We news junkies seem to be vastly dependent on polling in this most fraught of elections. That’s a big mistake. As has been many times demonstrated, the polls often conflict and are thus wrong. The “why so?” is complicated, as Robert Kuttner explained in The American Prospect. He cites Michael Podhorzer,
who astutely points out that all polling is “opinion journalism.” Why? Because pollsters make assumptions about who is a likely voter and how to weigh or overweigh different demographic groups. “The ‘opinions’ are not about issues or ideology, but about methodological approaches.”
There is a long history of presidential polls being wrong, some of which is explained here. The pattern has remained unchanged for about a hundred years. The polls now predict no better than they did then. Even so, the polling practice has proliferated. It’s a business, after all, and following polls can be addictive.
Last month, Pew came out with a study, “Key things to know about U.S. election polling in 2024.” It’s a little more positive than I’ve suggested, maybe because Pew is a major pollster. A big problem, they say, is predicting who will actually vote.
Roughly a third of eligible Americans do not vote in presidential elections, despite the enormous attention paid to these contests. Determining who will abstain is difficult because people can’t perfectly predict their future behavior – and because many people feel social pressure to say they’ll vote even if it’s unlikely.
Nate Cohn in a recent NY Times post says, “The newer opt-in [online] pollsters haven’t fared any better,” and newer ones keep popping up. So Why are they doing no better than traditional polls? The problem is, as always, “how to find a representative sample without the benefit of random sampling, in which everyone has an equal chance of being selected for a poll.”
Instead, the internet has made things messier and more difficult. So many problems in verifying the data, and so few solutions. I found another fascinating study that illustrates a difficulty other than what the critics have been talking about. Axios summarized it this way: There are stark gaps between what Americans say they think and what they really think about hot-button political issues.
I think the findings from that new study are amazing. To wit, how the general public [61% of all Americans] misrepresents its views:
-
- In general, I trust the government to tell me the truth: public response, 22%; privately, 4%
- In general, I trust the media to tell me the truth: public response, 24%; privately, 7%
- We live in a mostly fair society: public response, 37%; privately, 7%
- The government should close the U.S.-Mexican border: public response, 52%; privately, 33%
- The government should restrict the expression of views deemed discriminatory or offensive: public response, 26%; privately, 5%.
You can check out more of these results here (scroll to Key Findings). If indeed valid, what these outcomes plainly mean is that nearly all public opinion polling sampling may be invalid. Can pollsters ever really discover how people are going to vote?