انتخاب طرح، هزینه بیمه سلامت و به اشتراک گذاری حق بیمه

We develop a model of premium sharing for firms that offer multiple insurance plans. We assume that firms offer one low quality plan and one high quality plan. Under the assumption of wage rigidities we found that the employee's contribution to each plan is an increasing function of that plan's premium. The effect of the other plan's premium is ambiguous. We test our hypothesis using data from the Employer Health Benefit Survey. Restricting the analysis to firms that offer both HMO and PPO plans, we measure the amount of the premium passed on to employees in response to a change in both premiums. We find evidence of large and positive effects of the increase in the plan's premium on the amount of the premium passed on to employees. The effect of the alternative plan's premium is negative but statistically significant only for the PPO plans.

Over the last decade, health insurance premiums have increased much faster than general wage growth. Between 2000 and 2010, the hourly cost of health insurance doubled while wages only grew by 33 percent.1 As health insurance costs have grown to constitute a larger share of the total compensation package paid by employers, researchers have paid greater attention to the issue of access to health insurance through employer sponsored plans. Most of the existing literature has focused on the extensive margin of health insurance access, i.e. whether firms make employer sponsored plans available to their employees. Significantly less attention has been paid to the intensive margins – how much of the cost of providing health insurance is paid for by firms or the average quality of these insurance plans. This paper focuses on the first of these intensive margins, asking how the price of health insurance affects workers’ contributions when firms offer multiple health insurance plans. The literature on the firm's decision to include health insurance in the compensation packages for workers has concluded that this decision is typically price inelastic.2 In fact, while the percentage of firms that offered health benefits decreased from 68 percent in 2000 to 59 percent in 2009, this drop seems to be driven primarily by firms employing less than 10 workers (Kaiser/HRET, 2007, Exhibit 2.2). The percentage of large firms (i.e. firms that employed more than 200 employees) that offered health insurance did not change at all during the same period. Large firms almost universally offer health benefits (99 percent), and this rate has remained unchanged over the same time period. We argue that, although the premium size may not have an impact on a firm's decision to offer health insurance, it could have an effect on the employer's decision at the intensive margin, i.e. how to split the cost with workers. In fact, the percentage of firms paying for the entire cost of a single coverage plan decreased dramatically from 32 percent in 2001 to 18 percent in 2009, (Kaiser/HRET, 2007, Exhibit 6.15). In light of this evidence, the question of how premium sharing between employees and employers responds to increases in health insurance premiums becomes a pressing one since shifting the cost of health insurance onto workers may affect the probability a worker enrolls in the company plan. In fact, Cutler (2003) shows that most of the recent decrease in employer sponsored health insurance coverage can be explained by a decrease in take-up rates by employees due to an increase in the cost born by them rather than a decrease in offering rates by employers. In this study we want to disentangle what has affected this shift in the cost of health insurance on workers and specifically whether the size of the premium itself is a determinant of the cost sharing selected by the firm. We show that this is not a trivial question since, in a world of prefect sorting and flexible wages, health insurance premiums should not affect a worker's contribution, at least for the low quality plan. In addition to almost universally offering health insurance, large firms are also more likely to offer multiple plans compared to smaller employers. In 2001, 56.7 percent of firms with more than 50 employees offered more than one plan in 2001, while only 12.5 percent of smaller firms offer multiple plans (Crimmel, 2001). For this reason, we develop a model in which firms offer a choice of two plans to attract workers with heterogeneous preference for health coverage. This is an important contribution to the current literature, since previous studies all assume that each firm offers only one plan, which is an unreasonable assumption for large firms. Our theoretical model predicts that, under the assumption of wage rigidities, premium sharing should be a function of the own premium cost as well as the cost on any other plan offered by the firm. We estimate our model using information on firms that offer both Health Maintenance Organization (HMO) plans and Preferred Provider Organization (PPO) plans. We find that the own-premium elasticities, i.e. the percentage increase in the worker's contribution following a 1 percentage increase in the premium, are generally all greater than 1, implying that firms shift the burden of rising premiums onto workers. The rest of the paper is organized as follow: the next section reviews the literature on this topic. Section 3 describes the model and Section 4 outlines the methodological strategies adopted for the estimation. Section 5 describes the data while the empirical results are discussed in Section 6. We conclude with some final remarks on the main finding of this paper.

This paper developed a model of how firms determine employee contributions toward health insurance premiums when multiple plans of different quality are offered. We derive the conditions for premium sharing under two possible scenarios: (1) flexible wage and (2) wage rigidities. Our model predicts that when there are no wage rigidities and workers can sort themselves across firms according to their preferences for the wage/health insurance trade off, the employer should provide the low quality plan health insurance free of charge and ask individuals who prefer a higher quality insurance plan to pay the difference in the premiums of the two plans. When we include wage rigidities, the employee's contribution toward each plan becomes a function of the wage growth in the industry, the premium growth over time, and the level of both premiums in the previous period. Testing the models using firm-level data from 2005 to 2011, we reject the hypothesis of perfect sorting in the naïve model, after finding that the employees’ contribution to an HMO plan is a function of the PPO premium while the employees’ contribution to the PPO is not a function of the HMO premium, thus questioning the flexible wage hypothesis. In the model assuming wage rigidities, the results obtained via fixed effects estimation on a sample restricted to firms with consecutive observations show support for our predictions. Overall, the results highlight the importance of (1) running separate regression for the worker's contribution to each plan and (2) including both premiums in the regression rather than an average or just the own premium as done in previous research, since each premium has an opposite effect on the contribution level sets by the firm for each plan. Consequently, using only one premium and one contribution level would lead to an underestimation of the true responsiveness to the employee's contribution to the plan premium. Finally, assuming wage rigidities leads to bigger elasticity estimates than using a model with flexible wages. Hence the results of the specification based on the wage rigidity assumption better explain the recent shift of the cost of health insurance on employees. This result may explain why previous studies have failed to find that employer's contributions were responsive to premium increases. Overall, we found that the employees’ contributions to PPO plans are more sensitive to own premium increases than employee's contributions to HMO plans. This study focuses only on employee contribution's setting toward individual coverage plans only. Future research should extend this analysis to derive the conditions for employees’ contributions when a firm offers both individual and family coverage plans.