Skip to main content

A Weird Election to Forecast

Donald Trump would be the most hated nominee of the last four decades—less popular, even, than lice and Nickelback—but presidential elections tend to be referenda on the party in power.

By Seth Masket


Donald Trump. (Photo: Justin Sullivan/Getty Images)

What’s going to happen in this November’s presidential election? Political scientists have a pretty good idea of what factors drive elections, but applying them to this year’s likely match-up is surprisingly challenging.

Now, you may legitimately stop here and say, “Didn’t you and many other political scientists totally blow your forecast for the Republican presidential nomination?” And yes, I (and many others) totally did, and I won’t be offended if you want to ignore this and other future prognostications. But the point here is not to rehabilitate my skills as a soothsayer so much as it is to describe how elections tend to run and what makes this year so unusual.

Political science models for presidential elections (see Douglas Hibbs’ or Alan Abramowitz’s for examples, although there are plenty of others) suggest that an election will turn on a few key variables in the political environment, such as:

  1. The Economy: This outweighs all the others. Voters will tend to reward the party in control of the White House during times of economic expansion, and they will punish that party during recessions. Pundits tend to understate the role of the economy, suggesting that, say, Ronald Reagan beat Walter Mondale in 1984 because Reagan was charming and ran more effective advertisements and Mondale foolishly promised to raise taxes. All that may be true, but Mondale could have been a far better candidate and still gotten trounced by similar margins running against an incumbent president in a year the economy grew by seven percent.
  2. War and Peace: Democratic candidates got punished especially hard in 1952 and 1968 — years in which Democratic presidents were prosecuting unpopular wars in which many American soldiers were dying. Voters tend to turn against lengthy wars and punish the party in power for them.
  3. Time in Office: The longer a party has held the White House, the more voters will want to try a different one. Republicans managed to secure a third consecutive term in office in 1988. That was the only time that’s happened since the 1940s.
  4. Ideological Extremism and Moderation: There’s some evidence that voters prefer moderates to extremists, and they’ll reward the candidate they perceive as more moderate even in a tepid economy, as they did when Richard Nixon ran against George McGovern in 1972.

These four factors explain nearly all the variation in presidential election outcomes in modern American political history. Things like spending advantages, candidate personalities, campaign strategies, and scandals tend to vanish in importance when put up against these other factors.

So how do these apply for the 2016 cycle, assuming a match-up between Hillary Clinton and Donald Trump? Here’s where it gets tricky:

  1. The Economy: This is pretty straightforward. Forecasts suggest gross domestic product will probably expand at a bit over two percent this year. We’ll know more later this summer. But that represents decent, if not overwhelmingly strong, growth. On balance, this helps Clinton, but not by a lot.
  2. War and Peace: This is a fuzzier area. No, massive numbers of American soldiers aren’t dying abroad this year as they were in the early 1950s, the late ’60s, or the mid-2000s. At the same time, we’re not exactly at peace; the New York Times yesterday claimed that Barack Obama has been at war for his entire presidency. We’re living in a nearly permanent state of war, albeit one which much of the fighting, at least on our side, is performed by covert operatives and robots, making it appear less bloody to American eyes. Americans may be weary of foreign wars but not quite to the point of switching parties over it. This is probably a neutral issue for the fall election.
  3. Time in Office: This is straightforward. The Democrats are attempting to hold onto the White House for a third straight term, something they failed to do in 2000 with a stronger economy and a more popular incumbent. On balance, this helps Trump.
  4. Ideological Extremism and Moderation: This one’s really tricky. Americans have a pretty well-honed image of Clinton as a center-left Democrat. But where is Trump? On the issue he seems to care most about, immigration by Muslims and Mexican Americans, he has staked out very extreme positions. But on the bulk of issues the parties tend to focus on — abortion, tax rates, Medicare, Social Security, the minimum wage, and more — his stances have been far more tepid, and indeed have shifted a good deal in recent years (even recent weeks). This is why I suggested a few months ago that Trump’s ideal “point” was actually a probability cloud. So who’s the moderate and who’s the extremist in this election? Trump is definitely more extreme on immigration, and particularly in temperament, but, on average, Americans may perceive Clinton as further from their own ideological positions. So this may end up helping Trump.

So you’ve got one factor that marginally helps Clinton and two that help Trump. And yet that’s not quite enough information. Presidential elections, as the above analysis suggests, tend to be referenda on the party in power. But voters actually know a great deal about the likely Republican nominee and don’t like what they know. Trump would be the most hated nominee of the last four decades. (Clinton comes in second.) According to recent polls, he is less popular than lice, root canals, and Nickelback. He’ll lead a party that is having a hard time rallying behind him. His comments toward women have helped to produce a gender gap in excess of 20(!) points.

This may end up being a year in which personality and other candidate-specific factors actually end up mattering, especially when the big four factors of the fundamentals don’t necessarily point in a clear direction. All of this is why Clinton seems to be leading in polls by a consistent and substantial margin. Want to know who’s going to win this fall? You could do worse than just following polls. Or, you know, waiting until November.