Why Are MMT Critiques Generally Terrible?

The deluge of bad Modern Monetary Theory (MMT) critiques continues. Although it is possible that there are some diamonds in the rough, all the articles I read were downright stinkers. They were so bad that I refuse to dignify their existence by even citing them. However, as someone in the MMT camp, I just want to give a few pointers for my readers as to why those critiques stink.

There certainly are legitimate critiques of MMT, but they are hard to find. Even academics end up citing poor critiques written by non-MMTers; nobody can be apparently bothered to read the MMT academic literature. Which is a rather troubling commentary on the diseased state of modern academia (speaking as a hardline old school ex-academic). In this article, I just cover the ludicrous arguments; I will let the reader try to find some reasonable ones elsewhere. I am in the middle of developing an investment accelerator model within my Python SFC model framework, and I do not want to waste my time beating up on clowns.

Real World Versus Theory

Unlike modern academics in a certain discipline, we need to be careful with regards to what we are talking about. We really need to strongly distinguish between real world policy recommendations versus economic theory. The real world is complicated. Economic models have abstracted away that complexity so that we can make definitive statements about their simplified world. We cannot muddle those two modes of thinking.

I will discuss the real world critiques first (as that is of the most general interest), then digress to the discussion of theory, which really is only of interest to those of us who care about economic theory.

Inflation!

The "big idea" in Modern Monetary Theory that attracts the most attention is actually known as Functional Finance, associated with Abba Lerner (an early post-war Keynesian) - link to primer. For some exceedingly bizarre reason, many critiques of MMT by free marketeers ends up with invoking Venezuela, Argentina or whatever, and inflation. Let's look what MMT actually says:

The constraint on fiscal policy for a sovereign with a freely floating currency (and controls its central bank) is inflation.

(The disclaimer about the central bank is needed to deal with the case of the euro area; the euro floats versus other currencies, but the component countries have to follow the lead of their ECB master.)

Saying that "inflation is a constraint" is a fancy-pants way of saying "too much inflation is bad, so at some point, you gotta stop spending."

So arguing that "implementing MMT" will lead to inflation (or hyperinflation!) is ridiculous: the whole premise of MMT is that "too much" spending leads to "too much" inflation, and that is to be avoided. The only real debate is how much spending is "too much." As will be discussed next, actual policies advocated by MMTers are not particularly scary.

I Like Ike!

I had a discussion on Twitter with an author who had an argument that ran along the lines (paraphrasing):

1. Alexandria Ocasio-Cortez ("@AOC") is a MMTer.
2. Alexandria Ocasio-Cortez is also some kind of socialist.
3. Therefore, MMT is socialism.

This does not work; like most other economic schools of thought, MMT is not tied to a particular part of the economic spectrum. For example, both Paul Krugman and John Taylor are neoclassical economists, yet they are not exactly on the same page for economic policy. If we return to MMT, about the only part of the spectrum that will not be happy with MMT are hard money advocates; one of the early texts by Warren Mosler (a "MMT founder") is "Soft Currency Economics."

I am a Canadian Prairie Populist, and from the relatively free market of that spectrum. Almost nobody is going to have a clue about Canadian Prairie Populists, and I will not bother explaining. I will instead merely state that if I translated my views into the American lexicon, I would pass as Eisenhower Republican from the perspective of economic policy. If you believe that the Supreme Commander of  the Allied Expeditionary Force (SHAEF) is some kind of dangerous leftist, that is your problem, not MMT's.

The advocacy of a 70% marginal tax rate (as per @AOC) is not really a MMT policy; like practically everyone, MMT authors presumably favour progressive tax rates. (Even a flat tax is progressive; every system that I have seen proposed includes a 0% bracket for low incomes.) The level of those tax rates is not a point of MMT dogma. I live in Quebec, and the all-in marginal tax rate for almost any professional hovers around 50%. I think that level is about the highest rate that can be sustained politically; I have my doubts about very high rates aimed at extremely high incomes.

If we look at the MMT literature, there's only a few policy platforms that are tied to the theory.

• Floating currency. This is only controversial with gold bugs, post-Keynesians, and a lot of Europeans. On this side of the pond, floating your currency is a boring consensus opinion. (It is entertaining that gold bugs accuse MMTers of being socialist when they are advocating the government fixing the price in a key market, while the MMTers advocate staying out of the way of the market.)
• Job Guarantee. This is the most distinctive MMT platform. It is an extension of the welfare state, but it is actually replacing having people on welfare paid to do nothing - which used to be a big deal among conservatives. The wage paid is a key policy variable, so the effect of its introduction depends on what level is chosen. However, a Job Guarantee is far more sustainable economically than a Basic Income, since any overheating of the economy automatically drains the Job Guarantee pool.
• Eliminating the possibility of central government (Treasury) default. This is my personal hobby horse, but I think the idea floats around in MMT articles in various forms. This could be accomplished by just giving the Treasury an open-ended overdraft at the central bank; other options could be a more root and branch reform of government finance. It tells us all you need to know about the people who are allegedly worried about a fiscal crisis that they would scream very loudly against eliminating the possibility of government default by construction. It is a weaker variant of the next option.
• Possibly: Eliminating Government Bond Issuance, locking rates at 0%. This is hugely controversial with mainstream economists, who are deeply attached to interest rate policy. I would note that there is not a 100% consensus in favour of the policy among MMTers, although there is a consensus about how to analyse what the effects are. 

There are other policies that are advocated by particular MMTers, and/or MMT activists. From my perspective, I don't think that it is safe to say that everyone who is in the "broad MMT camp" will necessarily agree with policies.

The Theory in MMT

I normally stay clear of policy advocacy; I am more interested in writing about theoretical squabbling. The standard complaint about MMT by non-MMT academics is that there is nothing new there.

I would first note that I am a very ex-academic (although I may present papers at a MMT conference in my doddering old age). Academic turf wars are very exciting to the academics involved, but are nothingburgers to everyone else. I am obviously sympathetic to MMTers, but I lack the access to the research library that would allow me to easily do the reference chain analysis I did back when I was in academia. As I describe below, the contributions of MMT -- as presented by MMT academics -- seem plausible. If you are an academic in economics, go to the library and do your own %$& research.

The key point to understand is that MMT is part of a long line of post-Keynesian theory, and in academic circles, is described as such. The "MMT" moniker can either be viewed as a branding exercise, or an attempt to create a consistent body of thought within the wider post-Keynesian literature. This latter step was necessary, since we have to admit that the post-Keynesian project was a practical failure. The same people have been presenting the same papers yammering on about what Keynes really meant for decades, and they were roundly ignored by everyone else. The only "post-Keynesian" anyone outside that circle cared about was Minsky, and his relationship with the "professional post-Keynesians" was sketchy. (Minsky was the supervisor of MMT "founder" L. Randall Wray, and can be viewed as the godfather of MMT. The Job Guarantee is related to his Employer of Last Resort proposal.) 

In any event, most of the academic attacks on MMT came from post-Keynesians, who later authors then quite often cribbed.

If we put aside the uneasy relationship between "MMT" and "post-Keynesian" economics, we can then turn to the differences with the neoclassical tradition ("mainstream"). Speaking as a mathematically-literate outsider with no dogs in academic turf wars, the advantages of MMT/PK approaches over neoclassical are massive and glaring.

For example, I am working on a book on the theory of recessions. I want to be complete as possible. What can the neoclassical camp give me?

1. Econometric analysis that is just stock mathematics, and not tied to any theoretical school of thought.
2. Theoretical models that normally cannot generate recessionary outcomes. (To be fair, the neoclassicals claim to have fixed this in their post-2008 literature; I am agnostic about that claim right now.)

In other words, a theoretical contribution that is pretty close to nothing. Admittedly, one of the reasons why I decided to focus on the topic of recessions was that the neoclassical literature was so useless that I could in good faith largely ignore it, and so reduce the writing time for the book. However, recessions are hardly an obscure theoretical topic.

(Is there a "MMT theory of recessions" that is distinct from the post-Keynsian? I am not really aware of one. This is to be expected if MMT is part of a wider post-Keynesian tradition.)

Circling back to our main topic, we need to discuss fiscal policy. One could argue that MMT is not a huge advance over post-war Functional Finance. Meanwhile, the current neoclassical tradition is based on a growth that started in post-war Keynesianism. The argument is this: the neoclassicals believe that they progressed on the old Keynesians; MMTers would agree that they regressed.

I have written dozens of articles on this subject already; in fact, it is probably in my book, Understanding Government Finance. In my view, the debate largely revolves around two theoretical points: overlapping generations (OLG) models, and the inter-temporal governmental budget constraint. I discuss these in turn.

The neoclassicals appeal to OLG models as a way to "prove" that government debt "puts a burden" on future generations. (I discussed these models various times, such as here.) In summary, I believe these models are contrived, and just designed to give the desired result. 

One of the standard complaints I have seen by apologists for the mainstream (such as on Twitter) is that MMT is "not empirical." Meanwhile, the mainstream appeals to models that cannot be fit to data to prove that inter-generational debt burdens exist. Very simply: how do we fit an OLG model to the post-World War II government debt trajectory to determine which "generation" was "burdened" by debt?

As for the inter-temporal governmental budget constraint, it is a mathematical constraint that is imposed on dynamic stochastic general equilibrium models. It allegedly binds fiscal policy, and the presumption is that this constraint also binds in the real world. The objection is straightforward: it is extremely hard to find a convincing reason why the constraint should even be imposed on the mathematical framework in the first place. For 99% of papers, the constraint in imposed on purely arbitrary grounds.

I would note the article by Alexander Douglas about a game-theoretical explanation for the constraint. I have some concerns about the setup of the game -- does the private sector really have the discretion that the model suggests -- but that would require a longer explanation.

I think I can see the real reason for the existence of the constraint, and that reason is the Fiscal Theory of the Price Level (FTPL).

The short summary of the FTPL is that if government debt levels are not sustainable, the price level "goes to infinity" now. On a less extreme basis, the price level is determined by the stance of expected fiscal policy.

This causes three problems.

1. This is directionally the same as Functional Finance.
2. If the price level at time t is determined by the stance of fiscal policy, the price level at time t+1 is determined by the stance of fiscal policy at time t+1. The implication is straightforward: all the neoclassical argle-bargle about monetary policy is just plain wrong: inflation control is 100% a question of fiscal policy.
3. The FTPL is either trivial or wrong. It is trivial since expected primary surpluses cannot be observed, and could only be backed out by looking at the price level: so nothing is explained. Furthermore, we do not see massive instantaneous jumps in the price level every time a controversial tax package is voted on, or budgets announced in parliamentary governments.

Once we throw out these theoretical constructs, the mainstream is stuck with almost no viable theory about the limits of fiscal policy (beyond multiplier analysis, which ends up being hard to empirically distinguish from Functional Finance). All we are are left with are trivial analyses of debt trajectories in implausible economic models ("assume that debt levels rise by 6% forever, and nominal GDP growth rises by 4% forever , and there are no feedback loops between those variables"), or campfire stories about "bond vigilantes" or "unsustainable debt levels." The decades of embarrassingly bad predictions about Japan from various mainstream sources explains why bond market participants like myself never took mainstream assertions about fiscal policy too seriously.

Concluding Remarks

Although I am sure that legitimate critiques of MMT out there, good luck finding them.

(c) Brian Romanchuk 2019