Wednesday, May 1, 2024

The Conventional Wisdom Average

More than a week ago, I pointed out Sean Trende's blunder when he hopped on the bandwagon that "President Biden’s standing in the polls has improved". Another arbiter of the conventional wisdom, Byron York of the Washington Examiner, pointed out late yesterday,

Biden did rise a bit in the polls. According to the RealClearPolitics average of national polls, Trump has led Biden in head-to-head matchups since September 2023. (They were tied for one brief moment in October.) During his March-April “bump,” Biden’s deficit shrank, coming within two-tenths of a point of catching Trump. Democratic hopes rose.

“Biden’s position in the polls is improving against Trump,” declared the polling analyst Nate Silver on April 18. “It looks to me like Biden’s numbers against Trump have improved by a hair, probably to a degree that isn’t just statistical noise.” Even a slight improvement was like manna from heaven for Democrats and their allies in the media.

Alas, it didn’t last long. On Sunday came a poll from CNN that was simply devastating for the president’s reelection hopes. Remember, a poll is just a freeze frame of the race at this moment. It doesn’t mean things won’t change in the future. But it gives us an idea of where things stand right now. And the CNN poll showed Biden’s standing deteriorating before our eyes.

The same day, Sean Trende himself quietly backed off his earlier prognostication:

As a general rule, whenever people ask me about my thoughts on poll “x” or poll “y” my response is the same: Throw it in the poll average. Remember, not only are error margins real, but they’re intrinsic to sampling. An error margin of plus-or-minus three percent means that 95% of the time, the true population will be within three percentage points of the published “topline,” but that one time in twenty it won’t be. That’s not a pollster quality thing, that just comes with the territory of asking a fraction of the true population of interest.

So the fact that CNN’s latest poll has Trump up six points isn’t that interesting to me, in and of-itself. It’s marginally interesting that this is happening in the midst of the one criminal trial that is likely to occur before Election Day, but it’s more interesting that his standing in the poll averages has held with the trial under way.

But wait a moment. Trende himself in the piece I linked on April 19 relied on his own poll average to come up with just the conclusion he sorta-kinda now thinks was misleading:

As of today, Trump’s lead [in the RCP average] sits at just two-tenths of a percentage point.

And this was to substatiate his explicit view in the piece's title, "What To Make of Biden’s Rise in the Polls?" As of today, he's relying on his own poll average to say that well, after all, Trump has led in these since last year -- but then, of all things, he relies on the old chestnut,

Will this continue? Obviously it’s hard to say, as polls are just snapshots in time.

On April 22, still pondering that strange consensus that Trende had endorsed the previous week, I linked to this piece at The Hill, It’s time to retire the laziest cliché in election polling.

Opinion polling has no lazier cliché than “snapshot in time.”

The aphorism is intended to suggest impermanence — that polls taken weeks or months before an election have limited predictive value. The phrase has been repeatedly invoked as the 2024 presidential election race has unfolded. It will be heard many times before the campaign ends.

. . . [T]he phrase is a refuge or metaphoric shield for pollsters when their pre-election surveys misfire. In such cases, “snapshot in time” is cited in attempting to defend or rationalize polls that careen well off-target, as many of them did in the 2020 presidential election.

But it's worse for Trende, because in mid-April, he was claiming that not just an individual poll, but his own average of polls, was proving that Biden's lead was evaporating. But what's the cure? W Joseph Campbell, who denounced the "snapshot in time" cliche on which both Byron York and Sean Trende are falling back in the wake of new findings on the electorate's real-world views, offers no real remedy. He basically just says we should stop saying "snapshot in time":

[I]t may be time to retire the “snapshot in time” cliché, which polling analyst G. Elliott Morris has described as “more excuse [for pollsters] than anything else.” It’s been recognized as an irritating phrase, too: Several years ago, Politico ranked “snapshot in time” among the worst of America’s political clichés.

But doesn't this miss the point? What Prof Campbell is actually saying is that pollsters and prognosticators say a lot of dumb things that turn out not to be true. If we simply retire lame excuse A, "snapshot in time", with nothing else changing, all we'll get is lame excuse B, whatever that turns out to be, and we'll hear that one for decades until someone finally calls it out, and with no other changes, we'll just move to lame excuse C, having learned nothing.

I'm an Aristotelian, which means I look for causes. The situation where we're just permanently casting about for lame excuses for why pollsters and prognosticators are so consistently wrong says we aren't asking why this is the case. Let me propose an answer in the image of Sean Trende below:

This man is clueless, an empty suit, and a stuffed shirt. But let's see if we can tease out more about his background. Via his thumbnail at the American Enterprise Institute,

Mr. Trende has a law degree and a master’s degree in political science from Duke University and bachelor’s degree from Yale University. He is currently pursuing a PhD in political science at The Ohio State University.

Given the current state of unrest at many of the elite schools where people like Trende get their qualifications, there's been an increased tendency to question the value of the degrees they award. But this is nothing new. Looking back, I think one of the formative events in my life was arriving at an elite-school campus to begin my undergraduate studies and realizing, against everything I'd been led to believe by my parents, my friends' parents, my school counselors, the elite-school admissions staff, and all respectiable opinion, that my new schoolmates were anything but the crème de la crème.

If you got right down to it, every one of them was the soul brother or sister of the bizarrely clueless Sean Trende in the photo here, clueless empty suits and stuffed shirts in training. Within a couple of weeks, I made an appointment with the dean of freshmen to tell him I couldn't quite figure out what it was, but there had been a terrible mistake. I was in the wrong place. The dean calmed me down and convinced me to stick it out, but decades later, I basically think I was right.

Somehow our educational system is seeking out and rewarding the people who beome each generation's Sean Trendes.

0 Comments:

Post a Comment

Subscribe to Post Comments [Atom]

<< Home