Paul Krugman writes out a simple model (pdf), source here, and summary here: “Because America has its own currency and a floating exchange rate, a loss of confidence would lead not to a contractionary rise in interest rates but to an expansionary fall in the dollar.”
I see it differently. The confidence loss brings upward pressure on the real borrowing rates, and the Fed pumps out money to keep the short-run nominal interest rate down (have zero lower bound issues gone away?). Because of expected price inflation, the long-run nominal interest rate goes up and the medium-run interest rate goes up too. The long and medium real interest rates go up and stay up. There is no long vs. short-run interest rate distinction in Krugman’s model.
Exactly what happens with short-run interest rates depends on credibility and how much inflation the public is willing to accept. I would say this: the counterfactual of a bond vigilantes attack already means the public refuses to turn over much more of its wealth to the government. So most likely we have higher medium- and long-term real and nominal interest rates and chaos on the short rates, with no pretty scenario in sight. Eventually the short rates probably will rise too.
I do see that Krugman has written: “…inflation and expected inflation could matter, but I don’t think they do in this case, so that I suppress them for the sake of simplicity.” In essence, in his model, the central bank can push down “interest rates” (in general) without bearing any inflation repercussions. Of course that eases the problem but it seems oddly unremoved from the real world problems of central banking.
A few years ago, Brad DeLong wrote correctly:
International finance economists see a far bleaker future. They see the end of large-scale dollar-purchase programs by central banks leading not only to a decline in the dollar, but also to a spike in U.S. long-term interest rates, both nominal and real, which will curb consumption spending immediately and throttle investment spending after only a short lag.
To be sure, international finance economists also see U.S. exports benefiting as the value of the dollar declines, but the lags in demand are such that the export boost will come a year or two after the decline in consumption and investment spending. Eight to ten million workers in America will have to shift employment from services and construction into exports and import-competing goods. This cannot happen overnight. And during the time needed for this labor market adjustment, structural unemployment will rise. Moreover, there may be a financial panic: large financial institutions with short-term liabilities and long-term assets will have a difficult time weathering a large rise in long-term dollar-denominated interest rates. This mismatch can cause financial stress and bankruptcy just as easily as banks’ local-currency assets and dollar liabilities caused stress and bankruptcy in the Mexican and East Asian crises of the 1990’s and in the Argentinean crisis of this decade.
That is followed by a very good short discussion of inflation and monetary policy, consistent with my views above. Brad closes with an excellent bit, including:
Serious economists whom I respect enormously find themselves taking strong positions on opposite sides of this debate.
So I am baffled when Krugman writes: “But it’s really hard to create a scenario in which the bond vigilantes actually cause a contraction rather than an expansion when they attack.” And then he writes:
So what are the fiscal fear types thinking? Basically, they aren’t.
Only a few years ago, Krugman had a well-worked out and indeed quite admirable version of a problematic scenario, starting on p.454 and running through the conclusion, where it is presented as plausible, including by his commentators. Note that this isn’t a “does Krugman have the right to change his mind?” issue (of course he should and he did change his mind on the likelihood of such an attack coming soon). This is a theory question, where we consider the assumption of a vigilantes attack, and work through the theory, while admitting it probably is not imminent. The theory hasn’t changed in the last five years, and Krugman’s current discussion hasn’t caught up with the theory from five years ago.
If you would like data, here is Ricardo Hausmann et.al. on the Latin American move to floating systems and how it has affected their financial crises:
This paper attempts to assess the performance of alternative exchange rate regimes in Latin America relative to the benefits they are theoretically supposed to deliver. We will test empirically whether flexible systems allow for better cyclical management, more monetary autonomy and improved control of the real exchange rate. We find that flexible exchange regimes have not permitted a more stabilizing monetary policy but instead have tended to be more pro-cyclical. In addition, flexible regimes have resulted in higher real interest rates, smaller financial systems and greater sensitivity of domestic interest rates to movements in international rates.
Post eurozone crisis, the fixed rates probably look worse again in a broader data pool, but still there is plenty of well-known, mainstream evidence that floating rates don’t always give such an expansionary response to a financial crisis, even if they sometimes do. Again from the Hausmann paper:
…interest rates moved the least in countries with no exchange rate flexibility! In Argentina and Panama there were very small movements in the domestic interest rate. In contrast, countries with formally floating systems such as Mexico and Peru saw very large interest rate movements. The same can be said of Chile, which started the period with a very wide exchange rate band.
This doesn’t boil down to the usual ideological disputes. If anything, I am taking the position more skeptical of the claims of Milton Friedman. It’s simply that, with the fiscal cliff approaching, I see an Orwellian memory hole here.