Will Demographics Sink Donald Trump?
Can Donald Trump win the presidential election? A year ago most reporters and analysts would have scoffed at the question – at that time, many of us didn’t even think he would run. But with Trump poised to either win the Republican nomination outright or head into the convention with a strong plurality of delegates, it’s a question we should take seriously.
It’s difficult to put exact probabilities on a presidential election more than seven months out, and a Trump vs. Clinton matchup could be more unpredictable than most. The GOP front-runner’s mix of positions on the left and right could change the party’s coalition in unexpected ways. Hillary Clinton is far from a perfect candidate, and it’s unclear how her high unfavorable ratings will change the race. Additionally, it’s been hard to predict Trump so far – it’s possible that he does something that scrambles traditional general election models.
That being said, we can get an early read on his general election chances by examining demographic data and election fundamentals.
Demographics - How Steep Is Trump’s Climb?
Right now most analysts seem to think the real estate tycoon has a demographic problem. The thought process typically runs something like this: Trump might be able to turn out and win working-class whites at greater rates than past GOP nominees, but his less than sensitive rhetoric on race and immigration, his populist positions on taxes and trade and his general bombast would likely cause other large groups (e.g. Latinos, blacks, college-educated whites) to vote against him at higher than usual rates.
This analysis seems right to me, but it’s still early, and there are some voices of dissent here. So rather than trying to predict what will happen in November, I thought it would be more productive to look at some different scenarios to get an idea of just what would have to happen in order for Trump to win. Specifically, I generated four general election scenarios using the RCP Demographic Calculator (the third of which is interactive – it lets you play out your own customized scenario) to give readers a sense of how difficult it would be for Trump to put together a winning coalition under various conditions.
The first scenario – which I call the 2004 Time Machine – basically assumes that when President Obama is not on the ballot, African-Americans vote and turn out as it did before 2008 while every other group votes as it did in 2012. This scenario is somewhat unlikely – voters in every group shift at least a little bit in every election – but it helps us get a baseline for what the 2016 map might look like.
If Trump managed to stay around Mitt Romney’s numbers with every demographic but blacks (whose vote share is set at the average for presidential elections from 1988 to 2004, and whose turnout is set at 2004 levels) he would lose by about two points and the Electoral College vote would look almost identical to 2012 (Florida moves into the GOP column but the rest is the same).
But it’s unlikely that Trump will replicate Romney’s numbers exactly – if Trump were to assemble a winning coalition, it would likely include fewer Hispanics and more working-class whites. That brings us to the second scenario – The Trump Victory.
In this scenario, Trump does worse than Romney with Hispanics – his vote share is knocked down to 20 percent and Hispanic turnout increases by five points (assuming that Hispanics would be energized and turn out in large numbers to vote against Trump). We keep the reversion-to-the-mean with black voters, increase his white vote share to 64 percent and keep everything else at 2012 levels. Note that this is a pretty high bar for the Republican share of the white vote. The last candidate to exceed it was Ronald Reagan, who won 66 percent of the white vote in his blowout 1984 re-election. And if Trump has trouble winning college-educated whites, it will be difficult for him to get to that level with whites overall.
But if Trump were to get numbers like these, our calculator estimates that he would narrowly win the popular vote and either win or come close to winning in a number of blue-leaning Rust Belt states.
The third scenario – which I call the “Racial Minority Nosedive” – is probably the most commonly floated scenario among analysts today. In this case, Trump performs poorly with both African-Americans and Hispanics, but analysts haven’t come to a consensus on exactly how well or poorly he would do with white voters. That’s why I made this map interactive – so you can try out different scenarios and see what the results would look like.
The interactive is fairly easy to use. Just adjust the vote share and turnout in the control panel at the top, and the calculator will estimate the resulting popular vote and Electoral College vote. More details on the math and data can be found here, and the defaults are detailed below.
In this scenario, Trump gets 4 percent of the black vote – about what John McCain got in 2008 – and black turnout remains at 2012 levels. The Hispanic vote is knocked down to 20 percent and turnout is increased five points from its 2012 levels. The rest of the presets are left at 2012 levels.
In this case, Trump would lose by over seven points, and it would take roughly 66 percent of the white vote to get him to a narrow popular vote win.
I think something like the defaults in this scenario – a sizeable loss for Trump built on poor performance among racial minorities – is the most likely outcome of this race, but a lot can change in seven months and reasonable people can disagree on this. So feel free to play with the calculator and make your own scenarios. If you come up with one you think is particularly interesting or predictive, please email it to me or tweet it at me. My contact info is at the bottom of this article.
Finally, it’s possible that Trump loses in a blowout. Specifically, Trump could lose big with minorities and his losses with college-educated whites might more than offset his gains with blue-collar whites. That scenario might look something like this:
Turnout is unchanged from 2012 for every group except Hispanics (who tick up five points). Trump gets the same share of the black vote that McCain got in 2008, half of Romney’s share of the Hispanic vote and the same share of the white vote that George W. Bush got in 2004. The Asian and Other vote is left at the 2012 level.
In this case, Trump loses by about 10 points and is routed, 385-153, in the Electoral College. Our estimates put Georgia, Arizona and Indiana all within one percentage point of flipping to the Democrats – so slight differences between the model and reality could lead to an even larger Electoral College loss for Trump.
Fundamentals - Trump’s Other Route to Victory
The other route to a Trump win lies in the “fundamentals.” When analysts talk about fundamentals, we’re usually referring to a broad set of factors that campaigns have no control over yet set the stage for the election, like the state of the economy and the incumbent president’s job approval rating. If the economy is good and the incumbent party president is popular then that party’s candidate has the wind at his back (e.g. Bill Clinton in 1996, Ronald Reagan in 1984 or George H.W. Bush in 1988). On the other hand, if there’s a recession, a scandal or a disastrous war, voters tend to punish the incumbent party (e.g. John McCain’s loss in 2008 or Jimmy Carter’s in 1980). And while an especially good (or bad) campaign can occasionally buck the fundamentals, the candidate favored by the economy and other broad circumstances generally wins.
Right now, the fundamentals heading into 2016 don’t favor either party too heavily. The economy is a mixed bag with some good news (lower unemployment) and some bad news (stagnant wages for many). President Obama’s approval ratings are also okay – today’s RCP average has 48.6 percent of Americans approving of his job performance and 47.0 percent disapproving.
If the fundamentals stay roughly where they are now, I suspect Hillary Clinton (who will almost assuredly win the Democratic nomination based on current delegate math) would be the favorite, based on the demographic problems I think Trump will have. But if the fundamentals go south for the Democrats -- if there’s a recession, a scandal or an avoidable terrorist attack -- then Trump could overcome his disadvantages and win the White House.