Can the US Go Bankrupt? Yes, of course it can

A few days ago, the rating agency Standard & Poor’s changed its rating of US government bonds from the usual (highest possible) AAA to a similar one, but with a “negative outlook warning”. This caused havoc around the blogosphere and in policy circles. Some claimed that this was an untimely private, and politically motivated, action serving to undermine public spending programmes in the US. In any case, the market didn’t take much notice, as the interest on government bonds moved little. “Poor Standards” as Paul Krugman called it. He may be right. After all, S&P did funny ratings in the past (remember house-backed securities pre 2007?). And, by the way, it is not the first time, they have raised concerns about the continued AAA quality of US debt. I have no opinion about whether the warning is warranted according to objective criteria, but a government whose debt is around the size of GDP should in most books be worth a check.

What triggered me to write on this is therefore not the rating per se. Instead it is the fact that some has questioned the rating by the argument that the US cannot go bankrupt. For example, James Galbraith is quoted by Dave Lindorff to have said:

“US debt consists of bonds issued in US dollars, which I assume the S&P analysts know. How can the US possibly default on its own currency? The obligation is in nominal dollars, which is to say when the bond retires, the US issues a check in dollars to cover it.  Since the US prints its own currency (or actually just issues electronic payments to create new money) whenever it needs it . . . they will have the money to back their own bonds”

This is not necessarily correct. On every market there is a supply and a demand side. What Galbraith is forgetting is that people may not want to hold the money that the US prints. It has been well known since Cagan’s seminal paper “The Monetary Dynamics of Hyperinflation” from 1956 that there may be limits to how many resources a government can extract through the money press. Technically, financing debts and deficits by money creation is called seigniorage or using the “inflation tax”. And any tax revenue consists of a tax rate multiplied by a tax base. With monetary finance, the tax is the inflation rate, and the tax base is the real money stock held by the private sector (a liability of the government that is eroded in value by inflation; hence, it is “taxed”).

Now consider the thought experiment, as Galbraith does, that deficits and debt are so high that the US must rely on monetary finance to prevent default. By increasing the tax rate on currency (through the printing press), it will indeed for low to moderate rates of inflation be able to generate revenue. However, at higher inflation rates, the nominal interest rate will be higher, and people will want to hold less money (why stick on to something that gives you nothing in return when alternative opportunities get better and better?). In effect, the tax base is being deteriorated. The revenue from printing money at a faster pace may then go down as the increased tax rate (inflation) reduces the tax base (real money holdings in the private sector) sufficiently. Such a “Laffer-curve relationship” in monetary economics is well known, and probably much more well documented throughout history than the more conventional Laffer curve. It means in plain words that there may exist an upper bound on how much revenue a government can extract from money creation.

This is just a too important fact to be overlooked: A country can go bankrupt, even if it has monopoly power over money issuance. Inflation is not just an unpleasant side effect of inflationary financing, it is a phenomenon that may make an attempt to pay back public debt by printing money entirely futile. So yes, the US can go bankrupt. I don’t see it right around the corner, and definitely not because S&P decides to poke a bit to the AAA rating. But don’t think, please don’t think, that the printing press is always a certain, albeit bumpy, road out of fiscal trouble.

This entry was posted in Economists, Macroeconomics and tagged , , , , , , , , , . Bookmark the permalink.

6 Responses to Can the US Go Bankrupt? Yes, of course it can

  1. Morten says:

    Hi, I do not understand how it is not always possible for the US to repay its debt – could you please elaborate? The way I see it, if the US owes me 100 USD, nothing stops them from inflating the nominal money supply until 100 USD has no real value. Isn’t it only a problem for the debtors when they try roll over their debt?

  2. The scenario I consider, is one of anticipated actions. You seem to be thinking of a one-time unanticipated “devaluation” of the real value of outstanding bonds. That can be possible, but if carried out, I would consider that a de facto default.

  3. Jesper Lund says:

    Another possibility is that the US government may simply decide that a debt restructuring (i.e. sovereign default) is the lesser evil compared to starting a hyperinflation scenario in the economy.

    For the government bond holders (China and the oil-producing countries?), a debt restructuring and hyperinflation has the same effect, but for the rest of the economy there’s a difference (and for holders of CDS protection on US sovereign debt there’s a big difference!).

    In 1998 Russia defaulted on its domestic debt, although Russia could presumable have printed money instead.

  4. Jesper Lund says:

    It’s also relevant to point out that swap spreads are negative for long-term maturities in the US (swap spread = difference between swap rate and government bond yield for a given maturity). It started with 30Y and now the 10Y is negative as well.

    There are differerent interpretations of this, but the most obvious one is the declining credit quality of US government debt. Indeed, one could say that S&P is simply reacting to something that is already priced into the government bond market (instead of shooting the messenger, as Krugman and others do).

    This was completely unheard of a couple of years ago (every fixed income textbook will explain what factors contribute to a *positive* swap spread, usually a combination of the liquidity advantage of government bonds and the credit risk of the banking sector which through LIBOR is tranferred to the swap rate). But not anymore…

  5. seasaw says:

    I really can’t make sense of your article. The first comment asked you to explain and I can’t make sense of your response. If the U.S. owes money in dollars, and can print dollars, then they can pay off anything they owe. If you disagree, please explain more clearly. Please show that they can be a debt in dollars that the U.S. can’t pay to go bankrupt. So far, you have failed to show that the U.S. can go bankrupt, or, at least, had failed to demonstrate it in a way that I can understand.

  6. Of course, the government can substitute any values of a bond outstanding with pieces of paper money, which are essentially zero-interest bearing government debt. What I point to is the well documented fact (think Zimbabwe) that heavy use of the printing press will cause inflation. And as inflation erodes the value of paper money, people will be more and more reluctant to hold them. At some point, one therefore cannot rule out that printing more money actually generates a fall in real revenue as virtually nobody wants to hold the money. Then, if tax revenues are too low (not unreasonable as we are talking about high public debt), then eventually the government cannot pay for all of its obligations (public wages, public consumption) and it goes bankrupt.

    To emphasize: I don’t see this happening for the US around the corner, but it should be remembered that the monopoly power of printing money does not guarantee fiscal solvency.