Both Noah Smith and Scott Sumner have drank the Kool-Aid on Abenomics, both crediting the monetary component -- you know, the quantitative easing that doesn't do anything to the price level. Here's Sumner:

Abenomics has raised the price level in Japan, reversing a secular decline.

Actually, it looks like a pretty strong claim if you look at the data in a model-independent way:

Right when Abe takes over (vertical line), there is a big reversal in the direction of inflation. But that's a model independent take. Like this (rather hilarious) model independent take on global warming. What happens if we apply some knowledge of how economies actually work?

I looked at the required path of NGDP in the information transfer model to reproduce a continued fall in CPI from 2013 onward

In order to produce continued deflation, if NGDP was the only effect, NGDP would have had to continue on its downward trajectory. Now lets look at the difference between this counterfactual and actual NGDP and compare it to government expenditures in Japan:

Actually the reverse seems to have begun in 2008. The financial crisis gave the Japanese government a reason to expand spending -- it had been flat since 2002.

A Keynesian increase in government spending could explain the entire effect -- no need for market expectations, monetary policy or even Abe taking office (he did vow to continue the increase in government spending, so fiscal Abenomics, aka Keynesian economics, ~~works~~ continued to work).

Commenter LAL asked (below) if counterfactual deflation was smaller, would it lead to similar results. It does -- there is a slight difference in that the same increase in spending leads to a smaller increase in NGDP:

The difference is visible at the end after the vertical line (that is where the less deflation counterfactual differs from the more deflation counterfactual in the post above).

__Update 3/31/2015__Commenter LAL asked (below) if counterfactual deflation was smaller, would it lead to similar results. It does -- there is a slight difference in that the same increase in spending leads to a smaller increase in NGDP:

The difference is visible at the end after the vertical line (that is where the less deflation counterfactual differs from the more deflation counterfactual in the post above).

sorry you've lost me... i think your definition of Keynesian should be a little more clearly spelled out... Keynesian usually implies a relationship between the sub-components of national income...and your analysis completely ignores notions of monetary offset etc (which is i think the whole debate for scott sumner)...finally how did you pick how much counter-factual sustained deflation you needed to get this result...could I have picked other paths of less severe deflation to get similar resutls?

ReplyDeleteon the surface this analysis begs the question...

Hi LAL,

DeleteGood questions! The counterfactual CPI(t) [call it CPI*(t)] was chosen to be a linear extrapolation of (LOESS smoothed) CPI from data up to 2012, producing the green dashed curve in the second graph.

I then used the model fit of the price level to the CPI from the first graph

P = P(NGDP, M0)

And then solved the equation

CPI*(t) = P(NGDP*(t), M0(t))

for NGDP*(t). I then compared the relative change in NGDP(t) - NGDP*(t) to the relative change in final expenditures data in FRED (the last graph). I was actually shocked at how well it fit the curve!

If you had less counterfactual deflation, it's true that the change in NGDP would then likely be much smaller -- essentially choosing the fiscal multiplier. But any amount of deflation would still have a positive fiscal multiplier.

In general I agree with the mechanism monetary offset, however one of the more heterodox bits of the information transfer model is the information transfer index written as "kappa" or k in various places on this blog. When k ~ 0.5, you get something that looks like a pure quantity theory of money (with perfect monetary offset). When k ~ 1.0, you get something that looks like a liquidity trap. I wrote a lot more about this here that you may find interesting:

http://informationtransfereconomics.blogspot.com/2014/06/krugman-keynes-and-liquidity-trap.html

I also agree that my use of the word "Keynesian" above probably isn't sufficiently nuanced -- I tend to use the word Keynesian to refer to theories where (see e.g. here**)

G → G + δG implies Y → Y + c δG

to first order with c ~ 1. Monetary offset is in the class of theories where

G → G + δG implies Y → Y

There is no reason to assume one or the other a priori and in fact the information transfer model has both happen at different times (different valued of k, ~ 1 in the first case and ~ 0.5 in the second).

** At the link, these are essentially two views of Y = C + G + I + NX: "first order Keynesian" or "useless accounting identity". In the information transfer model, both views can be accurate depending on whether inflation is low (former) or high (latter).

Your multiplier I take it is "c" in the "c δG"? If so then by positive multiplier you mean c>0 ?

DeleteI think I would want to observe c>1 for a Keynesian multiplier. To me the essential piece of Keynesianism is that under relevant economic conditions...consumption is linear in current income...so in a time series we have to see output gaining more than equal to the increase in government expenditure...but i guess that might be what you meant by first-order...i might be talking about second order...either way it should up in the time series

my only remaining point would be that it looks close to 1 in the graphs, but with less counter-factual deflation would be closer to zero; how close could it have gotten? that would really help with formulating priors/hypothesis tests or whatever else people do to come to conclusions...

Yes, c > 0 is what I'm taking to be positive multiplier ... it appears that c ~ 1 when you look at the comparison of the relative size (i.e. log linearized): a 1% rise in G is a 1% rise in NGDP, but as G/NGDP for Japan is ~ 20%, that implies a multiplier bigger than 1 in terms of Yen ... actually it seems a bit too large at c ≈ 5 = 1/(20%).

Deletewell, i dont think Krugman would think that it is too large..but most time series folk might

DeleteHa! That's true ... actually I made a mistake because I didn't take into account that the additional spending comes along with higher deficits -- that is to say the 'stimulus' comes from both increased spending levels and reduced revenues so that deficit spending increases more than just spending. That means the multiplier is probably closer to 1-2 based on the deficit increasing about 3-5% of GDP and the total effect being 5-10% of GDP.

Deletenice catch, that looks broadly consistent with multipliers i hear about

Delete