Wisconsin recall voters "shouldn’t be basing their decisions on recent job growth, which owes far more to individual entrepreneurs and chance than to anything the governor has done."

"They should be making their choices based on the things the governor does control, not on short term economic data, which are neither perfectly measured nor under the governor’s control."

Says Harvard economics prof Edward Glaeser, who looks at the controversy between the Bureau of Labor Statistics numbers and the Wisconsin Department of Workforce Development numbers, which are being used, respectively, by Tom Barrett and Scott Walker.

One way to judge between the numbers is to look at alternative data sources and see whether they tend towards gloom or growth. The BLS itself provides an alternative estimate of job growth from its household survey, which is generally less accurate than the employer survey and suggests strong job growth between 2011 and 2012.  According to the household survey, Wisconsin’s unemployment rate has dropped significantly over the last year, and employment has increased by 22,000 between March 2011 and 2012, again not seasonally adjusted, which is closer to the figure Walker is promoting than the BLS’s own establishment survey.



The BLS Mass Layoff database also shows that between the first four months of 2011 and the first four months of 2012 the number of mass layoffs dropped by about 30 percent and the number of new unemployment insurance claims from those layoffs fell by 36 percent. Wisconsin’s total personal income grew by 4.5 percent nominally, or about 1 percent in real terms, between the fourth quarter of 2010 and the fourth quarter of 2011. Wisconsin’s rental vacancy rate was down to 6.6 percent in 2011, down from 8.6 percent in 2010.

My best guess is that the truth lies between the two figures, but closer to the upbeat Wisconsin state numbers than the BLS estimate.