to his total (out of about 3 million cast), while Sandoval added just about 4,000 votes (out of about 547,000 cast). Why? Because, the second time around, these governors’ Democratic opponents were so uninspiring that Democrats didn’t show up to vote. (Both Sandoval’s and Kasich’s 2014 opponents won about half the total votes of their predecessors.) And this led to a down-ballot disaster: In 2014, Democrats got blown out in every statewide race in Ohio, and Nevada Democrats surprisingly lost all the statewide offices in Nevada, as well as control of the state legislature.
We bring this example up to note that there’s reason to be skeptical of Republican efforts to insulate themselves from the top of their own ticket. Yes, gubernatorial elections are different than presidential elections, and midterm turnout is naturally lower than presidential-year turnout. But it’s not crazy to envision a similar situation unfolding at the national level. If Trump’s lackluster numbers remain poor, we could see a notable drop in Republican turnout—which would threaten GOP congressional candidates in key states and districts.
Given the increasing amount of straight-ticket voting, where voters pick the same party for president and for congressional races, and because the most competitive Senate races this year are mostly taking place in presidential swing states, it seems likely that the party that wins the White House will also win the Senate. It’s possible that the House could come into play, but Republicans have a nice 30-seat cushion; moreover, House Democrats have failed to recruit decent candidates in some potentially vulnerable GOP seats.
And that’s just if turnout remains pretty average.
If Republican turnout craters because of Trump, the damage to the GOP could be extreme, and many Democrats who look unelectable today could find themselves in office next year. Democrats might win Senate seats that the Crystal Ball currently see as leaning toward Republicans, such as Arizona and Missouri, creating opportunities for a large majority in the upper chamber. And in the House, depressed GOP turnout could swing the House, unseating Republicans thought to be safe and ushering in a number “accidental congressmen.”
5. If the numbers change, the outlook needs to change.
It’s clear that most analysts, ourselves included, did not take Trump’s candidacy nearly as seriously as we should have when he announced it back in June 2015. Once his numbers among Republicans improved and he held polling leads for months—far longer than some shooting star candidates did in 2012—observers should have recalibrated their expectations and given Trump more credence as a candidate. But, if you look back at polling at the start of his campaign, there were legitimate reasons to doubt Trump.
For one, Trump had flirted with running several times before, and it didn’t seem at all certain that he would be in the race for the long haul. Additionally, Trump’s favorability among Republicans was very weak: Quinnipiac University found it at 34 percent favorable/52 percent unfavorable in late May 2015 and Monmouth University found a 20 percent /55 percent split in June, right before Trump entered the race. Things got complicated when Trump quickly turned those numbers around and shot up the charts in Republican primary polling. Analysts didn’t make a mistake in pointing out Trump’s poor numbers when he entered the race; instead, our collective mistake was being too slow to adjust when the numbers did change and his durable level of appeal became apparent.
As much as many like to dump on polls, a sneaky story of this election season is that, taken together, the polls have been fairly decent. The many national polls that showed Trump leading throughout 2015 accurately picked up on his solid support from at least a plurality of Republican primary voters. And while national polls severely undersold Bernie Sanders’ potential—which was understandable considering that he was not a true national figure when he announced his candidacy—they consistently showed Clinton with a big lead, and she continues to lead Sanders by 14 points in the aggregate Democratic popular vote.
On the state level, 19 primary states had enough polling data to allow for HuffPost Pollster to create a polling average for both the Democratic and Republican contests. Of those 38 total primaries, the leader in the polling average won the most votes in 35 of them, the only exceptions being the Indiana and Michigan Democratic primaries (Sanders won both after significantly trailing in polls) and the Oklahoma Republican primary (where a big Trump lead turned into a comfortable Cruz win). Granted, several other poll averages were off by a significant amount—particularly on the Democratic side—but by and large the polls got the winners right. Those looking for pinpoint accuracy in polling might have been disappointed, but those using these polls as a rough guide to point themselves in the direction of the winner did get a good sense of where these races were headed.
Therein lies a lesson for the general election, especially because we’re at the point where these surveys are starting to have some predictive value for November: If the numbers start to change, pay attention. Trump’s horserace numbers against Hillary Clinton, both nationally and at the state level, might be poor today, and it might be unlikely that he’ll have as much success improving those numbers with a general election audience as he did with Republicans alone—but we should not rule this possibility out. And if the numbers do change in a consistent way across several reputable polls, we need to write a new narrative.
