One month ago, the progressive newspaper Vox warned readers that the Democrats' hopes for a positive election result could be dashed again. The big surprise was, of course, 2016. The polls showed Hillary Clinton as the clear winner. Nevertheless, although she won the popular vote, she failed to win the presidency against all odds.
Why did all forecasts fail? One possible explanation is that surveys are only an approximate science and that they can be wrong in either direction. Andrew Prokop, for Vox, had this to say:
My view is that it makes all the sense in the world to be deeply skeptical of polls showing big Democratic leads in states like Wisconsin and Ohio, where polls have consistently greatly overestimated Democrats across several election cycles. But the picture is less clear in other states, where polling error hasn’t been so clear or consistent. I wouldn’t blindly “trust” those polls, but I wouldn’t assume they’re likely wrong, either.
Underrepresented Republicans
It is a nuanced and judicious opinion, but perhaps insufficient. According to pollster Robert Cahaly of Trafalgar Group:
A certain segment of the Republican electorate is just not getting polled. An important chunk of people is what I call submerged voters that you cannot get on the phone and you cannot reach them any other way.
John Gizzy, NewsMax's White House correspondent, pegs the skew in the polls in favor of the Democrats at around 4-5 points. If true, it would be a huge bias, in an electoral market with only two parties in most elections, and as competitive as this one is. Gizzi relies on the scrutiny that he performs at PollingMonitor.com to the pollsters. For example, Quinnipiac has accumulated a bias in the last three elections (2016, 2018, and 2020) of 4.6 points in favor of Democrats.
Real Clear Politics does its study of polling bias. And, although it offers its average of the polls in each election, its forecasts correct the bias of each pollster, to obtain a more accurate result. So, there is a difference between 538, which does not correct for bias and expects Democratic control of the Senate, and RCP, which does correct for bias and predicts a Republican victory for the Senate, 52 senators versus 48.
Correcting for bias?
Cahaly defends the work of Trafalgar Group, in its effort to have the voice of that part of the electorate. And it suggests that the major pollsters are making an effort to do so, too.
We look at each of the states and see that their major polls are starting to get closer and closer to ours. I mean, when the major polls are so close to ours that Nancy Pelosi disavows the major polls, you have to start thinking they're getting at the truth.
If so, this bias may be lower in this election. But until the vote is held, we will not know.
