تقاضای کارکنان بیمه سلامتی کارفرما و انتخاب های برنامه درمانی

Although most private health insurance in US is employment-based, little is known about how employers choose health plans for their employees. In this paper, I examine the relationship between employee preferences for health insurance and the health plans offered by employers. I find evidence that employee characteristics affect the generosity of the health plans offered by employers and the likelihood that employers offer a choice of plans. Although the results suggest that employers do respond to employee preferences in choosing health benefits, the effects of worker characteristics on plan offerings are quantitatively small.

As the intermediary between health insurers and consumers in the purchase of the vast majority of private health insurance, employers play an important role in health insurance markets in the US. In 1999, over 88% of those with private health insurance, received their coverage through an employer (Mills, 2000). Underlying the prevalence of employer-sponsored coverage, however, is considerable variation in the health plans workers and their families receive. Employers vary in whether they offer health insurance, the number and types of plans they offer, and their premium contribution policies (Gabel, 1999). Differences among employees in the coverage they receive from their employers have generated concern over the incentives facing employers in choosing health plans for workers. My objective in this paper is to examine the relationship between employee preferences for health insurance and the plans offered by employers. My hypothesis is that employee preferences affect employer choices—employers have incentives to consider the preferences of their employees when choosing health plans, although they are constrained in their ability to cater to the diverse preferences of employees by the fixed costs of offering multiple plans. I empirically test this hypothesis by examining the relationship between the characteristics of workers which affect individual demand for health insurance and the generosity of health plans offered by employers. Although the results suggest that employers do respond to employee preferences in choosing health benefits, the effects of worker characteristics on plan offerings are quantitatively small.

The empirical results provide some evidence that employers respond to employee preferences for health insurance when choosing their health benefits. The average generosity of the plans offered by employers is higher in firms with a greater proportion of high-wage workers and variation among workers in their health risk and wages is positively associated with the probability of offering a choice of plans, although this last result is sensitive to the sample selection correction. Although the estimates from the selection-corrected models should be interpreted cautiously, they suggest that variation among workers in their demand for plans may be a barrier to employers in offering health insurance. The conditional effects of the workforce composition variables on the probability of offering a choice of plans are smaller than the population effects obtained in the selection corrected model. This implies that variation in worker characteristics would have had a larger effect on the probability of offering a choice of plans in firms choosing not to offer health insurance. The effects of the variables used as proxies for workers preferences, however, are generally small in magnitude, and not all variables demonstrated the hypothesized relationship. In particular, variation in worker characteristics did not explain variation in the set of plans offered conditional on offering a choice of plans. A number of possible explanations for this exist. First, although the survey is superior to other employer level surveys for the purpose of this study, it offers limited information on individual worker characteristics. Better data on individual workers within a firm, including variables such as family income, detailed measures of health status, and the availability of coverage from other sources, may generate stronger results. In addition, variation among plans along other dimensions, such as the availability of particular providers, may be related to worker preferences. This is consistent with the findings of Moran et al. (2000), who use an alternative measure of plan diversity and find a positive relationship between preference heterogeneity and plan diversity. Another possibility is that other objectives may also drive the decision of an employer to offer a choice of plans. For example, employers adopting a managed competition approach to designing health benefits would offer multiple plans with standardized benefits in order to create competition among plans for enrollees based on cost and quality (Enthoven, 1989). Finally, many employers may be reluctant to offer multiple plans varying in their generosity due to the potential for generating adverse selection within the firm. The results, however, have important implications for researchers using data from individuals purchasing coverage in the employer-sponsored market. In particular, the plan or set of plans offered to an employee is endogenous with respect to the characteristics of both the individual employee and the other workers in the employee’s firm. The endogeneity of the employee’s choice set potentially affects the results of studies of both individual choice among health plans based on data from the employer-sponsored market and the effect of employer benefit design on premiums. Although more precise estimates of the effects of workforce composition on employer choices are necessary to assess the magnitude of the potential bias, these results point to the importance of an issue often overlooked by researchers using data from the employer-sponsored market. The ultimate objective of developing a greater understanding of employer health benefit choices is to determine how these decisions affect the welfare of employees. Although this study provides evidence supporting the proposition that employers have incentives to respond to worker preferences when choosing health plans, it does not provide strong support for a particular model of employer behavior. For example, not all workers in the firm may be marginal in the employer’s decision, with variation in the preferences of the subset of marginal workers driving the employer’s choice of plans. In addition, the preferences of some workers may be weighted more heavily than the preferences of others in the employer’s decisions. Thus, this study does not allow us to determine whether employers choose health benefits for their workers in a socially optimal fashion. Many important questions remain regarding the relationship between employee preferences for health insurance and the decisions made by employers to evaluate the implications of employer choice for the welfare of workers.