Paul Krugman stirred some controversy in the economics blogosphere with a post detailing what would happen to the U.S. economy if bond vigilantes, fearful of rising government debt, sold off U.S. treasuries. Prof. Krugman’s post includes a small analytical model showing that such an attack would actually benefit the U.S. economy. The model is well constructed but incomplete because it leaves out the Federal Reserve, which would have to undergo a fundamental transformation to fit the hypothesis. Add the Fed to the mix and the model falls apart.
Prof. Krugman’s model finds that “with the Fed setting interest rates, any loss of confidence in U.S. bonds would cause not a rise in rates but a fall in the dollar – and a fall in the dollar would be a good thing, helping make U.S. industry more competitive.” This increase in competitiveness would, in his model, boost the U.S. economy.
Tyler Cowen criticizes Prof. Krugman’s model on grounds that it does not differentiate between short-run and long-run interest rates. In this context, however, I agree with Prof. Krugman that such a distinction is unlikely to change the outcome of the model. Prof. Cowen’s second point is far more important, that the model ignores that a currency devaluation would increase expected inflation.
Prof. Cowen links to a terrific piece by Brad DeLong that details the negative effects a currency devaluation would have on U.S. monetary policy. Prof. DeLong’s piece correctly points out that a Fed change in mandate is unlikely because the Fed has an “institutional memory of the 1970s, when inflation ran rampant”. Prof. Krugman’s model works because it implicitly assumes that the Federal Reserve will stop behaving as if it has a 2 per cent inflation mandate. Prof. DeLong does not have faith in that assumption, and neither do I.
What almost certainly would happen under the bond vigilante scenario is that a sell off of Treasuries would result in a significant devaluation of the U.S. dollar, as predicted by Prof. Krugman. Such a devaluation would cause a rise in inflation expectations, such as the Cleveland Fed estimates. The Federal Reserve would respond by tightening monetary policy, which would cause a reduction in both inflation expectations and economic activity.
Institutions matter. I see no reason why an attack of the bond market vigilantes would cause the Federal Reserve to alter its mandate. It is this implicit change in mandate, not the sell off of bonds, that is key to the Krugman result. The model works in a universe where the Federal Reserve has a different mandate. But if we lived in such a universe, the economy may not need to be revived in the first place.
Mike Moffatt is an assistant professor in the Business, Economics and Public Policy group at the Richard Ivey School of Business, Western University
Follow Economy Lab on Twitter