Can We Trust the Polls When Predicting the Winner of the 2020 Presidential Election?
With the 2020 US presidential election drawing close, some observers find themselves a little skeptical about how the media reports polling data.
And that’s absolutely understandable on account of Donald Trump’s shock victory in 2016. Despite mainstream media outlets all unanimous in predicting a big win for Hillary Clinton, the final results of the election were a lot different to what the polling data had suggested.
Clinton was expected to win… and to win well. She was at around -300 with bookmakers around the world, and it seemed that Trump was fighting a losing battle. The media appeared certain that the Democrat nominee would secure the Clinton family back into power, and supporters of the former First Lady looked supremely confident.
This raises the question — ahead of the 2020 presidential election, can we trust polls and the media to deliver a more accurate interpretation of pre-election numbers? I’m not so sure, and I’ll explain why Trump’s involvement in this election makes the gathering of such data difficult, at best.
Why Was the Polling Data So Wrong in 2016?
Can we say the greatest political shock in years went down when Trump was elected to office in 2016?
Almost a year later, many folk were left scratching their heads and wondering what the heck just happened. Trump supporters were delighted, of course, and the fact that his victory appeared to be one against all odds made it even sweeter.
But how were the polls SO wrong? Were they that wrong? I mean, they did predict that more people would vote for Hillary Clinton. And that’s what happened. The former First Lady won the popular vote.
But the inaccurate nature of the polls led to serious questions over modern data and its interpretation. It wasn’t as though we were in 1948, when almost every single reputable news outlet had Democrat Harry S. Truman behind New York Governor Thomas Dewey by a long distance, only for Truman to win at the death.
Even the New York Times reported that “Dewey’s election as President is a foregone conclusion,” only for things to turn out anything but.
But in a world where technology should be improving the accuracy of political prognostication or bettering our ability to predict who wins an election, 2016’s result proved you shouldn’t trust everything you read.
Polls Had Clinton’s Chances of Winning as High as 99%
That being said, there was a fundamental issue with the forecast of electoral votes. Almost every poll had Clinton in the ascendency, so imagine the shock when Trump crossed the line first.
Were people being lied to, or was the pro-Clinton data just so far off? That’s what people are still trying to figure out, especially those betting on Trump vs. Biden.
Before the election, the Princeton Election Consortium rated Hillary Clinton’s chances of beating Trump to the White House at 99%. That’s right, 99 percent, according to Princeton’s Bayesian model.
Just three days before the election, the prediction of 312 electoral votes for Clinton and 226 for Trump was not right, was it?
Are the People to Blame?
According to a 2017 report on the 2016 presidential election polling data by the American Association for Public Opinion Research, silent Trump supporters were found to be, in part, to blame for the inaccuracy of the poles.
Simply put, the Shy Trump effect could have played a big part in the results. In other words, those voting for Trump — but not willing to reveal that they were voting for Trump — means that the polling data was skewed.
This is something that could play a big part in the 2020 election, according to a study.
Back in 2016, another reason cited for the inaccuracy of polling data was as an underestimation of support for Trump in Pennsylvania, Michigan, and Wisconsin. The Democratic Blue Wall did not hold, and this was huge for the Republican candidate.
Educated Voters Were More Likely to Support Clinton
Also, a “strong correlation between education and the presidential vote in key states” was found to have played a part. The AAPOR report found that voters with higher education were more likely to vote for Clinton. They are also more likely to respond to surveys than those with “less education.”
Polling data was provided to the American public without accounting for an overrepresentation of college graduate voters, thus giving the impression that support for Clinton was higher than it really was.
Can We Trust Media Polls For Trump vs. Biden?
What has changed since 2016? Sure, we certainly have a better understanding of why polling data was off in the last election, but there is the same fundamental problem that makes it difficult to have faith this time around.
In essence, many Trump supporters are less likely to be upfront about their choice of candidate because of the stigma involved in doing so. And that makes things very interesting. Not only did it turn more than contrarians around to his cause, but it meant that predicting the political leanings of many people around the country was very difficult.
And of course, you have to ask yourself has there ever been a presidential candidate that has been lambasted like Trump in 2016? The answer is no. Vocal support of Trump is, therefore, something that many people wanted to avoid. That applied to sitting around the dinner table with partizan Democrat supporters or commenting on a news story on social media.
Social media, in particular, provided a platform for so many to express their negative views of Trump and created an environment where even suggesting sympathy towards the Republican candidate was enough to draw varying levels of criticism from friends, family, and relations.
Even More Silent Trump Supporters in 2020?
Imagine that you are the type of person to avoid conflict online. You might be supporting Trump in a family of Democrats. Or, in a state that is more liberal than conservative. You know that you can have an easier life by keeping your mouth shut and basically casting your vote without anyone knowing who you have opted for.
Not only does this make pre-election polling data questionable — it arguably makes it close to useless, in some respects. Trump’s supporters are even less likely to participate in surveys, either online, in person, or over the telephone, given how badly his presidency has been viewed by many quarters of the mainstream media.
Quite simply, it’s tough to use polling data to predict the 2020 presidential election. Those who followed it like gospel in 2016 will be conscious of the old proverb “fool me once, shame on you; fool me twice, shame on me.”
Of course, that’s my opinion. Would I bet on the 2020 presidential election polls being accurate? I don’t think that would be the wisest move, for the most part. However, you must remember that it was SOME elements of the polling data that were wrong, and not everything.
Still, I’m sure most of us don’t like getting bit by the same dog twice.
More on the 2020 Presidential Election
The 2020 US presidential election is arguably the most important political event of the century.
My colleagues and I are covering the election extensively and from all angles. You will find many informative pieces on Trump vs. Biden on our blog, so please do have a look through if you are interested in betting on the election, the result, or just politics as a whole.
The following are among our most popular posts.