Monday, May 9, 2011

Monetization of debt: what does it do? Krugman and Rognlie on MMT

Again in a previous post I suggested I would deal with the issue, which seems to be apropos, since there has been a certain discussion in the blogosphere about the so-called Modern Monetary Theory (MMT) approach (see here, here and here). First, I should clarify that the discussion to which I refer tends to conflate two different issues. One is the question that I will deal here, what are the constraints faced by the government in managing its budget, and the approach associated with MMT in this case is basically what used to be called functional finance, a tradition that harks back to Abba Lerner and Evsey Domar, and to which Keynes eventually agreed.

The other issue is related to the causality between money and economic activity, and is part of what in more modern times has been called endogenous money. The debates on this issue are older than the Bullionist/Anti-bullionist and Currency/Banking schools, and in modern times the endogenous money (anti-bullionist-banking) view was developed by Nicholas Kaldor and Basil Moore among other names. The reason the two issues tend to be conflated is that printing money is one way to finance government spending, and there is a traditional connection between sound finance and monetarist (exogenous money) views.

First, it is clear that if debt is denominated in domestic currency default per se is NOT possible, simply because the government can always monetize the debt. So the question is not whether the government has a budget constraint (what’s the meaning of a constraint if you can print money anyway), as Matt Rognlie says (he gets worked up by this, and affirms that: “MMT is wrong on money … The government does have a budget constraint”). Unless he denies monetization of debt is always possible (there might be political problems, but it is technically possible) in domestic currency, he needs to explain what is the constraint. Otherwise the question is really related to the consequences of monetization. By the way, that is the same confusion made by Krugman. He says:
“As I understand the MMT position, it is that the only thing we need to consider is whether the deficit creates excess demand to such an extent to be inflationary. The perceived future solvency of the government is not an issue.”
Again, what could be the solvency issue if the debt is denominated in a currency that the government controls? He may have issues with the consequences of printing money, but not with the fact that money can be printed. So solvency is NOT, and cannot be an issue. What are the consequences of monetizing debt then?

Long ago the authors of the Banking school (e.g. Thomas Tooke) noted that if more money is pumped into the system agents would spend the money or pay their debts (the second case is known as the reflux mechanism). If they pay previous debt, the money has no effect on the level of activity. On the other hand, if they spend and firms have extra capacity output must increase. Note that firms normally have extra capacity, and can produce more at the same price, contrary to the textbook (U-shaped) cost curves.

But what if the economy is at full capacity? Then it is clear that excess demand may have an impact on prices (or, if instead of monetization, the government prints more debt agents may demand higher rate of interest to hold it). The question then is whether the economy is at full employment and what determines full capacity, not whether monetization is always inflationary (which clearly would only be the case if the economy was always at full capacity; unless Krugman and Rognlie agree with Lucas, this could not be the case for them).

However, note that it has been accepted that the supply constraint is variable (the mainstream refers to it as the Time Varying Non Accelerating Inflation Rate of Unemployment or TV-NAIRU). The important question regarding monetization is what determines this supply constraint that imposes an inflation barrier to demand expansion, what Friedman referred to as the natural rate of unemployment (note that natural was meant to suggest that policy cannot affect it).

Here it is also important to note that one component of demand, investment, does have an impact on the supply side. In other words, investment has a dual effect, it is a component of demand (more sales of equipment) and it creates more supply capacity in the future (when the equipment is installed). So, as the economy grows, firms try to adjust their capacity to demand, so as to keep up with the economy, and avoid loosing market share to competition. This suggests that one of the reasons for the variation of the supply constraint is the expansion of demand itself.

It should be noted that this does not mean that the supply constraint is never reached, but it is clearly a rare phenomenon. The graph below shows unemployment in the United States from 1929 to 2011. Only in four occasions did the annual average unemployment fall below 4%, in the mid-1940s, the early 1950s, the late 1960s and the late 1990s (all during Democratic administrations, by the way, with the exception of 1953).

Inflation did accelerate in the first three, when the economy, because of war efforts (WW-II, Korea and Vietnam), was most likely close to full employment (I’m more skeptical about the 1960s, since inflation really picks up in the 1970s, and oil shocks where more important than full employment), but not in the last. Even in those cases, only in the first public debt was actually growing, and the size of the deficits (more than 20%) was incredibly large.  Interest rates did not increase significantly in any of these episodes either (this seems to be Rognlie bone with MMT).

This is the evidence that is used to suggest that governments have a constraint and beyond that inflation ensues? And that is taken as serious thinking on money and deficits! I would agree with Robert Vienneau that this is “unjustifiably arrogant” dismissal of MMT, to say the least.  Part of what I have referred to as the incredible persistency of monetarist views (even among more progressive economists).


  1. Interesting. I of course agree with everything. I am working on a proposal for allowing the ECB to buy bonds up to a certain amount. (at any rate it is already doing it on secondary markets).
    I am pretty positive that the inflationary consequences would be modest (liquidity trap+effective demand deficiency), an what is more, would be rather beneficial.

    I am less clear on the consequences for the euro. Would an important debt monetization mean a run on the euro? That was the case in Germany in the 1920s. But on the other hand, if I were a Chinese investor, would I rather hold safe zero interest euros, or risky Greek bonds?
    At any rate reading you is healthy in this depressing context!

  2. I don't think that there is any risk of that. Again, European debt (including Greek debt) is denominated in euros. No problem if the ECB buys Greek bonds. In fact, it's crazy they don't do it.

  3. Great blog & post. Have no idea why I haven't been hanging around here more. Was trepid about asking a question on an old post, but I see today is the day for replies to it! As I am irrationally fond of pretending to historical knowledge, I just wondered who you had in mind as endogenous money debaters before the Bullionist/Anti-bullionist and Currency/Banking schools.

  4. Traditionally Henry Thornton is cited as the anti-bullionist and David Ricardo as the bullionist, while Thomas Tooke is the quintessential Banking School author, and Robert Torrens his Currency School nemesis. The best book on these debates, in my opinion, is Roy Green's "Classical Theories of Money, Output and Inflation."


The economic and social consequences of the war on Europe and Italy

By Sergio Cesaratto With a certain pride I remember having already mentioned for some years, within the framework of my economic courses, po...