Must-read: Justin Wolfers: On Twitter
Must-Read: I have only one critique of Justin Wolfers’s series of tweets here. He says that they are simple math errors. It is true that they are conceptually simple. But getting all this right is not easy–I have seen Alan Blinder get it wrong (in the context of a $12B surge in inventories in a single quarter, and its effect on real GDP growth rates) at a blackboard in his OEOB office during the Clinton administration[1].
This does explain a puzzle. As somebody-or-other said in the conversation, if you believe with Gerry Friedman that all of the shortfall in real spending growth since 2007 can be easily recaptured via demand channels alone, then Sanders’s proposals are at most one-third the size that they should be–and that is the critique that Friedman should be making of Sanders given Friedman’s beliefs about aggregate supply. (Since I do not share those beliefs, I think Sanders’s fiscal stimulus proposals are about right-sized):
: On Twitter: “Romer and Romer find what look like…
…simple math errors–confounding level effects and changes–in Gerry Friedman’s analysis: https://evaluationoffriedman.files.wordpress.com/2016/02/romer-and-romer-evaluation-of-friedman1.pdf
Here’s the problem: Fiscal stimulus raises spending: ↑ΔG today → ↑ΔY. But then [once the government purchase] stimulus ends: That’s a big ↓ΔG tomorrow → ↓ΔY. Long run effect ≈ 0. Or if redistribution leads to more spending: Big ↑Δspending today. But then no further Δspending. Long-run raises level of GDP but not growth…. (Their economics checks out, obviously):
: Senator Sanders’s Proposed Policies and Economic Growth: “We have a conjecture about how Friedman may have incorrectly found such large effects…
…Suppose one is considering a permanent increase in government spending of 1% of GDP, and suppose one assumes that government spending raises output one-for-one. Then one might be tempted to think that the program would raise output growth each year by a percentage point, and so raise the level of output after a decade by about 10%. In fact, however, in this scenario there is no additional stimulus after the first year. As a result, each year the spending would raise the level of output by 1% relative to what it would have been otherwise, and so the impact on the level of output after a decade would be only 1%.
[1] Not, mind you, that I got it right back then–I had to go back to my Treasury office, think about it, and then do it three times. $12B was only 0.2% of what was then a nominal GDP level of $6000/year. But that $12B was four times as great a share of that quarter’s GDP: 0.8%. And if GDP jumps by 0.8% from one quarter to the next, that is a +3.2%/year jump in the growth rate of GDP. There is thus an amplification by a factor of 16 as we go from the raw change in inventory stocks in a quarter to the real GDP growth rate. But at the blackboard we were all saying it was either a factor of one or a factor of four…