-
Thursday, October 15, 2020

ANALYSIS/OPINION:

Right about this time in 2016, candidate Donald J. Trump said, “I don’t believe the polls anymore,” when nearly every national poll had Hillary Clinton with a significant lead over Mr. Trump. The polls in that election were dead wrong.

The president has standing to make the case that history is repeating itself, because polls were wrong in key swing states back in 2016. Those polls were a major factor in leading our nation’s thought leaders on both the right and left to declare Mr. Trump politically dead in the water. The polls we see today are similar to the polls we saw in 2016. Polling has proven to be an inexact science and the polling numbers we see today don’t seem to reflect reality. 


The election results and polls from 2016 are a case study in failure. Defenders of those polls argue that national polls ended up being close to correct on election day, yet we all know the national popular vote totals are irrelevant to the final count of the Electoral College. What the defenders will gloss over are the polls that led Americans to believe that Mr. Trump was doomed to failure.

The 2016 polls in swing states were way off. The site Five Thirty Eight projected that Hillary Clinton had a 71.4% chance of an Electoral College victory. The site projected big leads in Florida (2%), Wisconsin (5%), Michigan (4%) and Pennsylvania (4%). The New York Times projected an 85% chance of a Clinton presidency. Larry Sabato’s Crystal Ball projected Clinton would win 322 electoral votes. No pollster projected a Trump victory, because of faulty poll data in the critical states of Wisconsin, Michigan and Pennsylvania. It is possible that we are in a comparable situation right now with similar polling deficits for President Trump in these swing states.

Every four years, the media obsesses over every poll the moment they are released. In too many cases we find out on election day that the polls we believed for months were wildly wrong. Americans seem to have a collective case of amnesia, because we all continue to obsess over polls every four years and tend to believe them. There are other measures of the strength or weakness of a candidacy that helps Americans to see if a candidate has a fighting chance including a growing election prediction markets.

Polls can be volatile much like the voting public. Back in January of this year, many thought Mr. Trump was a lock for a second term because of the economy and the fact that he created 7 million new jobs with record-low unemployment numbers for African Americans, Hispanic Americans, Asian Americans and women. Even then, the polls had Mr. Trump down nationally by about 4%, but he was close enough in swing states that people had already factored in a Democrat bias in the weight of the polls. Many believed that the polls did not reflect the actual state of play as evidenced by betting markets having Mr. Trump either tied or favored to win.

Many have found alternatives to polls that help to predict the strength of a candidate’s chances that weigh polls as a factor, yet not the ultimate determiner of who is going to win. One thing we learned from 2016 is that pollsters missed Trump voters but the prediction markets seem to have baked them in. There is some evidence that Trump voters may be under-polled again this election. Mr. Trump has been pulling record breaking rally attendance dwarfing the numbers seen at Biden rallies.

While the premier sites like Five Thirty Eight give the president a 13 out of 100 chance of winning this fall, the prediction site PredictIt it has had Mr. Trump hovering at about 40 cents on the dollar with Polymarket giving Trump a .38 chance of winning with about a half million shares traded on that market. There is a big disconnect between the political predictors and the prediction markets.

The people putting money into the market versus political pundits who don’t gives the prediction markets a bit more credibility. The betting markets provide a way for a person to factor in the power of incumbency, the history of different states, polling data, the candidate’s political skills and breaking news when placing money on the outcome of an election. 

The American people need to seek alternatives when assessing the candidates rather than relying on pollsters to tell us who is going to win. They have been wrong too many times.

• Brian Darling is former senior communications director for Sen. Rand Paul.


Copyright © 2020 The Washington Times, LLC.