David Hume developed most of monetary economics back around 1750, partly by just sitting in a room and thinking. He discussed a thought experiment where everyone in England woke up one morning and discovered their purses contained twice as many gold coins as the night before. Hume argued that England wouldn’t be any richer in real terms; rather the price level would double. But he also noted that there would be a transition period where the extra money would be spent and trade would expand. So he understood the Quantity Theory of Money (QTM) and the Phillips curve. In the 1970s Friedman argued that in the past 200 years macroeconomics had merely gone “one derivative beyond Hume.” I.e., Hume looked at changes in the price level (as did Irving Fisher) and Friedman looked at changes in the rate of inflation.
Hume also understood the impact of an increase in money demand (or reduction in velocity):
“If the coin be locked up in chests, it is the same thing with regard to prices, as if it were annihilated.” David Hume — Of Money
Then people began to notice that when the supply of money increased there was a temporary decline in interest rates. This occurred because wages and prices are sticky in the short run, and some variable needs to adjust to equate money supply and demand in the period before prices adjust. Thus interest rates change until wages and prices adjust. This led economics to take a momentous wrong turn. Under the influence of Wicksell, and then later Keynes, macroeconomists began to see the change in interest rates as not some sort of epiphenomenon associated with an increase in the money supply, but rather as monetary policy itself. This culminated in the work of Woodford, who developed monetary models without money, where interest rates were all that mattered.
But Woodford did more than that, he showed that aggregate demand depended much more on the future expected path of monetary policy than on the current stance of policy. As we will see, that insight may have opened the door to a revival of Humean macroeconomics.
The QTM was based on the hot potato effect, the idea that if you double the supply of non-interest-bearing base money, people will try to get rid of excess cash balances. But in aggregate they cannot do so, as the central bank controls the supply of base money. Instead the attempt to get rid of excess cash balances drives up prices and NGDP until people are again willing to voluntarily hold the enlarged cash balances. That occurs when prices and NGDP have doubled, so that real money demand and velocity return to their original level.
There is a flaw in Woodfordian macro. No model lacking money can explain large changes in the money supply and price level. That’s why the post-WWI hyperinflations caused even Wicksell and Keynes to briefly return to the quantity theory of money. If you increase the monetary base by 87 times (and a number of middle income countries did this in the 1970s and 1980s) the monetary approach can give you a ballpark estimate of the price level (that it will rise by 87-fold), whereas the interest rate/economic slack approach to inflation is hopelessly lost. So there’s not much doubt that in some sense the quantity of money is still driving the price level, at least in the long run. But how can it be made relevant for our current situation, where inflation rates are quite low and velocity is quite unstable?
It seems to me that Woodford’s approach is capable of rescuing the QTM. Think about it. The QTM is criticized as only being able to explain price level changes in the long run. And yet Woodford says that the current level of aggregate demand is mostly determined by the expected future course of monetary policy; i.e. the long run. Keynes said in the long run we’re all dead, and now Keynes is dead and Woodford is saying in the short run AD is determined by (expected) long run changes.
For the moment let’s set aside the question of interest on reserves, and focus on non-interest-bearing currency. Before the recession (in 2007) currency was about 6% of U.S. NGDP. And if interest rates are positive in 2017 it will again be about 6% of NGDP, give or take 1%. That means the question of whether 2017 level of NGDP is 30% or 60% or 90% higher than in 2007, will mostly depend on where the Fed sets the currency stock in 2017. Not precisely, but approximately. Admittedly they don’t control the currency stock directly, they control the base. But in normal times the base is more than 90% currency. If the Fed wants that proportion to drop, they can increase IOR and inject enough extra base money to keep currency where they want it. If they don’t want it to drop they can lower the IOR, to zero if necessary.
The rate of growth in NGDP between now and 2014 will be strongly influenced by where people think NGDP will be in 2017. There is no interest rate path that the Fed can describe that would give people even a ballpark estimate of NGDP in 2017. But if they say they will set the currency stock at $1.6 trillion in 2017, (twice the 2007 level) then people will know that 2017 NGDP is also likely to be roughly twice 2007 NGDP. And if they do that, NGDP will grow very rapidly over the next two years.
Of course the Fed shouldn’t do that, for two reasons. First, that would create faster NGDP growth than is desirable. And second, it’s not as precise an instrument as targeting the forecast. Better to say they’ll provide as much base money as necessary in 2017 to produce an NGDP that is 30% higher than in 2012.
That’s the sort of monetary policy Hume would understand, and approve of.
PS. Why 30%? It would allow for 7% growth between now and 2013, and 5% thereafter. It’s a reasonable compromise that would dramatically speed up the recovery, but not create the politically unacceptable inflation likely to result from returning to the pre-2008 trend line.
PPS. Commercial banks? I don’t recall Hume having anything to say about them, so I never studied the subject.
I dedicate this post to MMTers everywhere.