Scott Sumner talks about his own work on monetary policy and a recent paper by Joshua Hausman (linked at the link) on the fiscal policy effects of a 1936 WWI veterans bonus payout in the context of the Great Depression. I've used the IT model to look at the Depression (here) as well as the impacts of fiscal stimulus (here)
So here's the 1936 bonus impact on GNP (I use GNP because data goes back far enough for this purpose); I show the counterfactual (i.e. no bonus) as a dashed line:
And here are the results for YoY CPI inflation and YoY RGDP growth using the monetary information equilibrium model with changing information transfer index described in the paper:
These are essentially the same results as here. I would like to pause to note: the IT model fit to RGDP growth is, in a word, choice. The model counterfactuals (i.e. without the 1936 bonus payout) are shown as dashed blue lines in those graphs.
We are able to calculate the impact on RGDP as well (using the same method as here):
Once again, the impact is comparable than the impact calculated by others, including Hausman (in gray, see his paper for other estimates). The stepped line is just an annual integral.
The effect is relatively small compared to the size of the other shocks, but positive.
PS In looking at the graph of nominal shocks, I noticed (although it could well be pareidolia) the tell-tale signature of of a pure entropy shock (i.e. non-ideal information transfer). One way to look at this is as the impact of a "hot potato effect" where high powered money is distributed to a few agents that equalizes over all agents through random market transactions. Here was an accompanying graphic:
Another way to achieve this same effect is through a "traffic jam on the Wicksellian roundabout" per these posts (here, here) on Nick Rowe's model. Anyway, I added the generic functional form to guide the eye. This post on how to think about the business cycle in the information transfer framework is also relevant.