Work-related accidents and the level of market competition: an analysis of worker injury rates at U.S. Steel Corporation, 1907-1939

Economic Inquiry, July, 2008 by Christopher S. Decker, David T. Flynn

I. INTRODUCTION

During the late 19th and early 20th centuries, manufacturing in the United States experienced substantial changes. There was tremendous growth in large-scale enterprises beginning with business consolidation during the trust movement of the 1880s. Technological advances generated substantial economies of scale, and the ensuing need for greater capital accumulation tended to favor large firms. Large firms came to dominate markets in industries such as steel and petroleum. As markets continued this trend into the early 20th century, worker safety became an issue of concern for workers, firms, and the country. Aldrich (1997), for instance, compiled a large amount of information demonstrating managements' increased concern over the issue of worker safety, motivated in some cases out of a desire for good public relations and in others as a profit-maximizing strategy. Fishback and Kantor (1998) found that the rapid spread of worker compensation laws during the early part of the 20th century was the product of coalitions between organized labor, insurers, and even employers who anticipated gains from replacing the existing negligence liability institution with a more structured and predictable legislative structure.

While a number of studies examined the development of worker safety activities by firms and investigated reasons for the development of such regulation, little attention has been given to the potential link between market structure and worker safety incentives. Our study focuses on worker accidents and competition in the U.S. steel industry at the turn of the century. In what follows, we develop a theoretical model demonstrating that increased competition generates greater efforts on the part of a firm to improve its worker safety efforts. An increase in the number of competitors reduces the dominant firm's market share, increasing incentives to seek and adopt cost-reducing strategies so as to mitigate profit losses, such as labor accident costs. Empirical analysis supports this result.

The existence and importance of a link between accidents and market structure have surfaced in several venues other than worker safety. While not entirely parallel to the turn-of-the-century steel industry, one might consider, for instance, the airline industry and its deregulation in 1978. Some argued at the time that the benefits of deregulation, primarily lower fares prompted by entry into the industry, would be offset by reductions in safety as increases in the number of flights could lead to increases in air congestion and airplane accidents. Risks would further increase as aggressive competition and tighter profit margins would prompt airlines to cut maintenance and equipment upgrade expenditures to unsafe levels. However, data presented in Kaplan (1986) and Walters (1993) show that fatal airline accidents actually decreased after deregulation. Moreover, in an investigation of the relationship between various measures of industry structure and the frequency of environmental accidents across a variety of manufacturing industries in the United States, Decker and Wohar (2006) found that higher accident rates are strongly associated with higher levels of industry concentration. Therefore, more competition is associated with fewer accidents.

We explore this relationship from a historical perspective, investigating the relationship between worker accident rates and the competitive environment in the United States' steel industry between 1907 and 1939. Since this industry was dominated by U.S. Steel Corporation (U.S. Steel), we focus particular attention on it (see below). Indeed, visual observation of the relationship between worker accident rates and firm market share highlights a compelling relationship. Utilizing market share data for U.S. Steel ingot production over time from 1901 to 1939 given by McCraw and Reinhardt (1989) as well as Schroeder (1953) and U.S. Steel injury rates published in Aldrich (1997), Figure 1 shows that there does appear to be a direct positive correlation between the company's market share and its worker accident rates. Indeed, there is an 85% correlation between U.S. Steel's market share and accident rates over the 1907-1939 period.

While this is striking, there are many other variables that could explain declining worker injury rates. To establish a credible relationship between these two variables requires (a) a theoretical foundation, consistent with profit-maximizing behavior, linking accidents with market share, and (b) empirical validation of the relationship. After controlling for other factors such as unionization rates, production levels, number of workers, and others, our empirical findings show that a lower market share results in lower worker injury rates, leading us to conclude that worker safety benefited from increased competition. (2)

The remainder of the paper is structured as follows: in Section II, we present a brief discussion of the state of the United States' steel industry with special attention to worker safety issues. In Section III, we develop a theoretical model relating worker accident rates to industry structure. In Sections IV and V, we present the empirical model, data, and related econometric issues. In Section VI, we present the empirical results, and in Section VII, as a robustness check on these results, we focus more broadly on the steel industry as a whole. In Section VIII, we conclude and offer suggestions for future research.