In this major new work, Michael J. Moore and W. Kip Viscusi explore the question, "How are workers compensated for exposing themselves to the risk of physical injury while on the job?" The authors detail the diverse nature of labor market responses to job risks and the important role played by compensation-for-risk mechanisms. Following an overview of the literature, they present a number of unprecedented results. Comprehensive and systematic discussions of issues such as wage-risk tradeoffs, the effects of workers' compensation on wages and risk, the role of unions, and the role of product liability suits in job-related injuries make the volume an essential work for all those interested in risk policy and workplace safety. Among the major results presented for the first time are the first estimates of the value of life derived from recently released occupational fatality risk data from the National Traumatic Occupational Fatality Survey. From these same data the authors also demonstrate that higher workers' compensation benefit levels significantly reduce fatalities on the joba finding that challenges virtually every other treatment of this topic.
Originally published in 1990.
The Princeton Legacy Library uses the latest print-on-demand technology to again make available previously out-of-print books from the distinguished backlist of Princeton University Press. These editions preserve the original texts of these important books while presenting them in durable paperback and hardcover editions. The goal of the Princeton Legacy Library is to vastly increase access to the rich scholarly heritage found in the thousands of books published by Princeton University Press since its founding in 1905.
Read an Excerpt
Compensation Mechanisms for Job Risks
Wages, Workers' Compensation, and Product Liability
By Michael J. Moore, W. Kip Viscusi
PRINCETON UNIVERSITY PRESSCopyright © 1990 Princeton University Press
All rights reserved.
Each week, 125 American workers are killed by on-the-job accidents. Although this performance record represents an improvement over earlier years, society's aspirations with respect to job safety have risen during this period as well, as increased wealth has created an increased demand for safety and a willingness to support government efforts to foster improvements in safety.
Discussions of such modes of intervention generally follow two themes. First, while many people feel that health and safety outcomes result from a random and capricious process, over the past decade, economists have established the importance of market responses to job risks, in terms of both compensating wage differentials and turnover responses. In this book we extend the economic analysis of these market responses. In particular, we present new evidence on both the wage and the quit effects of job risks, and we explore other aspects of the risk compensation package. Most notable is our emphasis on the role of workers' compensation insurance as it relates to the mix of the components of compensation for job risks.
Second, most safety policy discussions focus on direct intervention options. The Occupational Safety and Health Administration (OSHA) has been regulating health and safety conditions in the workplace for almost two decades. In addition, for most of this century, the workers' compensation system has provided income replacement for injured workers and their surviving families. Thus, direct regulations are the policy remedy for risk levels that are too high, and a social insurance effort provides the ex post remedy for any risks that remain.
There is a general sentiment that neither of these efforts has fulfilled its original promise. Policies of the Occupational Safety and Health Administration are generally viewed as ineffective, having but a small impact on workplace safety. Workers' compensation is viewed more favorably, though many claim that the benefit levels provided are inadequate from the standpoint of income replacement.
Our analysis suggests that matters cannot be isolated in as simple a fashion as this. For example, one cannot view workers' compensation as simply a social insurance effort. Viewed only as social insurance, the funding support of the program represents a lump sum cost to firms, which must then bear the burden of this effort with no corresponding safety incentives created for the firms. The workers' compensation system is not, however, a broadly based social insurance scheme. Rather, it constitutes an insurance program directed quite specifically at reducing the costs to workers of undesirable workplace characteristics. Since these characteristics vary across firms, hazardous firms will reap greater benefits from these efforts, because their workers are more likely to receive benefits. These greater benefits to workers in turn generate lower wage costs for more risky firms, since workers willingly accept a lower base wage in return for greater workers' compensation benefits. Similarly, to the extent that the funding mechanism for workers' compensation is linked to the risk performance of the firm, safety incentive effects lead firms to provide greater levels of safety, thereby reducing their premium levels. These safety incentives may be more effective than OSHA'S direct regulations in terms of promoting safety, and the substantial levels of insurance provided make workers' compensation benefits an important compensation package component.
This book emphasizes the exploration of interactions such as these, and the refining of estimates of previously analyzed key relationships that constitute the compensation mechanism. In some cases, we raise new classes of issues. No previous study attempts to address through a formal test whether the level of social insurance provided by workers' compensation is optimal. Similarly, no previous empirical work analyzes the overlap between workers' compensation and the tort liability system. This latter omission is particularly serious, given the emerging product liability crisis of the 1980s.
In Chapter 2 we establish the research context for our work. We begin by summarizing the major details of the literature that address several of the key relationships that we examine. These relationships include the link between job risks and wages, the effect of workers' compensation benefits on safety, and the effect of workers' compensation benefits on wage levels. We then provide a capsule overview of how the structure of our study and the nature of the results that we report later in this volume differ from those in the literature. Chapter 2 does not set out to summarize the entire book. It does, however, provide a survey of the earlier work that has been done in the field and an indication of how our results relate to that work.
Our analysis begins in Chapter 3 by analyzing the relationship between wages and workers' compensation. The conceptual model that we develop relates workers' compensation and wages in a manner similar to the relationship between job risks and wages. Under the classic compensating differential model, workers demand wage premiums for jobs that pose extra risk. Insurance benefits for hazardous jobs, on the other hand, constitute a positive job attribute. In return for workers' compensation benefits for the risks they face, workers will be willing to work on risky jobs for a lower wage. We expect to observe a negative compensating differential for workers' compensation benefits, just as we expect a positive compensating differential for job safety. These differentials are related even further since the wage offset workers are willing to accept in return for more generous workers' compensation benefits is an increasing function of the risk.
The most novel insight produced by this model is that we develop an empirical procedure that makes it possible to test for the adequacy of workers' compensation benefits. Whereas the earlier literature both on workers' compensation and more generally on social insurance raised the benefit adequacy issue, until now there has been no available test to ascertain the validity of the claims of inadequacy. We use as our reference point for the appropriate benefit level the efficient level of benefits that workers would select either under conditions of actuarially fair insurance or under conditions that reflect current rates of insurance loading. We then develop an empirical test of benefit optimality based on observed wage responses to workers' compensation.
The results presented in Chapter 3 from 1977 survey data indicate that at that time benefit levels were underprovided. This result corroborates the general consensus in the field, including that of the National Commission on State Workers' Compensation Laws. In particular, the wage offsets that workers willingly accept, as revealed by the data, greatly exceed the actuarial value of the benefits.
The wage equation estimates also provide estimates of the implicit value that workers place on on-the-job injuries. For the lost workday injuries considered in Chapter 3, workers' wage-risk tradeoffs imply valuations per expected injury in the range of $30,000 to $40,000. The more distinctive aspect of these results is that we are able to break down this valuation into the valuation of health loss and the valuation of income loss, thus providing the first empirical estimates of the costs of pain and suffering and nonwork disability in the literature. The valuation of injuries consists in large part of the nonmonetary losses, as the pain and suffering component ranges from $17,000 to $26,000 per injury.
In Chapter 4 we present an update of the 1977 results in an attempt to capture the influence of the dramatic increases in workers' compensation benefit levels in the late 1970s and in the early 1980s. Although the presumed inadequacy of benefits in the early 1970s was a major factor in an expansion in benefit amounts, there has been no examination in the literature of whether the subsequent benefit increases were inadequate or excessive, judged from the standpoint of optimal social insurance. The results that we present in Chapter 4 indicate that the change in the structure of workers' compensation benefits closed the benefit inadequacy gap. By 1982, the wage offsets that a worker accepted corresponded closely with what one would expect if the benefit levels were adequate.
These substantial wage offsets suggest additional implications as well. As workers' compensation premium levels passed the $20 billion per year mark, observers viewed this escalation in employer costs with alarm. Our analysis suggests, however, that the premium level is not an appropriate measure of the true cost of workers' compensation to the employer. In assessing the net cost of benefits to the firm, one must include the wage offset resulting from the benefits. In doing so, we find that the workers' compensation system more than pays for itself, on average.
In Chapter 5 we shift our focus from nonfatal injuries to fatal job accidents. Much of the interest in the compensating differential literature derives from its use in policy debates on the value of life. In particular, what do individuals' risk-dollar tradeoffs with respect to risks of death tell us about the value of risk reduction? These statistics do not yield the value of a certain death, but they do tell us something more generally useful, which is how individuals react to small probabilities of death in terms of their willingness to bear the risk or their willingness to pay for reduction in those risks.
Value of life numbers are used throughout the federal government to provide a yardstick for the cost-effectiveness of government policies. To assess the stringency of risk regulations, it is often instructive to compare the cost per life saved with the value that the beneficiaries of these programs implicitly place on the lives saved through these governmental efforts. Since available survey data on the U.S. labor market are so extensive, most analysts focus on estimates of the wage-risk tradeoff as a source of these implicit valuations.
The death risk data previously utilized in such studies constitute the weak link in the valuation process. Until recently, death risk data series based on spotty fatality reports have been extrapolated to obtain national fatality risk projections. Because occupational fatalities are not a particularly frequent event, this procedure is fraught with error.
In 1987, the National Institute of Occupational Safety and Health (NIOSH) released a new set of fatality risk data based on a census of all occupational fatalities that occurred during 1980–1984. The results in Chapter 5 provide new estimates of the value of life based on these data and then compare the estimates with those obtained from a more traditional data source, the U.S. Bureau of Labor Statistics data.
We generate quite striking results. The shift to the new data set roughly doubles the estimated value of life, from a figure in the $3 million range to one in the $5–$6 million range. The comparison of the two data sets suggests that the main reason for the increase stems from greater measurement error in the Bureau of Labor Statistics data.
Since individual lifetimes are finite, risk reduction policies do not save lives permanently. Rather, they extend them. This distinction is of substantial consequence from a policy standpoint, where often the choices among policies involve risk reductions for people in different age groups. One expects the value per expected life saved to increase with the expected longevity of the individual, but there has never been a concerted attempt to make such distinctions on an empirical basis.
From a conceptual standpoint, a closely related issue concerns the rate of discount that individuals use when valuing deferred risks of death. If one's life expectancy is extended by five years, then the fifth year of the additional life expectancy may well be valued differently than are the more immediate years. If markets were perfect, then we could use capital market rates as an index of individual rates of time preference with respect to long-term health risks. However, because of potential financial market imperfections and, in particular, because of the inability to trade health resources across time, many observers question whether rates of time preference for health coincide with financial rates of interest. At one extreme, some observers suggest that individuals may be myopic with respect to long-term risks so that they act as if they have high implicit rates of interest when making personal safety decisions, such as whether to wear a seat belt. On the other hand, governmental agencies are reluctant to discount health benefits at all because doing so decreases the attractiveness of many of their health risk reduction efforts. Some agencies urge that a zero discount rate be used.
The findings in Chapter 6 address each of these concerns. First, the estimated value per life year extended equals $175,000. The estimated rates of time preference with respect to these death risks range between 10 and 12 percent, which is consistent with financial market interest rates at the time of the employment survey. Thus, evidence does not indicate that rates of time preference with respect to deferred health risks diverge substantially from rates of time preference with respect to deferred dollar amounts.
The empirical results in Chapters 3 through 6 represent average estimates across an employment population. Attitudes toward risk differ across individuals in the population, however, suggesting important differences in the way in which the compensation mechanism works across different components of the labor force. Compensating differentials successful in attracting the worker to a job initially may not prove sufficient to retain the worker if he or she acquires very adverse risk information while working on the job. In these situations, learning about job risks induces worker quitting.
Although the job hazard–quit relationship is well established, a variant of this relationship is also of interest. Workers on the quit-no quit margin will tend to have a higher perceived risk of an on-the-job accident or death than will workers overall at a particular firm. Those workers on the quit margin consequently should place a higher value on workers' compensation benefits than do members of the work force in general By analyzing the wage–workers' compensation tradeoffs implied by a conventional wage equation as well as the wage–workers' compensation tradeoffs implied by workers' quit decisions, we can establish the different rates of tradeoff for each segment of the work force.
The results that we generate are consistent with theoretical predictions. Workers' compensation benefits reduce worker turnover significantly, a reduction in employment costs that should be taken into account when assessing the performance of this social insurance scheme. Furthermore, the results provide a new and more refined test of the role of worker learning and how it influences worker responses to the structure of the compensation package.
Labor unions represent another important institutional influence on the structure of compensation. Indeed, the high level of job accidents experienced earlier in this century provided the impetus for the initial unionization of many American workplaces. From a conceptual standpoint, labor unions provide the advantage to their workers of representing the preferences of the "inframarginal" workers, rather than those of new hires and the "marginal workers" — workers on the quit–no quit margin. Workers with substantial seniority rights and on-the-job experience are of less concern to the employer, because of these workers' limited mobility. In contrast, wages and job conditions must be made attractive to retain the marginal workers — new hires and those on the verge of quitting. Labor market evidence indicates that unions place a greater emphasis on the preferences of inframarginal workers than does the market, which leads to higher levels of fringe benefits and lower levels of job risks for union workers relative to nonunion workers.
Excerpted from Compensation Mechanisms for Job Risks by Michael J. Moore, W. Kip Viscusi. Copyright © 1990 Princeton University Press. Excerpted by permission of PRINCETON UNIVERSITY PRESS.
All rights reserved. No part of this excerpt may be reproduced or reprinted without permission in writing from the publisher.
Excerpts are provided by Dial-A-Book Inc. solely for the personal use of visitors to this web site.
Table of Contents
- FrontMatter, pg. i
- Contents, pg. vii
- List of Figures, pg. ix
- List of Tables, pg. xi
- Preface, pg. xiii
- Abbreviations and Symbols, pg. xvii
- One. Overview, pg. 3
- Two. The Research Context of the Analysis, pg. 12
- Three. The Performance of Workers' Compensation as a Social Insurance Program, pg. 34
- Four. Net Workers' Compensation Costs: Implications of the Wage Offset, pg. 53
- Five. Workers' Implicit Value of Life, pg. 69
- Six. The Value of Life: Quantity Adjustments and Implicit Rates of Time Preference, pg. 82
- Seven. Worker Learning and the Valuation of the Compensation Package, pg. 98
- Eight. The Role of Unions in Altering the Structure of Risk Compensation, pg. 111
- Nine. The Effects of Workers' Compensation on Job Safety, pg. 121
- Ten. Tort Liability Remedies for Job Injuries: Product Liability and Its Interaction with Workers' Compensation, pg. 136
- Eleven. Conclusion, pg. 162
- Appendix A. Estimation of the Value of Life Using Flexible Functional Forms, pg. 165
- Appendix B. A Conceptual Model of Worker and Firm Responses to Insurance Benefits, pg. 168
- Notes, pg. 173
- Bibliography, pg. 181
- Index, pg. 189