A STATISTICAL MYSTERY FOR ECONOMISTS.

A STATISTICAL MYSTERY FOR ECONOMISTS. Here is a mystery for economists to solve: According to United States Census Bureau data, the marriage rate for women at age 64 is 58%. The marriage rate for women at 65 is 50%. What socio-economic facts explain this sharp drop? Actually, the mystery has been solved. As this article in the Wall Street Journal for February 9 recounts, economist Betsey Stevenson at the University of Pennsylvania’s Wharton School discovered the anomaly and has co-authored an article arguing that the privacy-protection techniques of the Census Bureau created the anomaly. (As the article explains, he Census Bureau makes changes in data—for example, moving an the data for an individual to a different part of the country—so that individuals can’t be identified.) A more general explanation in the article: “Flawed software programming appears to be at fault.” This Freakonomics post deals with the same issue, using the example that the Census Bureau data seem to show—contrary to fact—that there are more very old men in the country than very old women.