One of the political analyses from the 2016 presidential election I hear most frequently is that “the pollsters got it wrong.”
That’s both true and false.
Polling is both an art and a science, and much like a computer, what you get out of a poll depends almost entirely on what you put into it. Some argue that the pollsters actually got the national polls right (within the margin of error), with Hillary Clinton winning the popular vote by a couple of percentage points.
But as we all know, the United States doesn’t choose a president by popular vote, so national polls are about as useless as wearing a mask while alone in your car.
Others alleged that pollsters were deliberately misreporting numbers showing an all but guaranteed Clinton victory, but that’s pretty hard to pull off as well. After all, polls are only as accurate as the data you feed them. That’s like adding up the wrong numbers then blaming the calculator for your incorrect answer. No, the polls weren’t wrong. The voter models used in most 2016 polls, on the other hand, were horrifically wrong.
One of the reasons that the 2016 Trump campaign knew that most of the polls were off came down to the samples, or specifically who the pollsters thought would turn out on election day. They assumed, for example, that African Americans would turn out for Hillary Clinton in similar fashion to their turnout for President Obama in 2008 and 2012. But even President Obama saw a slight dip in African American turnout from 2008 to 2012, as evidenced by his loss in North Carolina in 2012, a state he carried in 2008. Ultimately it didn’t matter, he won both elections.
But it was pretty arrogant to think that a white candidate, particularly one with the highest unlikability indexes in recent memory, would garner the same support from African Americans as the first African American president. Add in the fact that Donald Trump, who had famously addressed a room full of black pastors and reminded them that after 60 years of getting absolutely nothing from the Democrat Party, what did they have to lose? Trump got nearly 9% of the black vote in 2016, which isn’t anything to brag about except that it was double that of his recent Republican predecessors. Some 2020 polls show Trump polling in the high teens among African American likely voters, and that’s just the ones willing to participate in a poll.
Joe Biden is not without his own racially charged gaffes over the course of his half century in political office. But here come the pollsters, assuming that he will automatically receive 90% of the black vote merely because he was anointed by Barack Obama as Vice President for 8 years.
Despite polls showing Biden with a large lead nationally, few media outlets are reporting on the one index that kept Trump in the game 4 years ago: voter enthusiasm. There are two basic ways to determine the likelihood of someone voting; chronically (as in they vote frequently in local elections) and their own answer to a question like, “ On a scale of 1 to 10, 10 being the highest, how excited are you to cast your vote for (candidate).” In this index, President Trump is absolutely smoking Joe Biden by 20 points or higher in every survey. Trump also led Hillary Clinton by similar margins on this question in 2016. In fact, when given a list of options as to why the identified Biden voter is voting for Joe Biden, the answer chosen over 70% of the time is “I’m voting against Donald Trump.”
Nothing about Joe Biden.
Any way you slice it, most of the 2016 pollsters were embarrassed by how wrong they were. They can spend all day pointing to how close they got to the margin of error on the national vote, but that doesn’t excuse how poorly they anticipated what the actual 2016 electorate would look like, and how wrong they actually were on a state by state basis (ask Wisconsin). 2020 has shown that not only does the polling industry not want to be wrong again, they are apparently doubling down with flawed models that suggest one of two things; they are using their numbers as suppression techniques to drive down morale and ultimately depress turnout among Trump’s base (good luck with that), or that they are having legitimate problems getting Republicans and Trump supporters to respond.
I believe the latter, and I think the evidence is that in sample after sample I see the percentage of Democrats surveyed well over 60%, and in recent weeks sometimes over 70%. Identified Republicans in the sample only made up anywhere between 23% and 32% of the metrics I analyzed, even though they traditionally vote in higher numbers than that, and definitely did in 2016.
Going back to flawed models, some pollsters justify the low Republican response by pointing to turnout in the 2018 midterm elections.
Why is that flawed? Easy, Trump wasn’t on the ballot.
But going back to my earlier point, I don’t believe any of that. I think that Republicans or non-Republican Trump supporters have been so aghast at the visceral and often violent reaction from the left when confronted on their political choices that they are this time choosing to keep it to themselves.
So what happens if the pollsters get it wrong yet again in 2020? As I stated before, it would be disastrous for the industry should it be proven that 2016 wasn’t a simple outlier. But drilling down deeper, many congressional Democrats have been whispering about a plan to prevent future political upsets. The halls of Capitol Hill have been quiet lately because of the pandemic shutdown, but that hasn’t made it any less difficult to keep a secret in DC. The idea being floated is to “nationalize” political polling.
What does that mean? Well, much like the US Census, Democrats want your response to political polls to be required by law. You would no longer get to opt out or hang up on a pollster. You would not be allowed to keep your election day intentions private. And taking it even further, they want to make it illegal to lie to a pollster, much like it’s illegal to lie to the police or federal investigators.
Yes, some politicians believe that not being able to see the final score before the game is played isn’t fair to them, and that if you deprive them of this information, you should face legal consequences.
In any other year besides 2020, this would probably be chalked up to some Alex Jones conspiracy theory. But since my rumors generally enjoy a reputation of being premature facts, and the fact that 2020 has made all the impossible possible, don’t count it out. And if Democrats truly sweep the elections the way the pollsters and their Democrat-heavy samples want you believe, then get ready for the next generation of campaigns, where even ballots may become extinct. Cheers!
Brian Trascher is the founder of Gulf South Strategies USA, Co-Chairman of the Louisiana Trump Campaign, and serves on the Trump Victory Finance Committee.