I was going to swear off Casey Mulligan for the rest of the summer, but it can be hard to stick to a diet. He has a delicious new post and I just have to take a couple of swipes.
First, he criticizes me and Paul Krugman because we have praised studies of the impact of the stimulus that do not take account of displacement effects. Mulligan's example is the move of Yankee Stadium. He notes that the move created jobs in the new location but correctly points out that these was pretty much entirely offset by the jobs lost in the previous location.
This is a fair enough critique in the context of Yankee Stadium. In fact many claims of job creation by businesses (e.g. Boeing creating jobs in its proposed plant in South Carolina) are in fact stories of displacement. In the Boeing case, the jobs would be coming from unionized facilities in Washington State.
But how do we envision the job displacement in the stimulus story? If Illinois has more jobs in road construction due to the stimulus or more teachers in its classrooms, do we think this means fewer construction workers or teachers have jobs in Indiana or Wisconsin? It's a bit hard to see how that works. In the Yankee Stadium and Boeing example, we are taking something away and moving it to a new location. What exactly is being taken away in the stimulus story?
The other point Mulligan raises that deserves comment is his argument that more workers are employed in the summer because of the increased supply of young people looking for jobs. Mulligan sees this as evidence that the underlying problem is supply and not demand, since the additional supply of young workers seeking jobs led to more employment.
Of course there is a flip side to this picture that Mulligan has ignored. There is also increased demand in the summer. This is when families take vacations and go to beaches and resorts. That is both because the weather is better (not many lifeguards are employed on Coney Island in January) and because the kids are out of school. In other words, spending on tourism and recreation is highly seasonal.
We could use Mulligan's seasonal argument as a test of whether unemployment was driven by supply or demand factors if we could just isolate the supply side of the picture. But there is no easy way to do that and Mulligan certainly has not found one.