Recent Posts

Sunday, September 9, 2018

Jayadev/Mason Article On MMT

Arjun Dayadev and J.W. Mason recently published "Mainstream Macroeconomics and Modern Monetary Theory: What Really Divides Them?", which suggests that the gulf between Modern Monetary Theory (MMT) and mainstream macroeconomics is smaller than suggested, that MMT is much closer to orthodoxy than is normally portrayed.

I have been battling with installing a locking floating floor, which is an engineering task that I discovered that my doctorate in engineering provided little training. As a result, I have been somewhat distracted, and only quickly read the article.

I am unsure how the academic MMT community will view the paper; being portrayed as orthodox is perhaps not the way that they think of themselves. Furthermore, some of the theoretical points run into the thorny questions of academic originality. As an ex-academic, I understand concerns about originality and distinctiveness. Since I only browse the economic academic literature looking for useful tidbits, I cannot comment on the academic originality of MMT.

From my decidedly ex-academic perch, I am open to the idea that the gulf between the mainstream and MMT in policy terms is more a question of political economy and terminology. If mainstream economists actually paid attention to the mathematics that they say is so important, a lot of the distinctions with MMT would disappear. Essentially, they have built qualitative folklore around not particularly useful mathematics. Modern Monetary Theory is battling with the folklore, not the mathematics.

I just want to comment briefly on a few points that I saw some disagreements. (These are the points that stuck out to me when I was sprawled on top of a pile of building materials, and are not necessarily the most important parts of the paper.)


They state the following:

Output below this level implies unacceptably high unemployment and perhaps deflation; output above this level implies unacceptably high and/or rising inflation. This assumption can be represented as a Phillips curve, the same general form of which is used by MMT as in conventional textbook presentations. A corollary is that policy affects inflation only via the level of output.
From what I have seen, the MMT story on inflation is more complex than that text suggests. The "nominal price level anchor" provided by the Job Guarantee wage is constantly emphasised. My interpretation of the Functional Finance part of MMT as suggesting that "large enough" fiscal loosening leads to inflation -- but until we hit that "large enough" level, inflation may do whatever it does. The psychological anchoring of wages relative to a Job Guarantee wage may be far more effective tool for inflation stability than changing the overnight rate on risk-free collateral.

If the MMT inflation story was as simple as just a Phillips curve, why don't they say that themselves?

Aggregate Demand Management

The article was premised entirely on aggregate demand management. And MMT economists have discussed such aggregate management with respect to current events. However, the emphasis in MMT is on using a decentralised automatic stabiliser -- the Job Guarantee. The key advantage of the Job Guarantee is that it is spatially targeted at the areas of greater need.

The Debt Ratio (My Perspective...)

We noted above that while MMT advocates would probably agree that the debt ratio should not rise without limit, in general, they do not see the debt ratio as an important target for policy. 
I cannot speak for MMT academics, but I am far more worried about a Martian invasion than a country's debt ratio rising without limit.

If we start from a stock-flow consistent modelling perspective, we assume that the stock of private sector wealth enters into consumption decisions. As the government debt ratio rises, by definition, private sector wealth is rising. Eventually, the drawdown from saving will dwarf spending out of income if it is not otherwise budging. The increased spending creates a circular flow of income, driving nominal GDP above expectations. Since the duration of government debt is not zero, the debt-to-GDP ratio will fall -- as exactly happened in the post-war uplift in inflation.

As always, inflation is the limit for fiscal policy.

This allegedly does not happen in mainstream DSGE models, mainly because the mainstream economists make no real effort to solve the models. If they actually did the math, they would probably see the same effect.

(c) Brian Romanchuk 2018


  1. I think I will put my own spin on your closing comments. I am unsure of where this may go.

    "As the government debt ratio rises, by definition, private sector wealth is rising. "

    I agree. Government is spending to purchase labor and materials. Resources are consumed, used for the purposes of government.

    "Eventually, the drawdown from saving will dwarf spending out of income if it is not otherwise budging."

    I did not understand this claim. Labor is a renewable resource. Resources usually have limits that are not 'savings' referenced.

    "The increased spending creates a circular flow of income, driving nominal GDP above expectations."

    I don't think of debt funded spending as creating a circular flow. Instead, I think of an incremental one-time increase that is limited by cash withdrawals into savings mechanisms such as repetitive government annual borrowing.

    Japan seems to be an example of increasing governmental indebtedness with little inflation. I think you are suggesting that the Japanese government needs a more aggressive fiscal policy to cause inflation.

    1. "Eventually, the drawdown from saving will dwarf spending out of income if it is not otherwise budging."

      Means that the saving rate will decline and nominal consumption will be higher due to the wealth effect?

    2. Jussi - Nominal income/spending is higher due to the wealth effect, which lowers the ratio by raising the denominator. It might be possible to play games with the saving ratio; you could have lots of new saving, but the existing stock of wealth is “inflated away” relative to nominal income.

      Roger - This is in nominal terms. Try holding nominal wage income at $100 per person annual if they each have $1,000,000 in savings.

    3. Ah, okay. I thought that the saving rate would be also affected as consumption is someone's income (the other denominator). Of course that doesn't need to happen, people can save more along the higher incomes. I guess normally the saving rate is affected though.

  2. Also, as we explicitly say in the piece, by mainstream we mean the kind of models used in upper-level undergraduate courses and by policy makers, not DSGE.

    1. I missed that bit when I was battling the flooring... Sorry.

      I tucked in the references to DSGE because that’s my pet peeve, and I don’t see that big a distinction between formal DSGE and the undergrad econ textbook stuff in practice at the policymaking level. They fudge the DSGE models to get the same results as undergrad econ. That’s what I was getting at with my comments about “folklore”: the math is just a distraction along the way to the desired analytical result, which is based on folkloric policy beliefs.

  3. This comment has been removed by the author.

  4. Hi Brian, thanks for the comments.

    I'm a little puzzled by your criticism of our claim that functional finance is based on a Phillips curve type relationship between output and inflation, and your comment toward the end that "inflation is the limit for fiscal policy" which is simply a restatement of the Phillips curve.

    Of course the curve may have a flattish segment where the relationship is weak - nobody disputes that. The salient point is that you know output is "too high" when inflation starts rising, and this is shared by both sides.

    1. My point was fairly brief, but the point I was trying to make that in the flattish bit of the curve, the anchoring provided by the JG wage matters (according to MMTers). That distinguishes the MMT inflation theory from the mainstream (for that regime), even if both sides agree on the directional implications of very large demand impulses.

    2. Well, either it is true that:

      1) Large changes in inflationa re driven by the output gap, just as the textbooks say. Where the output gap is close to zero (so not much variation in inflation is to be expected in any case) a JG or similar government purchases at fixed price will provide some additional stability.

      2) Even where the output gap is large, the price-stabilizing effects of a JG (or price-destabilizing effect of a Universal Basic Income) will outweigh it. I.e. inflation is primarily a micro rather than a macro level phenomenon.

      If it's 1, then I think our summary was entirely reasonable for a big picture account that's abstracting away from secondary details. If it's 2, then I will concede our summary was not accurate. But I think adopting 2, if you're consistent about it, is going to cause you a lot of headaches!

    3. I am personally agnostic on inflation theories. My argument here remains that the summary is missing the emphasis on the JG wage. Warren Mosler certainly emphasises the price setting role of government purchases. That might directionally end up in the same direction as a Phillips Curve, but a lot of theoretical details seem to be missing from my vantage point.

      I’m certainly not the person to give a definitive answer to whether that is a fair summary; I am frankly waiting for the Mitchell/Wray textbook for the MMT “party line on inflation.”

    4. Well again, you do actually have to take a view on the relative magnitudes of the JG and output-gap effects on the price level. If the latter dominates -- which is definitely the view taken by Stepahnie Kelton, Randy Wray, etc. in much of their publisehd writing -- then I don't see anything wrong with abstracting from the former. If the former dominates, then you're now a monetarist.

      Actually I suppose that if the JG wage is fixed in nominal terms and you have a large positive output gap, the real value of the JG wage will erode to the point where no one takes a JG job. So perhaps the effect is asymmetric - JG sets a floor on inflation but not a ceiling.

    5. The output gao has been all over the place since 2008, but inflation rates hardly budged (ex-oil effects).

      I’m not the go-to guy for the MMT party line on inflation. That said, the JG has been constantly sold as a way to achieve full employment *and* price stability. Someone else will have to cover the details; I am waiting for the textbook to get the view from there.

    6. "Well again, you do actually have to take a view on the relative magnitudes of the JG and output-gap effects on the price level."

      Why? Why can't you hold that no one is sure exactly what causes inflation in every case but that the waste and costs that involuntary unemployment causes are very evident and large? That one problem is larger than the other and that we know how to alleviate it and that it may not be connected at all with the other in some circumstances anyways?

  5. Flooring is tough on the knees, hope yours aren't too sore. And those interlocking pieces can be trickier than they look.

    Seems pretty clear at this point that Bill Mitchell is not at all pleased to be portrayed as anything close to an orthodox economist.

    1. I just managed to read the first paragraph of the Mitchell comment (after I wrote my piece), and yes, “pleased” seems like a bad description...

      It took awhile, but I got a hang for the flooring. I put an underpad, and that really needed to be tight. I initially was trying to pin it on both sides, but realised that if I just pin down the side facing the existing flooring only, it was easier to tighten as I added rows.

      Although a couple of pieces must have had defects. I needed to lift one joint three quarters of an inch to allow the pieces to get into initial alignment. That took me about 45 minutes (and a beer) to figure out...

    2. Thank goodness there was a beer handy.

    3. It was our new extension, and our second fridge (a.k.a. “the beer fridge”) was literally sitting on the floor I was installing.

      (Having a second fridge is pretty standard in Canada, but I think causes gasps elsewhere...)

  6. JW Mason's skillful unification of NK and MMT arguments within a common theoretical model, which can show how that one model can generate both results, sits squarely in the post-Keynesian tradition of rapprochement initiated by the General theory. But I fear the ghost of Kalecki's famous 1943 article continues to haunt this particular machine. The Phillips curve was an attempt to smuggle 'social contest' back into the neo-classical apparatus, where all such questions are assumed to have been settled in advance at the point of equal exchange. For its part Functional Finance could also assume away that contest, because it was developed in the context of a Neurath type total war target economy, where the contest is subsumed in a broad collective effort. The Phillips curve thus represented an uneasy peace between the two sides, a kind of phoney war, which is now over. Building a consensus around it may now be too late in the day. If the curve still exists, its harder to see it in a world where labour market institutions have been so comprehensively flattened and the old dual model of unemployment / employment has been replaced by a complex spectrum of varying rates of participation, precarity and under-employment. It seems to me that the Bhaduri Marglin (1990) tradition is still the best game in town within academia, with MMT acting as an essential political disruptor to help break down what Brian calls economic folklore among the wider non-academic population.

    1. I came into this with the advantage of no economics education. I am not too concerned about the exact relationship between demand and inflation: sure, it is complicated. But it’s pretty clear that basic nearly linear models will not work for the environment we have now.

      I am not familiar with Marglin, probably should look it up. I am eventually doing something on inflation theories, but I am now concentrating on the business cycle side.

    2. Any realtionship can be locally approximated with a linear equation. And local approximations are all we can generally hope for, anyway. Trying to find the true model of inflation is a fool's errand.

    3. But the point is that the linear relationship is picking up a historically contingent set of social relationships. No relationship - no linear approximation.

  7. Note that Bill Mitchell has written a 3 part series on the J&M paper.

  8. "If the private sector is inflating, a tightening of fiscal and/or monetary policy shifts workers into the fixed-wage JG-sector to achieve inflation stability without unemployment." - Bill Mitchell

    1. Getting back to JW Mason's comments, I guess if we stretch definitions enough, that sort of looks like a generalised "Phillips Curve" argument, albeit we have to use the output gap, and not the unemployment rate that Phillips actually used. My issue with saying that this is "just the (generalised) Phillips Curve" is the importance of the fixed wage component of the JG. Mainstream theory (at all levels) insists that wages *should* be flexible, and assumes that everything is happening at a marginal rate. A pool of labour at a fixed wage does not fit into that framework very well.

    2. Mainstream theory could use the Treynor model to view the JG (or Basic income) as an outside spread floor to the private labour market at the inside spread, just like the old corridor system for targeting the overnight rate for central bank reserves. But that would be letting the cat out of the bag.

    3. The cat being that in order to exist 'markets' must be administered by institutions.

  9. Cat == markets
    Bag == Institutions

    MMT puts a price floor under wages with the Job Guarantee to curb deflation and it should extensively research the tax policies which would tend to automatically curb inflation as described in outline by Hyman Minsky. How those policies compare to buffer stock of unemployed workers to curb inflation and monetary policy to curb inflation (other then what I just said) is for the macro economists to debate in their games inside a game.


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.