Grasping Reality on TypePad, by Brad DeLong

The Economist asks what will happen to U.S. interest rates if the value of the dollar dives:

Economist.com: If the dollar dives, what will happen to America's interest rates? WHEN policymakers and pundits debate America's current-account deficit, the phrase "hard landin"D is never far from their minds. It sums up what might happen if foreigners tired of financing the gap, now over 6% of GDP: a sinking dollar, soaring interest rates, tumbling asset prices--all dragging America's economy into recession. Optimists, including Alan Greenspan, chairman of the Federal Reserve, think that all this is pretty unlikely. It is far more probable, they say, that America's imbalances will come right gradually, via a (gently) weakening dollar and slower demand growth at home.... But what if it did sink suddenly? Then, surely, the hard-landing logic would come into play: bond yields would rise sharply as investors demanded compensation for the risk of higher American inflation and further falls in the dollar.

Or maybe not. In a new paper*, Joseph Gagnon, an economist at the Federal Reserve, takes issue with this argument. It is true, he says, that currency crashes in emerging economies have sent interest rates soaring. And it used to be true of rich countries, too; but in the past 20 years their experience has been rather different. Since 1985, every industrial country whose currency has crashed--defined by Mr Gagnon as a depreciation of at least 8% in the first year and over 20% in two years--saw its bond yields fall, by 1.5 percentage points on average, the following year.... Mr Gagnon ascribes this change in rich countries' experience to the taming of inflation. A weaker currency might boost bond yields if investors fear higher inflation, or if they think monetary policy might be tightened to combat it. Yields might also be pushed up because investors fear further depreciation, or suspect that borrowers--in particular, the government--might default.... What really matters is the inflation risk, and this has become much less of a worry in the past 20 years. If inflationary expectations are falling, the link between tumbling currencies and rising bond yields may be weakened.... In 1997-98, Australia's dollar tumbled. Net foreign purchases of the country's bonds, worth 5% of GDP in 1996, turned into an outflow of 1% in 1998. Yet bond yields fell by more than two percentage points, partly because of the inflation-fighting reputation of Australia's central bank, but also because investors were worried about the country's growth in the wake of the Asian financial crisis.

The point is that there is no mechanical connection between currency crashes, long-term interest rates and the hardness of a landing. A fall in the dollar when America's economy was at full tilt ought to boost exports and might cause the economy to overheat; interest rates might rise, but no recession would ensue. Or a fall in the dollar might be the product of a slowdown--if, say, America's consumers stopped spending. Then lower, not higher, interest rates could ensue. Mr Gagnon has not refuted the idea that America could have a hard landing; but he has exposed some loose thinking on the subject.

Well, it depends on why the value of the dollar declines. There are a number of possibilities...

1. If foreigners--especially foreign central banks--stop buying dollar-denominated assets because they decide they have enough dollars, then U. S. long-term interest rates rise as the dollar falls because the lower demand for U.S. long-term bonds pushes their price down and their yield (unless the Federal Reserve takes immediate and strong action to try to keep long-term yields from rising)...
2. If bad macroeconomic news leads to expectations of a growth slowdown in the U.S., which leads people to expect easier Federal Reserve policy in the future, the dollar will fall (because dollar-denominated assets will be expected to have lower interest rates) and U.S. long-term interest rates will fall too (because financial institutions' attempts to lock-in higher yields boost demand for U.S. long-term bonds...
3. If the dollar declines because people expect the U.S. to follow a tighter fiscal policy in the future, U.S. long-term bond yields will fall for the same reasons as in (2)...
4. If the dollar declines because people give up hope that the U.S. will ever control its deficit and so expect a large increase in inflation, the resulting capital flight will push the dollar down and nominal yields on U.S. long-term bonds up (what will happen to real yields is a difficult question to answer)...

These are mixed up with the consequences of:

1. If after the dollar declines, people expect the dollar decline to continue--then they will sell U.S. long-term bonds and push their yields up...
2. If after the dollar declines, people expect the previous fall in the dollar to make it likely that it will appreciate and give them the opportunity to earn capital gains on their dollar-denominated assets--then they will buy U.S. long-term bonds and push their yields down...

This was, IIRC, the gravamen of Barry Eichengreen's discussion of Sebastian Edwards's U.S. current account paper at Jackson Hole. Sebastian presented correlations without inquiring as to what the impulse that caused the currency decline and the current-account reversal was. Barry wanted a taxonomy: if the currency declines because of impulse X, then result Y.