Wednesday, 11 August 2010
Could we monetise the entire national debt?
There has been a debate between Paul Krugman and James Galbraith on this issue recently.
Krugman accepts the need for a continued deficit in the U.S., and is happy with a certain amount of monetising, but doesn’t think the two latter can be taken as far as Galbraith advocates without inflationary dangers.
I accept that Galbraith and advocates of Modern Monetary Theory (MMT) are insufficiently concerned about inflation. In fact I think most MMTers have a thoroughly crude understanding of labour markets: they have no inkling of the way in which tight labour markets contribute to inflation.
But the basic idea behind MMT is good: that’s the idea that governments should abandon borrowing and (when stimulus is required) just run deficits funded by extra money or “printed money”.
I won’t deal with the above labour market point here, but I will deal with the Krugman v Galbraith debate.
Krugman introduces some very silly maths in his argument. He says “What determines the price level? Let’s assume a simple quantity theory, with the price level proportional to the money supply: P(t) = V*M(t)”
He then says “P is proportional to M”. He is simply saying that if the money supply doubles, then prices will double. Well, introduce that assumption at the start of an argument (or some maths) and lo and behold, the conclusion will be that money supply increases have a substantial effect on prices!
The above crude assumption doesn’t capture a subtlety pointed out by MMT advocates (and Keynes) which is that if there is an inadequate supply of money in private sector hands, the private sector will save money. It will deleverage. Or put it yet a third way, “paradox of thrift” unemployment occurs.
In the above scenario, a modest money supply increase should have virtually NO INFLATIONARY EFFECT.
Another subtlety not captured is the fact that money is very similar in nature to national debt. That is, there is no sharp diving line between money and non money, and national debt is one of the many “pieces of paper” which can, on a broad definition of the word money, be counted as money.
Quantitative easing (U.K. style) involves printing money and buying back some of this debt. There is little reason to suppose much of an inflationary effect (or any other effect) because cash and national debt are so similar in nature. Mish rightly refers to QE as “quantitative nothingness”.
To summarise so far, there are no theoretical reasons for thinking QE will have a HUGE effect. And this theory seems to have been born out by the fact that we have had unprecedented increases in the monetary base in 2009 thanks to QE, but rampant inflation has not ensued.
But obviously QE has SOME effect because cash is clearly more liquid than national debt.
But does that mean we cannot monetise the entire national debt? Not at all! QE alone is mildly stimulatory and possibly inflationary. Thus if we want to monetise the entire national debt, all we do is monetise the entire lot and then take suitable deflationary counter measures.
If the SOLE objective is to monetise the debt (i.e. do nothing else) the deflationary measure needs to be one that monetises the debt. Um, er, what’s that? Easy: collect more tax (and/or reduce government spending) and use the money collected (or saved) to buy back debt. (See Musgrave’s law).
One moral of all this is: beware of economists and others who are contracted to (or expected to) write a number of articles a week for some newspaper, whether it is Krugman and the New York Times or anyone else. We all have days (or weeks on end) when we cannot think of any stimulating ideas. In this situation quality bloggers normally just remain silent. In contrast, the above journalists just spew out waffle.