Economists of my generation recall a long boring debate over monetary policy, which frequently involved the term ‘velocity.’ Monetarists claimed the Fed should stop trying to fine tune the economy, and instead keep the money supply growing at a steady rate of about 4%. Keynesians countered that velocity was unstable, so the policy would not work. Monetarists countered that much of the instability of velocity was due to instability of money, and that if money growth was more stable, velocity would also be more stable.
And so the debate raged on and on. So here’s my question; why the heck does it matter what happens to velocity? Why should we even care? Why did the Keynesians work so hard to show velocity would be unstable, and why did the monetarists fight so hard to show it would be stable if only we would target M? Why?
You might think the reason is obvious—just look at the famous equation of exchange:
If velocity is stable, then a steady growth of M would yield a steady growth of P*Y. Yes, I get that. When we talk about the equation of exchange we are just talking about definitions.
Obviously in the long fruitless debate over velocity during the 1950s, 1960s, and 1970s, the implicit benchmark for a successful monetary policy was stable growth in P*Y. That’s obvious.
But here’s the mystery, almost no one in the 1950s, 1960s, or 1970s was actually advocating NGDP targeting! People have dug up quotations that seemed to show Milton Friedman opposed NGDP targeting. I don’t believe them. Friedman supported NGDP targeting, whether he knew it or not. Tell me which of the following imaginary scenarios would have been more likely:
Scenario A: Research using a sophisticated MIT supercomputer shows the Keynesians were wrong and Milton Friedman was right all along. Fluctuations in M2 velocity were 100% caused by fluctuations in M2 itself. That’s why the collapse of M2 in the early 1930s was accompanied by sharply falling velocity. In addition, the study showed that M2 velocity would be virtually constant if M2 were to grow at a steady rate of 4%/year. Friedman responds with a note of exultation. “Finally I’ve been proved right, and by an MIT study no less. There is no longer any reason to refrain from putting the Fed on automatic pilot, replacing Fed officials with a computer. No more discretion.”
Scenario B: Research using a sophisticated MIT supercomputer shows the Keynesians were wrong and Milton Friedman was right all along. Fluctuations in M2 velocity were 100% caused by fluctuations in M2 itself. That’s why the collapse of M2 in the early 1930s was accompanied by sharply falling velocity. In addition, the study showed that M2 velocity would be virtually constant if M2 were to grow at a steady rate of 4%/year. Friedman is dismayed by the findings. “Oh my God, that this means that if the Fed targeted the money supply growth rate at 4% a year, we’d have steady 4% growth in NGDP. That would be horrible, as it means that if RGDP fell by 1% then we’d have 5% inflation. I now see the folly of my ways; the Fed must not increase the money supply at a constant rate, but rather must use discretion to prevent the evils that would flow from steady NGDP growth .”
Believe it or not, I’ve had commenters who implicitly claim scenario B is more plausible.
For three decades economists debated monetary policy using the implicit assumption that stable NGDP growth is the benchmark for success, that NGDP targeting is optimal. And yet for some reason they never figured that out. Almost no one was advocating NGDP targeting from 1950-1980.