Recent Posts

Thursday, March 17, 2016

Modern Monetary Theory Not Modern - Sigh

I prefer to stay away from arguments about economic theory, however, there are some articles that are very hard to ignore. The article "MMT: not so modern" by Professor Simon Wren-Lewis of Oxford University is one such article. I believe that he mischaracterises Modern Monetary Theory (MMT), and so I wish to respond to those points.

I will start off by noting that he is probably correct in diagnosing that some online comments describing the details of monetary operations are not helpful for advancing the MMT case. In some contexts (as I note below), monetary operations matter -- but they can be a distraction for a great many substantive debates.

Think Of It As The Theory Of "Modern Money"

Saying that there is very little new about MMT is a standard mischaracterisation of MMT. The problem seems to be the use of the word "Modern" in the name; there are two explanations for it.
  1. This refers to a joke by Keynes regarding Chartalism: we have had "modern" government money for thousands of years.
  2. The wording is ambiguous, one could render it as "The Theory of Modern Money," where "modern money" is money that is not defined with respect to an external store of value. (Note the title of Randall Wray's book -- Understanding Modern Money -- which is consistent with this interpretation.) Since the major currencies were (implicitly or explicitly) pegged to gold up until the early 1970s, those of us a certain age can think of floating currencies as "modern." 
The main MMT authors state that MMT is a body of thought within the broad tent of post-Keynesian theory. One can argue that it is unfair to generalise about "mainstream economics" (Professor Wren-Lewis implicitly states this in his article, see below), as there are differences of opinion about policy within the "mainstream" tradition. The exact same thing is true for post-Keynesian economics, which is why MMT needed a label to distinguish itself from other post-Keynesians.

This is not a minor problem. MMT'ers and many other groups of post-Keynesians are seriously divided over the importance of the "external constraint," we need to have to have labels for these groups in order to make sense of such internal debates.

To what extent the MMT authors are developing new theory versus building on existing post-Keynesian thought is an interesting debate for academics, but is not of crucial importance for those of us who are not obsessed about citations in the literature.

Sigh. Mathematics.

Simon Wren-Lewis:
MMTers also seem curiously averse to equations.
This is touching upon a rather raw nerve for me. I do not like arguments from authority, and so I do not dwell too much upon my background. However, I will note that have an excellent academic pedigree in applied mathematics.

In my opinion, the post-Keynesian authors that I am most familiar with have a realistic attitude with regards to the application of mathematical models to economics, and the use of mathematics within papers meets the standards of applied mathematics (which is less formal than pure mathematics, but can be translated into a more formal version). I am much less impressed with the use of mathematics in the (mainstream) DSGE literature.

My writing here is informal, and I avoid the use of equations in such writing -- exactly as I was told by my supervisor (who had a training in pure mathematics). For example, I was able to give a basic understanding of how sampling destroys information with a graphical illustration (and no equations) in this article. It would have been easy to regurgitate pages of frequency domain mathematics to explain the concept (and it would be necessary for a formal treatment). However, I doubt that the mathematics would be helpful for anyone who does not already have a good grasp of such techniques. If I were writing an academic article, I might be forced to use more formal mathematics, but that would mean a much smaller potential readership.

Default Fears Not Just "Mediamacro"

This raises the question of why MMT seems to have quite a following. Perhaps it is a reaction to mediamacro’s often implicit assumption that a country like the UK or US could go bust through a forced default. And, to be fair, some mainstream economists seem to want to keep that misapprehension alive, while others take the existence of independent central banks as a binding constraint. 
I think he correctly pins down why MMT has a cult following amongst people in fixed income (like myself; Edward J. Delzio is another example, I reviewed his book The National Deb(i)t here) -- it explains why floating currency sovereigns do not default. Fixed income practitioners have seen people getting their faces ripped off for decades as they shorted the JGB market. In practice, MMT is the only visible school of thought that offers a coherent explanation.

To be fair to Professor Wren-Lewis, he's been largely correct on fiscal policy. (I am ambivalent towards his academic theories about fiscal policy, such as described in my summary of "optimal fiscal policy.") He was not worried about forced default. That just tells us that one should never make generalisations; not all "mainstream" economists agree with each other. That said, "mainstream" academic economists at universities are training almost all of the people who enter finance or supra-national agencies (which have produced innumerable reports predicting bankruptcy in various developed countries with floating currencies). If the majority of "mainstream" academics are on the right side of the question of sovereign default, there has been a massive failure of pedagogy. 

The Budget Constraint Is The Wrong Answer

For example MMTers seemed to think that they had discovered that a government with its own central bank need never default on its debt, but as far as I was concerned that was a standard and rather trivial implication of the government’s consolidated budget constraint. 
The standard governmental budget constraint is an equation that holds under the assumption that default cannot happen. If we allow for the possibility of default, the constraint no longer holds; we need to adjust the equation to allow for the loss of value on government bonds. It is straightforward that we cannot apply a mathematical model to a case that the model assumes is impossible.

To judge the odds of a floating currency default, we need to look at the institutional (operational) arrangements around sovereign borrowing. Which is exactly what MMT does, and the standard mathematical frameworks used in DSGE do not. (There are thousands of DSGE models out there; I would not be surprise if some researcher used one to make observations about government default. I have not seen such a model, but I have my doubts that they could get it to work. Sovereign default is a political judgement, not a rational decision.) Once we exclude the possibility of default, there is very little to distinguish MMT and the mathematics of DSGE models -- since both reflect the same underlying accounting reality. I have remarked upon that similarity in previous articles.

Central Bank Independence Debate Shows MMT Non-Empty

He wrote:
This seemed to lead to ideas that I thought were standard bits of macroeconomics

Professor Wren-Lewis had an article "Central Bank Independence and MMT" (published in September 2015). He notes that MMT proponents say that "central bank independence is a sham," whereas the mainstream is greatly excited about the importance of the concept.

Although he disagrees with the MMT view, the fact that he was having a debate about a rather esoteric theoretical topic tells us that MMT is not just following "standard bits of macroeconomics."

Concluding Remarks

I am not going to respond to his comments about how MMT thinks about money-financed expenditures. I would characterise the MMT position as not being massively different than how he might think about the issue, other than the terminology. The reality is that MMT questions a lot of hidden assumptions that are embedded in mainstream models. It would take some time to translate the MMT argument into terminology that he would be happy with. Although I believe that it can be done, I leave that task to someone else.

Although it would be nice for a dialogue between myself and some mainstream economists, I am not particularly hopeful that it will occur.  If I see anything useful that can be taken from mainstream theory, I will take advantage of it. Until such a time, I prefer to develop my own research programme within the broad post-Keynesian tradition.

See Also:

(c) Brian Romanchuk 2015


  1. To me, MMT agrees with the obvious: Government can print money if it desires. Agreed, government can make printing more complicated by borrowing from itself, but that is just good accounting.

    Economist seem to ignore the observable fact that printed money that is removed by taxation will (first) exist for a while and can be used to build GDP. There is a simple relationship between the amount of money printed and the potential GDP allowed. I will not repeat it here in the interest of brevity.

    To me, this relationship is justification for the MMT concept of money creation.

  2. Brian, can you describe what empirical evidence it would take to convince you that MMT and/or a post-Keyneisan framework in general should be abandoned?

    1. A more narrowly scoped version of my above question applies to this statement of yours in the post:

      "MMT'ers and many other groups of post-Keynesians are seriously divided over the importance of the "external constraint," we need to have to have labels for these groups in order to make sense of such internal debates."

      Is it possible to figure out a way to let the empirical evidence decide whether one, the other, or both are wrong rather than just branch the taxonomy further?

    2. A lot of post-Keynesian theory is empirical work. How do firms set prices? By contrast, mainstream macro assumes that supply and demand curves exist and have standard properties. That is, a lot of post-Keynesian economics is the result of empirical validation of theory.

      Could you invalidate SFC models? All you can do is show that certain types of behavioural relations break down. Those can be replaced by better ones. To "break" the models, you would need to invalidate accounting relationships. That's in Lovecraftian territory.

      As for the debate about the external constraint, I guess I would need to see some serious currency crises in the developed economies. The empirical lack of such crises is exactly why I am in the MMT camp; I do not fit into other post-Keynesian schools of thought mainly because of this one debate. At the same time, we only have the post-1970s history to work with, and so there is no statistical proof either way.

      This is not a question of whether we can predict what happens, rather what priority we put on risks when setting policy.

      I would largely ignore the "external constraint" when setting policy, but that does not mean that I believe that there cannot be a negative impact from the external sector. It means that I view that the odds of a problem are low, and that policy makers should focus on domestic concerns. That's a judgement that is going to be hard to "prove" one way or another.

    3. "Is it possible to figure out a way to let the empirical evidence decide whether one"

      The problem you have is that the empirical evidence is coming from a system that is running under policies dictated by the wrong model - essentially a form of monetarism.

      So the 'evidence' of mistakes is actually evidence that the current policies don't work. And we know that.

      So empirical data can only throw up mysteries that the current dogma cannot answer. It cannot shed much light on what a new set of policies will show.

      So when people start throwing around historical data about the 'external sector', then the response is always going to be "yeah, we know".

      Beyond that you're in the equivalent of Europe in the early 16th century and debating which new form of Christianity is going to replace the 'failing' Catholicism. A quick glance at history shows that the outcome of that sort of debate isn't a clean one.

    4. Neil, I'm not as interested in policy as in seeing if we can treat economics like a proper science. You have to crawl before you can walk. Policy based on such a science would come later. But of course we have policy now, and have had it in the past, because we HAVE to have some policy. This policy has (best case scenario) functioned as a series of natural experiments that may be useful for advancing the science. I'm NOT saying the mainstream has a lock on the science. I'm arguing that science IS the way we reliably figure out how reality works, and if we are to make any progress in economics, it should operate as a science does (independent of any policy prescriptions).

      One thing all scientists ask themselves about their theoretical frameworks and resulting models is "How would I know if I'm wrong?" If you can't answer that question, or your answer to the question is "I'd sit in an armchair and look for errors in my logic" ... without any reference to empirical data from the outside world, then I question whether or not you're doing science. Also, as John Handley and Jason Smith debate here on John's blog, there's a question of usefulness (Jason's comment in particular). Jason makes the point that a correct model (that you have no way to tell is or is not correct) may be rejected solely on the basis of being useless. For example, I can postulate an undetectable dragon lives in my garage, and I may be correct, but there's no way to know. It's a useless hypothesis: it doesn't really help me explain reality, which is what science is supposed to do.

      Part of that may be determining what it is reasonable TO know. For example, in thermodynamics it's not reasonable to expect to know the individual states (positions and velocities) of all the gas molecules in 1 mole of gas in a container. But it is reasonable to know two of the following: volume, pressure and temperature, and be able to predict the 3rd.

      So perhaps it's not reasonable to predict the onset of recessions... that doesn't mean there aren't other reasonable (and useful) things we can predict, even if just in a statistical sense.

      Evolution doesn't make many specific predictions about what new creatures will appear in the eons to come (at least I don't think it does), but it does predict what kind of fossils we'll find now: for example, it predicts we won't find any rabbit fossils in a pre-Cambrian layer. Those kind of predictions help advance the science, because they put the theory on the line: if we DO find lots of rabbit fossils in the pre-Cambrian, we will need to rethink things. But each time we check, and don't find any, we add a little more confidence that evolution is probably true.

    5. Brian,

      It bothers me considerably that you are only concerned about an external constraint in so far as it affects developed countries. Are you suggesting that MMT is an advanced-economy economic theory only? In which case, what economic theory is appropriate for the remaining two thirds of the world? If that is not your intention, then surely MMT DOES need to take external constraint into account.

    6. I am only familiar with developed economies, which is why *I* only talk about them.

      I believe that the equity markets are the buffer that stabilizes the developed currencies. The flows in practice are inherently stabilizing. Even if "everyone" hates a currency, liability-matching investors will go the other way. Yeah, the currencies tear around (CAD lost half its value versus USD over a couple of years) but with no visible impact on domestic policy. (Sure, fiscal conservatives use the currency to advance their views, but that's hardly surprising.)

      (My views on currencies might not be pure MMT doctrine, but in my view they are similar enough that I do not call it the "Romanchuk Theory Of The External Constraint", or something ridiculous like that.)

      What do we say about developing economies? They lack the equity markets, liability-driven investors to take advantage of this mechanism. They have an inherently difficult *real* problem. MMT says that the true constraints are real constraints, so there is no magic bullet to solve their problems. Various MMT academics have discussed the issues facing developing countries; my dismissive attitude does not necessarily reflect their scholarship.

      Certainly, developing countries should not borrow in foreign currencies, and they should discourage private sector borrowing. They probably have no choice but to have a certain "hard currency" reserve position. But after that, they should do their best to let their currencies float.

      They may lack the tax system to support a job Guarantee, but that's another real economy issue -- if you want to have a welfare state economy, you need an effective tax system.

      Look at what various nitwits at central banks did a few years ago. They jumped up and down saying that the Fed was causing a "currency war"', and grumble that they were going to create a new "reserve currency". That was an incitement to borrow in USD, which of course has since blown up. MMT'ers would have told them to shut up in the first place.

    7. Developing countries generally have a chicken and egg problem. They dont have a deep government bond market and their own populace does not necessarily consider their government bonds as the safest asset, especially in a crisis. So, fiscal policy is then constrained by capital flight or by inflation--even a country with massive reserves like China is not completely immune. How does one get to the point where there is a deep government bond market and people consider bonds as the safest asset--I don't know!

    8. Srini,

      The effectiveness of the tax system is probably a major issue. Taxation was minimal in the developed countries until they got ramped up for the world wars; the larger tax takes stayed in place after the wars ended. My impression is that this is one of the key policy problems in many countries.

      As for the bond market, you need a solid banking system in the local currency, and pools of capital like insurance and pension funds. Since pensions and insurance payouts are in local currency, it creates demand for bonds.

      In any event, it's a general question of how to best develop. No obvious answer, other than trying to emulate what other countries did. Not really my area of expertise.

  3. Brian,

    It's high time to rewrite a few neoclassical equations using noncommutative geometry, or some other highly esoteric mathematics that is completely unhelpful to economic analysis. Then we can sneer and dismiss them as ignorant oafs because of their lack of mathematical formalism. The geocentric Latin speakers snigger at the Copernican MMT revolutionaries for their table manners and diction. Progress, as they say, occurs one funeral at a time.

    1. I am an applied mathematician, so I am not the person to make mathematics more complicated. My DNA is to simplify the complex.

      One problem with DSGE math is the tendency for unjustified logical leaps. Frankly, simplification is what is more dangerous, as those leaps are more obvious.

    2. When I see Wren-Lewis these days I start to see Sir Thomas More. And we know what happened to him.

  4. I tend to agree: "monetary operations matter -- but they can be a distraction for a great many substantive debates"

    And here's a crazy thought: MMT + mainstream economics = a good conversation, perhaps even a constructive, eye opening one

    No need to limit oneself to one school of thought.

    1. I should be less negative about the mainstream. I tend to only note when I disagree, and not when I agree. The problem is that the mainstream views I agree with are so well known, that it does not seem particularly insightful to say that I agree with it. (For example, I am a "rate expectations fundamentalist," which is standard for mainstream modelling work.)

    2. I'll be honest and say that I think that is probably not going to be productive. The mainstream are political players and have shut out all other forms for a reason.

      I find this "let's have a big hug club" approach that you always get from the liberal side very amusing. It simply hasn't and doesn't work.

      People need an identity to get behind. Diluting that makes the whole thing politically inert.

    3. Neil - I agree with regards to academic politics. It's hard ball (as we North Americans say...), and I doubt that non-mainstream authors will ever be allowed into the "main" journals -- until political winds blow in another direction.

      Outside that, there is a need for coalition building. The Canadian prairie populist movements were somewhat like coalitions. They knew what particular policies they wanted, and they focussed on those objectives. It's not a method to build a ruling party, but you can achieve particular reforms which otherwise appear impossible to achieve.

    4. Neil: "I find this "let's have a big hug club" approach that you always get from the liberal side very amusing. It simply hasn't and doesn't work" The other way doesn't work either, and I'd say is even less productive.

      Brian: You are giving too much credit to the main journals. The large majority of the policy-oriented and policy-involved economists don't appear in those journals. I'd even argue that the journal crowd are not mainstream, but rather a very esoteric fringe that gets little or no attention from the more policy-inclined folks.

    5. Brian: Yes, your coalition view does seem to me an appropriate way of moving forward. I always thought that folks like Wray, Baker, Bernstein could probably agree on much policy proposals if ever they were to sit aside each other at a congressional committee hearing. Unfortunately, what we see instead is an insane amount of dispute about the operational realities, which make is seem that all these folks are farther apart (from a policy view) than they really are.

    6. Re: journals
      Yes, and there's a lot of non-macro economics -- which I am not interested in. I agree that it's unclear who pays really attention to mainstream DSGE economics, but at the same time, there's a clique of "serious" economists that go to things like the Jackson Hole shindig. I no longer care about the career advancement aspects of this dispute, but if I were a young academic, I would be quite unhappy. (Actually, I was an unhappy control systems academic.) If there are young academics reading this (and they want to stay in), my advice would be to specialise outside of macro, or in something technical like extracting information from options markets.

    7. "The other way doesn't work either, and I'd say is even less productive."

      Seems to have worked for the mainstream economists doesn't it.

      You need an identity to get behind. The mainstreamers found that in producing the supporting dogma for possessive individualism and corporatism.

    8. Given the wave of mud-slinging over the internets the last few days, I am moving into a "Can we all be friends?" mode. Mudslinging is somewhat entertaining to read (yay for page views), but it's unclear how constructive it is.

      As for Neil's points, I guess my weird political views (a 1970s Prarie Populist living in Quebec in 2016) makes me somewhat detached from these concerns. I live in place where the marginal income tax rate is 50% for the "middle class". I have a very different perspective on political economy than would someone living in the UK or US. (We have single payer - Nyah, Nyah! 🙂)

  5. I'm not averse to equations; I just think they should have logical operators and domain specifications attached to them.

    1. Generalisations are always wrong.

      Sure, equations have their place. But a lot of writing on the internet is informal, and it's generally a good idea to keep equations to a minimum in that environment. Some people can get away with it, but they usually have some other reason for people to read their work (they are a well known academic). A lot of the MMT writing falls into the informal category; the academic work is quite specialised, and are not useful as an introduction to MMT. ("What was the effect of job programme X in country Y?")

      I'll take a look at your article. I may have read it earlier, but I cannot tell just from the title.

  6. The fact of the matter is that if you handed Wray's book to a majority of Wren-Lewis' students the honest ones would say that it is fresh. The dishonest ones who stake the reputation on pretending to know everything about the economy (when in fact they know very little) would repeat Wren-Lewis' mantra.

    Of course Wren-Lewis is being wholly dishonest. Take a look at his policy positions for the UK> There is NO awareness of the need for permanent fiscal deficits in the case of structural current account deficits. So, no, he does not understand the theory and the theory says something fundamentally different to what he says about policy. The fact that Wren-Lewis cannot see this merely confirms that he is not well-read on the topic sufficient to engage.

    1. I have not read enough of Wren-Lewis' work to comment on that. The articles of his I have read are blogs about fiscal crises elsewhere (Greece,...), and I agreed with him. I looked at some of his academic work, and I have technical reservations (as seen in my linked article). I have not kept up with the domestic debate in the U.K. with regards to fiscal policy; my feeling is that I might agree with your assessment, but I do not have a strong feeling on the matter.

    2. Wren-Lewis is behind this:

      They don't promise surpluses. They want to split the capital and current account and run stimulus through the former while balancing the latter. Politically I think it's a stellar move. But I get no sense at all that Wren-Lewis is clued in on the implications of the CAD for the fiscal balance. That said, some prominent Post-Keynesians are working on this too.I had a Twitter exchange with Ann Pettifor a while back and she denied the sectoral balances pretty much altogether. It was weird.

      (It's Phil here by the way if you want to try to dig out the exchange...).

    3. The main difference of MMT is that it looks at things from a different point of view. Essentially it linearises the cycle from a different starting point. And that throws the light and darkness in a different pattern revealing things.

      As I said on Simon's blog I can give 5 million people a job and he can't. The response was hand-waving about Philips curves.

      You can't just dismiss the needs of 5 million people like that and expect to be taken seriously.

      The short twitter exchange between Simon and Pavlina/Scott showed that when pushed he just repeated mainstream dogma.

      The attempt here is the usual "nothing new here, move along" trick. The job is to constantly shine the light on those bits that mainstream want to sweep under the carpet - primarily the 5 million people that want work and can't get it.

  7. The real problem that the "mainstream" has, and I think you can see it in Krugman's recent wriggling, is that MMT *keeps being right* ..... My initial forays with the theory were for exactly that reason, the things it said would happen, happened. With time I expect that "being right" will be enough for its acceptance. Though I hope the timeframe is rather than shorter than it currently appears!

    1. From the perspective of people in policy positions (central banks, etc.), "mathiness" matters. They want a theory that they can use to crank out forecasts. Telling them that mathematical models are generally useless for forecasting (like I would when I am in a pessimistic mood) is not going to help your case.

      It takes time to build up policy credibility. Some of the MMT authors are doing precisely that, as well as some of the post-Keynesians. But it is unrealistic to expect immediate acceptance.

      On the academic side, the entire field of macro theory is dysfunctional. That is not surprising; most of academia is dysfunctional as well. I see little reason to be optimistic on that front.

    2. Srini: Alexander Hamilton is credited with the vision for a national banking system and a deep Treasuries market in the United States.

      Perry Mehrling promotes the view of a central bank as the dealer of last resort. Regardless of the economic models used the Fed knows its policies are transmitted to the markets via the Open Market Desk. During the financial crisis the Fed and Treasury teamed up to become a lender of last resort, a dealer of last resort, and an investor of last resort. The broker-dealer model of the financial economy applies also to the federal government in the United States. This model generates the complexity and unpredictability of the private sector and provides the clue to why the central bank and Treasury can be effective at fighting a credit-fueled inflation or fighting deflation in a financial market crisis. The DSGE models are largely exercises in mental masturbation.

  8. MMT lol,dismisses two thirds of of evidence because in reality he to lazy to observe the differences,in science you need a base level,in engineering a datum line but in MMT nothing,once the money is created what happens to it and when i say that i mean take a fire easy to create difficult to control and once out of control can act totally different,as for the maths always focusing on narratives & not the act & react ta narrative as on the rest of the equation,applying theory to one side of the equation & not the other,you must follow it through,governments can produce money but accounting them away isn't possible unless,Take Osborne 28% tax payments by those that lets say own 90% of the wealth,to other half of that equation is 10% of the wealth pays 72% of the tax,if prices are false and inevitable they are! the numbers are wrong the models are wrong because unless the models fulfil what is happening then they are wrong!you must search for the whole equation and you start that by first making sure that what you do to one side of the equation if you add to the others side balances it out,MMT doesn't it can but only if prices are perfectly aligned with true values because the difference between true values(the balanced position of everything within the sum) & it actual price is the damage it does to the economy and the inability of the model to rectify itself from the opposite of price tax & that where you treat the two so differently but you should not


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.