Workers' Compensation and the Level of Occupational Injuries

The unusual circumstance of a substantial decrease in the level of workers' compensation benefits for a well-defined group of employees permitted investigation of the relationship between benefit levels and injury rates. It was found that lower benefit levels were associated with a significantly lower frequency of injuries. There was no significant relationship between benefits and injury severity. The nature of the quasi-experimental data allows the inference that it was the benefit level change that brought about the change in frequency rates rather than the reverse. Workers' compensation has two important social functions. It is intended to provide income and medical security for injured workers and prevention incentives for employers [9]. Assessing the achievment of the security goal has always been relatively straightforward in that one need only measure the benefits paid to injured workers and compare them with whatever standard of adequacy is deemed appropriate [1]. Until quite recently, however, the contribution of workers' compensation to safety incentives has been a matter of speculation rather than analysis. Complex issues of causation, data availability, and statistical methodology long kept the prevention issue from being systematically addressed. Beginning with the National Commission on State Workmen's Compensation Laws (1972) and spurred along by the dramatic changes in workers' compensation during the 1970s, a program of research has begun to examine the role of workers' compensation in encouraging prevention [2,4,6,9]. It is the purpose of this paper to further understanding of the safety incentives of workers' compensation by analyzing data covering the previously unexplored circumstance of a decline in workers' compensation benefits. James R. Chelius is Associate Professor in the Institute of Management and Labor Relations at Rutgers University. He earned the Ph.D. degree at the University of Chicago. A former Director of the Occupational Safety and Health Center at Rutgers, he currently serves on a GAO Panel studying the veteran's disability system. Dr. Chelius authored Workplace Safety and Health: The Role of Workers' Compensation and edited Current Issues in Workers' Compensation. Karen Kavenaugh earned the M.S. degree in Industrial Relations and Human Resources at Rutgers University. This content downloaded from 157.55.39.127 on Wed, 29 Jun 2016 05:22:45 UTC All use subject to http://about.jstor.org/terms 316 The Journal of Risk and Insurance Workers' Compensation Safety Incentives Since the debate over the initial passage of workers' compensation laws at the beginning of this century, it has been asserted that an employer's automatic liability for work injuries would encourage greater attention to prevention. The predefined visible penalty of having to pay workers' compensation benefits has been thought to yield a greater incentive than the potentially large but uncertain liabilities of the previously used tort mechanism. The scant empirical evidence on this issue indicates that the introduction of workers' compensation was indeed associated with a lowering of industrial death rates [5]. It follows from the same rationale that the level of compensation benefits might influence the extent to which employers are willing to invest in prevention. Higher benefits would seem to yield a greater incentive for prevention. The presence of compensating wage differentials for risk, however, makes the relationship between benefits and incentives less straightforward. To the extent workers' compensation benefits merely substitute for wage premiums that otherwise would have been paid, a larger benefit simply changes the manifestation of the safety incentive rather than its magnitude. While there is some evidence of an offset between wages and workers' compensation benefits in the non-union sector [7], it is not so widespread or perfect so as to eliminate the potential relationship between benefit levels and employer safety incentives. The relationship between benefit levels and safety is also complicated by the impact of workers' compensation on employees. Just as a higher benefit increases the cost of an accident to the employer, it decreases that cost to the employee. A higher workers' compensation benefit thus may give less prevention incentive to employees. Since both employers and employees have a role in accident prevention, it is impossible theoretically to determine whether higher benefits cause the net level of accidents to be higher or lower. In practice the impact of higher benefit levels is further confounded because such benefits give employees an incentive to report real or imagined injuries that would go unreported if benefits were lower. While theoretically ambiguous, the empirical evidence on benefit levels and safety has been quite consistent. Studies of this relationship indicate that higher benefits are associated with higher levels of reported injuries [2,4,6,9]. This does not necessarily mean that employers are not encouraged to greater prevention efforts by higher benefits. It merely indicates that the influence of benefit levels on employee behavior is greater than their influence on employer behavior. Many uncertainties remain, however, as to the exact nature of this relationship. It is sometimes asserted, for example, that the positive relationship between benefits and injuries is due to high injury rates causing high benefit levels rather than the reverse the hypothesis being that high levels of injury cause governments to enact high benefit levels. Whether the This content downloaded from 157.55.39.127 on Wed, 29 Jun 2016 05:22:45 UTC All use subject to http://about.jstor.org/terms Workers' Compensation and the Level of Occupational Injuries 317 high injury rates associated with high benefits are due to more real injuries or simply more reported injuries is also a controversial issue. The unusual circumstance of a decline in workers' compensation benefits in a quasiexperimental setting provided an opportunity to analyze the robustness of the positive relationship between benefits and injury rates. The data also allow for some insight into the causal nature of the relationship as well as the type of injury most influenced by benefit levels.