ترجیحات مالیاتی برای مزایای جانبی و واجد شرایط بودن کارگران برای بیمه سلامت کارفرما

In this paper, I estimate the effect of the tax preference for employer-provided health insurance on the probability that an employee will be eligible for health insurance at work. I instrument for marginal tax rates with state-level state income tax rates and I control for unobservable state effects that could be correlated with the state income tax instrument by comparing the estimated tax effect on employer-provided health insurance, to the tax effect on sick leave, a fringe benefit that does not enjoy any tax advantage. The size of the tax effect on health insurance eligibility is positive, as predicted, and is both statistically and economically significant.

Recently there has been much public dismay over the large number of Americans who are uninsured. The fear that a lack of insurance means a lack of adequate health care has prompted policy proposals ranging from universal government-provided health insurance to mandated employer-provided insurance to alternative tax incentives such as medical care spending accounts. Most recently, legislators have introduced bills that would provide tax credits to the uninsured to pay for or offset the cost of health insurance premiums. Policymakers have often focused on insurance offered through the employer since employers are currently the most common source of group health insurance. At the same time, rising public and private expenditures on health care have focused policy debates on the effect that high levels of insurance may have on health care costs. Being insured lowers the price of health care to the individual, inducing individuals to consume more health care, possibly pushing up health care prices. The current tax policy of excluding employer-provided health insurance from individual taxable income and from a firm’s payroll base provides incentives for higher levels of insurance than individuals might otherwise choose, leading to suggestions that the tax exclusion for employer-provided fringe benefits such as health insurance be eliminated or capped. A change in tax policy designed to decrease possible overinsurance due to tax preferences for fringe benefits could also affect the probability that a person will be offered health insurance through the employer and thereby the number of workers who have access to group health insurance. Some alternative policies designed to increase insurance coverage rates such as the tax credit legislation currently on the table would also work through the tax system. To measure the effect of either type of policy change, we must first establish how taxes affect the probability that a worker is offered health insurance at the workplace. This task is more difficult than it might first appear since it is difficult to find variation in tax rates that is not associated with variation in unobservable variables that could affect the demand for health insurance. I offer a new method for estimating such tax effects using state-level state tax measures as instruments for marginal tax rates and through comparisons between fringe benefits that are and are not tax preferenced. Using the Survey of Employee Benefits from the Current Population Surveys of 1988 and 1993 I look at how income taxes affect the probability of being eligible for health insurance through one’s employer. In order to avoid the endogeneity problems that may arise from using individual marginal tax rates as the explanatory variable, I use the cross-state variation in state income tax rates to instrument for an individual’s marginal tax rate. In order to ensure that any estimated tax effects are not simply due to unobservable differences across states in demand for fringe benefits, I compare the effect of tax rates on employer-provided health insurance to their effect on the provision of sick leave, a fringe benefit that is not tax advantaged. I control for state-level non-tax-related differences in fringe demand across states using differences across states in taxable fringes. Under certain assumptions, described in more detail below, any difference in differences in tax excludable benefits can then be attributed to differences in tax rates. My approach avoids the bias caused by using endogenous measures of marginal tax rates as well as any omitted variable bias due to omitted state effects, allowing me to obtain unbiased estimates of the effect of the income tax preference on eligibility for employer-provided health insurance.

In this paper, I examine the effect of income taxes on the probability that a worker is eligible for employer health insurance. In order to avoid the problems associated with using individual marginal income tax rates based on taxable income as the explanatory variable, I instrument for marginal tax rates using income tax differences across states. To assure that this state-level instrumental variable is not simply picking up unobserved state effects that are correlated with the demand for health insurance, I compare the effect of taxes on health insurance fringe benefits to the tax on two other fringe benefits, one of which, sick leave, does not enjoy a tax-advantaged status relative to wages. The estimates imply that the tax preference for fringe benefits has a substantial and significant effect on the probability that a full-time worker has access to health insurance through the employer. Given the current importance of employers as a source of access to group health insurance, these results suggest that policymakers considering policies that decrease the tax preference for fringe benefits must carefully consider the cost of the loss of access to group health insurance that might accompany such changes or be prepared to establish structures other than the employment relationship through which group health insurance could be more widely offered.