How Could the Polls Have Been So Wrong… Again?
The 2020 election, much like that of 2016, was wracked with polls that missed the mark by a wide margin. Source.
With Joe Biden projected to be the President-elect, many are questioning why the results are not more conclusive. Although Biden’s victory is likely, Democrats were not as successful as polls forecasted. Prior to the election, polls projected Biden would win by a large margin — just as they predicted for Hillary Clinton in 2016. In addition, Democrats expected to control the House and gain the Senate. Both of these projections are now unlikely. In fact, Republicans are projected to hold their majority in the Senate and gain seats in the House, although Democrats will have the majority.
Many believed the polling this election would be more accurate after the 2016 polls were embarrassingly wrong, however, this year’s results were worse than four years ago. There is much debate as to why the polls could be wrong again, with many arguing that pollsters and the media do not take into account Donald Trump’s popularity. Although it is too early to pinpoint what went wrong, there are some possibilities. One idea is the mistakes corrected from 2016 were not the problem and the real issues were not fixed. Although a possibility, this is believed to not be the issue. A more compelling idea is that there were new problems this year that counteracted the issues fixed from 2016.
However, one common consensus is that pollsters and the media continue to misrepresent and underestimate Trump supporters. Whether the media refuses to accept that Trump is more popular than they thought or simply they don’t address the Trump voter because they don’t want them to exist is up for debate. Whatever the reason, there is no doubt the media and pollsters are out-of-touch with Trump supporters. This idea is why FiveThirtyEight was created by Nate Silver. Silver believed the media did not find the right people to take the polls and relied too heavily on “horse-race” journalism which focuses on anecdotes as opposed to data.
Some believe the Democratic Party is out-of-touch with its voters while others argue there is the possibility of a “hidden Trump”. In 2016, the idea of “hidden Trumps” was widely disputed, however, many now believe it is worth investigating. Trump has been an outspoken critic of the media and it is possible some voters do not trust the polls and choose not to respond. Meanwhile, the people who tend to answer polls are more politically engaged, which is more common among Democrats. Furthermore, there was a spike in support for Biden following the lockdowns and experts believe this is likely due to more Democratic voters taking polls and surveys while being home.
While nearly every poll in the country predicted Biden would win, there was one poll who predicted Trump would win, and at the very least be competitive. This company is called the Trafalgar Group and in 2016 they correctly predicted a Trump victory. In fact, they correctly predicted the number of electoral votes won. Many critics debate whether the Trafalgar Group is a legitimate polling organization due to the unwillingness of Robert Cahaly, the founder, to share polling methods. Cahaly is largely considered untrustworthy by other pollsters due to this, however, he has been closer than any major poll. Cahaly is a proponent of the disputed idea of the “silent Trump supporter”. It is Cahaly’s belief that conservatives are more likely to lie to a pollster or say what they think the pollster wants to hear due to a lack of trust or simply not wanting to be judged on their views. In fact, he claims it takes five times as many calls to get a conservative to answer a poll.
The fact the polls have continually proven unreliable is an issue. Polls are supposed to be a benchmark to gauge the outcome of an election. However, with the past polls being substantially wrong, many are beginning to lose faith in their reliability. The elections allow the polls to be corrected, but the problem is can we trust polling on subjects other than elections? Furthermore, when will polls stop underestimating voters and do the work required to fairly represent them? Although this remains to be seen, efforts will surely be made to correct past mistakes for the benefit of both parties.