Annals of derp: Chaotic sea of short-term trivialities

Arthur Brooks for the American Enterprise Institute sends me a note every once in a while, which you can't read if you're not on his exclusive mailing list, inviting me to come over—

If it’s been a while since you browsed our home page[paracom.paramountcommunication.com], take a look[paracom.paramountcommunication.com]. You’ll find an island of thoughtful, substantive takes on a host of issues that stands apart from the chaotic sea of short-term trivialities.

I'm not sure why he thinks he needs to put the name of his email marketing service provider in the links, or maybe he had the marketing service provider put in the links for him and they were anxious to get credit for their work, but there it is.

Anyway, for those of us who are actually in it for the short-term trivialities, he appends a "newsletter", and as you'd expect it's all about the Trump, and some ways of looking at the news and feeling a little less panicky, for those who happen to be Republicans.

First, if it makes you anxious to hear that the polls are all predicting outsize victories for Trump, you'll be glad to know that the polls are all wrong:

The question seems to be everywhere. In January, for example, The Boston Globe published a piece asking “What if the polls are all wrong?,” and The New York Times offered a take entitled “Why Polls Have Been Wrong Recently.” Apparently those fears were realized, because in February, The Washington Post offered a postmortem on the first caucuses: “Why were the Iowa polls so wrong?”

The first answer is—nothing new. The survey business has been out of whack for at least a few years now.

Unfortunately, Trump keeps winning in the Republican primaries, even though the polls saying he will win are evidently wrong. What's up with that?

Apparently it's that the polls themselves influence the voters:

But in a bizarre twist, just as the polls are historically unreliable, they also seem to have historic influence. One leading GOP candidate devotes more of his speeches to reporting poll results than talking substance. Sanders has reportedly done the same in several speeches, reading poll results from the podium.* These survey data, nothing more than imperfect second-order observations on the state of the race, are themselves taking center stage and inflecting the direction of the campaign.

In this way, even though the polls are wrong in predicting that that nameless leading GOP candidate will win, they appear to be not wrong, in that he wins.

Still more alarming is a third oddity:

Whatever you or I may think of the candidate, it is undeniable that Donald Trump has repeatedly shattered pundits’ predictions of his staying power and apparent appeal.

Just as the polls are wrong, it seems, the pandits are right. And yet even the rightness of the pandits is not enough to make their predictions come true. Luckily, Arthur Brooks is an economist and can help us with this issue:

In response, I’m experimenting with a kind of equation. (I am an economist, after all.) I tried this one out in one of my New York Times conversations with Gail Collins[paracom.paramountcommunication.com]:

A plurality of voters +

A fragmented opposition +

A massive majority of media attention

=  Victory (so far).

I think this means there are three main components to Trump's remarkable staying power:

• First, he has a plurality of the Republican voters, meaning that although he hasn't got a majority none of the other candidates have as many votes as he does. 
• Second, the field is fragmented into a group of other candidates who have fewer votes than he does, so that he gets a plurality.

You might think that these two facts taken together would account pretty well for the phenomenon, at least in arithmetical terms, or even that both of them are in fact the same fact, but that's obviously because you're not an economist. There's more!

• Third, the media are paying Trump incredible amounts of attention. For instance in this very newsletter from the American Enterprise Institute, of the three items it brings to its select readers' attention, altogether three are devoted to the Trump threat.

I think he's ignoring one last thing, which is that just as the polls, though wrong, influence the situation to make themselves look right, so the pandits, though right, might have an influence of their own, which causes the facts to flee, as it were, in the opposite direction, as if deliberately to make them look wrong. This would certainly explain Dr. William Kristol (who is not an economist).

Anyway, all these facts demonstrate clearly that although Trump cannot possibly win, he probably will, but that's no reason to despair:

Its tough to see how this self-serving negative feedback loop ends, but I am confident we’ll figure a way out. Perhaps after the carnage of this election, journalists will see the error of their ways. Or maybe the viewing public will vote with their wallets as people keep cutting the cord and forgoing cable TV altogether.

We'll be able to defeat Trump after he gets elected, once we've understood what we were doing wrong. Or at least stop watching cable, which amounts to the same thing. Doesn't it?

*Sanders, it should be noted in fairness, has good reason to read poll results showing his very high numbers from the podium, since the belief that he won't win is the most important reason, for many voters, for not voting for him. This is not the case for Trump, who simply doesn't like talking about anything else.