The U.S. trade deficit currently runs at the unusually highlevel of 5 percent of GDP, and something like 80 percent of the netdeficit has recently been funded not by private capital flows, butby foreign central bank purchases of dollars. U.S. netinternational indebtedness, which summed to less than 10 percent ofGDP in 1998 or 1999, now exceeds 25 percent of GDP and continues togrow. In late 2004 and early 2005 it became commonplace toanticipate that the dollar would plunge and interest rates wouldrise. Such prestigious figures as Paul Volcker and Joseph Stiglitzhave predicted that a dollar crisis is likely.

As this is written, that has not happened--the dollar hasstrengthened against the euro, the pound and several Asiancurrencies. After much American pressure, the Chinese have revaluedthe renminbi, but, at least initially, only by a minimal 2.1percent. Long-term dollar interest rates remain higher than euroand yen rates, but they were boosted only slightly by the Chineseaction and have continued to decline to their lowest levels indecades.

The trade balance is overrated as a driver of currency values;consequently, allowing currencies to float usually has littleimpact on trade deficits or surpluses. The market value of thedollar is driven also by all sorts of other factors, includingsupply and demand for its use as a national monetary reserve, as ashort- and long-term store of value, as an investment vehicle, ascirculating currency, and for facilitating trade between countrieseven where U.S.-produced goods are not involved. By emphasizingspecific national trade deficits or surpluses, we are led topropose the wrong solutions and may indeed aggravate contractionarypressure from elsewhere in the world economy--what I would term theslow-growth trap.

In part a consequence of the popularity of flexible exchangerates among economists, the post-World War II Bretton Woods regimeof fixed exchange rates was allowed to break down in phases duringthe early 1970s. Yet, despite many economists' embrace offree-floating currencies, Europeans and Asians subsequently movedin the opposite direction and sought to stabilize exchange ratesand re-establish currency blocs. The euro was introduced in 1999but was preceded by European exchange rate mechanisms dating backto the 1970s. By 1980 Hong Kong, Korea, Malaysia, Singapore, Taiwanand Thailand had more or less fixed their currencies to the dollar,creating a regional dollar standard, until the "Asian crisis" of1997 led some of them to devalue or float. The value of the Chineserenminbi hardly budged against the dollar from 1994 to 2005. Sincethe end of the 1997-99 crisis, a number of Asian governments andcentral banks have attempted to restore pre-crisis dollar pegging,this time using softer, changeable pegs. In the wake of the crisis,and facing opposition from the U.S. Treasury and the InternationalMonetary Fund (IMF)--both of which embrace flexible exchange ratesas doctrine--few would attempt a harder fix. Japan has joined thestabilizers in an effort to limit yen appreciation against thedollar.

Whatever floating-rate advocates may propose, there are avariety of reasons why developing countries often choose to pegtheir exchange rates, of which two stand out. First, mostdeveloping countries lack forward markets, which allow importersand exporters to hedge exchange exposure. A dollar peg provides apractical hedge, as it reduces the exchange risk in cross-bordertransactions. Second, many governments limit foreign exchangeexposure by domestic banks, for good, prudential reasons, but doingso prevents banks from being active dealers to stabilize theexchange rate. In this circumstance, floating would likely lead todisruptive volatility.

Much of the recent discussion about U.S. trade deficits hasfocused on fiscal deficits and inadequate U.S. savings. But growingtrade deficits and downward pressure on the dollar in part reflectcontractionary pressure arising outside the United States. U.S.exports may be low because of weak demand for consumption andbecause of investment, and consequent oversaving, elsewhere. U.S.capital inflows may be high because other economies providerelatively few opportunities for direct or portfolio investment. Weneed to consider the systemic consequences--that is, theconsequences for the world economy as a whole--of any measuresintended to address deficits. It is in this context that allowingother currencies, including the yen, renminbi and euro, to driftupward becomes problematic. By reducing the cost of imported goodsand materials, higher currency value puts relative downwardpressure on domestic prices. Where systemic price inflation ismodest, as has been the case for most of the past couple decades,downward price pressure in countries where currencies appreciatemight generate actual deflation.

Concern that the dollar may continue to decline, or, in anyevent, be volatile, leads investors to expect higher interest rateson dollar assets than on assets of surplus-country currencies.Where dollar rates are themselves declining toward low singledigits--that is, toward 4, 3 or 2 percent annualized--then interestlevels in trade surplus countries fall to close to zero percent,too low to make lending profitable. An illiquid financialenvironment constrains investment in both export and home sectorsand reinforces patterns of underspending and oversaving.Consequently, the U.S. trade deficit and overborrowing andoversupply of dollars are aggravated, which leads to furtherdownward pressure on the U.S. currency.

The Case of Japan

The nearly decade-and-a-half Japanese tailspin is usuallyassumed to have origins in some national economic malfunction--evenby financial sector economists who should know better. In fact, thecontractionary mechanism outlined above is most clearly at work inthe case of Japan, which has in turn been a trigger for East Asianinstability. The yen strengthened from 250 to the dollar in 1985 to80 to the dollar in 1995 before weakening. Since 1985, Japan hasexperienced mild but persistent deflation of wholesale prices. Thecombination of price deflation and an often-rising yen forcedinterest rates down, usually to several percentage points belowdollar rate levels. In a deflationary environment, firms do notwant to borrow for expansion, nor do they want to owe money in acurrency that has tended to gain external value. Bankers do notwant to lend at interest rates too low to generate profits or tocompensate for the risk of default. Investors prefer not to buylong-term bonds at very low interest rates; simple bond mathematicsindicates that their principal values would collapse if interestrates were some day to rise.

But despite the sharp depreciation of the yen from 1995 to 1997,and its essentially trendless behavior since, yen interest ratestypically remained 2 or 3 percent below dollar rate levels, whichsuggests that Japanese institutional investors now attach a largevolatility premium to holding dollar-denominated paper. In the faceof ongoing U.S. borrowing, dollar holdings in the portfolios ofJapanese financial institutions have inevitably risen. Importantempirical data indicate that the negative risk premium on yeninterest rates has increased as foreign assets become a largerportion of total financial-sector balance sheets. Since liabilitiesare in yen--in the form of bank deposits and pension and insurancepayout obligations--increased dollar holdings place at risk thesolvency of many institutions and even of the financial system as awhole. Japanese borrowers are thus inclined to convert dollarholdings into yen, and, in environments of low internationalinflation, demand for yen keeps Japanese interest rates close tozero.

A first consequence of Japanese instability is that the floatingyen has often been out of step with other currencies in East Asia,most of which have soft or hard pegs to the dollar. A decline inthe exchange value of the yen during the sharp dollar recovery of1995-97, for example, upset trade and investment patterns andhelped trigger the regional financial crisis that began in 1997. Italso forced several Asian countries either to devalue or to floatdownward against the dollar. The yen in the past was often forcedupward, not for market-induced reasons, but because U.S.politicians were exerting pressure to pry open Japanese trademarkets. During the past half-decade, Japanese authorities havemade an effort to manage the yen so that it tracks the dollarsomewhat more closely, so disruption from yen exchange ratemovements has been less. Even so, the yen appreciated sharplyduring the fourth quarter of 2004, and concern about futureappreciation and price deflation are embedded.

A second effect followed from the domestic deflation in Japan.Much of the trade advantage U.S. manufacturers would theoreticallygain through yen appreciation has been offset by lower Japanesedomestic prices. From 1950 through about 1977, Japanese wage levelsincreased far faster in nominal terms than did wage levels in theUnited States--reflecting greater Japanese productivity growth.Deflationary pressure, induced by yen appreciation, then led to anabrupt reversal so that, since 1977, U.S. nominal wage growth hasexceeded that in Japan in all but a few years. After decades of yenappreciation, Japan continues to run large bilateral tradesurpluses with the United States, which are the flip side ofongoing capital outflow. The underlying dynamic for capital exportshas gradually shifted from providing finance abroad for boomingexport-driven growth in Japan to (more recently) sending moneyabroad because of a lack of investment opportunities in Japan. Buteven this channel of adjustment is becoming blocked as manyJapanese investors now prefer to hold yen assets; hence, a growingportion of dollars are instead bought and held by the Bank ofJapan. Without the Bank of Japan's intervention, Japanese capitalsurpluses would be larger, and the logic of economic slowdown wouldagain push the yen higher, and domestic prices lower, in aself-amplifying way.

A third effect of the instability has been Japanese reluctanceto deregulate. In the face of rising exchange rates and shrinkingdemand, legislators and officials have not been willing to reducethe safety nets implicit in maintaining protected industries.Indeed, it is hard to identify important economic benefits gainedby either Japan or the United States from decades of flexibleexchange rates--exchange markets have driven fundamentals, ratherthan the other way around. The Japanese economy appears to havefallen again into recession during the fourth quarter of 2004, andthe volume of bank lending has contracted every year since1998.