A Modest Proposal
by JS
From this piece at TPM:
Next, since there are no controlled experiments in recessions and depressions, there’s no really concrete and dispositive evidence about what policies end or don’t end severe economic downturns.
Why don’t we change this? It seems like the easiest way to run a controlled experiment is to shape the stimulus by state. Certain states get little or nothing. Other states get massive stimulus packages. We then observe state level economic changes over the next couple of years. We wouldn’t be able to control for all the factors, but we’d be able to control for some of them. Moreover, this would be politically easy to do, just make sure a majority of states get stimulus and watch as representatives scramble for a piece of the pie.
The models would have to account for interstate commerce. Certainly as surrounding states recover, the increased economic activity will pull up the states that do not receive a stimulus. I would expect, though, that the data should show a “lag” for states that don’t receive stimulus, and a significant effect once estimates of interstate commerce are factored into the model.
