Too close to call: Polling — at all levels — takes a beating


Note: At the time of the submission of this column, multiple states have yet to be called for president.

Not only has President Trump trailed Joe Biden in the national polls for two years, he has lagged in most swing states and he has been behind in practically every issue. Whether job approval, personal favorability, character or the issues, Trump has terrible numbers. Only in his handling of the economy, strong leadership and whether he will win has Trump done well.

By any metric, Trump should not have stood a chance.

Yet, it is clear that the polls systemically underestimated Trump’s strength (not a surprise of faithful readers of this column). The final Electoral College tally according to the final 538 averages would have been 348-190. There is no chance Biden gets to that number. Polls in Florida, North Carolina and Ohio were off by up to 10 percentage points.

Among the greatest hits (or misses) were Quinnipiac, which appears to have been off by 12 points in Ohio and 8.5 points in Florida; Morning Consult’s margin appears to have been off by about 5 points in Ohio and about 9.5 points in Florida. At this writing, Trump leads in Georgia, North Carolina and Michigan — all states where the polling averages put him behind. The averages have Trump down 8 points in Michigan alone.

But the polling problems are not just on the ballot test. It appears to saturate the entire set of polling questions. I review the YouGov polls, as it is the only pollster that consistently releases the detailed crosstabs. When you look at all the questions and different demographic and partisan groups you can find out why people are making their choices. Sometimes there are oddities that call the poll’s validity into question.

Looking at the last YouGov poll, the problems in polling — particularly polling about Trump — look to be more severe than underestimating his support by a few points. The final poll changed little compared to the previous polls from YouGov. Trump’s underlying numbers are terrible and support the overall ballot test number. Trump had net disapproval on every issue and overall. He was under water on all personal qualities, and people continue to hate his Twitter use.

Normally the underlying numbers could tell us if the ballot test is not quite right, but Trump’s weak underlying numbers match up with a losing ballot test. How to explain this problem?

Pollsters may well be having a problem getting honest answers across the board.

It is not so much that people are trying to sabotage the polls with outright lying, it is more that people have a tendency to give people the information that they want to hear.

The legacy media has engaged in a crusade against Trump. Not only that, they oppose anything and everything he does and says. Trump is the most socially unacceptable President in living memory. Respondents may just be telling the pollster what they think the pollster wants to hear.

Consider the coronavirus. Trump scores very poorly on how he has handled the virus and has for months, with Biden polling better. This issue alone should sink him, given public concerns. But if you look more deeply into the YouGov poll, you will find that only 3 percent of respondents say they have tested positive and 21 percent say a family member has; 6 percent of respondents say they know a family member who has died and 14 percent, a friend (these numbers are suspicious as less than 0.1 percent of all Americans have died of the virus).

At the same time, unemployment is having a more direct impact with 11 percent of respondents having been laid off, 21 percent saying they know a family member who has been laid off and 21 percent knowing a friend. All this implies that lockdowns and economic problems are having a more direct impact on people’s lives than the actual virus. While Biden did not promise a new national lockdown, he has not absolutely ruled significant restrictions out, potentially leaving voters concerned.

It may be that voters feel that expressing greater concern over their economic circumstances as opposed to public health is too embarrassing and not appropriate. Yet, when they have to actually vote, economic concerns drive decision-making, not a vague threat they have not experienced.

The bottom line for this election, no matter who ends up president, is that the polling industry has generally done a terrible job. And it’s not just the ballot test, the polls appear to be saturated with problems.

The polling industry needs to take a hard look at how they are polling, who they are polling and what questions they are asking. Above all, they need to restore public confidence that they are reporting the objective facts. Until they restore credibility, respondents will continue to fudge their answers or not participate at all.

NOTE: This post has been updated to correct the apparent degree of error of the Morning Consult polls.

Keith Naughton, Ph.D. is co-founder of Silent Majority Strategies, a public and regulatory affairs consulting firm. Dr. Naughton is a former Pennsylvania political campaign consultant. Follow him on Twitter @KNaughton711.

Tags 2020 election 2020 presidential polls Donald Trump FiveThirtyEight Joe Biden Opinion poll Polling Presidential polling Public opinion YOUGOV

More Campaign News

See All
See all Hill.TV See all Video