Keynesianism, NeoFisherism, and MMT
By Scott Sumner
The title of this post lists three macro models that I believe are wrong. But they are not all wrong in the same way.
I’m not even sure I understand MMT, as when I try to engage with proponents of that theory they keep telling me that I’ve got it wrong. I say, “So you’re saying A, and here’s why that’s wrong.” They respond, “No, we aren’t saying A, we are saying B.” I respond, so you are saying B, here’s why that’s wrong.” And they respond, “No, we are not saying B, we are saying C.” Then I explain why C is wrong. It never ends.
Perhaps it’s just me. But here’s the problem for the MMTers. Paul Krugman is sympathetic to many of their policy preferences. He’s also “on the left”. He likes some politicians who like MMT. But he has exactly the same reaction to the model as I do:
Now, arguing with the MMTers generally feels like playing Calvinball, with the rules constantly changing: every time you think you’ve pinned them down on some proposition, they insist that you haven’t grasped their meaning.
I don’t expect everyone to be able to explain their models in a way that a slow mind like me can understand. But they should be able to explain it to one of the half dozen most brilliant economists in the world.
It seems to me that the problem with macro is that the underlying problems are so complex that there are a wide variety of ways to address these problems. For instance, just in the field of money you have the interest rate approach, the quantity of money approach, and the price of money approach. Within each of those you have varying assumptions about price stickiness, Say’s Law, crowding out, rational expectations, Ricardian equivalence, market efficiency, and a host of other issues. The possible approaches quickly multiply, each developing different frameworks and even different languages. C + I + G = PY. MV = PY. IS/LM. AS/AD. Etc., etc. You end up with a sort of Tower of Babel.
To me, MMT seems like a more extreme version of Keynesianism, having all of its flaws and none of its virtues. In contrast, NeoFisherism takes one of the flaws in Keynesianism, reasoning from a price change by assuming that lower interest rates are expansionary, and inverts it into the opposite error of assuming that lower interest rates are contractionary. Here’s Krugman:
Figure 1 illustrates my point. Suppose that the Fed or its equivalent in another country can set interest rates, and that a lower interest rate leads, other things equal, to higher aggregate demand. Then at any given point in time there is a downward-sloping relationship between the interest rate and GDP, as shown by the lines IS1, IS2, IS3.
Other things equal? Hold on there! Just how does the Fed achieve lower interest rates while holding “other things equal”? I claim that there are two ways for the Fed to drive rates lower, with an easy money policy or with a tight money policy. Neither involves holding other things equal. This post shows what each of those policy options might look like.
Now obviously Krugman meant “lower rates created by an easier money policy.” He’s a Keynesian. But then instead of simply saying lower rates, he might have said “Lower rates resulting from the liquidity effect from an increase in the monetary base.” In that case, it’s the bigger base that’s actually having the expansionary effect. Lower rates slightly reduce velocity, and make the expansionary effect smaller than otherwise.
Or he could have said “Lower rates produced by a policy that reduced the demand for base money, such as lower reserve requirements or lower interest on reserves”. That would work too.
If he doesn’t specify, how do we know that he doesn’t mean, “Lower rates via a Japanese style monetary policy that produces the slowest NGDP growth (over 25 years) in modern history.”? Or lower rates via the methods used by the Fed in 1921, when they contracted the monetary base so sharply that we had double-digit deflation.
The Keynesian model is more prestigious than the NeoFisherian model, but it’s not any better. Indeed during 2008-15, the NeoFisherian model was giving better advice as to whether monetary policy was too tight or not, even as the Keynesian model did a better job of explaining real time market reactions to unexpected changes in the Fed’s target interest rate.
Of course the market monetarist model can explain both, which is why it’s the best.