This Selected Issues paper on the United States analyzes problems in the measurement of output and prices. The paper examines income versus expenditure measures of national output. Sources of consumer price index and findings of the Boskin Commission are discussed, and mismeasurement of output and productivity is analyzed. Developments in productivity across industries in the United States are described. In particular, the paper focuses on the slowdown in aggregate productivity growth that began in the mid-1970s and examines whether this slowdown has continued in recent years and is common across industries.

Abstract

This Selected Issues paper on the United States analyzes problems in the measurement of output and prices. The paper examines income versus expenditure measures of national output. Sources of consumer price index and findings of the Boskin Commission are discussed, and mismeasurement of output and productivity is analyzed. Developments in productivity across industries in the United States are described. In particular, the paper focuses on the slowdown in aggregate productivity growth that began in the mid-1970s and examines whether this slowdown has continued in recent years and is common across industries.

III. Developments in Productivity Across Industries in The United States1

1. Since the mid-1970s, aggregate real output and productivity growth in the United States have slowed significantly. From 1960 to 1973, real GDP grew at an average annual rate of 3.8 percent, and business labor productivity rose by 2.6 percent a year; however, from 1973 through 1996, average annual real GDP and business labor productivity growth have declined to 2.5 percent and 1.2 percent, respectively. Studies of the U.S. economy have attributed the slowdown to several causes: an intersectoral shift of output and labor toward services and industries with lower productivity growth; a lack of technological progress in several mature industries; an increase in the obsolescence of capital owing to the regulatory environment and structural changes in the economy; a reduction in the benefits from increasing scale; and a decline in the growth rate of the capital-labor ratio.2 In addition, Griliches (1994), Gordon (1996), and Slifman and Corrado (1996) emphasize that measured output and productivity growth may be biased downward, and therefore, the slowdown may not be as large as suggested by the data.3 Moreover, since 1993, income-based measures of output have grown more rapidly than product-based measures, and thus, product-based measures may further underestimate recent output and productivity growth.

2. This paper examines the recent growth performance of the U.S. economy at various levels of aggregation. In particular, it focuses on the slowdown in aggregate productivity growth that began in the mid-1970s and examines whether this slowdown has continued in recent years and is common across industries. The paper assesses the extent to which the slowdown in productivity growth is concentrated in the services sectors and whether alternative measures of productivity (including total factor productivity) produce similar results. The analysis suggests that the slowdown in the growth of both aggregate labor productivity and total factor productivity after 1973 can primarily be explained by the rising share of the services sector in total output.

A. Measuring Productivity Growth

3. Two measures of productivity are commonly estimated, labor productivity and total factor productivity. Productivity reflects the efficiency of combining resources to produce output. It is usually measured by calculating the ratio of a weighted index of output to a weighted index of inputs. Labor productivity is calculated as the ratio of value-added GDP to homogenous labor hours. Because labor productivity measures output per unit of labor instead of output per unit of all inputs combined, growth in labor productivity may reflect growth in output due to the improved efficiency of all inputs (including labor) and an increase in the use of other productive inputs relative to labor. In other words, labor productivity measured in this way is a function not only of efficiency (which includes technology organization and management of the production process) but also of other inputs such as land, other natural resources, and physical and human capital.

4. Total factor productivity (TFP) attempts to account for the efficiency of all inputs in the production process. In practice, however, TFP is usually calculated as the ratio of value-added GDP to a weighted index of homogenous labor hours and physical capital. Hence such a measure of TFP may include as part of productivity growth changes in inputs other than labor and physical capital.4

B. Aggregate and Sectoral Productivity

5. Estimates of labor productivity growth for 12 sectors, including five private goods- producing sectors (agriculture, forestry, and fishing; mining; manufacturing; electricity, gas, and sanitary services; and construction), six private services-producing sectors (wholesale and retail trade; hotels and other lodging places; transportation; communication; finance, insurance, and real estate; and community, social, business, and personal services) and a government services sector, and four aggregate categories (goods-producing sectors, the services-producing sectors, private industries, and all domestic industries) are presented in Table 1. The first column shows averages for the period 1960–93, while the second, third, and fourth columns provide average growth rates for 1960–73, 1973–81, and 1981–93, respectively.5

Table 1.

United States: Growth in Labor Productivity

(Annual averages, in percent)

article image

6. Labor productivity growth slowed in all but one of the sectors and all four of the aggregate categories after 1973. The only sector for which the slowdown did not occur was government services.6 Since 1981, however, labor productivity growth has recovered in most of the sectors. In fact, average labor productivity growth rates after 1983 in four sectors (agriculture, forestry, and fishing; mining; construction; and hotels and other lodging places) are significantly higher than the corresponding average growth rates prior to 1973. For two other sectors (communication and wholesale and retail trade), average growth rates after 1981 are about equal to those before 1973, while for three other sectors (manufacturing; electricity, gas, and sanitary services; and transportation) average growth rates since 1981 are lower than those before 1973 but higher than those during 1973–81. For the remaining two sectors (finance, insurance, and real estate and community, social, business, and personal services), average growth rates since 1981 are even lower than those in the 1970s.

7. It is noteworthy that the average labor productivity growth rate after 1981 of the goods category is only marginally lower than its average rate prior to 1973, while the average growth rate of the services category is substantially lower in the later period. Although labor productivity growth for services was only slightly lower than that for goods during 1960–73, it was significantly lower during 1981–93.7

8. Tables 2 and 3 provide the average annual TFP growth rates using the growth accounting methodology. In Table 2, TFP is calculated using the gross capital stock, while in Table 3, TFP is calculated using net capital stock.8 The trends for TFP growth are similar to those for labor productivity growth. Other than for government services, generally, there is a slowdown in TFP growth between the periods 1960–73 and 1973–81. The exceptions are agriculture, forestry, and fishing; communication; and finance, insurance, and real estate when net capital stock is used. In addition, all sectors, other than finance, insurance, and real estate; community, social, business, and personal services; government services; and the services aggregate category have higher average annual growth rates after 1981 than during the period 1973–81. In fact, the four sectors which have higher labor productivity growth in 1981–93 compared to 1960–73 have higher TFP growth in the later period as well, For communication, government services, and the goods category, average TFP growth rates are also higher in 1981–93 than in 1960–73 for both measures of the capital stock. In addition, when TFP is calculated using net capital stock, average TFP growth rates of wholesale and retail trade and transportation are higher in 1981–93 than in 1960–73.

Table 2.

United States: Growth in Total Factor Productivity 1/2/

(Annual averages, in percent)

article image

Capacity utilization adjustment made only for mining and quarrying; manufacturing; and electricity, gas, and water using Federal Reserve Board measure.

Using gross capital stock.

Table 3.

United States: Growth in Total Factor Productivity 1/2/

(Annual averages, in percent)

article image

Capacity utilization adjustment made only for mining and quarrying; manufacturing; and electricity, gas and water using Federal Reserve Board measure.

Using net capital stock.

9. In general, the gap between labor productivity growth and TFP growth has narrowed over time. The difference between these growth rates is equal to the growth rate of the ratio of capital services to labor, multiplied by the capital share coefficient. Therefore, in part, some of the slowdown in output growth and labor productivity growth can be attributed to the slowdown in the accumulation of capital relative to the growth of labor.

C. Factors Explaining the Slowdown in Aggregate Productivity Growth

10. There are two potential explanations for the slowdown in aggregate productivity growth: the intersectoral shift of output and labor towards services and a slowdown in capital accumulation relative to labor force growth. The relative contribution of intersectoral shifts in the productivity slowdown can be estimated by decomposing measured productivity growth into the additional growth that would have taken place if the initial shares in total output of the various sectors had remained constant at their original levels and the productivity of the services sector had recovered to TFP growth rates that occurred during 1960–73.9 This contribution can be further attributed to the lagging recovery in services or to the services sector having lower productivity than the goods sector, by decomposing measured productivity growth into the additional growth that would have taken place if the initial shares of the sectors had remained constant at their original levels, while actual services sector productivity declined in 1981–93 relative to 1960–73. The relative contribution of the slowdown in the growth of the capital services to labor ratio can be measured by taking the difference between labor productivity and TFP growth.

11. These calculations show that intersectoral shift accounts for almost all of the slowdown in aggregate labor productivity growth between 1960–73 and 1981–93 (tabulation below), with most of this reflecting the measured slow recovery of productivity in services. A reduction in the growth of the capital services to labor ratio between the two time periods accounts for the remaining slowdown in labor productivity growth. The slowdown in TFP growth between 1960–73 and 1981–93 is equivalent to the slowdown in aggregate labor productivity growth less the slowdown in the ratio of capital services to labor. In the case of TFP growth, intersectoral shifts more than account for the slowdown.

Factors Explaining Productivity Slowdown Between 1960–73 and 1981–93

(In percentage points)

article image

12. The measured slow recovery in service-sector total factor productivity after 1973 could reflect several factors other than a slowdown in efficiency or productivity growth specific to this sector. These other factors include relative changes in the average level of labor skills (for example, due to new entrants into services sector jobs in recent years), in the underutilization of resources (for example, due to labor hoarding or incorrectly measuring capacity utilization), and in the obsolescence of capital. However, because the productivity slowdown is concentrated in the services sector and this is the sector for which output is most difficult to measure, the measured slowdown could be a result of mismeasurement, particularly because profitability in this sector has remained high.10

List of References

  • Baity, Martin N., 1982, “The Productivity Growth Slowdown by Industry,Bookings Papers on Economic Activity, Vol. 2, pp. 423459.

    • Search Google Scholar
  • Gordon, Robert J., 1996, “Problems in the Measurement and Performance of Service-sector Productivity in the United States,NBER Working Paper, No. 5519 (Cambridge, Massachusetts: National Bureau of Economic Research).

    • Search Google Scholar
    • Export Citation
  • Griliches, Zvi, 1994, “Productivity, R&D, and the Data Constraint,The American Economic Review, Vol. 84, pp. 123.

  • Gullickson, William, 1992, “Multifactor Productivity in Manufacturing Industries,Monthly Labor Review, Vol. 115, No. 10, pp. 2032.

    • Search Google Scholar
    • Export Citation
  • Morrison, Catherine J., 1992, “Unraveling the Productivity Growth Slowdown in the United States, Canada and Japan: The Effects of Subequilibrium, Scale Economies and Markups,The Review of Economics and Statistics, Vol. 74, pp. 381393.

    • Search Google Scholar
    • Export Citation
  • Salgado, Ranil, 1997, “Developments in Productivity Across Industries in the United States,IMF Working Paper, forthcoming.

  • Slifman, Lawrence and Carol Corrado, 1996, “Decomposition of Productivity and Unit Costs,Federal Reserve Board Occasional Staff Studies, No. 1 (Washington, DC: Board of Governors of the Federal Reserve System).

    • Search Google Scholar
    • Export Citation
1

Prepared by Ranil Salgado.

2

See, for example, Baily (1982), Morrison (1992), Griliches (1994), and Gordon (1996).

3

Griliches (1994), in particular, notes that the share of total output accounted for by the sectors in which output is easily measured (i.e., agriculture, mining, manufacturing, transportation, and utilities) has declined from 49 percent of U.S. GDP in 1947 to 31 percent in 1990. Gordon (1996) discusses the sources of bias in aggregate price indices, as well as other measurement problems affecting specific industries, particularly services. Slifman and Corrado (1996) question whether the ongoing measured slowdown in real growth in the services sector is an actual slowdown or reflects problems measuring output in the sector.

4

To calculate TFP, the growth accounting methodology (which attributes output growth to the growth in inputs and the growth in productivity) is employed:ΔTFP=ΔlnY-αΔlnH-(l-α)ΔlnK(1) where Y, K, H, and α are, respectively, output, capital services, labor services, and labor’s share of value-added output (which can be calculated by dividing labor compensation by output). Implicit in these calculations are assumptions of constant returns to scale, perfect competition, and profit maximization.

5

The rationale for dividing the post-1973 period into two periods is to abstract from the demand-induced slowdown due to the energy-price shocks of the 1970s.

6

Prices for government services are generally computed based on the cost of inputs, and thus, real output in the sector tends to grow with real inputs. Therefore, it is not surprising that labor productivity did not slow in this sector because changes in TFP (as noted below) and in the capital-labor ratio have been relatively small and stable since 1960.

7

These results are consistent with the findings of Slifman and Corrado (1996), among others.

8

Gross capital stock is defined as accumulated investment less accumulated discards (or scraps), and net capital stock is defines as accumulated investment less accumulated depreciation.

9

Note that TFP growth rates for the goods category and the government services sector are higher in 1981–93 than 1960–73. For labor productivity, the sectoral share is equal to the sectoral share of labor in the economy. For TFP, the sectoral share is equal to the weighted sectoral shares of labor and capital in the economy, where the relative weights are given by the labor and capital shares of sectoral compensation.

10

See Slifman and Corrado (1996).