Arjun Jayadev and I have a new piece up at the Institute for New Economic Thinking, trying to clarify the relationship between Modern Monetary Theory (MMT) and textbook macroeconomics. (There is also a pdf version here, which I think is a bit more readable.) I will have a blogpost summarizing the argument later today or tomorrow, but in the meantime here is the abstract:
An increasingly visible school of heterodox macroeconomics, Modern Monetary Theory (MMT), makes the case for functional finance—the view that governments should set their fiscal position at whatever level is consistent with price stability and full employment, regardless of current debt or deficits. Functional finance is widely understood, by both supporters and opponents, as a departure from orthodox macroeconomics. We argue that this perception is mistaken: While MMT’s policy proposals are unorthodox, the analysis underlying them is largely orthodox. A central bank able to control domestic interest rates is a sufficient condition to allow a government to freely pursue countercyclical fiscal policy with no danger of a runaway increase in the debt ratio. The difference between MMT and orthodox policy can be thought of as a different assignment of the two instruments of fiscal position and interest rate to the two targets of price stability and debt stability. As such, the debate between them hinges not on any fundamental difference of analysis, but rather on different practical judgements—in particular what kinds of errors are most likely from policymakers.
My two cents:
I don’t think that your piece reflects correctly what MMT guys think.
Inflation is a change in the price level and the most common way of thinking about inflation is a “quantity theory of money”, that works more or less this way:
a) there are 1000 muffins in the world;
b) there are 100 gold coins in the world;
c) therefore the price level is 1 gold coin = 10 muffins.
This is a pretty stupid way of thinking but it’s very common; since the quantity of money is a stock while the quantity of stuff is a flow (because it’s stuff that is produced and consumed during a certain time period) we have to introduce a “velocity” factor so that the price level P is:
P = (quantity of money) X (velocity of money)/(quantity of stuff [aka total real output])
This in my opinion a wrong way to pose the problem (as velocity of money is something like a magic asterisk), but it’s very common and certainly was in my macroeconomics 101 book 20 years ago*.
If we define the price level in this way, it follows that inflation can and can only depend on a change in one of the three factors (quantity of money, velocity of money, output).
In particular the price level should depend inversely on output (and not directly on output, as you say).
This difference stems from the fact that it is implicitly assumed in keynesian models that the velocity of money is the factor that causes inflation, and that the velocity of money increases during booms and falls during slumps more than proportionally than the change in output (otherwise we should have inflation in slumps and deflation in booms).
But the concept of “quantity of money” changed in time: during the gold standard period, the “quantity of money” was supposed to be the actual quantity of gold, whereas in chartalism government debt (which includes banknotes since during the gold standard period banknotes were redeemable in gold hence were debt) represents money.
In pratice what Knapp says is that government debt (inclusive of banknotes) substitutes gold, and hence that the “quantity of money” is exactly equal to outstanding govenment debt.
Thus the price level equation becomes, if we take Knapp seriously:
P = (government debt inclusive of banknotes) X (velocity of money) / (total output)
In Knapp’s time, people were anxious about the end of the gold standard, and therefore assumed that “inflation” would come by an increase in the quantity of money, and not by a change in the velocity of money, thus chartalism addresses inflation only as something that comes from an increase of the quantity of money: if the government deficit spends, it’s creating money (increasing government debt), if it taxes more than it spends it’s destroying money (decreasing total debt).
Furthermore, as nominal GDP is by definition equal to (quantity of money) X (velocity of money) we have this implicit relationship:
(government debt) / (nominal GDP) = 1/(velocity of money)
which is obviously empirically false, but this is caused by the fact that the whole quantity of money theory of the price level is a nonsense (or rather that the velocity of money concept, being a magic asterisk, does all the work in the theory).
However from this equation it’s obvious that an increase in government debt cannot cause an increase in the debt to GDP ratio, since this ratio depends only on the velocity of money, so the MMTers naturally ignore this problem.
____________________________________________________
[*] I have a degree in “scienze della comunicazione” that basically translates to “media studies”, however since we were supposed to study journalism too a macroeconomic exam was in the curriculum.
I think you are right that these debates were part of the context in which MMT got started, 25 or 30 years ago. But I don’t think a “quantity of money” has been an important part of orthodox macoreconomics or any major heterodox school for many, many years.
The suggestion that there might be an inverse relationship between output and the price level, along the loines you describe, is certainly found as late as Keynes and Schumpeter, but not I think seriously in anyone much later. It does still feature in *some* undergraduate textbooks as a downward sloping “Aggregate Demand” curve based on a fixed stock of moeny. But I think it’s pretty much uniformly acknowledged that this is just a pedagogical stand-in for what actually produces the relationship, a central bank following a policy rule that leads it to adjust interest rates in response to inflation.
Basically, once you say “the central bank sets the interest rate”, M disappears from the analysis.
So I respectfully disagree that the argument you sketch out here has much of anything to do with the debates over fiscal policy we are discussing.
Thanks for the answer.
I don’t think anybody really thinks that there is an inverse relationship between output and inflation, in facts it’s quite obvious that there is a direct relationship, and everybody knows it.
But, when I read about the MMT, I see generally 2 kind of pieces: either mainstream keynesians that say that MMT is really more or less just standard keynesianism, or people who just say that a government that is indebted in its own currency can never go bankrupt, and thus the only possible downside of deficits is eventually runaway inflation, that could be stopped by increasing taxes or otherwise decreasing the deficit or entering a surplus.
In my opinion this second position, that is what we could call the “political MMT”, is based on the implicit assumption that inflation is caused by an increase in the “quantity of money”, and that the quantity of money is actually outstanding government debt.
I don’t think that MMTers think consciously this, but I think they are working from theories that imply this, so their own implicit models end up implying this; in other words I think that they are actually using an outdated view, whitout realizing it (irony on the name “Modern Monetary Theory”).
I wandered a bit on Mitchell’s blog and I ended on this page:
http://bilbo.economicoutlook.net/blog/?p=2916
“Conclusion: we can dismiss this view immediately. Setting some debt to GDP ratio is a futile meaningless task. If you don’t want the private sector to have more financial asset choice then the sovereign government can always stop selling bonds. Simple as that. The net spending will continue as planned but the new net financial assets that are added by the deficits will still be held by the private sector – as bank reserves or in other ways they choose. The government will just deny them the ability to convert low or zero yielding bank reserves into interest-bearing government bonds.”
So I will amend my previous comments this way:
1) MMTers (or at least Mitchell) think in terms of the government issuing currency when spending and destroying currency when taxing.
They don’t see this issued currency (that I called “banknotes”) as part of the government debt.
2) I still think they see inflation as a problem of quantity of money: when the government deficit-spends it issues new currency, that generally will cause an expansion in employment and in the quantity of output (so both the quantity of money increase and the quantity of output increase, no or small inflation); if the real economy reaches some maximum output level or doesn’t expand fast enough government deficit cause an increase in the “quantity of money” that is greater than the increase in the quantity of output, hence inflation.
This is from another page on Mitchell’s blog:
http://bilbo.economicoutlook.net/blog/?p=2562
”
First, a currency might be valued for its intrinsic value (so gold or silver coins). This is a pure commodity currency system. In the C18th, commodity money systems became problematic because there was a shortage of silver and this system steadily gave way to a system where paper money issued by a central bank was backed by gold.
”
Note the idea that metallic systems became problematic because there was a “shortage of silver”, that is clearly a quantity of money kind of thinking.
3) But Mitchell sees government debt (bonds) as something different from money: governments “borrow” only because they want to give out interest-bearing assets, perhaps because they are ruled by capitalists.
4) so ultimately the government could and hould just stop to issue bonds (and have thus 0% government debt) and simply deficit spend or tax away money in order to achieve maximum employment and limit inflation, with the implicit idea that inflation is caused by excess growth in the money base relative to output.
as an aside, my personal opinion is this:
Personally, while I think that the government should, in fact, not issue government debt but just issue (print) money if needed, I don’t think that the price level (and thus inflation) depends on the quantity of the money base, but rather on a wage price spiral that depends on the level of employment and other legal/social issues, the quantity of silver or banknotes being totally irrelevant, and that the nominalwealth/nominal income ratio depends on a dinamic of “saving glut” endemic in the capitalist system, so I think that the MMT is wrong (although some of their political ideas might be correct).
Bill Mitchell seems to be lashing out about your piece because he believes you are saying that his analysis is the same as Mankiw’s textbook with it’s descriptions of loanable funds and crowding out whereby government deficit spending reduces the financial capacity of the private sector to invest in improvements in equipment, infrastructure, training etc. My impression is that there is a huge gulf between economists such as you (and even Simon Wren Lewis) and those descriptions in Mankiw’s text book. So perhaps the problem is that the “mainstream” spans between wackyness such as Mankiw’s text book all the way to Simon Wren Lewis and I suppose Meadway (John McDonnell’s adviser in the UK Labour Party). So I guess it is crucial to pin down which “mainstream” is being compared to MMT? (you do say that both the mainstream and MMT are broad and evolving but Bill Mitchell still got the impression he did).
I agree, it would have been nice to point to a specific textbook or similar to calrify what we meant by “the mainstream.” I don’t expect we are going to do anything further with this, but if we did do another version that’s something to fix.
The specific views we claim are shared by MMT and mainstream analysis of fiscal policy are the numbered premises reflected in the equations. Loanable funds obviously doesn’t figure there and is, I think, hard to reconcile with the idea that the central bank chooses an interest rate.
I can’t be responsible for Bill Mitchell’s impressions. If people like you found the piece useful, that’s good enough.
I have a bit more sympathy and understanding as to your reluctance to mention the name of that “One school of heterodox thought, which has gained increasing prominence over the decade” in your chapter described in the post from August 8. Despite that, I think that Bill Mitchell is right about the degree of difference of MMT from the typical ‘mainstream’ economics. But mainstream is a very, very broad category which could include economists like Dean Baker- who really is not that far off from MMT, at least in what he prescribes for policy.
How would you describe your own understanding of economics as far as the ‘mainstream’ view? Just out of curiosity. I mean what category (that I might know) would you say you are closer to? I would have guessed post-Keynesian, which probably puts you outside the ‘mainstream’ also. That all depends on whatever ‘mainstream’ actually is supposed to mean of course.
I have a very patchy comprehension of economics, I guess my deepest ignorance is around “mainstream” economics. I’m struggling to get my head around the contention in Bill Michell’s blog that New Keynsian economics is built from a foundation that includes loanable funds and financial crowding out and that, as you say, is incompatible with a central bank being able to set interest rates. Simon Wren Lewis and all central bankers believe central banks can set interest rates don’t they? Or do they believe that they can only set a “policy rate” that needs to be based on discerning and shadowing some mysterious underlying true interest rate that occurs due to loanable fund type phenomena? I guess media news presenters must believe something like that when they say that “we have a danger that interest rates could rise very rapidly” as though it was outside mankind’s control.
I guess what I’m getting at is the notion that when you state “a central bank able to control domestic interest rates is a sufficient condition to allow a government to freely pursue countercyclical fiscal policy with no danger of a runaway increase in the debt ratio” -perhaps many mainstream economists don’t believe the central bank has such a capacity. They agree that at an operational level the central bank sets a base rate. However in their view, the central bank is a slave to some mysterious underlying interest rate (perhaps arising as a result of some loanable funds type phenomenon). The central bank’s job is to discern that underlying interest rate and set the base rate as close to it as they can. If they screw up and miss, then the inappropriate base rate will cause a destructive instability in the currency that will outweigh any desired benefit that might be hoped from setting a different base rate. Perhaps the WWII (and WWI) examples of central banks imposing fixed interest rates are dismissed as financial repression that depends on the extraordinary level of control that war time governments are able to muster (I guess alongside requisitioning buildings, conscripting etc etc). -I have to stress I’m just trying to make out what the mainstream position is and I’m very ignorant and I’m just fishing for corrections/refutations etc.
Josh,
Given that MMT see no need for the issuance of government debt, is there then no need for the debt sustainability curve in your figure 1 in your paper? That is, both policy tools, i and the fiscal balance, can be directed to the stability goal without any other constraint.
Surely this is a major difference between possible MMT and orthodox policy modes?
Yes, this is explicitly stated in the paper.
The thing that should have been clearer (especially the title should have been different) is that **this is not a paper about MMT**. This is a paper about **a specific claim made by many people within MMT** – that the government budget can be set at any level necessary to achieve full employment without facing a financial constraint.
I do think that MMT people have not really come to terms with the fact that the actual policy they are calling for probably implies a declining path of debt-GDP.
“I do think that MMT people have not really come to terms with the fact that the actual policy they are calling for probably implies a declining path of debt-GDP.”
So what do see the consequences being?
Sounds like a good idea to me.