The other week, Peter Boettke was pretty impressed with Casey Mulligan. I hope he reads F&OST before he is impressed with this morning's post from Mulligan. This time, Mulligan is once again making the claim that summer labor supply trends raise doubts about demand concerns. Lots of people work in the summer - labor supply increases in the summer - ergo, labor demand is apparently not a problem.
People - this is why we seasonally adjust data. Of course labor markets behave differently over the course of the year. Employers know there are labor supply increases during the summer, so they shift activities accordingly. The correct comparison is not Summer 2011 to Spring 2011 - it's Summer 2011 to Summer 2007. And the last time Mulligan wrote about summer jobs, he made this comparison across years, and you know what? It provided evidence for the Keynesian story. The economy has regular seasonal variations, just as it has regular regional variations and regular variations across skill levels. Noting those variations doesn't tell you anything about how the macroeconomy is doing.
Take a look at this labor market:
See - I can make wages go down and employment go up in the summer too. Does anyone think we should second guess Keynesian demand problems in this labor market? Does anyone agree with Mulligan that in this labor market demand policies are the "wrong advice"?
Seasonal cycles happen. This is why we seasonally adjust data. Nothing in this cycle - nothing at all - changes the Keynesian story. It's a data artifact and it's amazing to me that so much has been made of it by Mulligan.
UPDATE: Perhaps this is another way to put it: if you blur the supply lines in the labor market above together, all you've done is seasonally adjust the data. Statistical agencies produce that information all the time because people find that information useful. What happens if you blur the demand lines? You've just eliminated the business cycle. The empirical content of the supply trends that Mulligan is citing here is considerably lower than the empirical content of the demand trends.