Appearing too often to be simple quirks or anomalies—especially most recently—election results seem no longer to slavishly follow scientific polling.

Let us consider what might be going on. First, assume that the voters are honestly answering the pollsters. 

Also, after so many years of accurate polling, let us give the scientific pollsters the benefit of the doubt and suggest their work is generally reliable and technically spot on.

A convincing resolution of this recent apparent polling-voting disparity might lie in the notion that the polls are actually measuring something other than what they claim to be. 

Suppose when an American voter is asked if they approve of a politician, they take it as a signal to vent their general dissatisfaction with government itself. The logic goes something like this: I am not very happy, life could be so much better, my upset is the fault of the person or people in charge. Thus, voters might declare their disapproval of a politician or even a party, but subsequently, despite what amounts to their complaining, vote in favor of that party. 

However precisely pollsters believe they are measuring voter sentiment, if the voter’s interpretation of the question goes beyond its mere objective language, then pollsters are not actually measuring what they think they are. 

Since 2003, sitting presidents have suffered the indignity of an underwater (below 50 percent) electorate approval rating 77 percent of their time in office (David Brooks, New York Times, November 10, 2023). Perhaps, in these less government-friendly times, a rejection of current leaders can be little more than a popularly professed anti-government sentiment which may not be quite the indicator it once was of what a voter’s ultimate electoral intent might be. 

A little humility, please

Perhaps more importantly, if pundits continue to take respondents’ disapproval responses at face value, then the confounded meanings that those responses imply will continue to be echoed and amplified by the pundits themselves. Pollsters thus unwittingly and unknowingly could be helping to cement a false narrative.  

Otherwise, explain to me why—in contradiction to most of the immediate pre-election polling—the Democrats lost the 2016 presidential race. Or perhaps you would rather prefer to clarify the nonappearance of the highly anticipated and solidly predicted Republican “red wave” of the 2022 midterms.  

Most particularly in politics, where heads can get very hot, however precisely and scientifically one measures the temperature of someone’s hat, it will not reveal how truly hot their head may be. 

John Pace is a retired math professor. He lives in Honesdale, PA.