Two Paper Trolling

I am tied up with Victoria Day Weekend gardening and outdoor maintenance activities, and was planning on skipping writing a blog article. (Happy Victoria Day to my readers who celebrate it.)  However, I saw an attempt at trolling MMT that was such an intellectual horror show that I decided to take the bait.

Noah Smith At It Again

Noah Smith started the show with his "Two Paper Rule" article. I would roughly summarise his argument that people who defend a school of thought saying that there is "a vast literature" is just an attempt to create a "mud moat" around the literature, protecting it from criticism. If you cannot cite two papers that are exemplary examples of the methodology in a field, it's not worth studying.

I have not spent a lot of time thinking about Noah's article; there are certainly some grains of truth to what he wrote. That said, I believe his views reflect his educational background: just because he studied physics, all other scholarly fields need to emulate physics. Try applying the "two paper rule" to Philosophy, History, Political Science, Control Systems Engineering, or whatever, and you would see that this is a rather reductionist methodology. An additional problem is the collapse of standards in academic publishing as a result of the "publish or perish" imperative: well over 90% of the article published after 1990 in almost all academic fields deserve to be ritually burned. In other words, there is a "vast literature" about practically anything that has a literature at all; and most of the literature is going to be chaff. All the key valid points within the "vast literature" may be buried in a multitude of papers, which may be of uneven quality, and using a variety of methodologies. In most cases, the sensible starting point is a graduate text/monograph, not journal articles; you only wade into the journal articles once you are attempting to build up a citation trail.

If I were to take the corner of the control systems literature that I mainly worked on ("anti-windup control"), the "two paper" criterion would have been meaningless. A lot of key results (in my view) were developed in the 1950s and 1960s, using mathematics that was sloppy. The "modern" (early 1990s, when I was working on it) was (largely) mathematically sound, but the application areas were of less interest, as the useful results were mined out. How could you represent that literature with two papers?

Furthermore, his comments on business cycle macro appear quite dubious.

When applied to economics, he says we can ignore the entire post-war literature because of the "Lucas Critique." This is convenient for people who wish to pretend that ancient debates are actually brand new and exciting, like the so-called "Neo-Fisherian effect." A somewhat less sympathetic view is that such an attitude is just a rationalisation by unscrupulous academics to bury the relevant literature in order to make said clique sound original.

Turning to the Lucas Critique, I have been forced to piece together the real mathematics that lies behind DSGE macro, as opposed to the hand-waving that passes for mathematics in most published papers. As a result, I have some appreciation for the critique, but my argument is that the critique does not say what most people think it says. If we take a narrow reading, it just suggests that a particular optimisation technique used for policy analysis was incoherent. Although interesting, any optimisation technique for policy analysis is inherently a stupid way to proceed. Therefore, the narrowly-defined Lucas Critique is not telling us very much. If we take a wide definition of the Lucas critique (which is what I usually see on the internet), most modern mainstream papers could be skewered by the same reasoning.

In any event, Noah is basing his argument on a dubious analogy to Physics.  "It's about evaluating the quality of the literature's methodology." There is a hidden assumption that the business cycle can be easily modelled. If this is not the case, any attempts at analysing the business cycle are going to be buried in struggles with the details of particular economics institutions. Any attempt at providing a clean and simple model for the business cycle is doomed.

Gerard Piling On

Like clockwork, Gerard MacDonell piled on, complaining about Modern Monetary Theory (MMT). Gerard's MMT criticism can be summarised:

1. He read Warren Mosler's "Seven Deadly Frauds."
2. He did not like it.
3. Like the Commander-in-Chief of his adopted country, he doesn't need to read any more books.

Needless to say, it is very hard to debate against that position.

I will once again repeat what I have written many times: MMT is a school of thought within the broad post-Keynesian tradition. It is an attempt to create a single internally-consistent world view within a literature that is fractured into multiple competing schools of thought. Although most "MMT" articles focus on a few topics that are of the most interest (Job Guarantee, free-floating currency economics, governmental financial operations), "MMT" implicitly covers all of economics. Meanwhile, many of those papers will have been written by authors who are not considered to be "MMT." (For example, MMT uses the stock-flow consistent modelling methodology, but the key papers in that area were written by people who were explicitly not "MMT.")

Realistically speaking, you need to specify what area of economics you are interested in before asking for the two papers. Furthermore, the papers you might be offered might not be written by "MMT" authors. If that disappoints you, you were not paying attention. Ask an academic philosopher which two articles summarise "Philosophy," and I imagine the response would be a rather curious expression.

Gerard's Attempt to Summarise MMT

Since he refuses to read the MMT literature, Gerard has a hard time finding what it says ("what did you expect?" is a question many might ask). As a public service, I will attempt to help.

Gerard attempted to summarise MMT's views towards fiscal policy as follows. 

The public sector budget constraint is either non-existent or a trivial accounting identity with no practical implication.  Accordingly, tax and spending policy should be set without  any regard to the deficit and long-run trajectory of the debt/GDP ratio.  Fiscal policy may occasionally need to be tightened, but the signal for that would be only an acceleration of inflation to an undesirable pace. Worrying about the trajectory of the debt itself is merely a reflection of a misunderstanding of how the payments system works, and is pointless.

I think it is close, but it needs some amendments. (My added text is in italics.)

What the mainstream refers to as "the public sector budget constraint" is either  non-existent (the infinite horizon statement, which is based on a raw inability to do stock-flow consistent model accounting*), or a trivial accounting identity (the one-period accounting statement) with no practical behavioural implications.  Accordingly, tax and spending policy should be set solely primarily based upon the inflationary implications of policy, as well as other real constraints, assuming that government is issuing liabilities in a currency tied to a central bank it controls. In other words, fiscal policy for a free-floating sovereign should be set without without  any regard to the deficit and long-run trajectory of the debt/GDP ratio.  Fiscal policy may occasionally need to be tightened, but the signal for that would be only an acceleration of inflation to an undesirable pace. Worrying about the trajectory of the debt itself is merely a reflection of a misunderstanding of how the payments system works, and is pointless, and reflects a complete lack of understanding of private sector behaviour with respect to savings accumulation.

Although this is a decent summary of Functional Finance, it is certainly not "all of MMT."

Footnote:

* The Fiscal Theory of the Price Level provides the only mathematically coherent mechanism to save the infinite-horizon version of the governmental budget constraint.

(c) Brian Romanchuk 2017