Explaining a Productive Decade

Abstract

Explaining a Productive Decade Stephen D. Oliner, Daniel E. Sichel, and Kevin J. Stiroh Productivity growth in the United States rose sharply in the mid-1990s, after a quarter century of sluggish gains. That pickup was widely documented, and a relatively broad consensus emerged that the speedup in the second half of the 1990s was importantly driven by information technology (IT).1 After 2000, however, the economic picture changed dramatically, with a sharp pullback in IT investment, the collapse in the technology sector, the terrorist attacks of September 11, 2001, and the 2001 recession. Given the general belief that IT was a key factor in the growth resurgence in the mid-1990s, many analysts expected that labor productivity growth would slow as IT investment retreated after 2000. Instead labor productivity accelerated further over the next several years. More recently, however, the pace of labor productivity growth has slowed considerably. [End Page 81] In light of these developments, researchers and other commentators have been intensely interested in the course of productivity growth since 2000. Distinguishing among the possible explanations for the continued strength in productivity growth is challenging, because much of that strength appeared in measured multifactor productivity (MFP), the unexplained residual in the standard growth accounting setup. Nevertheless, potential explanations can be divided into two broad categories: those centered on IT and those unrelated or only loosely related to IT. The simplest IT-centered story—that rapid technological progress in the production of IT and the induced accumulation of IT capital raised productivity growth—does not work for the period after 2000, because the contributions to growth from both the production and the use of IT declined. A second IT-related story that has received a great deal of attention is that IT investment proxies for complementary investments in intangible capital, and a growing body of research has highlighted the important role played by such intangibles.2 A third IT-related story identifies IT as a general-purpose technology that spurs further innovation over time in a wide range of industries, ultimately boosting growth in MFP.3 Because this process takes time, the gains in MFP observed since 2000 could reflect the follow-on innovations from the heavy investment in IT in the second half of the 1990s. Another broad set of explanations highlights forces not specific to IT. Gains in labor productivity since 2000 could have been driven by fundamental technological progress outside of IT production, as implied by the strong growth in MFP in other sectors.4 Alternatively, the robust advance in labor productivity could reflect broader macroeconomic factors such as normal cyclical dynamics, a decline in adjustment costs after 2000 as investment spending dropped back, greater-than-usual business caution in hiring and investment, or increased competitive pressures on firms to [End Page 82] restructure, cut costs, raise profits, and boost productivity. The profit-driven cost-cutting hypothesis, in particular, has received considerable attention in the business press.5 In this paper we try to sort out these issues using both aggregate and industry-level data.6 We investigate four specific questions. First, given the latest data and some important extensions to the standard growth accounting framework, is an IT-centered story still the right explanation for the resurgence in productivity growth over 1995-2000, and does IT play a significant role when considering the entire decade since 1995? Second, what accounts for the continued strength in productivity growth after 2000? Third, how has investment in intangible capital influenced productivity developments? Finally, what are the prospects for labor productivity growth in coming years? Our analysis relies in part on neoclassical growth accounting, a methodology that researchers and policymakers have used for many years to gain insights into the sources of economic growth. Notably, the Council of Economic Advisers, the Congressional Budget Office, and the Federal Reserve Board routinely use growth accounting as part of their analytical apparatus to assess growth trends.7 Of course, growth accounting is subject to limitations, and in recent years many analysts have leveled critiques at this methodology. For example, the standard neoclassical framework does not explicitly account for adjustment costs, variable factor utilization, deviations from perfect competition and constant returns to scale...