Printer Friendly

U.S. productivity performance in perspective.

This article examines U.S. productivity performance since World War II, by sub-period, in relation to its record over the century since 1889. The U.S. record then is compared with that of other OECD countries after 1950. Results of a recent growth accounting model published by this author (1) are used to help explain variations in sub-period rates of productivity growth, particularly the 1973-81 slowdown and subsequent pickup, as well as the differences among nations in productivity trends.

IS THE PRODUCTIVITY performance of the U.S. economy in recent decades really as poor as some critics contend? They point out that, of all the industrialized nations belonging to the Organization for Economic Cooperation and Development (OECD), the United States has had the lowest trend rate of productivity growth since 1950. They note that even in relation to its own historical trend, U.S. productivity growth slowed markedly after 1973. This is seen as a major contributor to subsequent large trade deficits. However, the inter-temporal and international perspectives are indespensable for a sound interpretation of recent U.S. developments and as a background for projecting the future growth of productivity and real gross product.

RECENT U.S. PRODUCTIVITY TRENDS

IN HISTORICAL PERSPECTIVE

Annual total and partial productivity estimates spanning the past century are available by linking the official series from the Bureau of Labor Statistics (BLS), which go back to 1948, to the series prepared by the present writer for the National Bureau of Economic Research (NBER) back to 1889. Over the ninety-nine-year period 1889-1988, a 1.6 percent a year average increase in total factor productivity (TFP) accounted for approximately half the 3.3 percent rate of growth in real gross product in the U.S. private domestic business economy shown in Table 1. TFP is a weighted average of real product per labor hour and per unit of capital, which grew at 2.1 and 0.5 percent average annual rates, respectively. The weights are based on gross factor income

Table 1

Real Gross Product and Productivity Ratios

U.S. Domestic Business Economy

(average annual percentages of change,

1889-1988 by subperiod
 Productivity -- ratios of product to:
 Real gross labor units of total factor
 product hours capital input (TFP)
1889-1988 3.3 2.1 0.5 1.6
1889-1919 3.9 2.0 0.3 1.7
1919-1948 2.8 2.1 1.2 1.7
1948-1988 3.3 2.2 -0.1 1.4
 1948-1966 3.7 3.3 0.7 2.4
 1966-1973 3.4 2.0 -0.9 1.0
 1973-1981 2.0 0.7 -1.7 -0.2
 1981-1988 3.7 1.7 0.4 1.3
Source: 1948-1988, Bureau of Labor Statistics, U.S. Department
of
Labor; 1889-1948, John W. Kendrick, Productivity Trends in
the United States, National Bureau of Economic Research,
1961.


shares, which averaged about two-thirds for labor and one-third for capital over the century.

According to economic historians (Abramovitz and David, 1973), productivity growth had been modest for most of the nineteenth century. It averaged 0.5 percent a year for 1800-55 for real gross product per labor hour, and 1.1 percent for 1855-90. So the 2.0 percent average annual rate between 1889 and 1919 represented a major improvement.

Sources of the accelerated growth after 1889 have not been quantified, but several major contributors should be mentioned. The development of a rational and systematic study of management is dated by the International Encyclopedia of Management (1978) as beginning in 1886 with a paper by Henry R. Towne. But it was in the decade before World War I that work by Frederick Taylor, Henry Gantt, Harrington Emerson, the Gilbreths and others produced a body of knowledge and principles that could be dignified by the term "science." Emergence of a professional management association, meetings and journals helped to disseminate findings of management studies.

Just as important during this period was the founding of schools of business administration, beginning with the Wharton School at the University of Pennsylvania in 1886. Increasing numbers of executives of the expanding ranks of large corporations were university trained.

Beginning in the late nineteenth century, large companies also started setting up research and development (R&D) laboratories employing scientists and engineers. For example, General Electric's corporate lab was established in 1900. By 1920, when the first Directory of Industrial Laboratories was published, many dozens were listed. It is estimated that they spent over $50 million on R&D at the compensation rates of that year. (2) Besides, until well into World War I, net foreign investment into the United States augmented growth of capital per worker here and promoted transfers of technology from Europe, which was still ahead of this country in a number of industries.

Over the three major subperiods shown in Table 1, rates of increase in labor productivity and TFP were close to those for the full century. The somewhat lower TFP growth 1948-88 reflected lack of net gains in capital productivity, more than offsetting a slightly above-average rate of labor productivity gains.

Within each of the subperiods, variations occurred in rates of change in the productivity ratios across cycle averages, describing what have been called "long swings." Since World War II, as shown in Table 1, the strongest growth was between 1948 and 1966, with labor productivity gains averaging 3.3 percent a year and TFP 2.4 percent -- about 50 percent above their long-term rates. It is generally believed that this strong performance resulted from a drawing down of the backlogs of innovational and investment opportunities that had accumulated during the depressed 1930s and World War II. The growth of real productive capital was strong. A decline in the average age of fixed capital enhanced the effects of factor saving innovations embodied in new plants and equipment.

This abnormally strong expansion was not to continue. Between 1966 and 1973, the growth rates of real product and labor productivity fell back close to their historical averages. TFP growth was somewhat under the long-term rate as real capital stocks increased faster than output. My article in this journal in 1971 was among the first to recognize and explore reasons for this early phase of a "productivity slowdown" (Kendrick, 1971).

The worst was yet to come. In the 1973-81 sub-period, TFP actually declined slightly, and labor productivity rose at only 0.7 percent average annual rate. Much has been written about this slowdown. There is considerable agreement about the main causes, although the weights assigned to each differ. The explanations include the oil price shocks of 1973 and 1979 and the associated acceleration in inflation generally, slower growth in capital (including real R&D stocks); unfavorable changes in the age-sex mix of the labor force and less favorable interindustry shifts, expanded government regulations that increased costs and inputs but not output as measured, increased slack in the economy and reduced economies of scale. (3)

Following the 1981-82 recession, almost all the unfavorable factors were reversed. The growth rates of labor productivity and of TFP accelerated by 1.0 and 1.3 percentage points, respectively, bringing them to over 80 percent of their historical averages. Many commentators have overlooked this dramatic recovery since 1981-82, considering the entire period since 1973 as a long slowdown. The recovery suggests that the long-term trend remains intact. The recent book by Baumol, Blackman and Wolf (1989) is one of the few that recognize this. As the authors state ". . . none of the data offer rational grounds for the fears that the economy has suffered a slowdown in its long-term growth rate." (4)

The pickup in manufacturing was even more dramatic. Relative to a trend rate of growth in TFP of about 2 percent a year, manufacturing dropped to 0.2 percent 1973-81, but then rebounded to an unprecedented 3.5 percent average rate between 1981 and 1989. The decline in the value of the dollar during much of the latter subperiod really put pressure on manufacturing firms to cut costs to meet international competition. Industries not sharing in the acceleration after 1981-82 were concentrated in production of services largely for domestic markets.

INTERNATIONAL COMPARISONS SINCE 1950

At the end of World War II, the United States was clearly the leading economic power in the world. Our GDP per employed person was by far the highest among the OECD countries in 1950 (see Table 2). Yet between 1950 and 1989, U.S. productivity growth was lower than that of any of the others, which ranged from average annual rates of 2.0 percent for Canada and 2.2 percent for the U.K. up to 4.3 percent for Canada and 2.2 percent for the Japan (see Table 3). Note that these rates for the entire domestic economy are lower than for the business sector alone, because output in government and households is moved by the employment estimates without allowance for productivity increases.

In the business economies, the productivity gaps with the United States were substantially reduced by 1989. More generally, there was a marked convergence in productivity levels, as the rates of growth for 1950-89 showed a significant negative correlation with the 1950 relative levels.

There is considerable agreement that the tendency towards convergence represents technological "catchup" with the United STates. Given that

Table 2

Real Gross Domestic Product Per Employed Person

Relative Levels in the United States

and 12 Other OECD Countries (a)

(United States = 100.0, Selected Years 1950-1989)
 1950 1960 1970 1980 1989
United States 100.0 100.00 100.00 100.00 100.00
Canada 76.00 78.9 82.9 91.5 94.2
Japan 15.2 23.3 45.7 62.7 72.7
OECD-Europe 43.7 49.6 61. 74.2 77.3
 Austria 30.8 39.2 53.7 68.2 72.2
 Belgium 47.1 50.6 62.5 79.9 83.6
 Denmark 48.5 53.0 59.5 65.9 65.4
 France 38.1 47.6 63.8 80.2 85.9
 W. Germany 34.5 49.1 61.9 77.5 82.0
 Italy 29.3 41.7 63.0 81.8 87.3
 Netherlands 52.3 57.5 69.9 81.1 77.2
 Norway 43.2 50.4 58.1 73.8 80.4
 Sweden 44.0 (b) 52.1 62.6 67.0 67.7
 United Kingdom 53.8 54.6 58.3 66.5 71.5
(a) Currencies converted based on purchasing power exchange
rates.
(b) Extrapolated from 1960 to 1950 by J. Kendrick.
Souce: Bureau of Labor Statistics, U.S. Department of Labor:
unpublished
tables dated April 1990.


[TABULAR DATA OMITTED]

scientific and technological knowledge is international, lagging nations have the opportunity to draw on that pool and increase productivity faster than the leaders. This assumes that they have sufficient "absorptive capacity," as Abramovitz (1987) terms it, and he adduces evidence that the other relatively industrialized nations were in a position after World War II to effect major transfers of technology from the United States. This was done through multi-national corporations, patent licenses, imports of sophisticated capital goods, international journals and meetings, advanced study in the United States, etc.

There was not sustained convergence of productivity levels among ninety-six other countries for which real product per worker estimates are available according to a study by Zagardo (1990). Thirty-nine of them had lower productivity growth than the United States. Lack of adequate absorptive capacity is shown by relatively high illiteracy rates and/or low percentages of students in secondary schools, according to the evidence presented by Zagardo (1990) and Baumol, et al. (1989). Workers in countries that had above-average productivity gains, such as S. Korea and Taiwan, generally had higher educational attainments, along with better infrastructure and more favorable institutional arrangements.

According to a growth accounting analysis for the period 1980-88 by subperiod, recently published by this author (Kendrick, 1990), there were two main reasons for the higher labor productivity growth rates in other advanced countries than in the United States. The first was more rapid growth of real capital stocks per worker, reflecting higher rates of saving and investment in those countries. The second was a more rapid increase in technological knowledge and know-how as indicated by the residual growth rates. The residual was obtained by subtracting from the real product growth rates the contributions of the tangible factor inputs and the nontechnical forces affecting productivity, as identified by Edward Denison (1986). The technological residual averaged 0.8 percentage point over the 1960-88 period in the United states, 1.7 percent in OECD-Europe, and 3.0 percent in Japan. These percentages represented almost half of the labor productivity growth rates in the several countries, and more than 80 percent of the growth rates of total factor productivity (TFP).

It is particularly noteworthy that, whereas the average rates of growth in TFP and in the technological residual rose in the United States between the 1973-79 and 1979-88 subperiods, they declined in OECD-Europe and in Japan. This supports the interpretation that the rates of technological catchup of Europe and Japan were slowing as these areas attained or surpassed U.S. productivity levels in an increasing number of industries.

The fact that the U.S. balance of trade was deteriorating between 1981 and 1987 while the productivity gap between the United States and other OECD countries was narrowing helps refute those who link overall relative productivity performance to international competitiveness. The main factor in the trade balance is the foreign exchange rate of the dollar.

The marked appreciation of the dollar from 1981 through early 1985, as high U.S. interest rates associated with high and rising federal budget deficits attracted foreign investment, was the main cause of the rising trade deficit. Although the dollar has depreciated since, the "J-curve" effect did not wane until late 1987. Since then, the U.S. trade and payments balances have steadily improved, even though the 1989-90 growth recession temporarily levelled productivity.

Even if exchange rates only reflected relative purchasing power of currencies, the latter changes not only because of relative changes in productivity, but also because of relative changes in prices of inputs. Thus, throughout most of the period since 1950, when U.S. productivity growth was less than that of our major trading partners, wage rates rose even less, relatively. Thus, unit labor costs and prices rose less here than abroad except in Germany and Japan during parts of the period.

Firms in particular industries whose productivity rates declined relative to their counterparts abroad generally experienced difficulties in international competition as their relative prices rose. But it is deceptive to point to such industries as indicative of a general productivity malaise. There are always some industries whose productivity is falling and prices rising relative to their foreign competition, and vice versa. These changes in comparative advantage affect the composition of exports and imports but not the overall balance.

Looking ahead, U.S. productivity should recover from its cyclical slowing as 1991 progresses. For the decade of the 1990s, the trend rates of growth should be close to the historical average of 2.1 percent for real product per labor hour and 1.6 percent for TFP. The rates could be higher if federal government policies were more pro-growth and stimulated higher rates of saving and investment, tangible and intangible, human and non-human. Strong increases in real R&D outlays lifted their ratio to GNP from 2.14 percent in 1978 to 2.68 percent in 1985. The cut in the incremental R&D tax credit in 1986, among other factors, slowed the growth and the ratio settled back to 2.56 percent in 1989. But given the long lags between R&D outlays and their productivity effect, the R&D growth since 1978 should have a positive effect on productivity for at least the first half of the 1990s. In the case of producer outlays for structures and equipment, the ratio to real GNP hovered around 12 percent for the entire decade of the 1980s, so that no boost to productivity from this source can be expected unless the investment rate increases. At least partial restoration of the investment tax credit that was repealed in 1986, and/or a drop in real interest rates, are not improbable events that could boost the contribution of investment to productivity growth.

Because capital is income- and output-producing capacity, faster increases in capital at least create the potential for stronger growth of real income and product. The dampening of inflation in the 1980s has set the stage for macroeconomic policies that should result in aggregate demand keeping pace with supply capabilities. (5) As in the 1960s, promoting investment will contribute to both objectives.

On the other hand, it will not be surprising if productivity growth in the other industrialized countries slows further as they come closer to U.S. productivity levels. Further gains will require more domestic R&D relative to technological transfers, which are less costly. Convergence in productivity growth rates towards the U.S. historical trend is not such a dismal prognostication if that rate can be sustained over many decades. After all, if factor inputs keep pace with population, a 1.6 percent average annual increase in TFP means a quintupling of real GDP per capita within a century -- which is what happened in America between 1889 and 1989.

FOOTNOTES

(1) See John W. Kendrick, "International Comparisons of Productivity Trends and Levels," Atlantic Economic Journal, Vol. XVIII, no. 3, September 1990, pp. 42-54.

(2) Estimates from Vannevar Bush, Science, the Endless Frontier, cited in John Kendrick, Productivity Trends in the United States (1961), Table 24, p. 109.

(3) Causes of the slowdown after 1973 are quantified and discussed in some detail in John W. Kendrick, "Productivity Trends and the Recent Slowdown: Historical Perspective, Causal Factors, and Policy Options," in William Fellner, ed., Contemporary Economic Problems, 1979, American Enterprise Institute, 1979, pp. 17-69.

(4) Baumol, Blackman and Wolff (1989), p. 81. See also the review of that volume by Kendrick in The Southern Economic Review, Vol. 57, no. 1, October 1990.

(5) The view of income and product as a function of the totality of capital stocks of all types is developed and statistically tested in John W. Kendrick, The Formation and Stocks of Total Capital, National Bureau of Economic Research, 1976.

REFERENCES

Moses Abramovitz, "Catching Up, Forging Ahead, and Falling Behind," The Journal of Economic History, Vol. XLVI, no. 2 (June 1986), pp. 385-406.

Moses Abramovitz, and Paul David, "Economic Growth in America: Historical Parables and Realities," Reprint no. 105, Stanford University Center for Research in Economic Growth, 1973.

William J. Baumol, Sue Anne Batey Blackman, and Edward N. Wolff, Productivity and American Leadership: The Long View, The M.I.T. Press, 1989.

Edward F. Denison, Accounting for United States Economic Growth, 1948-1969, The Brookings Institution, 1974.

Lester R. Bittel, ed., Encyclopedia of Professional Management, McGraw-Hill Book Co., 1978.

John W. Kendrick, "Productivity Trends," Business Economics, Vol. 6, no. 3 (August 1971).

National Science Foundation, National Patterns of R&D Resources: 1990.

Janice T. Zagardo, A New Look at Internationl Comparisons of Productivity from 1960 to 1985, George Washington University, doctoral dissertation, 1990.
COPYRIGHT 1991 The National Association for Business Economists
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 1991 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Kendrick, John W.
Publication:Business Economics
Date:Oct 1, 1991
Words:3195
Previous Article:Analyzing Modern Business Cycles: Essays Honoring Geoffrey H. Moore.
Next Article:Economic growth in the 1990s.
Topics:

Terms of use | Copyright © 2016 Farlex, Inc. | Feedback | For webmasters