By early March 2012, Americans were telling The Washington Post that they figured President Obama would be reelected. Obama had a slight lead in the RealClearPolitics average at that point, but the assumption that he would win came from a poll The Post conducted with ABC News in which Mitt Romney had a slight lead. Over the rest of the campaign, Obama and Romney jockeyed for a lead when we asked voters whom they planned to vote for. But there was no change in who people thought would win: The guy who won.
Right now, Americans think that Hillary Clinton is going to win the election in two weeks. CNN-ORC released a new poll to that effect Tuesday evening. Two-thirds of respondents think Clinton will win; a quarter bet Donald Trump will be the victor.
A 2012 study suggests that our polling from that year was not a fluke — and that Clinton's lead in the who-will-win question actually correlates to a higher likelihood of winning.
Researchers Justin Wolfers and David Rothschild looked at historical uses of the question in presidential polling at the state level and tallied how often the question of expected winner was correct vs. how often the top-line who-will-you-vote-for question was.
In most years, the expectation question ended up being correct in more state contests. In 2008, the numbers were the same; in 1972, the expectations question was wrong in one of the 38 states reviewed.
In his write-up of the research shortly before the election four years ago, the New York Times' David Leonhardt spoke with Wolfers, who explained why responses about the expected winner were shown to be more accurate.
The question allows people to consider not only their views but also those of their relatives, friends and colleagues, [Wolfers] said. Some voters may also give more-honest answers about their own plans, rather than naming a candidate who briefly intrigues them, as happened in the Republican primaries this year. And an expectations question allows people to take into account speeches, debates and news media reports.
The question also steps outside partisan lines, to some extent. Wolfers and Rothschild determined that Democrats were more likely to say that they figured the Democrat would win, but as the odds of the Democrats' victory increased, so did the willingness of Republicans to agree that it would happen.
CNN's poll reflects that.
Only 6 percent of Democrats and 5 percent of Clinton supporters think Trump will win. By contrast, 40 percent of Republicans and Trump supporters each think that Clinton will be victorious.
Trump can still win. Math and history aren't destiny. The nation thinks he'll lose, though, and it has a pretty good track record of predicting its own elections.