A new Quinnipiac University poll shows Hillary Clinton trailing Republican rivals in Colorado, which, being a swing state, gets lots of love and attention from pollsters. "Florida Sen. Marco Rubio bests Clinton 52% to 38%, the biggest gap," CNN explains. "Ben Carson wins a potential head-to-head matchup as well, with a 52% to 38% advantage. And Donald Trump leads 48% to 37%."
This seems like bad news for Clinton, the Democrats and frankly the nation itself, if all these clowns can best an experienced and competent politician like Clinton. But don't write off Clinton---or the concept of democracy itself---quite yet. The blunt fact of the matter is that, while primary polling is useful at this stage in the campaign, general election polling this far out from the election is useless.
In 2011, political science professor John Sides, in a q-and-a with the Columbia Journalism Review, explained why we should distrust these kinds of polls that are a full year away from the actual election. "The simplest fact about polling a presidential general election is this: early polls are much worse at forecasting the outcome than later polls," he explained, since the people being polled don't really have the same level of information that they're going to have closer to the election, when they start actually paying attention to the election coverage. (Most voters aren't breathless political junkies like professional pundits are. Shocking, I know.)
Sides followed up this interview with a post at his blog the Monkey Cage, which now lives at the Washington Post, with a chart from a book by political scientists Robert Erikson and Chris Wlezian. They compared poll data on presidential elections from 1956 to 2008, starting 300 days from the election up until Election Day, to see how accurate the polling was. Polling right before the election is, it turns out, pretty good at predicting outcomes. Three hundred days before, you might as well be reading tea leaves.
What changes from 300 days before to the night before? Voters actually start to learn stuff about the candidates. To political junkies, the idea of Ben Carson being more trusted than Hillary Clinton is bananas. The man thinks the pyramids were built to be grain silos! But the person who answers the phone when a pollster calls probably doesn't know anything about Carson at all, and so don't know enough to form a bad opinion of him.
"Carson is a tabula rasa who hasn't received the kind of scrutiny Clinton has, and isn't yet clearly identified as a partisan figure," Scott Lemieux, another political scientist who teaches at the College of Saint Rose, told me over email. "In the highly unlikely event that Carson becomes the nominee, he will be viewed less favorably by many Democrats and Democratic-leaning independents. Bill Clinton has approval ratings north of 60%, but this wouldn't hold up if he could run for president again and was the Democratic frontrunner."
Being perceived as a politician hurts you in the polls. While Clinton was the Secretary of State, she had high approval ratings, but as soon as she ran for office, the negative stereotypes about politicians kicked in and her approval ratings went down. Carson, Trump, and even Rubio, being less well known, are able to still have the glow of not being a "politician" in people's minds, mostly because they have no opinion about them at all. But as soon as you get the nomination, you get the hit of being known as a "politician."
This is also -- sorry, Sanders supporters -- the likely reason Sanders performs better than Clinton in general election match-ups. That's because no one knows anything about him and he gets the not-a-politician glow that would evaporate quickly if he got the nomination.
What can I say? People hate politicians.
None of this means that the Clinton campaign should blow the polls off completely. The polling data shows that voters currently think that Rubio and Carson are perceived as more honest than Clinton. That is an opinion clearly based on low information, but that means that there's a real opportunity here to educate voters on how untrustworthy her opponents really are.
But when it comes to predicting the outcome of the election this far out? No need to worry about it. We have many long months ahead of us and a lot can and will change before the 2016 election.