MMT Is Not Just "Theory X"

I keep running into arguments of the form: "Modern Monetary Theory (MMT) is just {mainstream theory X} with {qualifier Y}." (The loopiest recent variant of this was that "MMT is just the Quantity Theory with bonds as part of the money supply." My head hurts just thinking about that assertion.) I believe that such characterisations are largely inevitable, and it is not obvious how to deal with them.

It would be very easy to be annoyed at these characterisations of MMT; they almost invariably trivialise the theory, saying it is "just" some simple theory (the word "just" is always in the phrasing). However, having seen variations of this dozens of times over the years (with the theory it is compared to always changing, of course), I think a more constructive and proactive stance is needed. Unfortunately, I am unsure how to go about that.

Why These Comparisons Are Obviously Wrong

The first thing to keep in mind is that MMT is not a single idea, it is what appears to be an internally consistent economic school of thought, which lies in the "broad tent post-Keynesian" economics, following the definition of Marc Lavoie in his text Post-Keynesian Economics: New Foundations (link to review). The issue with (broad tent) post-Keynesian economics is that is the union of a number of distinct thinkers, who did not always agree with each other, and so it is hard to define. Modern Monetary Theory is an attempt to strip out one consistent theoretical narrative, tied to some policy views that are controversial within the broad post-Keynesian community.

(The most acute issue is the view of the external sector; the MMT preference is to let the external sector sort itself out, which causes some post-Keynesians to be very unhappy.)

Therefore, when someone compares MMT to some variant of a mainstream model, it runs into the immediate problem that MMTers reject practically every single core assumption of the mainstream model it is being compared to.

Saying that "MMT is just {mainstream model X with qualifier Y}" may make sense to the speaker in trying to understand some aspect of MMT, but it probably conveys no useful information to anyone else. It insinuates that MMT accepts the core assumptions of model X, which is incorrect.

Since everyone uses physics analogies, I might as well use one myself. For some mysterious reason, the electrical engineering department at McGill thought it was a good idea to require every honours electrical undergraduate to take a course in quantum physics taught by the Physics Department. Of course, it included a section on relativistic quantum mechanics, since there are so many electrical components that travel at a significant fraction of the speed of light. In order to pass the course, I developed a decent understanding of Special Relativity (which I have long forgotten, so don't bother asking me any questions about it).

One could think of special relativity as just Newtonian mechanics with some corrections thrown in for the speed of light. However, that sort of thinking probably doomed you to not doing well on the course. Special Relativity makes sense, so long as you follow the internal logic of the theory. The internal logic determines the form of the equations. Conversely, if you tried to memorise all the possible corrections from Newtonian mechanics, it would be extremely easy to do it wrong, since you have little idea why the corrections are being added in the first place.

In other words, although you might be able to get away with thinking that Special Relativity is just Newtonian mechanics with corrections for the speed of light, that is not a helpful way to explain it to others. (You can go the other way of course, taking Newtonian mechanics as the limit of Special Relativity where velocities relative to that of light are small.)

Furthermore, the breadth of the theory implies that it is not going to be easy for a MMTer to offer a short summary of what MMT is. Someone may be able to come up with a succinct theoretical summary, but that summary would be meaningless for anyone not already familiar with MMT in the first place. (For example, one could try to summarise DSGE macro as being "based on the optimising decisions of households over time," but that summary conveys almost no useful information to anyone who is not already familiar with the structure of the models.)

What Can MMTers Do?

The problem that MMT faces for acceptance by other economists is not really education, rather it de-education. The problem with mainstream macroeconomics is that it is built upon a series of wrong assumptions, and the objective of economic education is to ram those assumptions into students' ears until they come out of their mouths. Mainstream economists are very open to discussing models, but have almost no willingness to think about the assumptions they are using.

As a small telling example, take the government debt oeuvre of Reinhart of Rogoff. I first looked at it before I had ever heard of MMT, and knew it was obviously wrong: the data sets that they used commingled sovereign issuers with others with various external constraints on borrowing (they even threw a sub-sovereign into one of their lists of defaults). I never would have found the spreadsheet error in that one particular paper, since I knew it was a priori not worth reading -- never mind testing.

After various theoretical embarrassments, various mainstream economist argue that they knew all along that there is a difference between currency sovereigns and other issuers. However, if that were true, how did Reinhart and Rogoff's work get taken seriously in the first place? The reality is that mainstream theory is hopelessly deficient in its analysis of government debt, and that they have been forced to bolt on an ad hoc correction (an epicycle) to better fit reality. The problem is not that particular mainstream economists made a mistake, rather they are battling with an assumption that the currency framework does not matter when discussing sovereign debt. Since the assumption is never examined, it is extremely difficult to have a useful conversation about government debt.

Since abandoning core assumptions would be a de facto admission that the entire DSGE literature was a failure, I would not hold my breath waiting for such an event. Instead, the course of action that may have a chance of success is keep discussions very concrete. As soon as theory is brought up, the clash of assumptions will drown out any useful discussion, (Although there appears to be a 100% chance that the mainstream side will invoke hyperinflation at some point of the debate -- even though they cannot offer an explanation of how the hyperinflation would occur even in the context of a mainstream model.)

As a result, there is probably room for primers to MMT written at an advanced level. (One would think that this would be the role of the academic literature, but such is the state of the "science" of economics is that the best way to guarantee that a mainstream economist will not read something is to publish it in a heterodox journal.) There is little point aiming at mainstream academics, but one can plausibly reach out to people working in finance or even places like rating agencies. The larger the advanced technical audience is for MMT, the more ridiculous the attempts to explain it away will appear.

(Update: I should have noted this when I first wrote this, but the blogs by Bill Mitchell -- the Billy Blog -- are often at the more advanced level of analysis that I am discussing here. The problem is that many critics of MMT have obviously not read any of his material. The new MMT textbook -- which I have long argued was a critical missing link for MMT -- will also help matters. However, many people who already have an undergraduate/graduate education in economics would probably pretend that reading a new undergraduate text is beneath their station. )

(c) Brian Romanchuk 2018