Abstract
This Selected Issues paper on the United States analyzes the measures of potential output, natural rate of unemployment, and capacity utilization. Traditionally, measures of resource utilization have been used as indicators for the potential build-up of inflation pressures, and hence as guides for the formulation of macroeconomic policy. The paper highlights that the most commonly used indicators of resource utilization in the United States are the output gap, the employment gap, and capacity utilization in industry. The paper also analyzes the wage and price determination and productivity trends in the United States.
III. Productivity Trends in the United States1
1. Productivity in the United States has experienced significant gains during the past few years following a prolonged slowdown beginning in the early 1970s, after the first oil-price shock. Standard measures of total factor productivity (TFP) based on the growth accounting methodology indicate that TFP slowed from an average annual rate of increase of 1½ percent in 1960–74 to an annual growth rate of ⅛18; percent in 1975–90. During the 1990s, TFP has experienced a recovery, growing at an annual rate close to ½ percent (Figure 1).
United States: Productivity Measures
Citation: IMF Staff Country Reports 1999, 101; 10.5089/9781451839579.002.A003
Sources: Bureau of Economic Analysis; and IMF staff estimates.United States: Productivity Measures
Citation: IMF Staff Country Reports 1999, 101; 10.5089/9781451839579.002.A003
Sources: Bureau of Economic Analysis; and IMF staff estimates.United States: Productivity Measures
Citation: IMF Staff Country Reports 1999, 101; 10.5089/9781451839579.002.A003
Sources: Bureau of Economic Analysis; and IMF staff estimates.2. In order to better understand the nature of the slowdown, it is useful to decompose TFP into two components: investment-specific productivity change (ISP) and technologically neutral productivity change (TNP). ISP captures technological improvements embodied in new equipment and machinery and is closely related to the notion of technological progress. TNP largely captures the changes in productivity associated with the organization of capital and labor in productive activities. This decomposition illustrates the dependence of TFP not only on technical advances but also on how these advances are adopted.
3. A sharp decline in TNP growth after the first oil-price shock in the mid-1970s more than accounted for the slowdown in TFP growth during the period through the 1980s. After increasing at an average annual rate of 1½ percent in 1960–74, TNP growth is estimated to have declined on average by ¼ percent a year during 1975–90. TNP began to grow again at an annual rate of ½ percent in the 1990s, perhaps reflecting efficiency gains from the corporate downsizings and restructuring that took place in the late 1980s and early 1990s. In contrast, ISP growth held relatively steady at an annual rate of 2 percent in the 1960s, 1970s, and 1980s. In the 1990s, however, it has picked up sharply, averaging 3½ percent a year.
4. The growth-accounting framework to measure TFP was first introduced in Solow (1957). The main assumption is that growth in the production of goods is equal to the weighted average of the growth in inputs of the aggregate production function plus the growth of TFP (also referred to as Solow residual). In this framework, TFP captures both the state of technology and innovation, as well as how efficiently capital and labor are organized in the production process.
5. One of the problems associated with the measurement of technical progress in the growth-accounting framework is that all vintages of capital equipment are treated alike in terms of their productivity—one unit of new capital has the same value as one unit of old capital. However, advances in technology tend over time to be embodied in the latest vintages of capital equipment (which is the definition of ISP). Therefore, new machines are more productive than the ones they replace. Consequently, each new unit of investment can be thought of as increasing the capital stock by q units if measured in units of the previous vintage of equipment. The price of a new unit of capital also can be thought of as being q times the price of an old unit of capital Thus, growth in ISP can be tracked by movement in the relative price series q. This series can be approximated by the ratio of the implicit price deflator for personal consumption expenditures on nondurable consumption goods and services (excluding housing services) and the implicit price deflator for producer durable equipment.2 Estimates of ISP were constructed following the methodology developed by Greenwood, Herkowitz, and Krusell (1997).3 q was estimated using the chain price indexes for personal consumption expenditures for nondurable goods and services and producers durable equipment from the U.S. National Income and Product Accounts.4
6. TNP represents sources of productivity growth that affect the organization of capital and labor in the production process, including such factors as the skills of the labor force and the nature of its training (which alternatively can be classified as human capital) and organizational structure and management skills.5 TNP estimates were obtained after solving a general equilibrium vintage capital model that incorporated ISP in the aggregate production function. At first glance, it appears puzzling that the significant gains in ISP have not been accompanied by corresponding gains in TNP. However, the efficient utilization of newly introduced technologies historically has tended to be preceded by an adoption and learning period during which efficiency decreases as changes in the organizational structure of the production units are being introduced.6 The learning process can last a considerable period of time during which TFP growth slows down or even declines as the gains in ISP are offset by the losses of TNP,7 a phenomenon widely documented at the industry level.8 Hence, it is not surprising that the “downsizing” of firms and the adoption of new information technologies in the late 1980s were associated with negative or stagnant TNP growth.9 At some later stage of the reorganization and learning period, TNP should start to show signs of recovery, as production processes combine labor and equipment more efficiently, as has generally been the case in the 1990s.
List of References
Baily, M.N. and Gordon R., 1998, “The Productivity Slowdown, Measurement Issues, and the Explosion of Computer Power” Brookings Papers on Economic Activity, pp. 347–420.
Cullison, W., 1989, “The U.S. Productivity Slowdown: What the Experts Say” Federal Reserve Bank of Richmond Economic Review 75, pp. 10–21.
David, P. A., 1991, “Computer and Dynamo: The Modern Productivity Paradox in a Not-Too-Distant Mirror” in Technology and Productivity: The Challenge for Economic Policy (Paris: OECD).
Gallman, R.E., 1992, “American Economic Growth before the Civil War” in Gallman R.E. and Wallis J.J., eds., American Growth and Standard of Living Before the Civil War, (Chicago: The University of Chicago Press).
Gordon, R., 1990, The Measurement of Durable Goods Prices, NBER Monograph Series (Chicago: University of Chicago Press).
Greenwood, J., Herkowitz Z., and Krusell P., 1997, “Long-Run Implications of Investment-Specific Technological Change” American Economic Review 87, pp. 342–62.
Greenwood, J., and Jovanovic B., 1998, “Accounting for Growth,” Working Paper 6647 (Cambridge, MA: NBER).
Greenwood, J., and Yorukoglu M., 1997, “1974,” Carnegie-Rochester Conference Series on Public Policy 46, pp. 49–96.
Hornstein, A., and Krusell P., 1996, “Can Technology Improvements Cause Productivity Slowdowns,” in Bernanke B. and Rotemberg, J., eds. NBER Macroeconomics Annual 1996 (Cambridge, MA: NBER).
Hulten, C., 1992, “Growth Accounting when Technical Change is Embodied” American Economic Review 82, pp. 964–80.
Jovanovic, B., and Nyarko Y., 1995, “A Bayesian Learning Model Fitted to a Variety of Empirical Learning Curves,” Brookings Papers on Economic Activity, Microeconomics, pp. 247–49.
Krusell, P., Ohanian, L.E. Rios-Rull, JV. and Violante, G.L. 1997, “Capital-Skill Complementarity and Inequality: A Macroeconomic Analysis,” Research Department Staff Report 329 (Minneapolis: Federal Reserve Bank of Minneapolis).
Mokyr, J., 1994, “Technological Change, 1700–1830” in Floud R. and McCloskey, D. eds.? The Economic History of Britain since 1700 (New York; Cambridge University Press).
Solow, R., 1957, “Technical Change and the Aggregate Production Function” The Review of Economics and Statistics 39, pp. 312–20.
Yorukoglu, M., 1998, “The Information Technology Productivity Paradox” Review of Economic Dynamics 1> pp. 551–92.
Prepared by Jorge A. Chan-Lau.
From the definition of ISP and, for simplicity, by assuming a one good economy, one unit of investment can be seen as producing q units of capital in the next period, and the relative price of capital equipment in terms of the consumption good is equal to 1/q in a competitive equilibrium, q then can be expressed as the price of nondurable consumption goods and services relative to the price of capital equipment. See Greenwood et al (1997), Greenwood and Jovanovic (1998), and Hornstein and Krusell (1996) for details.
Estimates for ISP in Greenwood, Herkowitz, and Krusell (1997) were made through 1992 and were based on NIPA data prior to the substantial revisions made to the data in 1997.
Altematively, ISP can be estimated using the Tornqvist index methodology proposed by Gordon (1990) and used in previous studies of ISP (Greenwood et al., 1997, and Krusell et al, 1997). The Tornqvist index is a cumulative exponential index of growth rates, each of which aggregates the underlying subcomponent growth rates by a weighted average of the expenditure shares in the two periods used to compute the growth rate. In contrast, the chainprice index is based on the Fisher formula and equal to the geometric mean of a Laspeyres price index for the previous period and a Paashe price index of the current period. Both indexes avoid the problems related with fixed-weighted indexes. The Tornqvist ISP index grows consistently faster than the chain-price ISP index. Nevertheless, both indexes show that ISP has grown at a higher rate since the late 1980s.
See Lindbeck and Snower (1995) for a review of the economic literature on organizational change.
See Mokyr (1994) for a detailed analysis of the Industrial Revolution in Britain, Gallman (1992) for a similar study of the United States during the Antebellum period, and David (1991) for a chronicle and analysis of the introduction of the electric motor in United States.
Jovanovic and Nyarko (1995) document learning curves for a variety of industries and activities.
