People forget this now, but back in 2012, there was a major cottage industry on the right that believed that all the polls showing President Obama leading Mitt Romney were skewed. Fox News devoted many hours to “analysis” on how to unskew the polls to find the “real” outcome of the race in the weeks leading up to it.
And after the election, what do you know, the polls were skewed — in the other direction, towards President Obama. So he ended up winning by an even wider margin that the national and state polls would have us believe.
Of course, we all remember what happened in 2016, and I wrote yesterday about how the 2020 race is different so far and why we shouldn’t be so freaked out about a recurrence. And as Nate Silver has pointed out many times in the past four years, the national polls in 2016 actually turned out to be highly accurate, it was the state polls, particularly those in Pennsylvania, Michigan and Wisconsin that were off. Pollsters have studied what went wrong in 2016 and have adjusted by 1) increasing the number of non-college white voters in their samples and 2) polling more frequently and planning to continue to do so right up until election day, because many key states had a dearth of quality polling in the final two weeks and failed to pick up a dramatic move towards Trump that was reflected in national polls.
Knowing this, the Upshot, Nate Cohn’s New York Times-based election prediction and poll averaging shop, now includes a feature on its poll averaging where it points out that if state polls turn out to be as much off in 2020 as they were in 2016 and 2012 in the same directions, here’s what the current polls might look like. It’s an interesting exercise … but it’s also right back to the voodoo poll unskewing of Fox News in 2012. It’s not based on any kind of serious analysis of how these polls are being conducted or what the demographic mix might be, it’s just an exercise in looking for a way for Donald Trump to win.
They do this partly to cover their ass — hey, don’t blame us if he pulls it out, we showed you the scenario for how it would happen. They also do this because there’s a public outcry for it — conservatives by and large believe that Trump will win and the New York Times has no interest in being their ideological punching bag while liberals fear Trump might win, so why not feed the anxiety a bit if it increases click rates?
Personally, I wish Biden supporters would spend a little less time fretting about how this election could be lost and a little more time contacting everyone they know who can’t stand Trump and making sure they have a solid voting plan. If they plan to vote by mail, have they requested a ballot yet? When do they expect to receive it? Where should they call if they haven’t received it a month before the election? Once they receive it, is there a secure dropbox where they could drop the ballot? If not, what post office location do they trust most to deliver the ballot if taken there? If mail balloting won’t work, when does early voting start? What can they do to protect themselves from COVID at the early voting site? And if they absolutely must vote on election day, can they plan to go with a friend just to lower fears of voter intimidation at the polling place, which could be a major issue this year? And can they plan on taking the day off, if necessary, to make sure their ballot is cast?
Worrying or not worrying will have no impact on the outcome of this race. Taking personal ownership of this election and becoming a self-appointed voting plan project manager for everyone you know could have an enormous impact. So, get to work. You have no control over how good a job the pollsters will do this cycle. You have more control than you might think over your friends and family and their ability to cast a legal, safe, ultimately counted ballot.