John Taylor has a longer post that can be explained by my title (see previous episodes here, here, here, and here in a variety of different contexts). If you believe fiscal policy works, you have a model that says fiscal policy works. When you estimate the effects of fiscal stimulus, you will say it works.

If you believe fiscal policy is ineffective, you have a model that finds it doesn't work, and so when you estimate the effects of fiscal stimulus, lo and behold, you show no effect and say it doesn't work.

Taylor says "the models have had their say. It is now time to look at the direct impacts using hard data and real life experiences." Yes. Robert Hall, for one (ungated copy? think so) suggests that while you could have a new Keynesian model which makes the purchases portion of the stimulus work, it was simply too small ($62.5 billion) to close the GDP gap ($1.2 trillion.) Paul Krugman would agree. But you have to assume a larger deficit would still have zero effect on real interest rates to argue for one big enough (in Hall's case, more than ten times the size for 2009.) I for one don't buy that premise.

Labels: economics