February's Snowstorms and U.S. Economic Data

The weather in October 2009 (the third-coolest October in the 115 years of records, and the wettest) and December 2009 (14th-coolest December, 11th wettest) ranks as more extreme than in February 2010 (29th coldest, 70th wettest). What makes February the outlier were the snowstorms. According to the National Climatic Data Center, the U.S. agency that monitors weather:

"Major snowstorms plagued the Atlantic states throughout the month, most notably back-to-back storms on Feb 4th-7th and Feb 9th-11th. Separately, these storms ranked as Category Three and Two storms on the Northeast Snow Impacts Scale (NESIS), respectively. However, if combined and treated as one storm, the resulting combined total would become only the third Category Five storm (the most extreme category) of the NESIS record. A third storm, also ranking as a category Three on the NESIS scale, occurred across southern New England on Feb 23rd-28th. February 2010 is the first month during the NESIS period of record (since 1956) to place three storms of Category Two or greater."

The storms were especially disruptive because they hit big cities not accustomed to major snowstorms. Over 46 inches fell on Washington D.C., during February, breaking a previous record of 34.9 inches, and shutting down most federal government offices for four-and-a-half days. New York City was hit with 10 inches on February 10 and another 21 inches on February 25-26.

Seasonal adjustment will not correct for February's data distortions, because this procedure irons out spikes attributable to normal weather.

Some economists may work around the data problems by averaging February and March numbers, pretending that the month was a long one. Another shortcut, where regional breakdowns are available, is to focus on a region, such as the West, that experienced normal weather.

The housing and construction data will be the hardest to read. Housing starts will take a hit in February, but then bounce back in March. Home sales will take a hit in March and possibly April, and then bounce back smartly—but by how much, is anyone's guess. Auto sales also suffered in February, so they should get a payback boost in March. The employment numbers clearly took a hit, although it is difficult to pinpoint a number. According to Nigel Gault, IHS Global Insight chief U.S. economist:

"More than 1-million people reported that they were employed but not at work because of bad weather, 740,000 more than the usual figure for February. But that did not seem to translate into a big hit to payrolls—bear in mind that workers are only removed from the payroll employment count if they received no pay at all for the survey period, which was the pay period including February 12. Overall, the storms may only have cut 50,000 or so from payrolls, rather than the feared 100,000–150,000. But all such estimates remain educated guesswork."

Snowstorms, like major hurricanes, are not disruptive enough to dent the GDP numbers. GDP measures production, not destruction. While storms destroy work, they also create work. Indeed, most work lost during the storm is made up later.

The snow has melted, and it is up to the economic analysts to make sense of the mucky data left in the storms' wake. Unfortunately, we may have to wait until the second half of April, when the data for March flow in, to get a good reading on recent economic activity.