Relying on conventional wisdom or the so-called “appeal to authority” is like walking straight into a trap, especially when the authority in question isn’t exactly reliable. The narrative that the 2024 presidential race is a nail-biter comes from polls, and we all know how well that worked out in the past. Even pollsters themselves are waving red flags, urging people not to trust their numbers as gospel. John Anzalone, who has done polling for the likes of Joe Biden and Hillary Clinton, even admits there are “a lot of bad polls out there.” And to make matters worse, as Republican pollster Greg Strimple points out, many hardcore Trump supporters aren’t even bothering to respond to these polls—especially the online ones.
After the polling disaster of 2016, where the so-called experts completely missed Trump’s win, pollsters have been scrambling to figure out a better system. They’ve tried everything from landlines to cell phones, text messages to knocking on doors. Yet, it’s not just Trump voters avoiding them—young women, Latinos, and Black voters aren’t picking up the phone either. The result? A skewed mess that’s more guesswork than science.
What was once a respected statistical analysis has now become little more than a shot in the dark. Figuring out how to properly “weight” responses has become an exercise in wishful thinking. So when we hear about these polls showing a close race, one has to wonder—what if it’s not close at all? What if one candidate is far ahead and the polls just aren’t picking it up?
Conrad Black over at the New York Sun puts it bluntly. He points out that the Democrats’ massive wins in California and New York give them millions more votes than the GOP wins in places like Texas and Florida. Strip those out, and Trump is leading by about 8 million votes across the other 45 states. While swing states remain close, Trump’s support is consistently underestimated, and momentum seems to be on his side. Even commentators like Megyn Kelly have started to peel back the curtain, noting how Harris, the current Democratic candidate, is someone most Americans were embarrassed by when she was vice president just a few months ago.
State polls, especially in battleground areas, might offer a slightly clearer picture since the data set is smaller, but even these are far from perfect. Pew Research notes that even polls with a so-called margin of error under 3% give people a false sense of precision. That margin only accounts for one source of error and doesn’t consider other potential mistakes lurking behind the numbers.
Of course, this issue cuts both ways. The undercounting of young people, minorities, and other Democratic-aligned groups could mean that Harris is doing better than expected. Or maybe both Trump and Harris are being underestimated, leaving the whole polling process a jumbled wash.
Let’s not throw all polls out the window just yet, though. Some state polls are reasonable snapshots of where things stand at a particular moment. But when it comes to forecasts from places like FiveThirtyEight or RealClearPolitics? You might as well flip a coin.
