Recent Posts

Tuesday, December 17, 2019

A Skeptics Guide To Mankiw's Skeptic's Guide To MMT

N. Gregory Mankiw (writer of one of the best-selling economics textbooks) has recently written "A Skeptic's Guide to Modern Monetary Theory" (MMT). He discussed several assertions made in the textbook Macroeconomics, by William Mitchell, L. Randall Wray, and Martin Watts (link to article with my initial comments on that text). His conclusion was "In the end, my study of MMT led me to find some common ground with its proponents without drawing all the radical inferences they do."

Professor Mankiw explains the genesis of his article as follows:
[...] So I recently tried to figure out what MMT was all about. I wanted to identify the key differences between this new approach to macroeconomics and the approach in mainstream textbooks, such as my own.
Fortunately, there was an ideal vehicle for this endeavor. In 2019, Red Globe Press published a new textbook, simply titled Macroeconomics, written by three MMT proponents: William Mitchell and Martin Watts (both of University of Newcastle, Australia) and L. Randall Wray (Bard College). This brief essay explains what I have learned about MMT from this textbook treatment.
The most natural way to approach a body of new theory is to pick up a textbook. It is exactly my mode of operation. The only issue here is that Macroeconomics is an undergraduate textbook. As was noted by Scott Fullwiler on Twitter, neoclassical academics get testy when people use the simplified models from introductory undergraduate textbooks as representing the entirety of neoclassical macroeconomics. As such, I would use a graduate level textbook  (or at least senior undergraduate level) when I needed to dive into a new subject.

Correspondingly, I think the implied claim in the title "Guide To Modern Monetary Theory" needs to be viewed to be closer to "A Discussion of the MMT Macroeconomics Textbook." Since the textbook is an academic work that needs to stand up to scrutiny, this narrowing of focus is legitimate.

The Difficulty with MMT Being a School of Thought

There is no doubt that it is difficult to get a handle on MMT if one does not want to dive into journal articles, at least if one wants a more advanced treatment. (If I had any commercial sense, I would have made an explicit primer. My book Understanding Government Finance (link) is a primer at the target reading level I am discussing, but it is not really a true primer for broad MMT theory.) As a result, Mankiw's description of difficulties of getting a handle on the topic is common.

 From an academic standpoint, MMT is extremely broad at this point, as it gone multi-disciplinary. My focus is on the "hard macro" core, but even that covers a lot of ground.

There are at least three main subjects of discussion within the MMT literature.
  1. Constructive. Theoretical statements about the economy, which is the "MMT framework."
  2. Criticism. Theoretical critiques of neoclassical theory. Although this is obviously related to the MMT framework, the critiques may be true even if the MMT theory is incorrect.
  3. Policy Proposals. Policy proposals that are presumably related to the MMT framework, but may make even be viewed as good policy proposals from other frameworks (including neoclassical theory).
Having to cover all these bases is an imposing task, particularly when one considers that the various elements can stand on their own (as my italicised comments suggest). The breadth of the topics involved creates the main limitation of the Mankiw article, as they are missed.

What Does Mankiw Say?

The Mankiw piece is easy-to-read seven pages (plus references). He draws the following quotes from Macroeconomics (page numbers are for that text, denoted MWW); Mankiw's bridging text is in [brackets].

The financial constraint.
The most important conclusion reached by MMT is that the issuer of a currency faces no financial constraints. Put simply, a country that issues its own currency can never run out and can never become insolvent in its own currency. It can make all payments as they come due.” (MWW, p. 13) [As a result] for most governments, there is no default risk on government debt. (MWW, p. 15) 
Money and the price level.
[They assert that] “no simple proportionate relationship exists between rises in the money supply and rises in the general price level. (MWW, p. 263) 
Conflict theory situates the problem of inflation as being intrinsic to the power relations between workers and capital (class conflict), which are mediated by government within a capitalist system. (MWW p. 255). [Mankiw summary of p.264-265: "MMT advocates see these guidelines, and even government controls on wages and prices, as a kind of arbitration in the ongoing class struggle.]
Mankiw's summary of MMT's views on inflation and growth.
Mainstream theories of inflation emphasize not class struggle but excessive growth in aggregate demand, often due to monetary policy. This idea also appears in MMT. Its proponents admit that “all spending (private or public) is inflationary if it drives nominal aggregate demand above the real capacity of the economy to absorb it.” (MW&W, p. 127) The advocates of MMT, however, make this possibility seem more hypothetical than real.
We are also told that “capitalist economies are rarely at full employment. Since economies typically operate with spare productive capacity and often with high rates of unemployment, it is hard to maintain the view that there is no scope for firms to expand real output when there is an increase in nominal aggregate demand.” (MW&W, p. 263) 
That is the sum total of input from MMT. The remaining text in the seven pages is a discussion of the history and results of neoclassical (New Keynesian) macroeconomics. One may note that this guide to MMT has one (1) referenced MMT text, as compared to eight (8) neoclassical citations, including two (2) from N. Gregory Mankiw.

Can One Respond to the Mankiw Critique?

I would paraphrase Mankiw's criticism of MMT as follows: if we assume that neoclassical theory is correct, MMT is either incorrect (where it contradicts neoclassical theory) or trivial.

This is an obviously true statement, but it begs the question: is neoclassical theory correct? As my earlier discussion noted, one of core topics of discussion of the MMT literature are critiques of neoclassical theory. We need to investigate whether those MMT critiques are out to lunch. The concern with the Mitchell, Wray, and Watts text is that is very much an introductory text, without the complex discussions that would be needed to critique neoclassical theory at an advanced level, and unfortunately, the bibliographic information only has a limited number of citations to the more advanced critiques.

Nonetheless, if we dig further into the MWW text, some of the concerns appear to be addressed.

Financial Constraints

I will take an important example, Mankiw's discussion of the financial constraint on governments. I will strip out a few key points raised by Professor Mankiw.
First, in our current monetary system with interest paid on reserves, any money the government prints to pay a bill will likely end up in the banking system as reserves, and the government (via the Fed) will need to pay interest on those reserves. That is, when the government prints money to pay a bill, it is, in effect, borrowing. 
This is related to an issue raised by Jeffray Gundlach (link to my comments). The obvious rebuttal is that the government can lock the risk-free interest rate at zero. What matters is whether MMT and neoclassical theory disagree with the implications of such a step.  I will return to this below.

The next observation by Mankiw:
Faced with these circumstances, a government may decide that defaulting on its debts is the best option, despite its ability to create more money. That is, government default may occur not because it is inevitable but because it is preferable to hyperinflation. 
I turned to my copy of MWW, and on pages 566-567, I found the following index entries:
hyperinflation, 255, 342-346
  definition 343
  MMT view, 343-4
  Quantity Theory of Money and, 343
  Weimar Germany, 344-5
  Zimbabwe, 345-346
Without even going back to pages 342-346, it suggests that there might be at least something in the textbook that MMT takes a slightly view on the subject.

Effect of Locking Rates at 0%

If we look at page 334, we see that MWW discussed the zero interest rate policy, which was not picked up by Mankiw.
This could happen if, for example, government adapted a zero interest rate target [emphasis in original]; in that case excess reserves would drive the overnight interbank lending rate towards zero and the government would not need to do anything further. [...] The decision depends on whether government wants to target a positive overnight rate, or a zero overnight interest rate.
To be fair, there is no extended discussion of the complex theoretical issues around this decision (at least that I could find).

We could summarise the neoclassical view on locking interest rates at 0% as being really, really, bad. This is an "interest rate peg" that shows up in graduate level dynamic stochastic general equilibrium (DSGE) introductory texts, and the implication is that the price level becomes indeterminate (whatever that means).

Obviously, the policy rate can be locked at 0% (or even lower) for years at a time, the question is whether such a peg can be maintained "forever." General equilibrium models with such a reaction function tend to result in an indeterminate price level.

The reason why this is an inadequate critique is clear to anyone with a background in mathematics. A DSGE model is a mathematical "system" (using terminology from applied mathematics). One starts with a set of variables, and assumed relationships between those variables. We then do some mathematical manipulations to derive a set of implications.

One might think that this implies that there is a true process of discovery behind these steps. This is not exactly the case. A set of internally consistent mathematical relationships is just a logical construct. In most cases, one can start from a wide variety of initial assumptions, and re-derive the entire logical relationship.

That is, we can do the following. Pick a set of "implications", assume them to be true, and the do proofs "backwards" to derive the initial assumptions as being "implications." To spell this out, assume that we can do the following.
  1. Assume A is true.
  2. Do N pages of mathematical operations ("proof").
  3. Demonstrate that B is implied.
We typically will be able to accomplish the alternative.
  1. Assume B is true.
  2. Do M pages of mathematical operations.
  3. Demonstrate that A is implied.
(Why do this? We may be able to derive proofs in a more elegant fashion from a different starting place.)

The effect is straightforward: one could view DSGE models that exhibit price level instability as being the models that result if we assume that fixed interest rates result in an indeterminate price level. If we do not make such an assumption, the models are invalid.

So we cannot pull out some random DSGE textbook and argue that locking interest rates at 0% causes hyperinflation in the real world, since those models are essentially built on the assumption that is the outcome. We need to see whether there is any empirical backing for the assertion about the disastrous effects of an interest rate peg.

I am not going to attempt to answer that debate here. However, this debate is what matters, and is not discussed in the Mankiw piece.

Quantity Theory of Money

Mankiw writes:

[...] They assert that “no simple proportionate relationship exists between rises in the money supply and rises in the general price level.” (MW&W, p. 263)
This assertion overstates the case against the mainstream view. In U.S. decadal data since 1870, the correlation between inflation and money growth is 0.79. Cross-country data exhibit a similarly strong correlation. (Mankiw, 2019, pp. 109-110) 
Correlation is not causation.


I will not cover further Mankiw's discussion of MMT views on inflation. I think the quotations provided above indicate that one could see some overlaps in the operational effect of theories (making MMT redundant if we assume neoclassical theory is true), and obviously in trouble where it disagrees (if we assume that neoclassical theory is true).

Concluding Remarks

Mankiw's piece at least offers readers a pointer into the MWW text on theoretical points of dispute. However, anyone wanting to undertake a scholarly discussion of the topic would need to spend far more time grappling with the criticisms of neoclassical theory in order to gauge whether MMT has contributions to make.

Note: The textbook Macroeconomics is available online at Amazon here (affiliate link).

(c) Brian Romanchuk 2019


  1. "Without even going back to pages 342-346, it suggests that there might be at least something in the textbook that MMT takes a slightly view on the subject."

    I'm guessing it is a slightly 'different' view? (on the subject of hyperinflation or government default)

  2. That is, we can do the following. Pick a set of "implications", assume them to be true, and the do proofs "backwards" to derive the initial assumptions as being "implications." More compactly,

    - If A can be shown to imply B,
    - then if we assume B is true, we can derive that B implies A.

    Can you explain this part a bit more clearly please. If A implies B, it doesn't necessarily mean that B implies A. Would need to be 'if and only if' for this to work no?


    1. (Strange, my comment disappeared.)

      I re-wrote the section, hope it is clearer. Thanks for pointing out the issues with the text.

    2. Thank you so much for the clarification!

  3. Mankiw says here about government creating money- " Second, if sufficient interest is not paid on reserves, the expansion in the monetary base
    will increase bank lending and the money supply. Interest rates must then fall to induce people to
    hold the expanded money supply, again putting upward pressure on aggregate demand and

    Is he using the old money multiplier idea when he claims the new reserves will increase bank lending? Seems that way to me. And wouldn't interest rates have to rise (rather than fall) to induce people to hold the expanded money supply?

    1. (Once again, an earlier reply disappeared?)

      I'm not digging into his logic. The locking rates at 0% is a MMT policy proposal, and makes his story here pretty much a non-issue.

  4. Mankiw- "Mainstream theories of inflation emphasize not class struggle but excessive growth in aggregate demand, often due to monetary policy. This idea also appears in MMT." The idea that inflation is 'often due to monetary policy from causing 'excessive growth in aggregate demand' is not often found in the MMT I am familiar with. But he is right about excessive aggregate demand causing inflation according to MMT. I guess I would say he was half right here. But half wrong also. I wonder how he grades his students...

    1. The aggregate demand story is pretty similar from 30,000 feet, but the devil is in the details. E.g., the use of output gaps (as defined by a mechanical procedure on GDP), NAIRU, etc. If all one does is take a few out-of-context quotes from an introductory undergraduate textbook, those quotations will not advance things that much. One would need to look at the sections on inflation in way more depth to contrast and compare properly.

    «the equation Wren-Lewis used needs additional assumptions or clarification.»
    One of these I may be able to provide here... :-)
    If that is the *definition* of capital, it embodies some very powerful political points:
    * All capital is the result of "investment".
    * Therefore there is no such thing as "land" or natural resources in general, stuff that produces income largely without "investment" or "depreciation".
    * That since "capital" is soley the result of "investment" minus "depreciation", then income from capital is "deserved" by that "investment". There is no such things as economic rent at least in relation to capital.
    The politics of the definition above are those of the neoclassical reaction to the classics.
    Interestingly there is a good argument here that the neoclassical reaction was against Georgism rather than Marxism or socialism:
    The thesis is that Georgism was a defense of profit on capital actively invested against rent on property passively owned, and JClark created neoclassical Economics in order to erase the distinction between "land" and productive capital. As in the *definition* above.
    «it embodies some very powerful political points:»
    Oops I forgot one of the most important "benefits" of that definition of capital:
    * There is no question as to initial or current endowments. It vanishes from the model.
    «There is no such things as economic rent at least in relation to capital.»
    This BTW is the gateway to Rand/Mankiw/... style economics.
    Because then only labor may receive economic rent, and thus exploit investors who create capital.
    Because any income for labour that exceeds subsistence is economic rent, and only the creative and capitalists "deserve" income higher than subsistence level, because they create all value.

  6. Regarding DSGE models, one could alternatively appeal to SMD, implying that a general equilibrium of sufficiently small dimension to NOT effectively be arbitrarily large in number with arbitrarily bad dynamics is categorically different from any GE that would apply to our actual economies ...
    ... and conclude that a DSGE that is tractable enough to yield a conclusion regarding whether the price level is indeterminate is not a model of the economy in which we live and so its conclusions are of no particular relevance to discussion of policy for the economy in which we live.

    1. The only issue with that line of attack is that the neoclassicals long stopped caring about the properties of solutions of DSGE. Every single discussion ends with “well, it was covered ”, and that passes the buck as well. Essentially the view is that the mathematics is so trivial and that everyone knows it, so there is no point in writing it out in full. The models say whatever the text in some abstract says it says, and there is no further discussion.

  7. I don't think the argument that since Mankiw only looked at an introductory MMT textbook therefore his criticisms can be dismissed is a very wise argument to make. I'm actually impressed that he read and referenced the darn thing before describing MMT- that is far, far more than the typical mainstream econ critic manages to do.

    And it is not like Mankiw's own more famous intro textbook hasn't been regularly used to criticize mainstream econ thought by MMT economists. That don't throw stones if you live in a glass house thing.

    Anyways, one of the main purposes of an introductory textbook is to provide an introduction to that field of study. If a student couldn't come up with a reasonably accurate description of the major tenets of that field after reading the book, then the book is probably not serving its purpose. Or you have a stupid or biased student. If that was the case then it is important to point out where the student went wrong by criticizing his arguments rather than just pointing out that it was only an intro textbook.

    Not necessarily saying Mankiw is biased. I'm sure he is not stupid though.

    1. I explicitly said that discussing the textbook is fair game. The question is whether he actually discussed even the contents of the textbook in a comprehensive fashion. After all, all he needed to do was look at the index to see if there was any discussion of hyperinflation. In fact, there was a lengthy discussion of the meaning of “financial constraint” that probably addressed issues raised.

      The reality is that there is almost nothing to debate in his critique - other than the inflation discussion. His points about financial constraints evaporate if the policy rate is locked at zero, and he wrote nothing about that. All you can say is that if the central bank is let free to set interest rates wherever it wishes, it can torpedo fiscal policy. But MMTers would probably not disagree with that.

      The inflation debate is interesting, but it would be hard to find people right now who will defend the Quantity Theory of Money. You would really need to dig into details to see where MMT inflation views differ from the neoclassical consensus in 2019, and not Mankiw’s idiosyncratic views on the subject.

      I don’t think it is a good idea to waste too much time going after the contents of intro textbooks. Heterodox authors have a point about the embedding of mythology, but that tells us little about the state of research. So I don’t waste my time with those critiques. (The only mainstream “econ 101” text I own was part of the CFA programme, and it didn’t have all the stupid stuff heterodox authors complain about.)

    2. I misunderstood your earlier responses to my comments. Sorry. I see your point now. Thanks Brian.

    3. In light of Bill Mitchell's most recent post on this issue, I retract my statement about being sure Mankiw is not stupid. It's not an 'either or' situation- people can be biased and be stupid at the same time. And there is probably a great deal of correlation there.

  8. Blowing smoke about bipartisan failure
    Comment on Brian Romanchuk on ‘A Skeptics Guide To Mankiw’s Skeptic’s Guide To MMT’*

    Brian Romanchuk summarizes: “I would paraphrase Mankiw’s criticism of MMT as follows: if we assume that neoclassical theory is correct, MMT is either incorrect (where it contradicts neoclassical theory) or trivial. This is an obviously true statement, but it begs the question: is neoclassical theory correct? As my earlier discussion noted, one of core topics of discussion of the MMT literature are critiques of neoclassical theory. We need to investigate whether those MMT critiques are out to lunch.”

    NO, there is absolutely NO need for further investigations of neoclassical economics: “The moral of the story is simply this: it takes a new theory, and not just the destructive exposure of assumptions or the collection of new facts, to beat an old theory.” (Blaug)

    Neoclassical economics is dead for 140+ years. Is MMT the valid new theory? Again NO, MMT, too, is refuted on all counts.#1 By consequence, the comparison of standard textbooks of both approaches, i.e. Neoclassics vs MMT, is an exercise in smoke blowing.#2-#4 More specifically, both approaches get the foundational macroeconomic balances equation wrong.

    Because the conceptual foundations are false the whole analytical superstructure is false and, as a result, the respective textbooks are scientifically worthless.#5-#11 Neither Neoclassical nor MMT policy guidance has sound scientific foundations. Neither mainstreamer nor MMTers have something worthwhile to say about how the monetary economy works.

    Egmont Kakarot-Handtke

    * Bond Economics

    * Gregory Mankiw A Skeptic’s Guide to Modern Monetary Theory

    #1 MMT: cross-references

    #2 Get it econ suckers: microfoundations = false, macrofoundations = true

    #3 Neoclassics and MMT ― much like pest and cholera

    #4 Heterodoxy ― an axiomatic failure just like Orthodoxy

    #5 To this day, economists have produced NOT ONE textbook that satisfies scientific standards

    #6 Refuting MMT’s new Macroeconomics Textbook

    #7 The father of modern economics and his imbecile kids

    #8 False on principle

    #9 Economics textbooks ― tombstones at the Flat-Earth-Cemetery

    #10 CORE: more lipstick on the dead economics pig

    #11 Macroeconomics and the fake History of Economic Thought


Note: Posts are manually moderated, with a varying delay. Some disappear.

The comment section here is largely dead. My Substack or Twitter are better places to have a conversation.

Given that this is largely a backup way to reach me, I am going to reject posts that annoy me. Please post lengthy essays elsewhere.