# Recent changes in the growth of U.S. multifactor productivity.

Since 1979, multifactor productivity growth has recovered
completely only in manufacturing,for the rest of private business,
growth has recovered partially, but remains below the 1948-73 trend rate

The dramatic slowdown in productivity growth that began about 1973 has been the focus of much analysis and discussion. In recent months, this discussion has taken on new life, as analysts continue to probe the influence of U.S. productivity growth on the country's competitiveness in world trade. Other commentators have raised the possibility that the slowdown has ended and productivity growth has resumed its earlier pace.

The present article contributes to the discussion by presenting and analyzing recent Bureau of Labor Statistics measures of growth in multifactor productivity in the United States. It also presents preliminary results of BLS studies of factors that have affected productivity change. Special attention is given to the productivity growth record for recent years, to examine whether productivity might have resumed its pre-1973 pace.

BLS multifactor measures

In 1983, the BLS introduced measures of multifactor productivity for three major sectors of the U.S. economy-private business, private nonfarm business, and manufacturing. Since that year, annual news releases have provided current multifactor productivity measures. Because the methods used to develop these measures have been described in two 1983 publications,' only a brief summary is presented here.

In the BLS measures, growth in multifactor productivity is measured as the difference between the growth rate of output and the growth rate of combined capital and Labor inputs.2 Growth in multifactor productivity reflects increase in output due to factors other than growth in capital and Labor inputs.sup.3 Multifactor productivity calculated in this way provides a numerical answer to the question: What is the portion of the growth rate of output that cannot be accounted for by the growth rate of combined inputs?

The measured multifactor productivity growth rates reflect changes in all influences on output other than the inputs. They reflect changes in technology, including changes that result from research and development activities; economies of scale; changes in the management or organization of resources; and changes in the skills and efforts workers bring to the job.

Multifactor productivity is closely related to the commonly used concept of labor productivity, or output per unit of labor input. It can be shown that, under certain assumptions,.sup.4 growth in labor productivity is equal to growth in multifactor productivity plus another factor. That factor is the growth in capital input per labor hour times the share of capital income in the value of output-or capital's share, for short. It follows that growth in labor productivity can be decomposed into two parts, the part due to growth in multifactor productivity and the part due to growth in capital input per labor hour or the capital-labor ratio. Over long periods in most economies, all three of these terms will be positive, so that a positive growth rate in labor productivity is the sum of positive growth in multifactor productivity plus the positive growth rate in the capital-labor ratio times the capital share in output. (Of course, it is possible for one or more of these growth rates to be ne ative.

This relationship can be expressed in an equation: (1)([the rate of change of Q] / Q) - ([the rate of change of L] / L) = [the rate of change of A] / A + S.sub.K ([the rate of change of K] / K) - ([the rate of change of L] / L) where Q is output, L is labor hours, K is capital input, S.sub.k is capital's share, and A represents the state of technology. A dot over a variable indicates the rate of change of the variable with respect to time. The ratio ([rate of change of K] /K) therefore the percentage rate of change of capital. The term [the rate of change of Q] / Q - [the rate of change of L] / L is the rate of change over time in labor productivity, Q/L (output per hour. Similarly, [the rate of change of A] / A is the percentage rate of change in multifactor productivity, or output per unit of combined labor and capital inputs. The term [the rate of change of K] / K - [the rate of change of L] / L is the percentage rate of change in capital input per hour, the capital-labor ratio, while S.sub.k is capital's share, (Capital's share, S.sub.k, and labor's share, S.sub.1, both are fractions and their sum exhausts the total value of output; it follows that S.sub.k + S.sub.1 = 1.)

Equation 1 expresses the relationship described earlier: The percentage rate of change in labor productivity, [the rate of change of Q] / Q - [the rate of change of L] / L, is equal to the percentage rate of change in

multifactor productivity, [the rate of change of A] / A, plus the percentage rate of change in the capital-labor ration, [the rate of change of K] / K - [the rate of change of L] / L, times the share of capital in output, S.sub.k.

Equation 1 decomposes the change in the familiar labor productivity ratio, Q/L, into the role of multifactor productivity and the role of capital input relative to labor input (the capital-labor ratio).'

Long-term trends

Annual measures of productivity change often are sensitive to cyclical effects. It is helpful to look at longterm trends in order to minimize the effects of cyclical disturbances. In addition, long-term trends help to provide a benchmark for gauging the relative performance of productivity growth for shorter periods. Such a gauge has become increasingly important with the deceleration of productivity growth over the last 15 years.

From 1948 to 1986, multifactor productivity in the private business sector increased at an average rate of 1.4 percent per year (table 1). This growth reflects a 3.2percent average annual increase in output and a 1.8percent rate of growth of combined inputs of labor and capital services. Labor services (hours) in private business increased at a rate of 0.9 percent per year, and capital services grew 3.4 percent annually. The capital-labor ratio grew 2.4 percent per year.

The nonfarm business sector had a pattern of productivity growth for the 1948-86 period that was similar to that of private business. Multifactor productivity increased 1.1 percent annually for the entire period as output rose 3.3 percent yearly and combined inputs grew at a 2.1 -percent annual rate. Hours increased 1.4 percent annually in private nonfarm business, a faster rate of growth than in private business, reflecting a decline in hours in farm production. Inputs of capital services increased 3.6 percent per year in private nonfarm business.

In manufacturing, multifactor productivity grew at an annual rate of 1.9 percent between 1948 and 1986. Output increased at a 3.3-percent rate and combined inputs grew 1.4 percent yearly. Labor hours rose at an annual rate of only 0.6 percent, and capital services grew 3.4 percent per year.

The measured relationship between multifactor productivity and labor productivity was approximately the same for the three sectors. Multifactor productivity growth accounted for 60 to 70 percent of labor productivity growth during the postwar period, with the remainder arising from an increase in the capital-labor ratio and changes in capital's share (the share of capital income in the value of output, S.sub.k). Because capital's share was fairly stable over time (table 2), changes in this share had little effect on labor productivity. In manufacturing, multifactor productivity growth accounted for the largest proportion of labor productivity increase, 70 percent. Manufacturing also had the fastest rate of growth of the capital-labor ratio; however, capital's share was generally lower in manufacturing than in the other sectors, and this tended to dampen the influence on labor productivity of growth in the capital-labor ratio. For private business and private nonfarm business, growth in multifactor productivity accounted for about 60 percent of labor productivity increase during the period.

The productivity slowdown

The post-1973 decline in the growth rate of both multifactor and labor productivity has been a source of persistent concern for business leaders, economists, and policymakers. A plethora of books and articles has appeared analyzing the probable causes and consequences of the productivity slowdown. 6 The following discussion examines the magnitude of the slowdown, using the most current measures,

Between 1948 and 1973, the United States enjoyed a sustained period of strong productivity growth. Multifactor productivity in the nonfarm business sector rose at an annual rate of 1.7 percent during this period (table 3). This, coupled with the increase in the capital-labor ratio, produced an annual rate of growth of labor productivity of 2.5 percent. For the private business sector, the productivity advance was even greater: Extra growth occurred because of a shift of workers out of the farm sector, which had relatively low productivity, to the nonfarm sector, which had higher productivity..sup.7

The manufacturing sector also experienced sustained high growth rates of both multifactor and labor productivity in the postwar period. The rapid advance in labor productivity was a reflection of both a high rate of multifactor productivity growth--2.0 percent annually-and a high rate of growth in the capital-labor ratio-2.6 percent per year,

Between 1973 and 1979, the Nation's rate of productivity growth changed drastically. Output per hour in private business increased only 0.6 percent annually and multifactor productivity rose only 0.1 percent per year. (See table 3 and chart 1.) In private nonfarm business, output per hour increased at an annual rate of 0.5 percent, while multifactor productivity did not grow at all.

Labor productivity continued to show some advance in these two sectors only because of continued, though slower, growth in capital intensity (the capital-labor ratio). In private business, the average annual increase in the capital-labor ratio fell from 2.6 percent in the previous period to 1.5 percent in the 1973-79 period. Similarly, in nonfarm business, the growth rate fell from 2.2 percent annually over the period 1948-73 to 1.6 percent in 1973-79. Only in manufacturing did the capital-labor ratio continue to increase rapidly. In fact, this ratio accelerated to a growth rate of 3.3 percent annually, up from a 2.6-percent rate in the earlier period.

Of the total deceleration in labor productivity growth in private business (2.3 percentage points), over 80 percent (1.9 percentage points) was a result of the deceleration in multifactor productivity growth. Less than 20 percent of the reduction in the rate of labor productivity increase was due to the decline in the growth rate of the capital-labor ratio. The same pattern held in private nonfarm business.

In manufacturing, the deceleration of multifactor productivity growth accounted for more than the full decline in the labor productivity growth rate. Because the rate of increase in the capital-labor ratio actually accelerated during the period, labor productivity growth slowed less than multifactor productivity growth.

The recovery: 1979-86

During the most recent period, 1979 to 1986, there has been a partial recovery in productivity growth. The recovery must be considered incomplete because growth rates for both multifactor and labor productivity in private business and private nonfarm business have risen above the low rates of 1973-79 but have not reached the pre-1973 rates. In manufacturing, however, multifactor and labor productivity growth rates have surpassed the pre-1973 rates.

In private business and private nonfarm business, the rate of increase in labor hours in 1979-86 declined relative to the 1973-79 rate, while capital services grew steadily, resulting in an acceleration of growth in the capital-labor ratio. These developments contributed to the more rapid rise of labor productivity. A modest increase in multifactor productivity growth also assisted the growth of labor productivity. In manufacturing, however, the rise of more than 2 percentage points, from 1.4 percent to 3.5 percent, in labor productivity growth reflected an increase in multifactor productivity growth of 2 percentage points, with no noticeable assistance from the capital-labor ratio.

Elements of multifactor productivity change

It is noted above that the measured growth rates of multifactor productivity reflect changes in technology (the processes used to produce output); economies of scale; organizational or management changes; and the skills workers bring to the job (generally acquired through schooling and experience). In addition, multifactor productivity reflects any errors made in measurement of the inputs (hours and capital services) and the output. BLS has conducted and continues to conduct research in these areas to gain understanding of the multifactor productivity changes and also to eliminate possible measurement error. The present areas of research include the measurement of the effects of research and development (R&D) expenditures on productivity growth; the measurement of the changing amounts of education and experience that workers possess and the subsequent effects on productivity growth; and a better measure of hours, which reflects the actual number of labor hours spent at the workplace as opposed to the hours (including vacations and other leave time) for which workers are paid.

Increased expenditures on R&D are considered to be a prime factor in the creation of more efficient technologies. Relative expenditures on R&D in nonfarm business slowed substantially in the 1970's. The total stock of R&D,.sup.8 which had increased 7.8 percent annually between 1948 and 1973, rose 4.0 percent per year between 1973 and 1979. From 1979 to 1985, the growth rate of the R&D stock accelerated moderately to 4.4 percent..sup.9 However, results of the empirical work also show that the post1973 changes in the growth rates of the stock had almost no effect on post-1973 rates of change in multifactor productivity. The research results for manufacturing are quite similar to those for nonfarm business.

Preliminary work on the productivity effects of worker experience and education shows that a small but measurable portion of the slowdown was attributable to changes in the composition of the work force. Increase in workers' education and experience is positively correlated with growth in output and productivity. During the late 1960's and throughout the 1970's, the work force expanded rapidly, resulting in an increased proportion of younger, inexperienced workers. At the same time, more and more workers attained higher levels of education, tending to counter some of the consequences of the influx of inexperienced workers.

Preliminary results indicate that the contribution to multifactor productivity growth of changes in labor composition in private business dropped from an average of 0.2 percent per year in 1948-73 to zero in 1973-79..sup.10 The rise in the proportion of less experienced workers ceased after 1979, although there continued to be an increase in the average years of schooling of the work force." After 1979, labor composition changes contributed about 0.3 percent per year to multifactor productivity growth-about as much as they had before 1973..sup.12 Similar preliminary results hold for private nonfarm business, while results for manufacturing are not presently available.

Beginning in 1982, BLS began collecting data on the ratio of hours at work to hours paid for production and nonsupervisory workers in nonagricultural establishments. One purpose of this survey is to develop new data on labor hours for use in productivity measurement. Only 15 percent of total labor hours, as presently measured in the labor productivity and multifactor measures, are based "hours at work" concept, the concept most consistent with productivity measurement. This portion of total hours is derived primarily from the Current Population Survey, a survey of households conducted for BLS by the Bureau of the Census. The remaining 85 percent of the hours are based on an "hours paid" concept; this portion of total hours is derived from the Current Employment Statistics survey, a BLS establishment-based program. Hence, the current productivity measures, as presented in tables 1 and 3, reflect predominately an "hours paid" concept.

The growth rates of hours at work and hours paid may, of course, differ over time. If, for example, hours paid increase faster than hours at work, then productivity growth will be underestimated. Old estimates of the historical trend of the hours-at-work/hours-paid ratio show a small divergence in the growth rates of the two measures. Between 1948 and 1973, the ratio decreased 0.1 percent annually, and from 1973 to 1979, it fell 0.2 percent per year. These computations suggest that the slowdown in labor productivity growth is 0.1 percentage point less than recorded in table 3. From 1979 to 1986, the ratio of hours at work to hours paid remained virtually unchanged..sup.13 Hence, productivity in nonfarm business has increased as presently reported. The results for manufacturing indicate that changes in the ratio of hours at work to hours paid had effects on productivity, as presently reported, of 0.1 percent or less for the periods under examination. Summary

BLS measures show that, after a period of strong growth in both multifactor productivity and output per hour from 1948 to 1973, there followed a period of little or no increase from 1973 to 1979. Since 1979, productivity growth has recovered partially in the private business and the private nonfarm business sectors. Only in manufacturing has the recovery been complete.

An analysis of the 1973-79 slowdown in labor productivity shows that the major part of the slowdown cannot be explained by any of the factors examined to date in the BLS research program. This conclusion holds for all three of the sectors examined here-private business, private nonfarm business, and manufacturing. In this regard, the BLS results coincide with the analyses of most private researchers. Three of the four factors discussed in this article-growth in the capital-labor ratio, changes in labor composition, and a decline in the ratio of hours at work to hours paid-contributed in modest ways to the slowdown, while the fourth, the decline in the growth rate of R&D expenditures, did not. The major component of the deceleration in labor productivity was a slowdown in multifactor productivity that was not explained by these four factors.

The partial recovery of labor productivity growth in private business and private nonfarm business after 1979 can be attributed largely to increases in the capital-labor ratio and to changes in the composition of the labor force. The other two factors-R&D expenditures and hours at work-did not contribute to this recovery. For manufacturing, the complete recovery in labor productivity growth after 1979 was due predominately to increased growth in multifactor productivity.

The dramatic slowdown in productivity growth that began about 1973 has been the focus of much analysis and discussion. In recent months, this discussion has taken on new life, as analysts continue to probe the influence of U.S. productivity growth on the country's competitiveness in world trade. Other commentators have raised the possibility that the slowdown has ended and productivity growth has resumed its earlier pace.

The present article contributes to the discussion by presenting and analyzing recent Bureau of Labor Statistics measures of growth in multifactor productivity in the United States. It also presents preliminary results of BLS studies of factors that have affected productivity change. Special attention is given to the productivity growth record for recent years, to examine whether productivity might have resumed its pre-1973 pace.

BLS multifactor measures

In 1983, the BLS introduced measures of multifactor productivity for three major sectors of the U.S. economy-private business, private nonfarm business, and manufacturing. Since that year, annual news releases have provided current multifactor productivity measures. Because the methods used to develop these measures have been described in two 1983 publications,' only a brief summary is presented here.

In the BLS measures, growth in multifactor productivity is measured as the difference between the growth rate of output and the growth rate of combined capital and Labor inputs.2 Growth in multifactor productivity reflects increase in output due to factors other than growth in capital and Labor inputs.sup.3 Multifactor productivity calculated in this way provides a numerical answer to the question: What is the portion of the growth rate of output that cannot be accounted for by the growth rate of combined inputs?

The measured multifactor productivity growth rates reflect changes in all influences on output other than the inputs. They reflect changes in technology, including changes that result from research and development activities; economies of scale; changes in the management or organization of resources; and changes in the skills and efforts workers bring to the job.

Multifactor productivity is closely related to the commonly used concept of labor productivity, or output per unit of labor input. It can be shown that, under certain assumptions,.sup.4 growth in labor productivity is equal to growth in multifactor productivity plus another factor. That factor is the growth in capital input per labor hour times the share of capital income in the value of output-or capital's share, for short. It follows that growth in labor productivity can be decomposed into two parts, the part due to growth in multifactor productivity and the part due to growth in capital input per labor hour or the capital-labor ratio. Over long periods in most economies, all three of these terms will be positive, so that a positive growth rate in labor productivity is the sum of positive growth in multifactor productivity plus the positive growth rate in the capital-labor ratio times the capital share in output. (Of course, it is possible for one or more of these growth rates to be ne ative.

This relationship can be expressed in an equation: (1)([the rate of change of Q] / Q) - ([the rate of change of L] / L) = [the rate of change of A] / A + S.sub.K ([the rate of change of K] / K) - ([the rate of change of L] / L) where Q is output, L is labor hours, K is capital input, S.sub.k is capital's share, and A represents the state of technology. A dot over a variable indicates the rate of change of the variable with respect to time. The ratio ([rate of change of K] /K) therefore the percentage rate of change of capital. The term [the rate of change of Q] / Q - [the rate of change of L] / L is the rate of change over time in labor productivity, Q/L (output per hour. Similarly, [the rate of change of A] / A is the percentage rate of change in multifactor productivity, or output per unit of combined labor and capital inputs. The term [the rate of change of K] / K - [the rate of change of L] / L is the percentage rate of change in capital input per hour, the capital-labor ratio, while S.sub.k is capital's share, (Capital's share, S.sub.k, and labor's share, S.sub.1, both are fractions and their sum exhausts the total value of output; it follows that S.sub.k + S.sub.1 = 1.)

Equation 1 expresses the relationship described earlier: The percentage rate of change in labor productivity, [the rate of change of Q] / Q - [the rate of change of L] / L, is equal to the percentage rate of change in

multifactor productivity, [the rate of change of A] / A, plus the percentage rate of change in the capital-labor ration, [the rate of change of K] / K - [the rate of change of L] / L, times the share of capital in output, S.sub.k.

Equation 1 decomposes the change in the familiar labor productivity ratio, Q/L, into the role of multifactor productivity and the role of capital input relative to labor input (the capital-labor ratio).'

Long-term trends

Annual measures of productivity change often are sensitive to cyclical effects. It is helpful to look at longterm trends in order to minimize the effects of cyclical disturbances. In addition, long-term trends help to provide a benchmark for gauging the relative performance of productivity growth for shorter periods. Such a gauge has become increasingly important with the deceleration of productivity growth over the last 15 years.

From 1948 to 1986, multifactor productivity in the private business sector increased at an average rate of 1.4 percent per year (table 1). This growth reflects a 3.2percent average annual increase in output and a 1.8percent rate of growth of combined inputs of labor and capital services. Labor services (hours) in private business increased at a rate of 0.9 percent per year, and capital services grew 3.4 percent annually. The capital-labor ratio grew 2.4 percent per year.

The nonfarm business sector had a pattern of productivity growth for the 1948-86 period that was similar to that of private business. Multifactor productivity increased 1.1 percent annually for the entire period as output rose 3.3 percent yearly and combined inputs grew at a 2.1 -percent annual rate. Hours increased 1.4 percent annually in private nonfarm business, a faster rate of growth than in private business, reflecting a decline in hours in farm production. Inputs of capital services increased 3.6 percent per year in private nonfarm business.

In manufacturing, multifactor productivity grew at an annual rate of 1.9 percent between 1948 and 1986. Output increased at a 3.3-percent rate and combined inputs grew 1.4 percent yearly. Labor hours rose at an annual rate of only 0.6 percent, and capital services grew 3.4 percent per year.

The measured relationship between multifactor productivity and labor productivity was approximately the same for the three sectors. Multifactor productivity growth accounted for 60 to 70 percent of labor productivity growth during the postwar period, with the remainder arising from an increase in the capital-labor ratio and changes in capital's share (the share of capital income in the value of output, S.sub.k). Because capital's share was fairly stable over time (table 2), changes in this share had little effect on labor productivity. In manufacturing, multifactor productivity growth accounted for the largest proportion of labor productivity increase, 70 percent. Manufacturing also had the fastest rate of growth of the capital-labor ratio; however, capital's share was generally lower in manufacturing than in the other sectors, and this tended to dampen the influence on labor productivity of growth in the capital-labor ratio. For private business and private nonfarm business, growth in multifactor productivity accounted for about 60 percent of labor productivity increase during the period.

The productivity slowdown

The post-1973 decline in the growth rate of both multifactor and labor productivity has been a source of persistent concern for business leaders, economists, and policymakers. A plethora of books and articles has appeared analyzing the probable causes and consequences of the productivity slowdown. 6 The following discussion examines the magnitude of the slowdown, using the most current measures,

Between 1948 and 1973, the United States enjoyed a sustained period of strong productivity growth. Multifactor productivity in the nonfarm business sector rose at an annual rate of 1.7 percent during this period (table 3). This, coupled with the increase in the capital-labor ratio, produced an annual rate of growth of labor productivity of 2.5 percent. For the private business sector, the productivity advance was even greater: Extra growth occurred because of a shift of workers out of the farm sector, which had relatively low productivity, to the nonfarm sector, which had higher productivity..sup.7

The manufacturing sector also experienced sustained high growth rates of both multifactor and labor productivity in the postwar period. The rapid advance in labor productivity was a reflection of both a high rate of multifactor productivity growth--2.0 percent annually-and a high rate of growth in the capital-labor ratio-2.6 percent per year,

Between 1973 and 1979, the Nation's rate of productivity growth changed drastically. Output per hour in private business increased only 0.6 percent annually and multifactor productivity rose only 0.1 percent per year. (See table 3 and chart 1.) In private nonfarm business, output per hour increased at an annual rate of 0.5 percent, while multifactor productivity did not grow at all.

Labor productivity continued to show some advance in these two sectors only because of continued, though slower, growth in capital intensity (the capital-labor ratio). In private business, the average annual increase in the capital-labor ratio fell from 2.6 percent in the previous period to 1.5 percent in the 1973-79 period. Similarly, in nonfarm business, the growth rate fell from 2.2 percent annually over the period 1948-73 to 1.6 percent in 1973-79. Only in manufacturing did the capital-labor ratio continue to increase rapidly. In fact, this ratio accelerated to a growth rate of 3.3 percent annually, up from a 2.6-percent rate in the earlier period.

Of the total deceleration in labor productivity growth in private business (2.3 percentage points), over 80 percent (1.9 percentage points) was a result of the deceleration in multifactor productivity growth. Less than 20 percent of the reduction in the rate of labor productivity increase was due to the decline in the growth rate of the capital-labor ratio. The same pattern held in private nonfarm business.

In manufacturing, the deceleration of multifactor productivity growth accounted for more than the full decline in the labor productivity growth rate. Because the rate of increase in the capital-labor ratio actually accelerated during the period, labor productivity growth slowed less than multifactor productivity growth.

The recovery: 1979-86

During the most recent period, 1979 to 1986, there has been a partial recovery in productivity growth. The recovery must be considered incomplete because growth rates for both multifactor and labor productivity in private business and private nonfarm business have risen above the low rates of 1973-79 but have not reached the pre-1973 rates. In manufacturing, however, multifactor and labor productivity growth rates have surpassed the pre-1973 rates.

In private business and private nonfarm business, the rate of increase in labor hours in 1979-86 declined relative to the 1973-79 rate, while capital services grew steadily, resulting in an acceleration of growth in the capital-labor ratio. These developments contributed to the more rapid rise of labor productivity. A modest increase in multifactor productivity growth also assisted the growth of labor productivity. In manufacturing, however, the rise of more than 2 percentage points, from 1.4 percent to 3.5 percent, in labor productivity growth reflected an increase in multifactor productivity growth of 2 percentage points, with no noticeable assistance from the capital-labor ratio.

Elements of multifactor productivity change

It is noted above that the measured growth rates of multifactor productivity reflect changes in technology (the processes used to produce output); economies of scale; organizational or management changes; and the skills workers bring to the job (generally acquired through schooling and experience). In addition, multifactor productivity reflects any errors made in measurement of the inputs (hours and capital services) and the output. BLS has conducted and continues to conduct research in these areas to gain understanding of the multifactor productivity changes and also to eliminate possible measurement error. The present areas of research include the measurement of the effects of research and development (R&D) expenditures on productivity growth; the measurement of the changing amounts of education and experience that workers possess and the subsequent effects on productivity growth; and a better measure of hours, which reflects the actual number of labor hours spent at the workplace as opposed to the hours (including vacations and other leave time) for which workers are paid.

Increased expenditures on R&D are considered to be a prime factor in the creation of more efficient technologies. Relative expenditures on R&D in nonfarm business slowed substantially in the 1970's. The total stock of R&D,.sup.8 which had increased 7.8 percent annually between 1948 and 1973, rose 4.0 percent per year between 1973 and 1979. From 1979 to 1985, the growth rate of the R&D stock accelerated moderately to 4.4 percent..sup.9 However, results of the empirical work also show that the post1973 changes in the growth rates of the stock had almost no effect on post-1973 rates of change in multifactor productivity. The research results for manufacturing are quite similar to those for nonfarm business.

Preliminary work on the productivity effects of worker experience and education shows that a small but measurable portion of the slowdown was attributable to changes in the composition of the work force. Increase in workers' education and experience is positively correlated with growth in output and productivity. During the late 1960's and throughout the 1970's, the work force expanded rapidly, resulting in an increased proportion of younger, inexperienced workers. At the same time, more and more workers attained higher levels of education, tending to counter some of the consequences of the influx of inexperienced workers.

Preliminary results indicate that the contribution to multifactor productivity growth of changes in labor composition in private business dropped from an average of 0.2 percent per year in 1948-73 to zero in 1973-79..sup.10 The rise in the proportion of less experienced workers ceased after 1979, although there continued to be an increase in the average years of schooling of the work force." After 1979, labor composition changes contributed about 0.3 percent per year to multifactor productivity growth-about as much as they had before 1973..sup.12 Similar preliminary results hold for private nonfarm business, while results for manufacturing are not presently available.

Beginning in 1982, BLS began collecting data on the ratio of hours at work to hours paid for production and nonsupervisory workers in nonagricultural establishments. One purpose of this survey is to develop new data on labor hours for use in productivity measurement. Only 15 percent of total labor hours, as presently measured in the labor productivity and multifactor measures, are based "hours at work" concept, the concept most consistent with productivity measurement. This portion of total hours is derived primarily from the Current Population Survey, a survey of households conducted for BLS by the Bureau of the Census. The remaining 85 percent of the hours are based on an "hours paid" concept; this portion of total hours is derived from the Current Employment Statistics survey, a BLS establishment-based program. Hence, the current productivity measures, as presented in tables 1 and 3, reflect predominately an "hours paid" concept.

The growth rates of hours at work and hours paid may, of course, differ over time. If, for example, hours paid increase faster than hours at work, then productivity growth will be underestimated. Old estimates of the historical trend of the hours-at-work/hours-paid ratio show a small divergence in the growth rates of the two measures. Between 1948 and 1973, the ratio decreased 0.1 percent annually, and from 1973 to 1979, it fell 0.2 percent per year. These computations suggest that the slowdown in labor productivity growth is 0.1 percentage point less than recorded in table 3. From 1979 to 1986, the ratio of hours at work to hours paid remained virtually unchanged..sup.13 Hence, productivity in nonfarm business has increased as presently reported. The results for manufacturing indicate that changes in the ratio of hours at work to hours paid had effects on productivity, as presently reported, of 0.1 percent or less for the periods under examination. Summary

BLS measures show that, after a period of strong growth in both multifactor productivity and output per hour from 1948 to 1973, there followed a period of little or no increase from 1973 to 1979. Since 1979, productivity growth has recovered partially in the private business and the private nonfarm business sectors. Only in manufacturing has the recovery been complete.

An analysis of the 1973-79 slowdown in labor productivity shows that the major part of the slowdown cannot be explained by any of the factors examined to date in the BLS research program. This conclusion holds for all three of the sectors examined here-private business, private nonfarm business, and manufacturing. In this regard, the BLS results coincide with the analyses of most private researchers. Three of the four factors discussed in this article-growth in the capital-labor ratio, changes in labor composition, and a decline in the ratio of hours at work to hours paid-contributed in modest ways to the slowdown, while the fourth, the decline in the growth rate of R&D expenditures, did not. The major component of the deceleration in labor productivity was a slowdown in multifactor productivity that was not explained by these four factors.

The partial recovery of labor productivity growth in private business and private nonfarm business after 1979 can be attributed largely to increases in the capital-labor ratio and to changes in the composition of the labor force. The other two factors-R&D expenditures and hours at work-did not contribute to this recovery. For manufacturing, the complete recovery in labor productivity growth after 1979 was due predominately to increased growth in multifactor productivity.

Printer friendly Cite/link Email Feedback | |

Author: | Dean, Edwin; Kunze, Kent |
---|---|

Publication: | Monthly Labor Review |

Date: | May 1, 1988 |

Words: | 3084 |

Previous Article: | The declining middle-class thesis: a sensitivity analysis. |

Next Article: | Wage adjustments in contracts negotiated in private industry in 1987. |

Topics: |