Jim Clark: How did the polls get the election so wrong? (opinion)
It was Election night 2016, at 8 p.m. Eastern Time and east coast polls had just closed.
Real Clear Politics, which averages results of the major polls, showed Hillary Clinton at 45.5%, Donald Trump at 42.2%, Libertarian Gary Johnson at 4.7% and Green Party’s Jill Stein at 1.9%.
At the New York Times, Nate Silver’s election prediction model at FiveThirtyEight.com, which uses polling averages to produce probabilities of winning, showed Clinton at 71%. At 10:30 Eastern Time the major networks announced that Trump had won swing state Florida, and Clinton’s probability dropped to 60%.
Then Iowa and Ohio went to Trump, and the Clinton meter went to 40%. Early Wednesday morning, Pennsylvania and Wisconsin were called for Trump, and Clinton’s odds sank to zero.
What happened with the polls? Stalwarts like Bloomberg Politics, CBS News, Fox News, Reuters/Ipsos, USA TODAY, Quinnipiac, Monmouth, Economist/YouGov and NBC News/SM all got this election seriously wrong.
To better understand, we need to take a peek at the shrouded world of pollsters. The basic premise is that it is possible to forecast the future actions of a huge number of people by taking a small sample of individuals which reflect the characteristics of the large group.
So to forecast election results a sample of say a thousand people with the same percentage of Republicans, Democrats, independents, etc. as the electorate is identified and their opinions are asked.
Usually the sample includes only likely voters (those who have voted in past elections). A “fudge factor” called the “margin of error” is mathematically calculated and reported with the results.
When a campaign-shattering event occurs (such as Trump’s salty language tape or Clinton’s on-again, off-again FBI investigation) pollsters generally select a new representative random sample of voters to measure swings in opinion of the candidates.
How has this all worked? Not great. The granddaddy of all polling gaffes was in 1948 when the then only polling company, Gallup, predicted Dewey would crush Truman. The opposite occurred and Gallup announced the error was because they stopped polling a week before the election.
Gallup actually had it easy. Today’s pollsters rely largely on “robocalls” over land lines to contact their subjects but in a market when nearly 60 per cent of Americans rely exclusively on cell phones pollsters tend to get through to senior citizens; federal law prohibits robocalling cell phones (thank goodness).
But for polls to work they have to communicate with voters in order to ascertain their preferences.
Who got the 2016 election right? Only three pollsters accurately predicted a Trump victory — Rasmussen, Investor Business Daily/TIPP and the LA Times/USC poll.
Rasmussen is frequently dismissed by liberal media, which has branded it a “Republican” poll. Investor Business Daily (“IBD”) consistently had Trump slightly ahead. Its final poll had Trump up by 1.6 points over Clinton.
How did they get it right? IBD recognized more enthusiasm for Trump by Republicans and independents and predicted that GOP turnout would equal Democratic turnout despite Clinton’s vaunted “ground game.”
They also recognized a cohort of voters who didn’t vote in 2008 or 2012 but supported Trump in primary elections. These were ignored by other polls as “unlikely voters” but were Reagan Democrats looking for someone to “clean up the swamp.” IBD/TIPP correctly forecast the last four presidential elections.
LA Times/USC also got the outcome right. The newspaper partnered with the University of Southern California to come up with a different polling method.
They carefully selected their sample of voters and communicated by email every day. They never changed their sample but measured changes in sentiment among the same voters. LA Times/USC showed Trump the winner in every poll from the beginning of the general election campaign.
At least we found out where to go before placing election bets.
Jim Clark is president of Republican Advocates. He has served on the Wshoe and Nevada GOP Central Committees. He can be reached at firstname.lastname@example.org.