Monday, March 31, 2003

Moynihan and statistics 

Mitch Pearlstein from the Center of the American Experiment sent me a copy of his retrospective on the life of Daniel Patrick Moynihan, who died last week.
It goes without saying that Moynihan himself deserved far better, as the report that quickly came to bear his name didn't contain a racially suspect punctuation mark, much less a single sentence lacking in goodwill.

He was prescient, but he was pilloried. Or, more accurately, in order to forewarn what was sadly in store for our country, he did what good scholars and public servants are supposed to do: He bravely faced hard facts.

(Just two numbers: About two-thirds of all African-American boys and girls are now born out of wedlock. About one-third of all American children now come into this life out of marriage.)

But it was also our nation that deserved far better than the politically correct ambush, followed by the timorous silence, which succeeded the report.

In contrast, back here on the home front Miss Median struck again, this time with a reference to "overrepresentation" and "underrepresentation" in the military, citing this article. As any labor economist worth her salt will tell you, aggregate statistics will not tell you a damned thing about discrimination. And it absolutely ignores the simple fact that we have a volunteer military, so that what MM is criticizing is really the career choices that Latinos and blacks make. But citing aggregate numbers is all the diversity pushers have to offer, and when the numbers go against the story, as they did in Moynihan's work, they prefer to villify the researcher than confront the data.