Sunday, March 16, 2014


The government runs on predictions for everything from next weeks demand for frijoles to next century's average global temperature. Most of us have a pretty good idea about just how dependable these forecasts are as witness the use and overuse of the word "unexpected" when perusing reports of last weeks economic activity or yesterdays precipitation.

The Weekly Standard has a piece discussing the impact of uncertainty on the field of economics and includes a story at the bottom on the denouement of weather forecasting when exposed to statistical analysis:
In an autobiographical essay published 20 years ago, the left-leaning economist Kenneth Arrow recalled entering the Army as a statistician and weather specialist during World War II. “Some of my colleagues had the responsibility of preparing long-range weather forecasts, i.e., for the following month,” Arrow wrote. “The statisticians among us subjected these forecasts to verification and found they differed in no way from chance.”
Alarmed, Arrow and his colleagues tried to bring this important discovery to the attention of the commanding officer. At last the word came down from a high-ranking aide.
“The Commanding General is well aware that the forecasts are no good,” the aide said haughtily. “However, he needs them for planning purposes.”
 Today we predict the weather not just one month in advance, but 10, 25, 50, or 100 years into the future with the same certainty of the weathermen of WWII. All agree there is nothing in next century's weather that can't be properly fixed with a big tax increase and an expansion of government regulating authority.

To paraphrase: “The Committee chairmen are well aware that the forecasts are no good,” the aide said haughtily. “However, they need them for planning and budgetary purposes.”

No comments: