December 2005, Vol. 128, No. 12
Download the PDF (36K)
Book reviews from past issues
Adequacy of Earnings Replacement in Workers’ Compensation Programs. Edited by H. Allan Hunt. Kalamazoo, MI, W.E. Upjohn Institute for Employment Research, 2004, 176 pp., $16/paperback.
To evaluate the adequacy of a social insurance program, we have to measure results against policy objectives. Cash benefits of workers’ compensation are intended to ensure some level of earnings for injured employees, but whether or not this level is adequate depends on the extent of loss. How do we measure that loss? How much more would the worker be earning had an injury not occurred? A growing body of research attempts to answer these questions. Lest you think that the effort to quantify reduced earnings is purely theoretical, California lawmakers in January 2005 implemented a revised workers’ compensation schedule for rating permanent disabilities in awarding benefits. Intended to "promote consistency, uniformity, and objectivity," that new schedule adjusts for "diminished future earning capacity," as estimated by a wage loss study.
In the late 1990s, the National Academy of Social Insurance (NASI) commissioned a panel to analyze similar wage-loss studies and review current provisions to evaluate the level of workers’ compensation benefits. The study panel presents its findings in Adequacy of Earnings Replacement in Workers’ Compensation Programs, edited by H. Allan Hunt, Assistant Executive Director of the W.E. Upjohn Institute and Vice-Chair of the NASI Workers’ Compensation Steering Committee.
Despite the intentionally narrow focus of the analytical discussion, the book begins with a comprehensive, easy-to-understand description of workers’ compensation. In less than 70 pages, the NASI Study Panel of Benefit Adequacy goes beyond simply introducing the basic features of workers’ compensation to explain the conceptual and methodological issues involved with adequacy determinations and wage-loss measurement. Intended primarily for policy stakeholders and the research community, this volume may have broader appeal, because it provides an excellent introduction to issues that might not be so readily obtained from reviewing any particular wage-loss study.
Restricting the scope of adequacy evaluation to an analysis of cash benefits and wage loss was deliberate for three basic reasons. First, potentially large noneconomic losses, such as a diminished quality of life due to impairment, are difficult to measure. More importantly, as the study panel asserts, "Workers’ compensation was not developed as a system intended to replace noneconomic or nonwork disability losses, but as a no-fault system for rapid and certain compensation for the economic consequence of workplace injuries and illnesses." Further justification for limiting the discussion to wage benefits is that workers’ compensation programs generally ignore nonwage compensation. Note, however, that this means injured workers may still bear considerable losses, even if wage replacement adequacy is achieved.
Earnings loss occurs for a variety of reasons, including reduced or lower-paid work schedules due to physical restrictions following recuperation from injury. Pre-injury earnings are not necessarily a good proxy for post-injury earnings, especially when there are lengthy recuperation periods. Future earnings often differ from current earnings, especially for entry-level and older workers. In addition, there may be other influences on future wages, such as job changes, changing economic conditions, and unemployment.
Before estimating wage loss, the study panel reviews statutory provisions. Using national disability claim distributions, the panel computes what an average cash benefit would be, by State and for the Nation, for the period of 1972 through 1998. The results of this analysis are mixed. The panel found that the average weekly Temporary Total Disability (TTD) wage benefit for the Nation as a whole increased from 80 percent of the poverty threshold in 1972 to almost 110 percent in 1998. However, with substantial variation among States, this benefit, received by workers who are unable to work temporarily due to injury, was below the poverty level in 16 States. Overall, for all benefit types, including those for permanent disabilities, NASI found that the average expected workers’ compensation benefit rose by 30 percent in real expected benefits, with most of this occurring between 1972 and 1976. The panel noted that this initial benefit increase probably resulted from the 1972 report of the congressionally-mandated National Commission on State Workmen’s Compensation Laws.
That 1972 report, in addition to finding that benefits in most States were "not adequate," contained a number of far-ranging recommendations intended to improve workers’ compensation program design and administration. The report also provided a "model" benefit structure, and the NASI study panel used that formula as another benchmark for State comparison. Although not all members of the panel approve of using it as a standard of measurement, it does provide a different perspective to the adequacy discussion. For example, in terms of the expected 1998 statutory benefit, New York was the third most generous State in the Nation—but when compared to model benefit structure, the State ranked 14th. Nationally, TTD benefits improved relative to the 1972 report, reaching nearly 90 percent of the model in 1998. Benefit adequacy for permanent disability benefits, however, stagnated, or even declined, since the mid 1980s.
With a simulated wage loss based on actuarial estimates, the panel computed a wage replacement rate—the ratio of expected statutory benefits to expected wage losses. Although the replacement rates for TTD grew to 55 percent, other types of claims only achieve a 20-percent or lower replacement rate, according to this assessment. If the assumed wage loss would increase over the course of a permanently disabled claimant’s work life, these estimates are further reduced. Considering that both the 1972 National Commission report and 35 States use a two-thirds target replacement rate for TTD cases, this analysis points to inadequacy.
The NASI panel found wage replacement performance of Permanent Partial Disability (PPD) benefits to be "markedly lower than for other types of benefits." These benefits are received by workers who never fully recover from their injury, but they are able to work. Such cases, amounting to less than one-third of workers’ compensation claims, account for almost two-thirds of all workers’ compensation costs. Permanent disabilities complicate the earlier statutory analysis, because they are highly dependent on assumptions about long-term loss. Complexities surround estimates of disability duration, especially with respect to nonscheduled benefits. In addition, benefit levels often depend on physician disability evaluations and other subjective factors that cannot easily be captured in a formula. Thus, "there has never been any consensus about the optimum level of wage replacement" for PPD benefits.
NASI notes that a major drawback of its analysis of statutory provisions using a simulated wage loss is that it standardizes losses by using the same distribution of injured workers in each State. It does not take into account differences in benefit structure or program design. Also, a standardized claims distribution assumes that workers in different States face the same losses. But, as BLS data show, the rate and duration of work-related injury varies among States.
Because of a discrepancy between statutory benefits and those realized, along with the limitations of the statutory analysis, recent empirical studies of wage loss use actual claims data—coupled with administrative records for large populations of workers—to estimate earnings loss. Researchers use matching or regression analysis to estimate a profile of earnings over time, before and after injury, to control for additional characteristics and more accurately estimate the effect of injury on earnings. The panel notes two methodological issues affecting the quality of these analyses. First, the choice of comparison groups is critical, because some characteristics, such as occupation and industry, may affect both the probability of injury and the level of earnings. Some covariates, such as age and tenure, may affect the duration of and recovery from injury. In addition to control group issues, researchers also have to contend with missing earnings data in the administrative data for periods subsequent to injury. Missing data may be attributable to many factors, including workers moving out of State. Up to now, some researchers have treated these factors as injury-related or as nonrelated, unusable data, with other researchers providing a range of values to account for both possibilities.
The wage studies reveal several common findings. There is much variation in replacement rates for permanent disabilities among different jurisdictions and among workers within jurisdictions. Also, the length of the observed period after the injury impacts any assessment of wage replacement benefit adequacy—for example, a RAND study found that workers injured in 1991 experienced a wage replacement rate of 58 percent at 3 years, and 48 percent at 5 years. This difference may be partially explained by lump-sum payments or other front-loaded compensation.
To date, researchers have studied only 5 States. Full comparability does not exist, and part of the problem for the future is that of available data. Data on worker characteristics as well as coding practices may differ between compensation data systems and administrative data sources, such as unemployment insurance files. These data may also differ between one State and another. Part of the purpose of this book, as the panel asserts, is to improve the comparisons among States, as well as to further refine and standardize study methodology: "Our hope is that this volume will help to stimulate development of such a standard, through further discussion and debate." The adequacy measures presented in this book are useful for State comparisons and benefit trends, but more data about subgroups of the injured population (for example, by age, type of injury, or occupation and industry) will surely enhance the adequacy discussion, as will more information about post-injury outcomes. The panel concludes by listing 14 questions for further study in the context of a research agenda.
The question of adequacy is likely to continue even with improved wage-loss studies. The panel admits, "The policy remedy for a mismatch between the wage losses injured workers experience and the benefits delivered under workers compensation programs requires attention to more than the basic level of benefits." Aside from the manifold assumptions used to develop empirically-based target wage replacement rates, the statutory benefit formula used in State programs "of necessity includes more factors than the wage replacement rate." It can be argued that these extra factors are actually partially responsible for the adequacy shortfall—target rates are subject to adjustments and cut-offs, as they are in the new California benefits schedule.
Nevertheless, "The study panel believes that wage loss studies are the best yardstick to measure the adequacy of benefits." This work explores a variety of approaches to evaluating adequacy, and it is both a useful summary of recent research, as well as groundwork for future studies. With these "groundbreaking" wage loss studies, the controversy surrounding workers’ compensation is not going to go away, but hopefully the results will help us reduce the adequacy shortfall.
Bureau of Labor Statistics,
Boston-New York region
Within Monthly Labor Review Online:
Welcome | Current Issue | Index | Subscribe | Archives
Exit Monthly Labor Review Online:
BLS Home | Publications & Research Papers