Commentary Magazine


Contentions

Lies, Damned Lies, and the New York Times

As Disraeli famously observed, statistics are a source of untruth so rich that they deserve to be a special category of mendacity all by themselves.

If you would like a classic example of this, I recommend Iowahawk’s utter evisceration of Paul Krugman’s February 27 column in the New York Times. Krugman had argued that low-tax, anti-union Texas was condemning its children to a life of hamburger-flipping because its low-spending ways were failing to educate them. To add artistic verisimilitude to this narrative, Krugman cites the fact that Texas ranks 43rd in state rankings of educational achievement.

As Iowahawk points out, those rankings are completely meaningless unless you correct for each state’s ethnic makeup. Wisconsin is 4 percent black and 4 percent Hispanic. Texas is 12 percent black and 30 percent Hispanic. If you look how each state does within each ethnic group, guess what: Texas beats Wisconsin like a rented mule. The National Assessment of Educational Progress tests kids around the country in the fourth and eighth grades for progress in reading, math, and science. Of the 18 ethnically controlled comparisons, Texas beats Wisconsin on 17 of them and is above the national average on all 18 (Wisconsin is 8-8, with two tying the average).

Iowahawk’s critique is not only dead-on but also laugh-out-loud funny, as he usually is.

Of course, Disraeli lived before the era of public polling, which today generates politically motivated statistics the way machine guns generate bullets. I recently discussed the New York Times poll supposedly showing overwhelming support for the union position in Wisconsin. It did so by the simple expedient of oversampling households likely to support that position.

But the art of phrasing the questions in ways that will produce desired results is just as good a way to cook the statistical books and long ago reached a high degree of development and refinement. The Times poll asked respondents: “Some states are trying to take away some collective bargaining rights of public employee unions. Do you favor or oppose taking away some collective bargaining rights of the unions?” But “rights” is a powerfully loaded words in American politics. And as the Heritage Foundation points out, collective bargaining for public employees is not a right; it’s a privilege. Far from being constitutional, it is often not even statutory. In Indiana, for instance, collective bargaining for government workers was granted by mere executive order, so Governor Mitch Daniels was able to revoke it by executive order, which he did on his first day in office. Further, the pollster Frank Luntz said on Fox and Friends yesterday morning that the phrase “public employees” polls much better than “government employees.”

So had the Times phrased the question as, “Do you favor or oppose taking away some aspects of collective bargaining for government workers?” it would probably have gotten a completely different set of numbers for and against. Perhaps that was just sloppy polling technique on the part of the Times; perhaps it was a carefully crafted question. In either case, the result was a junk statistic.