Who is Thomas B. Edsall and what is the complaint.  Thomas B. Edsall is one of the best things that happens in a week of The New York Times.  He publishes a column on (most) Wednesdays.  The column is about, says the identifying blurb that comes with it, “politics, demographics, and inequality.”

He has a method for preparing his columns that I like very much.  First, he comes up with really good question.  Here’s an example: “Aggressive, insulting campaigning is said to be good for turning out your partisans, but does it turn out just as many people to vote against you?”  Is there, in other words a “net turnout gain?”

Take these two paragraphs, for instance, from the same column I am going to complain about.

“We identify three possible negative outcomes for democracy,” the political scientists Jennifer McCoy and Tahmina Rahman of Georgia State and Murat Somer of Koç University Istanbul, wrote in their 2018 paper, “Polarization and the Global Crisis of Democracy.”

The three negative outcomes, according to the authors, are gridlock; democratic erosion or collapse under new elites and dominant groups; and democratic erosion or collapse under old elites and dominant groups.

A really interesting question is posed.  A paper is cited.  It is, you will notice, hyperlinked and the whole article is available is you want to pursue it further.

Having a good question in hand, he writes a bunch of people who have studied the question he has asked.  He writes really good people; he writes people whose books and articles I know.  They write back to him and sometimes he writes back to them to clarify or challenge some point.

The result is a column with a lot of quotations in it.  Edsall provides the question and the narrative and his “guests” provide the experimental data that the question requires.  Edsall actually does more than that and I value the additional things he does—until this last week—but I am trying to present a general model of how he works.

He is in touch with a very considerable range of informed opinion.  He doesn’t “cover” the range from left to right because most of the questions he asks are not left/right questions, but he does use research that relies on different models of data aggregation and different kinds of experimental research and that helps me be confident in his conclusions.

However.

This last week, he cited a paper by Joshua Kalla and David Brookman, who had done an experimental study of “exclusionary attitudes”—prejudice against outgroups, for instance.  They describe a kind of conversation that can produce “durable reductions” in such attitudes.  That sounds to me like a good thing to pursue.

To this hopeful study, Edsall makes two objections: one serious, one snarky.  It will not surprise you that it was the snarky one I object to.  The serious objection was that to the extent the problem is misinformation, the increase in newspaper fact-checking should help.  I don’t think so.  I haven’t been able to access the whole article yet, but the overview appears to have very little interest in  “misinformation.”  And you would not expect it to, given that the focus, as given in the title, is on “exclusionary attitudes.”  So I think Edsall just missed that one.

But he follows that up by saying that to the extent the solution to this problem is what Kalla and Broockman say it is, i.e. the non-judgmental exchange of narratives in interpersonal conversation, it isn’t going to work.  And maybe if won’t.  But Edsall ridicules it by taking the language of the study—language borrowed from therapy—and using that language in a setting where it would obviously fail and where Kalla and Broockman did not propose that it be used.

It is hard, Edsall says:

to conceive of circumstances under which Chuck Schumer and Mitch McConnell, or Nancy Pelosi and Donald Trump, would “non-judgmentally” exchange “narratives in interpersonal conversations.”

Imagining such conversations between high-ranking politicians noted for their combative style is, indeed, a challenge and I am quite sure—I’ll tell you for sure when I locate the whole article—that Kalla and Broockman did not propose such a conversation.  I also think that Edsall knows they did not propose such a conversation and that he cited the language of the study not to criticize it, buy only to ridicule it.

That’s my complaint.

There is no need to ridicule this study, particularly if you have no better solution yourself.  Edsall does not have a better solution.  I am quite sure that if I asked him what solution he proposes, he would say that proposing solutions is not really his line of work.  I accept that as a valid excuse.  In his studies of “politics, demography, and inequality,” there is really no reason for him to be proposing solutions.  

On what basis would he propose them?  That they have been shown to work is small studies at universities?  That they worked when they were tried in South Africa?  That, as tenuous as the methodology seems, it is the best one proposed so far?  None of those really work for Edsall.  They are not his line of work.

So why should he go out of his way to ridicule a method when a)it’s not his job and b) he doesn’t have anything better?

That’s my complaint.Having said that—and hoping he doesn’t do it again—I want to remember that he is a wonderful source of good questions and solid research that bears on those questions.