Reallocation and Technology: Evidence from the US Steel Industry

Abstract

We measure the impact of a drastic new technology for producing steel-the minimill-on industry-wide productivity in the US steel industry, using unique plant-level data between 1963 and 2002 Identifying the sources of productivity growth of firms, industries, and countries, has been a central question for economic research. There remain, however, many empirical obstacles to credibly identifying the underlying sources of productivity growth. First, the measurement of productivity at the producer level typically requires an estimate of the production function and, therefore, has to confront both the endogeneity of inputs and unobserved prices for inputs and outputs. Second, it is difficult to observe potential explanatory variables at the producer level, such as technology, competition, and management practices.1 Finally, in order to establish causality, exogenous shifters of such variables are required in order to trace out their effects on productivity.1 See Syverson (2011) for an excellent overview of the various potential determinants of productivity at both the producer and the industry level. Two prominent studies on the triggers of productivity growth are Schmitz (2005) and Olley and Pakes (1996) who study the role of two such triggers: import competition in the iron ore market and deregulation in the telecommunications market. Hortaçsu and Syverson (2007); Bloom, Eifert, Mahajan, McKenzie, and Roberts (2013); and Jarmin, Klimek, and Miranda (2009) show that factors such as vertical integration, management, and large retail chains lead to systematic differences in productivity between plants and, consequently, have implications for aggregate industry performance. * Collard-Wexler: Economics Department, Duke University, 233 Social Sciences, Durham, NC 27708 (e-mail: collardwexler@gmail.com); De Loecker: Economics Department, Princeton University, Fisher Hall, Princeton, NJ 08540 (e-mail: jdeloeck@princeton.edu). We would like to thank Jun Wen for excellent research assistance, and Jonathan Fisher for conversations and help with census data. We would like to thank three anonymous referees for excellent comments and suggestions. Furthermore, we thank Nick Bloom, Rob Clark, Liran Einav, Ariel Pakes, Kathryn Shaw, Chad Syverson, Raluca Dragusanu, and seminar participants at many institutions.

Health Care Exceptionalism? Performance and Allocation in the US Health Care Sector

“…This relationship has been analyzed extensively in a variety of industries and countries as a proxy for the role of competition in these settings (e.g. Olley and Pakes, 1996; Pavcnik, 2002; Escribano and Guasch, 2005; Bartelsman, Haltiwanger and Scarpetta, 2013; Collard-Wexler and De Loecker, 2015). Intuitively, competitive forces exert pressure on lower productivity firms, causing them to either become more efficient, shrink, or exit.…”

“…If ${\beta }_1^s$ is zero or negative, it indicates that lower quality facilities are the same size or larger than their high quality counterparts and suggests that forces beyond quality competition are driving the allocation of market activity. In the manufacturing productivity literature, ${\beta }_1^s$ ≤ 0 has been found in some former Soviet-bloc countries in the early 1990s (Bartelsman, Haltiwanger and Scarpetta, 2013) and in the U.S. steel industry circa 1960-70 (Collard-Wexler and De Loecker, 2015). 2 …”

Healthcare Exceptionalism? Performance and Allocation in the U.S. Healthcare Sector

“…If is zero or negative, it indicates that lower quality facilities are the same size or larger than their high quality counterparts and suggests that forces beyond quality competition are driving the allocation of market activity. In the manufacturing productivity literature, ≤ 0 has been found in some former Soviet-bloc countries in the early 1990s (Bartelsman, Haltiwanger and Scarpetta, 2013) and in the U.S. steel industry circa 1960-70 (Collard-Wexler andDe Loecker, 2015). 2 (2) where Δ h is a measure of the hospital's growth rate in admissions and all other variables are defined as in equation (1).…”

How Do Electricity Shortages Affect Industry? Evidence from India

“…Recall that the CEA separately reports the two components of shortages: Assessed Demand and Energy Available. 27 Online Appendix Figures A6 and A7 illustrate this. For example, the economy in Andhra Pradesh grows rapidly over the study period, while Uttar Pradesh grows relatively slowly.…”