Misunderstanding Modern Monetary Theory
Paul Krugman wrote a post today regarding MMT called "I Would Do Anything For Stimulus, But I Won’t Do That (Wonkish)." The gist of Krugman’s post was to refute Modern Monetary Theory’s view on money and deficits. Krugman writes:
Right now, the real policy debate is whether we need fiscal austerity even with the economy deeply depressed. Obviously, I’m very much opposed — my view is that running deficits now is entirely appropriate.
But here’s the thing: there’s a school of thought which says that deficits are never a problem, as long as a country can issue its own currency. The most prominent advocate of this view is probably Jamie Galbraith, but he’s not alone.
Now, Jamie and I are, I think, in complete agreement about what we should be doing now. So we’re talking theory, not practice. But I can’t go along with his view that
So long as U.S. banks are required to accept U.S. government checks — which is to say so long as the Republic exists — then the government can and does spend without borrowing, if it chooses to do so … Insolvency, bankruptcy, or even higher real interest rates are not among the actual risks to this system.
Krugman goes on to use a model with strongly monetarist/neoclassical embedded assumptions to make his points. Jamie Galbraith responded in the comments and I am posting his comments here. But, first, a few words.
I agree that deficits matter. But I take a more Austrian/austerian view in general – so of course I would say that.
However, as I understand MMT, Krugman’s post mischaracterizes both MMT and Galbraith’s statement. There are two separate issues here that should be disaggregated and treated in isolation. The first issue is about money and government’s source of funding. A separate but related issue is deficits.
On the funding side of things, it sounds like Krugman is trapped in a gold standard view of money as he assumes the government must issue bonds to fund itself. He forgets that we live in a fiat world and that taxes don’t fund government spending, requiring government to issue bonds for a shortfall. Remember, a fiat currency is one that is created by government. Government can satisfy any commitment in that currency if it so chooses. It could theoretically credit accounts electronically to fulfil its commitment, laws permitting – no bonds necessary.
From the government’s perspective, there is no functional difference between any of its obligations like bank notes, electronic credits, or treasury bills and bonds. As the Ten pound note says, “I promise to pay the bearer on demand the sum of [fill in the blank sum][fill in the blank fiat currency].”
So, the U.S. government could legitimately stop issuing bonds altogether if it wanted to. When people complain about the admittedly enormous government debt, they don’t think of the mechanics of the issue. As I see it, in a fiat money environment, the first function of the Treasury bonds is to serve as a vehicle to add or subtract reserves in the system to help the Federal Reserve hit a target Fed Funds rate. The second is to give holders of government obligations a return on their investment. After all, bank notes or bank reserves don’t pay much if anything.
This is where the austerian in me says "government could simply pay for things with money it prints electronically out of thin air." You may not like this fact but that’s operationally how fiat currency works. I would argue that this eventually leads to currency revulsion (Krugman talks about using lumps of coal as money) and inflation.
[W]hile there is no operational constraint on government because of the electronic printing presses, there is an effective constraint in the form of debt and currency revulsion and price instability (large measures of deflation or inflation). On countries like Greece or Portugal in the Eurozone, the operational constraint is a lot more real than it is on the U.K. because of currency union. The same is true for countries with a currency peg or large foreign currency debts like Latvia, Hungary or Dubai.
So the problem for deficits is not national solvency but inflation and currency depreciation. That makes me worried about deficits. If that makes me an inflation hawk and anti-deficit, then so be it. Nevertheless, MMT does say the same thing about deficits, namely that they can lead to inflation. But MMT also says that inflation is not a problem when you have an enormous output gap from 17% underemployment. MMT proponents recommend deficit spending to close that gap. But you can’t spend at will under MMT; eventually the output gap closes and inflation becomes a big problem.
Notice that Galbraith never specifically mentions deficits in his statement. I don’t think he’s talking about deficits at all. His statement goes more to how government funds itself i.e with fiat money that it creates. He also speaks to his view that U.S. banks are forced to accept the government’s money because they want to do business in the only legal tender currency unit of taxation. While Galbraith may be more sanguine about the prospect of currency revulsion in the medium-term than I am, he never mentions deficits.
So Krugman clearly misunderstands MMT because it sounds to me like he’s saying the same thing. Someone correct me if I’m wrong.
Here’s what Galbraith said in the comments in response:
July 17th, 2010
Paul’s argument is that *infinite* inflation is a theoretical possibility. Well, yes. It happened in Germany in 1923.
There is no reason to cut Social Security benefits or Medicare now, with effect in the future, in order to avoid the theoretical possibility that some combination of policies might at some time in the future give us the economic conditions of post World War I Germany.
Those conditions were desperately resource-constrained.
In the actual world we live in, government does not have to "persuade the private sector to release real resources." In the actual world, the private sector has already released those resources by the tens of millions of people.
All the government has to do, in the actual world, is mobilize those resources, which it does by issuing checks, preferably to pay people to do useful things.
There is no reason why this should be considered "costly." Done correctly, in economic terms it amounts simply to the reduction of the waste that is associated with unemployment.
Nor is it necessary, when the government issues a check, that it issue a bond to "borrow" the money behind that check. The check creates money in the first place. (Yes, it does this from thin air, by changing numbers in bank accounts.)
Operationally, this is a free reserve in the banking system. The reason the government issues a bond later, is that the banks like to have a higher rate of interest than they can earn on reserves, and the government likes to oblige them.
This is why Treasury auctions don’t fail: the government has already created the demand for the bonds, by issuing checks to the banking system.
If the government spent but declined to "borrow," what would happen? Nothing much. Banks would hold their reserves as cash rather than bonds, and their earnings would be a bit lower. It is *not* true, as a rule, that people (or banks) move readily to substitute lumps of coal for dollars, unless the price level is already moving up and out of control.
It is very difficult to get other people to accept coal in place of dollars!
Paul’s logical error here is that of assuming-the-consequent. He assumes the inflation which causes dumping of money. But if there is no dumping of money, the inflation will not generally occur.
Yes, again, it’s technically possible that the banks and others would start dumping dollars and buying up oil, wheat, rubber, and so forth (and leasing storage facilities for the stuff) thereby driving up the price level.
I wrote — correctly and deliberately — that bankruptcy, insolvency and high real interest rates were not risks. Inflation *is* a risk.
By this, to be clear, I mean an ordinary garden-variety increase in the inflation rate is a risk — not the *infinite-inflation* scenario.
Inflation, though unattractive, is not remotely comparable to bankruptcy or insolvency, unless you get to Paul’s *infinite* inflation scenario. So what about that?
In his model, it is driven by his monetarist (quantity-theory) simplification, that the increase in money flows directly into prices. But this is just a modeling error. In the real world, especially in broadly deflationary conditions, people — and banks — simply hang on to cash. There is a Paul Krugman who understands this, from close study over many years of the Japanese stagnation.
However, and again, in the present state of the world economy, and for the foreseeable future — and except for the energy sector — surely a small rise in the inflation rate is a trivial risk.
My position is that the government should focus on real problems: unemployment, care for the aging, energy, climate change, and the disaster in the Gulf of Mexico.
The so-called long-term deficit is not a real problem. And the capital markets demonstrate every day that they agree with this judgment, by buying long-term Treasury bonds for historically-low interest rates.
I may amend this post with links to prior posts at some future point today.