Monetarist vs Keynesian: Velocity Is the KeyCommentary by Bob McTeer
June 02, 2014
One of the neat (in my opinion) things I used to do in elementary macro or money and banking class was to show how you can explain the basic equalities of the macro economy using two languages: the language of the equation of exchange (MV=PQ) and the Keynesian language built into our national income accounts (C + I + G + X – M). In both cases, total spending equals GDP (GNP way back then). If you want to focus on the role of money and monetary policy, your better bet is probably the equation of exchange. Otherwise, the Keynesian framework would probably be more useful, especially if your view of monetary policy focuses on interest rates rather than the money supply. Using the Keynesian framework, does not—repeat not—make you a Keynesian in terms of your policy preferences. It’s like talking about the economy in French and talking about the economy in English.
Our monthly GDP reports, as indicated, are structured along Keynes’ spending categories of consumption, investment and so on. The second estimate resulted in a total of the spending categories of minus one percent at an annual rate. Roughly speaking, we’ve learned that the positives of consumption and exports were slightly more than offset by the negatives (in an arithmetic sense, not a desirability sense) of weak investment spending and strong imports.
Just for old time’s sake, I decided to translate that into money and velocity terms and was surprised at the result. The process is simply to take the GDP number and the money growth number and calculate the appropriate velocity number from those two. Using M2 as the preferred money supply, then the velocity number would be the velocity of M2. How many times does M2 turn over to generate GDP?
According to the Federal Reserve’s H-6 money stock series, M2 grew at an annual rate of 6.9 percent in the three months from January 2014 to April 2014. (That number is typical of M2 growth in recent years.) Nominal GDP rose only 0.3 percent in the first quarter. Since money growth plus velocity growth equals nominal GDP growth, M2 velocity must have declined by 6.6 percent. So a 6.9 percent growth in money was almost entirely offset by a 6.6 percent decline in velocity to yield a 0.3 percent growth in nominal GDP. That is quite an astonishing result.
I’m begging the question of why velocity declined so much and whether faster money growth would have produced a bigger impact on GDP or whether it would have been offset by an even larger decline in velocity. I grew up as a Milton Friedman monetarist, and predicting the impact of faster or slower money growth depends on velocity remaining somewhat stable, or at least predictable. Of course, we have all learned that velocity is a reflection of the demand for money. The first quarter could be described as an increase in the demand to hold money by the public.
The first quarter results are just an extreme version of what has been going on for some time now. Money has been growing around seven percent per year and that has been partially offset by significant declines in velocity. This time it was almost completely offset. Certainly, the components of the national income accounts show the recent performance of the economy in greater detail than what happened to money and velocity. To remain in your good graces, I’ll try not to let that make me a “Keynesian.”