The 20-Year Surprise

Roughly every 20 years a sizable portion of the U.S. electorate gets a surprise.

In some presidential elections the outcome is obvious (think 1964 or 1972), while in others the outcome is in doubt (think 1960 or 2000). But in all years there remains a chunk of voters who firmly believe their candidate is going to win, this despite all evidence to the contrary.

For these folks, I offer Karl Rove as your patron saint.

You remember Rove, especially his infamous 2012 election night Fox News meltdown as he argued Mitt Romney still had a chance to win even as Fox’s own stats nerds told him otherwise, that Barack Obama was re-elected. Rove was wrong, but he wasn’t alone. Nearly three-fourths of those who said they preferred Romney before the election predicted he would win.

For this to make sense, it helps to think of election “losers” as belonging to one of two categories — expected losers, those who rightly predicted their preferred candidate would lose, and surprised losers, those who insisted their preferred candidate would win (but didn’t). It’s these surprised losers who are the most fascinating.

Later I’ll get to the “so what” of why this matters. Back to the 20-year surprise. In 1960, in 1980, and in 2000 we had significant increases in surprised losers. Check out the graphic below. The solid line represents the percent of surprised losers in presidential elections since 1952, the dotted line the percent of expected losers (all based on ANES cumulative data). There’s a lot of up and down, a seeming random walk in the data, but that walk isn’t so very random after all. Yes, in close elections those surprised by the outcome naturally increases, and in runaway elections fewer are surprised. You’d expect that, but 1980 was hardly a close election, though it was an odd one.

So every 20 years, like electoral clockwork, we have a large number of surprised losers.

So what?

Democracy, political theorists tell us, rests on the consent of the losers. Not the winners; they’re happy. As political scientist William Riker put it: “the dynamics of politics is in the hands of the losers. It is they who decide when and how and whether to fight on.” Losers are not only less happy about the electoral outcome, they’re also less trusting of government, more skeptical about how the election was conducted, and more doubtful about democracy. Look at those last ones, skepticism of how an election was conducted and satisfaction with democracy. When you’re looking for the “so what” in research, those are good ones to consider.

So understanding losers, that matters, and I argue it’s the surprised losers who hold the key, the ones who expected their candidate to win but on election night discovered otherwise. They have reason to be even more skeptical, more doubtful. After all, they expected to win, and didn’t.

Now look harder at that graphic above, look specifically at the most recent elections. There’s something going on there after the 2000 election that we don’t see in 1960 or 1980, when the differences between surprised and expected losers bounced back to wander closer to one another. We have separation. Another way to look at this is below where I examine the percent of people across these elections who expected their preferred candidate to win (regardless of electoral outcome).

As you can tell, rarely do fewer than 70 percent believe their candidate is going to win. See that jump in later elections, starting in 1996? Something’s going on, something interesting, something perhaps tied to the partisan nature of recent elections, or tied to the partisan nature of cable news media, or tied to … well, I don’t know quite what, yet. That’s why you do the research, but clearly more and more people since a low of 1996 believe their own candidate would win, and 2012 was the highest presidential election year in the data.

I’ve published one study on the question of news media and surprised losers and I think there’s more room here to work from not only a news media angle but also looking at other explanatory factors. But the real “so what” is whether or not this blip in recent elections also represents a significant increase in skepticism about electoral integrity and democracy, because as those kinds of doubts increase, the stability of that democracy comes into question.

Show your support

Clapping shows how much you appreciated Barry Hollander’s story.