Why Default on U.S. Treasuries is Likely
Liberty Fund has just featured my article on "Why Default on U.S. Treasuries is Likely." Here is the first paragraph:
"Almost everyone is aware that federal government spending in the United States is scheduled to skyrocket, primarily because of Social Security, Medicare, and Medicaid. Recent"stimulus" packages have accelerated the process. Only the naively optimistic actually believe that politicians will fully resolve this looming fiscal crisis with some judicious combination of tax hikes and program cuts. Many predict that, instead, the government will inflate its way out of this future bind, using Federal Reserve monetary expansion to fill the shortfall between outlays and receipts. But I believe, in contrast, that it is far more likely that the United States will be driven to an outright default on Treasury securities, openly reneging on the interest due on its formal debt and probably repudiating part of the principal."
comments powered by Disqus
James - 9/11/2009
You note that the government has gigantic unfunded liabilities: social security, medicare, etc. If the US defaults on its debt (or inflates its debt away), these liabilities won't go away, will they? People are still going to expect their handouts, and vote for politicians who promise them. Is there any financial strategy that can solve the government's problem of too many grasshoppers, not enough ants?
Jeffrey Rogers Hummel - 8/5/2009
Bill: Your comments are always well thought out and often challenging. I partly agree with you about the fiscal theory of the price level, although I think you (and other critics of the theory) are mistaken about the claim that it simply assumes inflationary default. The best defense of the theory, in my opinion, is John Cochrane's article in the Journal of Monetary Economics, 52 (2005), "Money as Stock," in which section 4.1 explicitly incorporates the possibility of open default on government securities into the model. As for your argument about the Fed's simply buying up outstanding Treasury debt, I'm not sure I follow and need to think about it more. Are you suggesting that this scenario represents something other than seigniorage? Or are you simply suggesting that it demonstrates that the real demand for base money does not constrain seigniorage? If the first, I strongly disagree. If the second, I suspect we may only have a semantic quibble over how strongly we would interpret "constrain."
Bruce Ramsey - 8/4/2009
When has a government ever defaulted on debt in its own fiat currency? It's difficult to imagine this. More likely, interest rates would rise, and legislators would be under pressure to enact balanced budgets. They run deficits now because credit is so cheap--which is to say, because the world lets them.
Bill Woolsey - 8/4/2009
I favor explicit to inflationary default. Perhaps it is because I favor the privatization of money, and so, short of renationalizing it, inflationary default is off the table.
I also find the fiscal theory of the price level, which assumes inflationary default, wrongheaded. Inflationary default is something that should be analyzed using an adequate monetary theory rather than being assumed as its basis.
Still, I think Hummel fails to emphasize the key mechanism that makes inflationary default possible. That is, the possibility of using newly created money to pay off government bonds as they come due. Of course, it could be paid off early by purchasing the debt on the market. And while it is easy to see how this works all at once, it can (and is) done partially.
This isn't the same thing as sengiorage. Or rather, thinking about the amount of goods and services government can obtain through money creation fails to account for its effectiveness in paying off nominal debts. The amount of goods and services government can purchase with money creation in equilibrium is the rate of money creation times the real quantity of base money. The real quantity of base money depends on the real demand for base money. The increase in the price level due to the money creation is the key mechanism by which the government’s command over goods and services from the inflation tax is limited. But the higher price level doesn't reduce the ability to pay off existing debts.
Hummel's points about how the real demand for base money is small under modern financial institutions are correct. Money creation can do little to fund government purchases of goods and services and reduce future deficits. Further, the logic of real balances does tell us how much those who owned the government debt end up with. How much goods and services the newly created money will buy. But it tells us nothing about the ability of the government to pay off those debts with newly created money.
Suppose the government borrowed in terms of Euros. And the question was whether the government could use money creation to pay off those debts. They create money, buy Euros, and pay off the euro debt. Because the dollar value of the euro would rise in this scenario, the amount of debt the government could repay would depend on the real revenue raised by the inflation tax. The fact that the goverment is printing dollars to repay dollar denominated debt matters.
It is also not the same thing as the transfer from creditors to debtors because of unexpected inflation. If current nominal interest rates don't include a sufficient premium for the inflation that occurs, the higher nominal tax revenues will be able to discharge more real principle. This would apply to all creditors and debtors in a scenario of inflationary default.
However, the money issuer can pay off debt as it comes due. The way expectations of this impact public finance is that fear of inflationary default results in the government only being able to borrow at higher nominal interest rates. If the government does not default (with inflation or otherwise,) then the real interest burden rises. But, it can always pay off the existing debt with nearly created money and discharge as much as it wants, leaving the former debt holders with currency of perhaps miniscule real value.
Imagine that monetary authority printed money and dropped it from helicopters. An unexpected increase in inflation would transfer wealth from creditors to debtors. The government, as debtor, could benefit. But only if the increased rate of newly dropped money was unanticipated.
But the reality is that the monetary authority can use newly created money to pay off debts that are denominated in terms of the money that the monetary authority is creating. That matters.
The ability of the money issuer to pay off existing debts with newly created money doesn't suffer from the ordinary seignorage limitation because the nominal principle doesn't increase as prices rise. The debt is denominated in terms of the money being issued.
The money issuer is paying off the debt to discharge it, rather than seeing its real value fall, and so the nominal interest rates it would have to pay to borrow in an inflationary environment are irrelevant. The fact that new money is used to pay off the debt and not introduced by some other mechanism matters.
One point Hummel makes does point to this key mechanism of inflationary default--he discusses the interest paid on reserve deposits at the Fed. Presumably, the Fed could purchase the entire national debt without there being any inflationary effect if it was willing to pay sufficient interest on reserves. The Treasury would continue to pay the Fed, the Fed would pay the banks, and the banks would pay the depositors. The national debt would all be held indirectly by those holding deposits in the banking system.
That would defeat the point of inflationary default. The whole point would be for the Fed to not pay interest on reserves, and transfer the interest income it earns from the Treasury back to the Treasury. It is unlikely that in such a scenario, that much would be funded by reserve balances anyway, but rather it would end up being currency.
So, we are not protected from inflationary default by a limited real demand for base money. Nor does a rapid response of nominal interest rates to expectations of inflation do much good. Using newly created money to pay off debt does discharge the debt. And, of course, creates massive collateral damage to the rest of the economy.