Academic journal article Journal of Risk and Insurance

Economic Determinants of Workers' Compensation Trends

Academic journal article Journal of Risk and Insurance

Economic Determinants of Workers' Compensation Trends

Article excerpt

Why Costs Rose: Frequency and Severity Components

Estimates indicate that workers tend to increase the frequency and severity of workers' compensation claims as benefits increase and as the waiting period for receiving benefits falls. But this can be either because they are willing to undergo more risk ex ante which results in more injuries ex post--behavior that I call risk bearing moral hazard--or because higher benefits induce them to report more injuries and increase the duration of their claims--behavior that I call claims reporting moral hazard. This article adds to recent literature that aims to sort out the significance of these two types of behavior in explaining cost trends in compensation costs (Butler and Worrall, 1991b, for example) by looking at differential incentive responses in two different data sets.

Employee costs in workers' compensation have risen sharply in the last 15 years, escalating from something less than 1 percent of total payroll costs to over 2 percent. These average costs per employee can be factored into two components: claim frequency or the number of claims per employee, and claim severity or the average cost per claim.

This article examines indemnity costs from 1954 to 1984 for 39 states in a sample of National Council on Compensation Insurance (NCCI) states and all states reporting to the Occupational Safety and Health Administration (OSHA) (43 states) from 1976 to 1986 (though data for all states for all years are not available in either sample). This data set ignores medical costs, which are also reimbursed under the workers' compensation system but which deserve a separate treatment and do not enter into the actuarial "trend calculation" that is the focus here.

Indemnity costs per worker can be decomposed in a natural way as follows:

Indemnity Costs/Employees = Indemnity Costs/Number of Claims [center dot] Number of Claims/employees = Severity [center dot] Frequency.

Taking logarithms of the terms on both sides of the equation leads to

ln Indemnity Costs/Employees = ln(Severity) + ln(Frequency).

The logarithms of frequency and severity will be the dependent variables in the statistical analysis employed in this article. The estimated coefficients from the regressions of these variables on the logarithm of benefits measure the percentage response in the dependent variable to a 1 percent increase in benefits. These estimates are, respectively, the benefits elasticity for frequency and severity.

Review of the Literature

Severity Studies

Much initial severity research has examined the transition from claimant to nonclaimant status using the NCCI's Detailed Claim Call data base. These transitions are modeled as a function of the wage, the worker's weekly compensation benefit, and other claim information. Worrall and Butler (1985) used proportional hazard models in their estimates, indicating benefits significantly increased claim severity. Since the technique they employed assumes nothing about the distribution of nonwork spells of disability except that they are separable in time, benefits elasticities are not immediately recoverable and were not calculated.(1) They found a significantly positive influence of benefits on the duration of claims but were unable to calculate the benefits elasticity. Hence, Butler and Worrall (1985) and Worrall et al. (1985) fit a general class of distribution functions to duration data, including controls for inherent differences in claims. Using general parametric and nonparametric controls for unobserved heterogeneity, Butler and Worrall (1991a) find benefit effects on duration elasticities consistent with their earlier research for the generalized gamma family of distributions. Other duration elasticities were provided by Meyer, Viscusi, and Durbin (1989) by looking at average durations before and after a real increase in benefits in a given state.

The principal finding of these studies is that the duration of disability varies directly with workers' compensation benefits and indirectly with the wage. …

Search by... Author
Show... All Results Primary Sources Peer-reviewed


An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.