Bring Back the Bean-Counters

Felix Salmon has a good piece up about economists wrong-footed by the data, and by the general decline in the utility of "economic numbers" like, e.g., the Gross Domestic Product. He offers a good menu of possible reasons:

the fragmentation of the economy, with a large rise in the number of small companies and self-employed individuals; the globalization of the economy, with its attendant blurring of the idea of what counts as "domestic production"; the fact that first-rank economists simply don't become statisticians any more; and a general fiscal neglect of the statisticians that the government does have.

All true, and here's another: there are just too damn many of them. We've learned how to measure everything (at least badly) and so it is almost always possible to pull out a plausible counter-datum to defang any particular data point. Makes me pine for the day when the bean counters really counted.