Thursday, June 6, 2019

MMT, With Japan in the Middle


In the squabbling on the internet over Modern Monetary theory, Japan has emerged, at least in some people’s minds, as the prime test case.  Considering the Japanese economic experience through the 80s into the lost decade of the 90’s, and on into its own version of Quantitative Easing, and its inability to stimulate much in the way of increased output or much in the way of inflation, it’s a country which seems well suited to providing data for testing conventional monetarism, MMT, and probably also the Fiscal Theory of the Price Level.  That would, of course, be a lot easier if there were a well defined body of MMT theory which could be used to develop formal statistical tests.

As it is, we’re forced to rely on websites and newspaper reports: see here, as well as here and here for example.  The New York Times has a longish piece on the Japan question here which we’ll deal with in a later post – I still have some posts on MMT itself to go up.  Be warned that the NYT piece is really just a list of who says Japan is following MMT and who says it isn’t – the most interesting thing in many of the pieces I’ve linked to is when MMTers are cited as saying that no, Japan isn’t implementing MMT.


What I want to look at in this post is an anti-MMT WallStreet Journal piece (behind a subscription wall) by John Greenwood and Steve H. Hanke.  Hanke is at Johns Hopkins University, and he’s one of those guys whose writings on monetary matters are worth taking seriously, even if you wind up disagreeing with him.  In this one Hanke and Greenwood write:
“‘Modern monetary theory” is the latest craze to circulate among the chattering classes. It is a fuzzy, post-Keynesian theory that has caught on in antiausterity circles. The MMT doctrine states that fiscal deficits don’t matter as long as countries borrow in their own currencies and inflation stays in check. This is like manna from heaven for proponents of more government spending and larger fiscal deficits.
 Not surprisingly, MMT has reared its head in Japan, where debt is denominated in yen, inflation is nowhere to be found, and the ratio of government debt to gross domestic product has gone to the moon. MMT advocates claim Japan provides proof for their theory, but nothing could be further from the truth.”
Actually, as I noted above, there are MMT theorists who deny that Japan is in their camp, but we’ll leave that aside for this post.  What I want to look at is a bit where Hanke is writing about conventional monetarist theory and where I’d argue that, unusually, he’s misfired completely.
His basic theoretical tool is the Quantity Theory equation, MV = PT, where M is the money supply, V is the velocity of money, P is the price level and T is a measure of real economic activity.  Taking logs and differentiating, to get percentage changes, we have %ΔM + %ΔV = %ΔP + %ΔT.  Then if we hold velocity constant, so that %ΔV = 0 we get %ΔP = %ΔM - %ΔT.  It’s from this expression that we reach the conclusion that the rate of inflation (the percentage rate of growth of the price level) is equal to the excess of the rate of growth of the money supply minus the rate of growth of real economic activity.

Now consider the following excerpt from Hanke and Greenwood’s WSJ piece:

“Separate from Japan’s overall savings surpluses, its broad-money metrics have grown at a snail’s pace. Since Japan’s bubble burst in 1990-91, broad money has grown at a paltry 2.6% a year, as measured by M2. In addition, since the 1950s money velocity has been negative, decreasing at an average rate of close to 2% a year. This is one of the most striking and consistent macroeconomic relationships on record. Japan’s slow broad-money growth mixed with a contracting money velocity has held down nominal GDP growth.”

Obviously, in the second sentence, he means that since the 1950s the rate of growth of velocity has been negative.  God only knows what a negative Velocity would mean.  In the good old days the WSJ copy-editors would never have missed that one.  That aside, Hanke goes on to say:

“Faced with a contracting money velocity, Japan would have to increase its M2 money supply at a minimum of 5% a year, which is double its trend rate, to hit its inflation target of 2% a year and reach its potential annual growth rate of 1%. In practice, inadequate M2 growth over nearly three decades, combined with a declining money velocity, has translated into average real growth of just 0.9% a year. This has put Japan into a deflationary straitjacket, with prices decreasing by an average of 0.6% a year, as measured by the GDP deflator.”

But think about this.  Hanke’s using the growth rate version of the Quantity Theory equation and saying that the reason Japan hasn’t seen inflation was because %ΔV has been negative.  The trouble with this is that it treats V as a thing with an existence of its own.  But from the Quantity Theory equation, V = PT/M.  Note that in the quantity theory interpretation of inflation which we gave above, we assumed that V was constant – i.e that %ΔV = 0.  But since V is really just PT/M, when we say that the rate of growth of velocity is negative are we really saying anything more than that the value of nominal output (PT) is rising more slowly than the value of the money supply (M)? IS there actually any information content in V, or is saying that velocity is dropping just another way of saying that money is being pumped into the system faster than nominal spending is growing?  We could undoubtedly get a very good statistical fit out of this relation, but does it actually tell us anything useful?

If you read the macro literature which came out before Keynes published the General Theory, one thing which comes through is the way Velocity played into much more than just the quantity theory equation of exchange.  In particular, business cycle downturns were often explained as being caused by a drop in the velocity of money.  One thing which Keynes (who in his 1923 Tract on Monetary Reform had been a convinced quantity theorist – that’s part of the reason why Milton Friedman regarded the Tract as Keynes best work) tried to do in his 1930 Treatise On Money  Vol. 1  and Vol. 2 was to get away from some of the logical problems with the concept of velocity.  He was only partly successful at best, and had dropped the concept by the time he wrote the General Theory.  

We generally accept that institutional change can increase the velocity of money – perhaps the most obvious case was the invention of fractional reserve banking by 17th Century Scottish Goldsmiths, although there was also the emergence of cheques as the primary means of payment in Britain during the Napoleonic Wars, but in that latter case the increased velocity argument only holds if you define money to be notes and coin and not to include chequing deposits.  And most of the institutional change examples which we invoke are cases where velocity increased, not cases where it decreased.

I’m inclined to conclude, then, that to argue that the Japanese case fits classical monetary theory so long as we assume that some factor is causing velocity to decline exogenously, and that the decline is not just as a result of the definition of  V = PT/M, simply isn’t convincing.  

No comments:

Post a Comment

Note: Only a member of this blog may post a comment.