Election Polls: Fool Me Once, Shame On The Pollsters. Fool Me Twice…

 

In the calm before the electoral storm of November 3rd, one of the fundamental questions on many people’s minds was whether we can trust the polls. The polls were infamously inaccurate in 2016, or as The Daily Wire put it at the time, “a vast majority of the key polls were not just wrong, they were humiliatingly wrong,” with some lowlights listed below:2016 Polls: Trump +3%

2016 Result: Trump +9.5%

Correct – Under by 6.5%

Michigan

2016 Polls: Clinton +3.4%

2016 Result: Trump +0.3%

Wrong – Off by 3.7%

Nevada

2016 Polls: Trump +0.8%

2016 Result: Clinton +2.4%

Wrong – Off by 3.2%

New Mexico

2016 Polls: Clinton +5%

2016 Result: Clinton +8.3%

Correct – Under by 3.3%

Ohio

2016 Polls: Trump +3.5%

2016 Result:  Trump +8.1%

Correct – Under by 4.6%

Pennsylvania

2016 Polls: Clinton +1.9%

2016 Result: Trump +0.7%

Wrong – Off by 2.6%

Wisconsin

2016 Polls: Clinton +6.5%

2016 Result: Trump +0.7%

Wrong – Off by 7.2%

The most egregious polling errors occurred when pollsters predicted the wrong winner, often by large margins. They projected that Clinton would easily win Michigan, and Trump actually won narrowly. They projected that Trump would narrowly win Nevada, and Clinton actually won comfortably. They were also wrong about Pennsylvania and Ohio, with Trump defying predictions by 2.6% and an unforgivable 7.2%, respectively.

With some battleground states reporting results, were the pollsters more accurate this time?

Note: At time of writing, Arizona, Georgia, Nevada, North Carolina, and Pennsylvania have not been called by multiple outlets.

Florida

2020 Polls: Biden +0.9%

2020 Result (with 96% of estimated votes reported): Trump +3.4%

Wrong – Off by 4.3%

Iowa

2020 Polls: Trump +2.0%

2020 Result (with 92% of estimated votes reported): Trump +8.2%

Correct – Under by 6.2%

Michigan

2020 Polls: Biden +4.2%

2020 Result (with 97% of estimated votes reported): Biden +1.2%

Correct – Over by 3%

Minnesota

2020 Polls: Biden +4.3%

2020 Result (with 94% of estimated votes reported): Biden +7.1%

Correct – Under by 2.8%

Ohio

2020 Polls: Trump +1.0%

2020 Result (with 90% of estimated votes reported): Trump +8.1%

Correct – Under by 7.1%

Texas

2020 Polls: Trump +1.3%

2020 Results (with 96% of estimated votes reported): Trump +5.9%

Correct – Under by 4.6%

Wisconsin

2020 Polls: Biden +6.7%

2020 Result (with >98% of estimated votes reported): Biden +0.6%

Correct – Over by 6.1%

Based on current data alone, with some battleground states still to be decided, it seems like the polling has — if anything — become more inaccurate, particularly regarding states Trump ended up winning. Pollsters incorrectly predicted that Biden would narrowly win in Florida, a state Trump won comfortably. Even in states where polling correctly projected the winner, the errors went far beyond the expected margin of 2-3%. In Iowa, polls were off by 6.2%. In Texas, it was 4.6%, in Wisconsin, it was 6.1%, and in Ohio, polls missed the mark by a staggering 7.1%.

As discussed during The Daily Wire’s Election Night coverage, pollsters had claimed that they were making changes to address their failures in 2016. However, as we predicted, they seem to have failed — yet again — to account for the central variable of “shy” Trump voters; voters who, due to understandable concerns of societal pressure, don’t feel comfortable admitting that they plan to vote for Trump to their friends or family — let alone an anonymous pollster on the telephone.

Polling data is an important part of the electoral process, both for campaigns and for those following them. However, polling data is useless if it is also demonstrably inaccurate. The unfortunate fact is that — despite their supposed efforts — pollsters don’t seem to have learned from their mistakes, and have failed to provide us with the information we need. After two elections with appalling projection data, have we been fooled twice?