Remarks at the Annual SIEPR Summit Stanford University

March 3, 2006

The Case for a Competetive Dollar

By MARTIN FELDSTEIN

(PDF Version)

Tonight I want to discuss an important subject: the value of the U.S. dollar. My theme this evening is that America needs a competitive dollar. More specifically, we need a competitive exchange rate relative to the other major currencies of the world an exchange rate that will make American goods more attractive to foreign buyers and that will cause American consumers and firms to choose American made goods and services.

For more than a decade, under both Democrats and Republicans, official Washington has emphasized a very different theme. That message has been that "a strong dollar is good for America." I think it is time to change the message. The message should be that we need a competitive dollar abroad and a strong dollar at home.

It is important to distinguish between the strength of the dollar at home and the value of the dollar relative to foreign currencies. A strong dollar at home is one that maintains its overall purchasing power in the domestic market. Stated differently, a strong dollar at home is one whose purchasing power is not eroded by domestic inflation.

The key question is whether these two goals a strong dollar at home and a competitive dollar internationally are compatible in practice. And even if they are compatible in the long run, can the dollar shift to a more competitive level without generating inflationary pressures that would cause the dollar to be weak at home?

I believe that they are compatible. We can have an internationally competitive dollar without inflation and we can shift to a competitive level of the dollar without raising the rate of inflation. Before I present the evidence to support this conclusion, let me discuss why a competitive dollar would be a good thing for the United States.

We now have an enormous trade deficit. American imports exceeded American exports last year by about $725 billion. That's twice the trade deficit that we had in 2001 and more than four times the trade deficit that we had in 1998.

To finance this trade deficit, the U.S. has to borrow from the rest of the world or sell American assets like stocks, businesses, and real estate to the rest of the world. In addition to borrowing to pay for our trade deficit, we must also borrow to pay the net interest and dividends that we owe on the amounts that we have previously borrowed or on the assets that we have sold to foreign buyers. Unless the trade deficit shrinks, the combination of the trade deficit and the interest and dividend payments to foreigners will grow ever more rapidly. This combined total, i.e., the current account deficit, was $800 billion for 2005. This was more than 6 percent of GDP and an increase of more than $100 billion from the previous year.

Until a few years ago, most of the funds from abroad came in the form of equity as private investors around the world thought that the risks of cross-border investing in the United States were outweighed by the higher potential return. Now the capital inflow is no longer in the form of equity funds. The $800 billion inflow last year was essentially all in the form of foreign purchases of US bonds or deposits in US banks, including the deposits in the Federal Reserve bank by foreign central governments. Although there is no way to know who is providing the very large capital inflows, I think it is likely to be foreign governments and organizations acting on behalf of such governments.

This form of capital inflow poses a serious risk to the American economy. Foreign governments may decide to shift their funds from dollars to other currencies. Similarly, private investors who are current holders of US stocks and bonds may decide to make such a shift. But they cannot collectively shift their funds out of dollars in the near term and cannot even reduce the flow of dollars to the United States. The trade deficit has to be financed and that requires a corresponding inflow of funds. To bing those funds would require higher rise in interest rates (or a fall in the value of the dollar). A shift in preferences away from the dollar could cause the interest rate in the US could rise rapidly. I don't have to tell you about the damage that would be caused by such a large rise in interest rates.

The only way that we can reduce our financial dependence on the inflow of funds from the rest of the world is to reduce our trade deficit. And the only way to achieve a substantial reduction of the trade deficit is by having a more competitive dollar. I won't speculate on how far the dollar has to fall to make the dollar competitive enough to reduce the current account deficit to a sustainable level. Back in 1985, when the current account deficit was less than 3 percent of GDP, the real tradeweighted value of the dollar fell 25 percent in 12 months and then continued to fall until it was down nearly 40 percent. Some experts predict that the dollar's fall would now have to be even larger in order to reduce the trade deficit to a level that can be comfortably financed.

I sometimes hear the claim that a decline of the dollar's value is not needed because the the US trade deficit would be reduced substantially if other countries would just grow more rapidly. It is certainly true that faster growth in Europe and Japan would lead to more imports by those countries and therefore a smaller US trade deficit. But the magnitude of the reduction in our trade deficit that could be achieved in this way would be quite small. Even if the growth rates throughout Europe and Japan were to double for the next five years, our trade deficit would be reduced by less than two percent of U.S. GDP less than one-third of the current deficit and we would have to wait five years until even that limited reduction was achieved.

There are moreover good reasons to start a major reduction of the trade deficit as soon as possible I will explain why in a few minutes.

But first I want to return to the question of whether the move to an internationally competitive dollar is compatible with maintaining a strong dollar at home. Or would lowering the dollar's value relative to other currencies cause inflation in the United States?

The good news is that a competitive dollar in the global market and a strong dollar at home are compatible in both the long run and during the transition to a more competitive dollar. This is shown by experience as well as being implied by economic logic.

Consider what happened in the 1980s, the last time that the dollar was substantially overvalued and fell sharply. Between 1981 and 1985 the trade deficit rose from $16 billion to more than $120 billion as imports surged and exports fell. Although these amounts are not large by the standard of today's trade deficit, they were the first annual trade deficits of more than $30 billion and a major reversal from the trade surpluses that the US had experienced until the mid-1970s. The primary reason for this trade deficit was the high value of the dollar relative to the yen and the European currencies. The real (i.e., inflation adjusted) trade weighted value of the dollar relative to other major currencies had increased by 40 percent between 1980 and the start of 1985.

But then in April of 1985 the dollar began a sharp decline. The dollar's trade weighted value fell 23 percent in just 12 months and by a total of 37 percent by the beginning of 1988.

The more competitive value of the dollar turned around the trade deficit. Exports started to rise, with merchandise exports up more than 40 percent in two years while the corresponding imports rose at only half that pace. Since imports were initially much larger than exports, it took a few years for the trade deficit to decline. But by 1989 it was down by more than 40 percent from its peak. So the lower dollar was unquestionably effective in shrinking the trade deficit.

Although the rapid fall of the dollar caused the price of imports to rise sharply, the overall inflation rate did not rise. The CPI inflation rate, which was 3.9 percent rate in 1984 (and almost exactly the same in the previous two years) actually declined to 3.8 percent in 1985 and to 1.1 percent in 1986. Between 1985 and the start of 1988, while the dollar fell 37 percent, the inflation rate averaged only 3.1 percent a lower rate than in the previous several years. Although a decline in oil prices contributed to this lower rate of inflation, the core rate of inflation that excludes energy prices also fell during this period. That's not a guarantee that a dollar decline won't raise inflation, but is shows that it was possible to have a sharp dollar decline with no adverse effect on inflation.

The good behavior of domestic inflation while the dollar was falling should not come as surprise. Domestic inflation reflects domestic monetary policy. With sound monetary policy, the increases in import prices are offset by lower inflation for domestically produced goods and services. So we can have a strong dollar at home and simultaneously a competitive dollar abroad. This is true not just in the long run but, as the experience of the 1980s demonstrated, during the transition as well.

The decline of the overvalued dollar in the 1980s also had no adverse effect on economic activity in the United States. The rise in national saving to free up the resources needed to reduce net imports happened without higher unemployment The unemployment rate came down year after year from 7.5 percent in 1984 to 5.5 percent in 1988. Bond yields fell and the stock market rose sharply, with the S&P index up 40 percent from 1985 to 1988.

In short, both experience and economic theory imply that the US could now t to a more competitive dollar without experiencing either increased inflation or decreased economic growth.

This favorable US experience was of course quite different from the recent experience of the emerging market countries in which sharp currency declines occurred. In Korea, Thailand, Argentina and other emerging market countries, currency declines in the late 1990s led to declines of GDP and rising inflation. The primary reason for their poor performance is that the business firms in those countries had substantial amounts of outstanding debt denominated in dollars rather than in the currency of their own country .

For example, when the value of the Thai bhat fell relative to the dollar in 1997, Thai businesses saw the value of their debts (measured in bhat) rise substantially, often forcing them into bankruptcy. Thai banks that had borrowed dollars on the world market and loaned those dollars to local businesses were unable to collect on their loans, causing a collapse of those banks. Similar things happened in other emerging market countries that were burdened with a serious currency mismatch. Faced with declining output and employment, the central banks in those countries did not pursue the kind of tight monetary policies that would have been needed to avoid a rise in inflation.

But because we in the United States finance our current account deficit by borrowing in our own currency, we can move to a more competitive dollar without the adverse effects that followed currency declines in other countries.

Some skeptics argue that a more competitive dollar would not reduce our large trade deficit. Despite the favorable impact of the lower dollar on our trade deficit in the 1980s, they fear that US exporters can no longer compete with foreign products and that American manufacturers could not compete with imports from low wage countries even if the dollar declined again by nearly 40 percent.

I think these skeptics are wrong. The United States is a major exporter even with today's overstrong dollar. In 2005, the US exported $892 billion worth of goods, including $450 billion of capital goods, $116 billion of consumer goods and $230 billion of industrial supplies and materials. A lower dollar would raise the volume of sales in all of these categories.

The response of imports to a more competitive dollar would be more complex. Although it is true that U.S. manufacturers would still generally be unable to compete with producers in low wage countries, it is important to realize that nearly half of our imports come from high wage countries in Europe, Canada and Japan. A dollar that is more competitive relative to the currencies of those countries will cause Americans to buy more US made products instead of imports from those countries.

An increase in the relative price of products from the low wage manufacturers in Asia and Latin America will also make those products less attractive to American consumers. Even if the same kinds of products are no longer made in the United States, American consumers will respond to the higher price of those products by buying less of them and spending more of their money on other goods and services made in the United States.

For example, in my own case, I buy camera equipment and like to have the latest technology. Much of that equipment is made in China and Thailand. If the prices of that equipment were to rise by 30 percent, I would probably buy new equipment less frequently. Some of the money that I now spend on cameras and lenses would go instead to restaurants and other services here in the United States. And when the dollar decline makes foreign travel much more expensive, I will do more of my vacation traveling in the United States.

Although economists have studied the sensitivity of import and export volumes to changes in the exchange rate, there is still much uncertainty about just how much the dollar must change to bring about any given reduction in our trade deficit. But that should not be interpreted as saying that the deficit reduction will not happen in response to a lower dollar. Rather it says that we cannot know how large the fall in the dollar must be to bring the trade deficit down to a sustainable level.

I noted earlier that it would be good for the dollar to become more competitive now. Why not wait until next year or the year after? After all, an overvalued dollar gives us the ability to buy foreign goods at lower prices. And the existing volume of exports brings more yen and euros than they would if the dollar were more competitive.

There are two reasons to hope that the dollar will become more competitive sometime soon. First, the longer we wait, the larger will be our accumulated debt to the rest of the world. And when that imbalance is larger, it will take a larger fall of the dollar to repay the debt incurred by today's imports. The gain in purchasing power that we enjoy as a result of the overstrong dollar today will be offset by the extra loss in future purchasing power of the dollar that will be needed to service and repay the future debt.

But the primary reason for wanting the dollar to become more competitive in the near future is that we may need an increase in exports this year and in 2007 to sustain the economy's current pace of expansion. Although forecasters generally believe that the most likely outlook for the US economy in 2006 and 2007 is a continuation of solid economic growth, there is a serious risk that a sharp decline in consumer spending will cause an unwanted GDP slowdown or even an economic downturn. The principal reason for this concern is the very low level of the household saving rate.

Let me explain. The household saving rate has been falling since the early 1990s, declining from 7 percent in 1990 to just 2.4 percent in 2002. The primary reason for this decline has been the increasing wealth of households that resulted from the rise in the stock market and in the value of owner-occupied homes. Despite the stock market fall in the year 2000, the level of the S&P share price index was nearly three times as high in 2002 as it had been in 1990. During the same dozen years, home prices also rose very significantly.

This rise in wealth caused savers to save less and dissavers to dissave more, bringing down the net household saving rate. Households in their 40s and 50s, who would normally be the primary savers, looked at the increased wealth in their IRAs and 401k plans and the increased values of their homes and concluded that they did not have to save as much for their retirement as earlier cohorts did at the same ages. And the retirees in their sixties and seventies who looked at their own net worth concluded rightly that they could afford to dissave more than previous generations of retirees. With less saving by the savers and more dissaving by the dissavers the overall household saving rate fell.

The gradual decline in the saving rate that began in 1990 suddenly became a sharp drop in the past few years. The saving rate fell from 2.5 percent in the summer of 2003 to a negative minus 1.6 percent in the summer of 2005. This negative saving rate and the corresponding sharp rise in consumer spending was made possible by the massive refinancing of home mortgages. That mortgage refinancing was the result of the combination of falling mortgage interest rates (from over 7 percent in 2002 to 5.5 percent at the start of 2004) and rising home prices.

As a result, total mortgage borrowing exploded from about $800 billion in 2002 to an annual rate of more than $1.5 trillion in 2005. The $700 billion rise in mortgage borrowing was equivalent to more than 7 percent of disposable personal incomes. Refinances are running at about 40 percent of total mortgages, implying an increase in refinancing equal to about 3 percent of disposable incomes. Additional net equity withdrawal from real estate also occurs as retirees sell their homes and become renters.

Refinancing allowed households to take out cash to spend on other things while also reducing their monthly mortgage payments. The enormous magnitude of the increase in refinancing made it possible for the saving rate to turn negative.

This rise in refinancing and the associated negative saving rate are now likely to be reversed. The principal driver of refinancing lower mortgage rates has turning around with the mortgage interest rate up from 5.5 percent to more that 6.25 percent. For most homeowners, there is no longer the incentive to refinance. The declines in the sales of new and existing homes and the increased inventories of unsold homes suggest that the rise in house prices will slow in the months ahead, removing the other source of refinancing. The actual volume of mortgage refinancing has already started to decline.

The reduced level of mortgage refinancing will shift the household saving rate from negative to positive. Even if the saving rate returns only to the low 2.5 percent level that we had in the summer of 2003, the rise in saving will be about 3.5 percent of disposable income or about 2.5 percent of GDP. This relative fall in consumer spending would be a very large drag on overall aggregate demand.

What can take the place of the lower consumer spending to maintain overall aggregate demand? Business investment is unlikely to rise faster when slower consumer spending is reducing total sales. Housing construction is already in decline.

The key to maintaining aggregate demand, i.e., the key to our continued expansion, must be a shift in our trade balance increased exports, lower imports, and more spending on goods and services produced in the United States. For this to happen, the dollar must decline to make U.S. goods and services more attractive.

Even if the dollar does decline during the coming months, the delays in the response of exports and imports to the more competitive dollar will mean that the increase in aggregate demand from this source may not happen for a year or more.

That brings me back to my main theme. The United States needs a more competitive dollar. And we need to see the shift to a more competitive dollar starting as soon as possible.

What would it take to make that happen? The 1985 Plaza Accord suggests one possible approach. At that meeting, in September 1985, the finance ministers of the G-5 countries acknowledged that the dollar was overvalued and needed to decline further. Monetary and fiscal policies had adjusted in the US and foreign countries, just as they are doing today. With the official statement that the dollar needed to decline relative to our trading partners, the markets did the rest, confident that the US government did not have a plan to punish those who sold dollars by engineering a sudden reversal of the dollar.

The G-5 finance minsters (or the G-7 successor group) may not be the appropriate group now since much on the U.S. trade imbalance is with countries that are not part of the G7. It might be best to start with a meeting with major Asian trade surplus countries convened by China in which there is an agreement for a coordinated appreciation. But the specific institutional arrangement is less important than the official US acknowledgment that the dollar has to come down and that there is no program or desire to support it at its overvalued level.

If the Federal Reserve pursues a strong dollar at home while the dollar becomes more competitive in global markets, we can achieve both price stability and a more balanced path of economic growth.

The writer is professor of economics at Harvard University and president of the US National Bureau of Economic Research.