Hoisted from Archives from September 2005: The Coming Dollar Crisis?
Grasping Reality with Both Hands: The Coming Dollar Crisis?: As is often the case, the most interesting things I learned at last August's round of conferences came not in the formal conference sessions but in the informal small-group conversations before, around, in the interstices of, and after the conference.
Take the Federal Reserve Bank of Kansas City's "Greenspan Era" conference. It was held in Jackson Hole, at the Jackson Lake Lodge in Grand Teton National Park, in the shadow of the Grand Tetons, which are perhaps the most impressive mountain range in North America. ("Perhaps" because the Canadian central bankers present pointed out the Canadian Rockies from Lake Louise to Jasper, while Federal Reserve Bank of San Francisco president Janet Yellen sang the praises of the mountains of her own Federal Reserve District: the east face of the Sierra Nevada as seen from the Owens Valley, to be specific.) You spend the mornings in windowless conference rooms, and the afternoons outside--on the Snake River, hiking, climbing, looking for moose, looking for elk, hoping that bear are not looking for you.
But the afternoons--and the formal and informal breaks in the mornings when you flee the windowless conference room for the fresh air of the west lawn of the lodge to stare at Mount Moran and company across the lake--are filled with arguments. Go with Federal Reserve Governor Don Kohn on an afternoon hike up Cascade Canyon, I am told, and expect to gain 3000 feet in two hours while being quizzed intently about technical details of monetary policy. Try to keep from grounding yourself on sand bars in the Snake River, and you will hear ex-senior executive branch officials of both political parties give their assessments of why neither of the Bush II Treasury Secretaries has been able to make effective use of the thousand or so people who work directly for him.
This time the most interesting rounds of break-and-afternoon conversations I heard were sparked by Sebastian Edwards's paper about the U.S. current account deficit. My conversations quickly exposed a deep fault among the conference attendees. Those who analyzed or forecast the U.S. domestic macroeconomy agreed that a steep decline in the value of the dollar sometime in the next five years was overwhelmingly likely, but by and large they did not think that such a decline would pose a big problem for the U.S. economy. (They agreed that it might well pose a very big problem for some of America's trading partners.) By contrast, those who analyzed or forecast the international economy as a whole were typically terrified by the prospect of a steep (30% or more, perhaps much more) decline in the value of the dollar: they thought a severe U.S. recession was a definite possibility, and that the situation would require exceptionally skillful handling to keep from becoming a serious economic problem.
Why this disjunction?
The domestic macroeconomists would typically argue more or less like this:
Yes, the dollar is likely to decline steeply either when foreign central banks stop buying dollar-denominated assets to keep the values of their currencies down or when international speculators lose confidence or both. But so what? The fall in the value of the dollar will boost foreign demand for U.S. exports. Workers will be pulled out of other sectors into the export sector. The effects of the dollar decline are much more likely to be a plus for employment rather than a minus, a boom rather than a recession.
To this, the international economists would respond more-or-less like this:
When foreign central banks stop buying or international speculators lose confidence in the value of the dollar and thus stop buying U.S. long-term bonds, two things happen: the value of the dollar falls, and the rate of interest on dollar-denominated long-term bonds spikes. The spike in long-term interest rates discourages investment spending directly, and also discourages consumption spending because higher interest rates mean lower housing and stock prices and thus lower consumer wealth. The fall in domestic spending happens now. The rise in exports as the falling dollar makes U.S.-made products more attractive to foreigners happens two years from now. In between, a lot of people are unemployed--and as they are unemployed, they cut back further on their spending. Plus there is the risk that the fall in the value of the dollar and the fall in long-term asset prices generated by the interest rate spike will cause enough bankruptcies among financial institutions to cause a flight to quality--which will further raise non-safe interest rates, and further discourage investment and consumption spending
This then puzzled the domestic economists:
Why should interest rates spike? The Federal Reserve controls American interest rates. If it wants to keep the price of the ten-year Treasury bond high, it can simply start buying bonds until the price of ten-year Treasuries is what the Fed wants it to be. There's no reason for employment in construction and other interest rate-sensitive sectors to fall before employment in exports and related sectors rises--at least not unless the Federal Reserve makes a big mistake and allows rising interest rates to shoot the economy in the head.
And at this point the response of the international economists fragmented:
Some said that the falling dollar would create inflation--with imports at 1/6 of GDP, a 40% fall in the dollar would, if fully passed through to import prices, add 6% to the U.S. price level. The Federal Reserve would feel honor-bound to maintain its reputation as an inflation-fighter, and so would allow interest rates to go high enough to produce enough unemployment to push nominal wages down far enough to offset this rise in import prices. Thus the Federal Reserve would welcome the spike in interest rates as appropriate, and take no steps to offset it.
Others said that the adjustment to the fall in the dollar would require that ten million workers shift out of construction, retail, and consumer services occupations and into export and import-competing manufacturing industries. You cannot move ten million American workers from one sector to another in a matter of a year or two without creating lots of structural unemployment.
Still others said that financial stress would be the key: perhaps some major Wall Street firms would discover big unhedged risks in their derivative books; perhaps perhaps others would find that the values of their portfolios were more responsive to changes in long term interest rates than they had thought. In either case, it is financial distress and chaos that really triggers the recession.
And the domestic side had rebuttals to each of these three points:
If the Federal Reserve announces now that it is targeting a measure of inflation that is not grossly affected by import prices--that it is targeting nominal wage growth, say--there is no need for the Federal Reserve to defend its credibility by attacking the economy. Just as the Federal Reserve has trained observers that it is more important to worry about 'core inflation' than 'headline inflation', so the Federal Reserve ought to be preparing observers to recognize that inflation produced by rising import prices is a one-time event, not an inflationary spiral that needs to be fought by triggering a deep recession.
A large structural shift will cause high unemployment only if the transition is quick and brutal, and only if workers are pushed out of job-losing rather than pulled into job-gaining sectors. Whether it is quick or gradual and whether it is push or pull depends, once again, on the path of interest rates. Only if the Federal Reserve fails to do its job and allows for a massive interest rate spike is there a problem.
Financial stress is something that can be managed: if the Federal Reserve keeps the path of interest rates smooth, great financial stress is unlikely.
And the domestic side of the argument pointed to the historical experience of the U.S. from 1986-1990: Between 1985 and 1989 the value of the U.S. dollar declined by 40%. Between 1986 and 1990 the U.S. trade deficit declined from 4.0% of GDP to 0.5% of GDP--without a big recession, or significant macroeconomic distress.
Before dinner one evening I was lectured by a prominent Washington-area international finance economist about all the reasons that the 1986-1990 U.S. experience was likely to be a bad guide to the future:
1986-90 began with a 50% decline in world oil prices, a powerful stimulus to the world economy. This time the process is beginning with a doubling of world oil prices.
1986-90 saw Europe growing rapidly. Europe has a high propensity to buy U.S. exports, and the European boom meant that U.S. exports grew much faster in the late 1980s than anyone had expected. This time it is Asia that is booming, not Europe. And Asia has a relatively low appetite for U.S. exports.
The Japanese government was willing to buy very large amounts of dollar-denominated assets in the late 1980s to keep the decline in the value of the dollar "orderly." In so doing, it inflated its domestic credit base and touched off its own property bubble. No foreign government is going to risk this again just because the U.S. would rather that the decline in the dollar was slow and orderly.
The problem then was half as big relative to the size of the U.S. economy as is the problem now.
One way I found myself thinking of the argument is that the domestic-side economists look at the goods market and think of a decline in the value of the dollar as a supply shock, and as not that big a supply shock: if half of the adjustment in import prices is taken in reduced margins by producers abroad, and if the shock is spread out over four years, then 40% / 2 x 16% / 4 = 0.8% increase in inflation relative to baseline over three consecutive years. The Federal Reserve could easily allow that to happen without--providing it explained its causes well--running any risk of damaging the credibility of its commitment to effective price stability. No big deal. International finance economists, by contrast, look at the asset markets. A 40% decline in the dollar over four years is a decline at the rate of 10% per year. Once financial markets convince themselves that such a decline is coming and that they need to be compensated for it, that ought to drive a 400 basis point wedge between U.S. and foreign long-bond expected returns. And that is a very big deal.
Martin Feldstein said something very smart just after we had both taken off our shoes at Jackson Hole airport. He said that the domestic-side economists were keying off the past experience of the U.S. after 1985 and of Britain after 1982, and so were saying "no big deal"; while the international finance economists were keying off of the experiences of developing countries that had run large current-account deficits--Mexico 1994, East Asia 1997, Argentina 2001. Each side had its own preferred models that functioned very well at explaining the past historical cases that they focused on. But there was no way right now of settling, empirically, whether a model built to explain the U.S. in 1985 or Korea in 1998 was more applicable to the U.S. in 2006--you had to make a bet, either that continuities in U.S. economic structure were important, or that financial globalization was important, in choosing your model and your terms of analysis.
It was very interesting. And very disturbing. Brilliant economists, thinking hard, unable to reach even the beginnings of analytical agreement about how to model the distribution of possible futures.
