growth, ranging from increases of about 3 percent in France to one-half of 1 percent in Germany. Japan, with a drop of 5 percent, was the only country studied to experience a decline in productivity in 1992. (See table 1.)

The U.S. manufacturing productivity increase in 1992 represents a substantial improvement over the previous year's performance and is the largest increase the United States has experienced since 1987. Productivity growth also improved in 1992, relative to 1991, in many of the foreign countries studied. The exceptions to this pattern were Germany and Japan, where productivity performance was worse in 1992 than in 1991, and Denmark and the Netherlands, which experienced about the same rates of increase in both years.

While productivity growth was improved in many countries, these gains were accompanied by declines in employment and hours worked in virtually every case. The productivity increase in the

Recent exchange rate changes

As of September 1993, the currencies of 12 of the 13 foreign economies studied had depreciated substantially. The exception was Japan, where the yen appreciated throughout the year and was up 20 percent against the dollar in September, relative to the yen's average 1992 value.

The Korean won and the New Taiwan dollar depreciated slowly but steadily throughout the year and by September, the won was down about 3 percent while the New Taiwan dollar had depreciated about 7 percent. The Canadian dollar depreciated about 9 percent.

The currencies of the European economies generally followed a pattern of depreciation in the first 3 months of 1993, a moderate appreciation in April and May, further decline through August, and some strengthening in September. As of September 1993, the German mark and the Dutch guilder were down about 4 percent, relative to their 1992 averages; the Belgian, Danish, and French currencies were down 7 percent to 9 percent; the Norwegian krone and British pound were down about 13 percent; the Italian lira, 21 percent; and the Swedish krona, 27 percent.

In the first three quarters of 1993, U.S. unit labor costs were 2 percent below both their level for the same period in 1992 and the annual average for 1992. Consequently, these relative exchange rate changes suggest that U.S. manufacturing competitiveness probably improved substantially relative to Japan, but may have deteriorated, compared with Canada and Europe.

United States resulted from a combination of a 3percent increase in manufacturing output and a drop of 1 percent in labor input (as measured by hours worked). Increases in productivity for Canada, Belgium, Denmark, and France also were brought about by rising output and falling labor input, while Norway experienced an increase in output, but no change in hours worked. Despite declines in output, Germany, Italy, the Netherlands, Sweden, and the United Kingdom all achieved productivity increases because the hours worked measure fell inore. In Japan, a 1-percent decrease in hours worked was not enough to offset a 6-percent drop in output.'

Output and labor input. U.S. manufacturing output was up 3.2 percent in 1992, following 2 consecutive years in which output declined. The U.S. output increase was exceeded only by Korea (5 percent) and Taiwan (3-1/2 percent). Canada, Belgium, Denmark, France, and Norway had smaller output increases of less than 2 percent, while Italy, the Netherlands, Sweden, and the United Kingdom experienced declines of less than 1 percent. Sharper output declines occurred in Germany and Japan.

Employment in manufacturing fell in 11 of the 12 countries for which this measure was calculated. Japan was the only country to increase manufacturing employment (nearly 2 percent) in 1992. Manufacturing employment declined about 2 percent in the United States, a larger relative decrease than occurred in Denmark, Germany, the Netherlands, or Norway, but smaller than the employment reductions in Canada, Belgium, France, and Italy, which posted declines ranging from 2-1/2 percent to 4-1/2 percent; the United Kingdom, which experienced a drop of about 5-1/2 percent; and Sweden, where employment was down by 9 percent. All of the countries, except Japan and Germany, underwent employment declines also in 1991.

Although employment rose in Japan in 1992, average hours worked were reduced by 3 percent, resulting in about a 1-percent decline in total hours. Total hours also fell in all of the other countries except Norway, where hours were unchanged. As with employment, hours fell in all countries except Japan and Germany in 1991.

Hourly compensation costs. U.S. manufacturing hourly compensation costs-which comprise wages and salaries, supplements, and employer payments for social security and other employerfinanced benefit plans-increased 4-1/2 percent in 1992. This figure was about in the middle of the range for the countries studied. Hourly compensation costs increased at about the same rate in Japan and Belgium as in the United States, while

Table 1.

Canada, Denmark, France, Norway, and Sweden all experienced smaller increases. Germany, Italy, the Netherlands, and the United Kingdom each recorded increases of between 5-1/2 percent and 8-1/2 percent.

All of the countries studied had smaller hourly compensation increases in 1992 than in 1991, except the Netherlands, where the rate increased only slightly. The reduction in the U.S. rate of increase, relative to 1991, was less than those in most of the other 10 countries.

Unit labor costs. U.S. unit labor costs in manufacturing were virtually unchanged from 1991 to 1992. Canada also had about the same unit labor costs in 1992 as it had the year before. Swedish unit labor costs fell 3-1/2 percent, but the other

European countries had unit labor cost increases, ranging from less than 1 percent in Belgium and France to more than 5 percent in Germany. The three Asian economies sustained the highest increases; 6 percent in Taiwan, 7 percent in Korea, and 10 percent in Japan-the largest increase in that country since 1975. The underlying causes for the substantial unit labor cost increases in Korea and Taiwan differ from those leading to the sharp increase in Japan. In Korea and Taiwan, the increases occurred in the context of relatively strong output growth and are not out of line with recent experience. The Japanese unit labor cost increase in 1992 was primarily the result of an economy in recession. That Japanese firms, particularly the larger firms, traditionally have been reluctant to shed regular employees during cyclical downturns

is probably an exacerbating factor. This practice hinders them from reacting quickly or fully to cut total labor costs in the face of declines in output, despite reducing the rate of increase in hourly compensation costs.

Unit labor costs in U.S. dollar terms. The U.S. competitive position vis-à-vis most other economies, as measured by unit labor costs adjusted for changes in the exchange rate, improved in 1992. The U.S. position was enhanced by flat unit labor costs and a weaker dollar relative to the currencies of Japan and most of Europe. In only two of the economies studied did the home currency depreciate relative to the U.S. dollar between 1991 and

1992. The Canadian dollar fell 5 percent against the U.S. dollar-the first decline in 6 years-and the Korean won fell 6 percent. The British pound was unchanged relative to the dollar, and the Italian lira appreciated less than 1 percent. Currencies in Norway and Sweden appreciated in the 4-percent to 4-1/2 percent range, while Japan, Taiwan, and the other European countries saw their currencies appreciate between 6 percent and 7 percent against the U.S. dollar.

The weaker dollar meant that, for many economies, unit labor costs increased substantially more when measured in U.S. dollars (that is, adjusted for changes in exchange rates) than when measured on a national currency basis. The most notable exception to this pattern was Canada, the only country to experience a decline in unit labor costs on a U.S. dollar basis in 1992. On a national currency basis, Canadian unit labor costs were unchanged over the year, as were unit labor costs in the United States. When measured in U.S. dollars, however, Canadian unit labor costs fell 5 percent, solely because of the depreciation of the Canadian dollar.

Aided by the depreciation of the won, Korean

unit labor costs measured in U.S. dollars rose only

one-half of 1 percent. The United Kingdom, where the pound was unchanged relative to the dollar over the year, showed an increase of about 3-1/2 percent. Swedish unit labor costs, which fell on a national currency basis, were unchanged on a U.S.-dollar basis. The remaining economies found their unit labor cost increases exacerbated by exchange rate movements. Unit labor costs in U.S. dollars rose between 5 and 7-1/2 percent in Belgium, Denmark, France, and Norway. Japan, Taiwan, Germany, and the Netherlands faced U.S. dollar-basis unit labor cost increases in the double digits, with Japan experiencing the greatest increase, at about 17 percent.

Comparative trends, 1979-92

Comparable U.S. manufacturing output data currently are not available before 1977. (See the box

on the U.S. output measures, page 54.) Consequently, the analysis of long-term trends is restricted to the 16-year period from 1977 forward. However, it is more useful to choose 1979, a peak year for U.S. manufacturing output, as a starting. point for analysis. The year 1979 also provides a convenient starting point for analysis of the foreign economies covered, because most of them also recorded manufacturing output peaks in 1979 or 1980. Japan, which did not experience a decline in output until 1986, is the exception to this pattern.

It is useful to divide the 1979-92 period studied at 1985 because U.S. manufacturing productivity growth accelerated in the second half of the pe

riod. In addition, the trade-weighted value of the dollar rose strongly between 1979 and 1985, then reversed itself and fell even further between 1985 and 1992. This makes 1985 an especially relevant breaking point for the analysis of changes in competitiveness stemming from movements in unit labor costs and currency exchange rates.

Productivity. U.S. manufacturing productivity rose at an average annual rate of about 2-1/2 percent between 1979 and 1992. This rate of increase placed the United States about in the middle of the range of the 11 foreign economies covered, although 7 of the 11 had growth rates that were slightly to substantially higher than the U.S. rate. Canada, Denmark, and Germany had average annual productivity growth rates of between 1 percent and 2 percent; France, the Netherlands, Norway, and Sweden had average annual gains in the 2-1/2 percent to 3-percent range; and Japan, Belgium, Italy, and the United Kingdom experienced the highest rates of increase-between 3-1/2 percent and 4-1/2 percent per year.2

Between 1979 and 1985, U.S. manufacturing

productivity grew at a rate of only 2 percent per

year. While Denmark and Germany recorded similar average annual rates of increase, all of the remaining countries had more rapid productivity growth. Belgium led the group during this period, with growth of 6 percent per year, followed by Italy, with a 5-percent rate of increase, and the United Kingdom, with 4-1/2 percent.

U.S. manufacturing productivity growth improved substantially in the 1985-92 period, rising to 2.8 percent per year. In contrast to the earlier subperiod, U.S. productivity growth between 1985 and 1992 was exceeded only by that of Japan and Italy, at 3 percent to 3-1/2 percent per year, and the United Kingdom, at 4-1/2 percent. Japan's average growth rate of 3-1/2 percent for this period was pulled down substantially by a particularly poor showing in 1992. The U.S. rate of increase was matched by Belgium, France, and Sweden. The other five countries had lower productivity

growth rates during this period. They ranged from about 1 percent annually in Canada and Denmark to 2 percent per year in Germany.

Average productivity growth rates for the 1979-85 and 1985-92 periods are shown in chart 1 for the United States, Canada, Japan, Europe (trade-weighted average), and selected European countries.

The United States was the only country to experience a higher productivity growth rate in the 1985-92 period than in the 1979-85 period. However, the United Kingdom maintained its high growth rate of 4-1/2 percent over both subperiods. Japan and France saw small dropoffs of about one-half of 1 percentage point in their average rates of increase, and Germany's productivity growth remained about constant between the two periods. At the other end of the spectrum, Belgium's average growth rate fell to about 3 percent, from 6 percent, in the 1979-85 period, Canada's growth rate dropped from 2-1/2 to less than 1 percent, and the Netherlands saw a decline from more than 4 percent per year to about 1-1/2 percent. The drop in Japan's average rate of growth was entirely attributable to that country's economic performance in 1992. If 1991 is used as the endpoint for the analysis instead of 1992, the

difference in Japanese growth rates between the earlier and later subperiods becomes a 1-percent increase, rather than a decline.

Rising output with stable or increasing labor input is the image typically called to mind when productivity gains are discussed. The 1979-92 productivity gains recorded by most of the countries, however, resulted about as much or more from declining employment and hours as from output increases. The United Kingdom achieved its 4-1/2-percent productivity growth over the period primarily through a 4-percent annual reduction in hours worked; manufacturing output rose less than one-half of 1 percent per year. Norway's 2-1/2-percent productivity growth rate resulted almost entirely from a decline in total hours, as output remained almost unchanged over the period. Productivity gains for Canada, France, and Sweden resulted more from reducing labor input than from gains in output, and hours reductions were about as large a factor as output gains in the United States, Belgium, Germany, and Italy. There were two exceptions to the pattern of productivity gains resulting partly from declines in labor input: Japan, where both output and hours rose, and Denmark, where output rose and employment and hours were about unchanged.