Recent Posts

Sunday, January 22, 2017

Postscript To Misunderestimating MMT

Gerard MacDonell has a follow up article to his earlier critique of MMT (my previous article was my response to his critique). I just want to expand on a couple of points, and respond to his response...

Policy Space and MMT

Gerard wrote:
I would invert the second part to say that MMT does not open up any obvious policy space, even though it does change the analytics.
Although I cannot say this is wrong, it does depend upon what you think about as "policy space." My comments in the original version of my article perhaps undersold MMT.

I will note that some people have interpreted some MMT arguments as saying that we can abolish taxation outright (which would obviously open up policy space). This interpretation is incorrect. I cannot remember seeing text written by one of the MMT academics that I interpreted that way; I only saw this interpretation on the internet. There is no value in discussing incorrect interpretations of a theory.

However, there are two policy proposals by MMT, which could be interpreted as "opening up policy space," although it is possible to argue that they can be understood using conventional analysis. This is a semantic point I do not care to discuss; what matters are the merits of the policy proposals.

  1. The elimination of bond finance. The central government just issues base money (see technical appendix below). Interest rates are locked at 0% forever. One might argue that this step eliminates policy space, as monetary policy no longer exists, but that is the objective (the mainstream focus on monetary policy is part of the problem, not the solution). With interest rates locked at zero, arguments that fiscal policy is "unsustainable" themselves are (ahem) unsustainable. I discussed this proposal in the original article.
  2. The Job Guarantee. This is the signature MMT proposal. The proposal relies on MMT analysis of the labour market. I did not discuss this in my original article, but I would argue that it is a better policy than the existing framework, and so would open up "policy space" (since it is not being considered by the mainstream).
Gerard did not discuss the Job Guarantee, and I argued that his attempt to deal with the elimination of bond finance was unsatisfactory. (In fact, he was originally unsure whether that was the proposal.) As a result, he is not going to see a lot new in MMT policy proposals -- since he was unaware of (or ignored) the ones that existed.

I will note that there are a lot of other economists (pretty much the entire mainstream) that would agree with Gerard's view that we need interest rate policy to contain inflation; that said, a lot of people used to believe in witches. Whether or not an economy can reasonably function with a permanent ZIRP under all circumstances is an interesting open economic debate.

However, my discussion so far about policy space undersells the policy relevance. Not being utterly wrong adds a lot of room to policy space. 
  • The euro area would never have been created without a strong central fiscal agency if European policy makers followed MMT principles. This would have saved a spectacular amount of misery for those unfortunates stuck in the euro area periphery.
  • The entire disastrous austerity drive after the Financial Crisis would not have happened.
  • People would not have wasted time looking for magical debt-to-GDP ratios.
  • We would see that the drive to create fiscal watchdogs as just being another attempt by unelected technocrats to grab power from elected politicians, on the basis of mainstream economic mumbo-jumbo.
  • We would be spared the idiots calling for imminent hyperinflation in Japan. (Although I do find they provide a great deal of entertainment value.)
  • The idiocy around the debt ceiling in the United States could be avoided.
  • The big idea of DeLong and Summers -- that we should invest in infrastructure because interest rates are low -- would be seen as silly. Instead, the debate would be: does the United States need infrastructure investment, and are the real resources available to make the investment possible?
  • The entire premise of hydraulic Keynesianism -- we just need "aggregate demand" to deal with unemployment -- would be seen as questionable. We instead need to look at the real side of the economy, and ask whether the demand created will match the available pool of unemployed/underemployed labour. (This was Minsky's critique in the 1960s, and explains why the Job Guarantee is a lot more sophisticated than brute force "Keynesian" approaches like brute force infrastructure spending.)
In summary, we would have been spared practically all of the verbiage dumped by mainstream economists on fiscal policy over the past decade. This is an advantage that cannot be minimised.

Permanent ZIRP

Gerard had a section "My post wasted time considering the case of the Fed paying interest on reserves...."

He wrote:
 It would be so if MMT advocates were never to offer US fiscal policy advice, except in the context of being crystal clear that they assume that the institutional set-up at the Fed is entirely different from what it actually is.
This is wrong. The MMT policy proposal is to abolish bond finance; but it can analyse the current system. The MMT analysis is that Gerard's first case (paying interest on reserves) is exactly the same as QE, and MMT says that QE is completely ineffectual since it is functionally equivalent to bill finance.* Since MMT says the policy is ineffective, an analysis that tells us it would not work is useless as a critique of MMT.

I believe that Gerard argues that QE is ineffective, because it is equivalent to bill finance. Since his analysis is exactly the same as MMT's, I think his complaint here is somewhat silly.

More generally, MMT consists of analysis and policy proposals. Just because MMT prefers floating currency regimes, it does not mean that you cannot use MMT principles to analyse a fixed exchange rate system. The points of failure of a fixed exchange rate system are exactly why MMT advocates a floating rate system.

Saying that economists should not give policy advice if the real world does not conform to their theory's assumptions is pretty dangerous for someone who thinks Paul Krugman is worth reading. I would stack up MMT's assumptions versus mainstream assumptions any day of the week.

At this point, I need to emphasise that I am including broader post-Keynesian analysis as being "MMT" -- which is exactly what the MMT academics do. Modern Monetary Theory inherits analysis (such as Stock-Flow Consistent models) which were developed by post-Keynesians who are not considered part of the MMT school of thought. MMT is an attempt to create a single cohesive world view out of a broader, somewhat self-contradictory post-Keynesian literature. It has a narrow focus, in order to keep everyone on the same theoretical page, but we cannot lose sight of the broader literature if we are discussing topics outside of that focus.

Arguments about the MMT Canon

The bulk of Gerard's article revolved around his understanding of the MMT literature. The obvious solution: he should buy Understanding Government Finance (conveniently available in ebook and paperback forms!).

My less flippant answer is that we need to focus on more concrete substantive examples, and not worry about terminology. I have attempted to deal with what I see as Gerard's concrete objections, but I am not attempting to address his complaints about how MMT puts itself forward.

I doubt that Gerard wants to read hundreds of academic primers on MMT, but at the same time, criticising a theory without actually reading the literature seems like a strange exercise. My solution to this is to discuss how the MMT analysis differs from the mainstream with regards to concrete issues, and then let the reader find the primers if interested.

Appendix: Eliminating Bond Finance

One possible point of confusion about the MMT proposal to eliminate bond finance revolves around Treasury bills. The people behind the development of MMT are very well aware of the institutional framework for fixed income (Warren Mosler made a reasonable amount of money by setting up a fixed income hedge fund.) This institutional awareness means that the proposal is slightly more complex than what would be suggested by the simple version used in an economic model.

The simple version of the proposal is that the government just issues "base money," with an interest rate of 0%.

In the real world, this policy would need to be supplemented by the issuance of Treasury bills. This is necessary as there are non-banks that need to hold safe assets on their portfolios, and they cannot realistically stockpile currency, nor leave deposits at the central bank.

The difference is the issuance scheme: the central government issues Treasury bills at a fixed price (I believe that a yield of 0.25% is the suggestion), and the private sector can buy as many bills as it wishes at that fixed yield. As a monopolist, the government can issue Treasury bills by allocating on a price or quantity basis; the hidden assumption that the allocation has to be done on a price basis is wrong.

As a result, one might technically need to say that interest rates would be locked at 0.25%, and not 0%, forever.


* Why does MMT say that increasing excess reserves is the same as bill finance? The preferred MMT analytical approach is to consolidate the central bank with the rest of the government (regardless of the institutional framework; the only exception might be if you wanted to model a sovereign default). Once you consolidate the central bank, we see that:

  • excess reserves are a government liability that pays the short-term rate; and
  • Treasury bills are a government liability that pays the short-term rate.
In other words, they are functionally equivalent, and so lead to equivalent outcomes.

(c) Brian Romanchuk 2017


  1. "Gerard did not discuss the Job Guarantee"

    Then he did not discuss MMT. The Job Guarantee is what links the nominal to the real in the MMT analysis. It's the anchor system that keeps things in check and what makes ZIRP viable.

    It's sort of interesting that economists can't see the difference between killing demand and business activity with ex-employees left doing nothing, and killing demand and business activity with ex-employees left doing something.

    And that is likely down to having a one dimensional view of value.

  2. "This is necessary as there are non-banks that need to hold safe assets on their portfolios"

    Or you guarantee bank deposits fully - which has the same effect.

  3. Great Work Brian.

    Id just like to add that I think the last part of your post here is the most important part, IOW, the last part is the key MMT insight that overturns the whole logical artifice of mainstream economics wrt Govt operations.

    "excess reserves are a government liability that pays the short-term rate; and
    Treasury bills are a government liability that pays the short-term rate.
    In other words, they are functionally equivalent, and so lead to equivalent outcomes."

    Once you accept as obvious and true that Govt financial Liabilities come in essentially two forms issued by two separate branches of Govt (Reserves and TSY CDs). All sorts of confusion go away.

    Why is one type of Govt IOU "debt" and dangerous while the other type is just "money". Contract term? Then why is a 12mo CD at Chase "money" and a 12-mo CD at the Fed "debt"? Debt is interest bearing and money is not? Well then in 2009 when Congress gave the Fed the authority to pay IOR all of a sudden all that reserve "money" become "debt". What kind of sense does that Make? demand deposits at commerical banks pay interest, so are those not money ?

    If you just consider all Govt IOUs as money then all the QE confusion goes away, all the Govt debt confusion goes away. Things become much more manageable.

  4. I'm glad you did a follow up and a very good one at that. Thanks.

    I thought that MMTers being "utterly humorless" might be Mr. MacDonell's best critique. I have also wondered about that myself. But "utterly" is an overstatement of the case as is demonstrated even in your comments on the last post by Detroit Dan. And just so he knows, I did enjoy his Woody Allen clip. And you probably did too since you thoughtfully re-formatted the link to it.

    But I also liked "This claim is like MMT itself: perhaps true strictly speaking [Thanks!], but also very misleading." [no Thanks]. But if it is true strictly speaking it is difficult to see how it is also misleading. I find MMT fairly easy to understand and far more logical in its structure than what I know of mainstream economics. (Which is a whole lot more than your average person does.)

    That MMT has such a logical structure should mean that critics should have an easy time of dispatching with the theory- just show a few of the premises are wrong and you could topple the whole thing. But no one has managed to do that so far.

    On the other hand we got mainstream economists. You show that the loanable funds theory is wrong, for instance, but they don't let that stop them from basing all sorts of arguments and models on it. Its ridiculous at times.

    Finally, Gerard should be reassured that yes Bill Mitchell is "of the cloth" as far as MMT goes.

  5. Critics of MMT apparently resist being taken to the non-government tin shed as shown in the diagram under the link below re Deficit Spending 101 part 3 on Bill Mitchel's blog:

    In the United States the financial assets insured by the Fed/Treasury battery extend to liabilities of depository institutions covered by federal deposit insurance and to the insured liabilities of other government sponsored enterprises (GSEs). The executives at these banks and GSEs are given perverse incentives to grow firm balance sheets, to take out salary, bonus, and stock incentives, while the government sponsorship keeps cash flowing in the financial system. Therefore extending a government guarantee to bank and nonbank GSEs can create significant moral hazard if the regulators are unwilling to impose strict regulations on asset portfolios and unwilling to place strict limits on payouts to financial firm executives subject to clawback provisions.

  6. I thought we left the golden age of MMT critiques years ago.

    1. JKH, some like me are still learning. And some just like to argue, err discuss (might also be me). I remember reading some of your critiques of MMT some years ago and remember them as very accurate in the portrayal of MMT ideas, at least as I understood them at the time. But I never understood what the significance of your differences with MMT theory amounted to. It always seemed to me that you seemed to agree with most of it, and the importance of the things where you had a disagreement was never apparent to me. Maybe I should track some of them down to re-read now that I know a little more about the subject.

    2. JKH, I actually did recently ask Bill Mitchell about one of the areas where I think you had a disagreement with MMT. And got a nice reply which I am still working on understanding.

      That is at

      I believe I may have mentioned your initials there. In case you were interested. And of course any help with improving my understanding would be appreciated.

    3. Jerry,

      I should have put a smile emoticon after my comment. It was intended that way.

      Bill Mitchell likes to refer to “the top end of town” when it comes to income distribution. I like to think of Marc Lavoie as the “the top end of town” when it comes to critiques of MMT. Steve Waldman has been insightful as well. Google either of those along with ‘MMT’ and you should find their critiques. That was a few years ago. That’s the “golden age” I refer to. No disrespect to what is ongoing now.

      My own problem with MMT had mostly to do with its exposition style. I think this is important, because so much of the entire exercise has to do with effective explanation so that people can learn in the most satisfying and non-frustrating way. I think MMT has a great start in seizing the accounting bull by the horns and riding it aggressively. Accounting is a logical subset of economics. Mainstream economics simply doesn’t get that. It’s a profound epistemological failing in my view.

      I’ve had a couple of points of objection to what used to be the MMT exposition style at least. The first had to do with the meaning of ‘S’ in referring to saving. The idea of a closed economy interpretation or even a global interpretation is a derivative. It’s not the right starting point. The starting point is the national income identity:

      C + I + G + X = C + S + T + M

      Everything flows from this. The MMT ‘sector balances’ flows from this. ‘Net financial assets’ flows form this. The chopped closed and global economy versions flow from this. This is the starting point for what S is.

      Manipulate it anyway you want, and you come to an interpretation of S as private sector saving. The problem I had with MMT going back was their persistent presentation of a concept of ‘net saving’ which they then elided quickly into ‘saving’. This is expositional suicide bombing. In the former, they were referring to private sector NFA flow, or the private sector financial balance flow, which derived properly is:

      (S – I) = (G – T) + (X – M)

      That left hand item is obviously not the same as private sector saving. This equation says that private sector net financial saving ‘finances’ (so to speak, without causality interpretation) the government deficit and the capital account deficit (adjusted a bit for some international income components, but …).


    4. ...

      But MMT would elide the two. A number of us interpreted that as obscuring the importance of I in the calculation of private sector saving. That’s ideological if that’s the case, which seems consistent with MMT’s liking for government deficits as a source of ‘net financial assets’ and therefore as a marginal contribution to private sector saving.

      To which I responded:

      S = I + (S – I)

      That was to make a point of pure algebraic decomposition, illustrating the two components of private sector saving, one of which MMT tended to omit in its language style of exposition.

      To which one Randall Wray responded with a post of nothing but juvenile venom and drivel, along with an apparent complete ignorance of the usefulness of Boolean algebra in explaining the logic of the point being made, which was one of decomposition of private sector saving into its two components, only one of which was net financial saving.

      My other problem with MMT similarly had to do with exposition. To cut to the chase, MMT explained a view of ‘consolidation’ of Treasury and the central bank as if the manager of the System Open Market Account at the Fed was the guy who also cut social security checks and other expenditures in the government budget. One need just check out the expositions of the thing at that time and think about what was being said to see this. To which I responded that understanding the financial effect of the consolidated operations of the Fed and the Treasury is not the same thing as describing a hypothetical consolidation of these two functions at the operational level. This is a point they never got, deflecting the thing into the aspect of OMO around bond issuance and that sort of thing, in order to defend their presentation. But that aspect is left stage to the larger point about expositional confusion.

      Having said all that, I’ve always thought MMT started out on the right track by recognizing the importance of accounting as a foundation for understanding economics. Looking back, I would now agree with words Warren Mosler once used to describe mainstream economics – a ‘disgrace’. I just didn’t like the way MMT unnecessarily mangled stuff after that in explaining their core approach to government deficits.

    5. JKH is discussing saving S (verb) and investment I (verb) which are economic flows associated with estimates of Gross Domestic Product (GDP). I believe these concepts are defined in the production and income accounts in the system of macroeconomic accounts.

      System of Macroeconomic Account Statistics Overview (117 pages)

      Table 1 on nominal page 20 shows a set of vertical accounts linked to a set of horizontal accounts. Production and income, in the vertical path, are efforts to record economic flows, and balance sheets are adjusted via the capital account (non-financial assets) and the financial account (financial assets and liabilities) to link GDP to changes in stocks or levels of capital and financial instruments in the economy.

      The MMT argument is that if the federal Treasury spends G and taxes T such that G-T > 0 then the domestic and foreign sectors record a net increase in financial assets equal to the federal deficit. But if G-T < 0 then the domestic and foreign sectors record a net decrease in financial assets equal to the surplus. The federal government does not have a savings account and the central bank which purchases and sells only Treasury securities does not alter the net financial position either.

      Note carefully the box called "other economic flows" in Table 1 of the reference above. This box represents balance sheet adjustments associated with a casualty loss or capital gain or capital loss. When doing stock-flow analysis I would assume, without an in-depth study, that federal deficit or surplus appears to be directly coupled to the changes in the capital and financial accounts, however the debt to GDP is not at all coupled over time due to entries in the box called "other economic flows."

      To understand the boom-bust cycles in capitalism, and the reason that federal debt is not directly coupled to changes in GDP as Brian argues, one needs to investigate the prevailing customs that generate accounting conventions in the asset and liabilities re-valuation accounts. These re-valuation conventions are a feedback into economic decisions by agents in the economy and cause federal debt to GDP to decouple over time.

    6. Thank you Joe Leote. I will look up that IMF book, but am late for work now.

  7. Ah yes, the use of emoticons is essential when dealing with the humor-challenged in the econ tribe. :) You should have read my first comment on this thread. :) Abundant use of smiley faces can often save oneself from biting replies that can lead one to wonder too much about the nature of humanity, or at least if the members of the econ tribe belong there. :)

    Seriously, thanks for the response JKH. I think I remember asking you to go ahead and just shoot me if I ever questioned you about S=I. Thanks for not doing that this time.:) I will go re-read the Glasner posts (from 2015?) and your comments (which is what my question to Bill Mitchell was based on) and attempt to come up with a better understanding. I will also look up the Lavoie piece again. I am not sure I am up to reading through the entirety of the Waldman piece and all the comments again, but we will see.


    1. JKH, I had not seen your mega comment from 4/24/15 on David Glasner's Uneasy Money (the one that began with your initials) dated 4/14/15 at the time. That comment has cleared up the answer provided to my question from Bill Mitchell for me at least. Thanks.

      Stripped down 2 sector model Income = Expenditure
      E=Y and by derivation, I=S

      On the other hand (S-I)=(G-T)+(X-M) refers to S and I in an economy with a government and where foreign trade is possible. So it is not inconsistent to say I=S in some cases but maintain that the sectoral balances equation is not the same as saying 0=(G-T)+(X-M) in all cases either. Like I said, I am still learning :)

    2. A way of framing this for a realistic, multi-sector economy is to ask what the sources of saving are for domestic investment.

      That becomes:


      = private sector saving + government surplus + capital account surplus

      = S*

      = S + government surplus + capital account surplus

      (S* thus defined; and consistent with national accounts usage of S)

    3. So S* = S = I when you assume a closed economy (or equivalently a balanced capital account) along with a balanced government budget

      ... something like that

  8. The best exposition of so-called modern monetary theory that I have ever read is contained in the nine page article under the link below:

    Michael Kalecki - Political Aspects of Full Employment

    I believe the older term for this type of reasoning is called functional finance. Section I describes how full employment without inflation was 'financed' by the federal government issue of Treasury securities to purchase goods and services with wartime price controls to avoid excessive inflation. This "corner case" can be used to study how the system operates in peace time and to propose a full employment or job guaranty scheme with financial stability via federal government.

    Footnote 2 reads in part: "Another problem of a more technical nature is that of the national debt. If full employment is maintained by government spending financed by borrowing, the national debt will
    continuously increase. This need not, however, involve any disturbances in output and employment, if interest on the debt is financed by an annual capital tax. ... "

  9. The above article misses a trick I think, namely a basic logical flaw in interest rate adjustments. See:


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.