اثر یارانه های بیمه درمانی خود اشتغالی در شغل آزاد
|کد مقاله||سال انتشار||مقاله انگلیسی||ترجمه فارسی||تعداد کلمات|
|27335||2010||13 صفحه PDF||سفارش دهید||محاسبه نشده|
Publisher : Elsevier - Science Direct (الزویر - ساینس دایرکت)
Journal : Journal of Public Economics, Volume 94, Issues 11–12, December 2010, Pages 995–1007
This paper estimates the effect of an increase in the deductibility of health insurance premiums for self-employed individuals on the probability of being self-employed. Using a panel of tax returns from 1999 to 2004, we estimate fixed effects instrumental variable regressions for the probability of being self-employed, entering into self-employment, and exiting from self-employment. Our results suggest that this policy increased the probability of being self-employed by 1.5 percentage points, and increased the probability that a taxpayer would be primarily or exclusively self-employed by 1.1 and 0.35 percentage points respectively. These effects explain about a third to a half of the total increase in self-employment by these definitions over the sample period. We also find that the probability of entering self-employment increased by 0.8 percentage points and find suggestive evidence that the probability of exit decreased by 2.8 percentage points.
When potential entrepreneurs are deciding whether to leave their job and strike out on their own, the cost of health insurance can be an important consideration, since leaving a job often carries with it the loss of employer-sponsored health insurance. In addition, direct purchases of insurance while self-employed do not receive the same favorable tax treatment that is provided for purchases through an employer, and this differential treatment may decrease the likelihood that an individual would want to enter self-employment, or to remain self-employed, ultimately resulting in a lower level of entrepreneurship. This paper examines the extent to which making the cost of health insurance more favorable for the self-employed through changes in the tax code increases the level of self-employment in the United States. In the United States, the after-tax cost of health insurance depends on whether the insurance is purchased through an employer or not for several reasons. First, purchases of employer-sponsored insurance (ESI) generally carry lower administrative costs due to lower expenditures on underwriting, advertising, and enrollment. Second, ESI premiums reflect greater risk pooling, which lower the variance of medical claims and reduces the risk insurers face. Third, the types of policies available are often very different, with non-ESI policies generally providing less coverage and requiring larger out of pocket expenses. Finally, the tax treatment of ESI purchases is more favorable, since purchases through an employer are generally excluded from both payroll and income taxes, whereas purchases by the self-employed are not excluded from payroll taxes, and until 2003 were not fully deductible from income taxes.2 In recent years, policy changes have served to decrease the tax differential between employer-sponsored health insurance and health insurance purchases by the self-employed. The Tax Reform Act of 1986 (TRA86) included a provision that allowed the self-employed to deduct 25% of health insurance premiums. More recently, legislative changes in 1998 increased the deductibility of health insurance premiums for self-employed individuals for federal income tax purposes from 60% in 1999 to 70% in 2002 and to 100% in 2003, which further reduced the after-tax price of health insurance while self-employed. A number of papers, including Gruber and Poterba, 1994, Selden, 2009 and Heim and Lurie, 2009 each examine the effect of a subset of these tax changes on the take-up of health insurance among the self-employed and find significant effects on the decision to purchase health insurance.3Heim and Lurie (2009) additionally find a significant effect of the tax change on the amount of insurance purchased. The exogenous change in the cost of health insurance induced by these changes in tax policy also provides a natural experiment to assess the sensitivity of the self-employment decision to health insurance prices.4 Two working papers have examined the effect of these tax policy changes on the propensity to be self-employed. Velamuri (2008) uses CPS data to examine the effect of TRA86 on the decision of employed women to become self-employed, and finds that the rate of self-employment among women who were not covered under a spousal plan rose by between 14% and 25% after TRA86. Gumus and Regan (2009) also use CPS data, and find small effects of the 1997–2006 expansion of the self-employed health insurance deduction on entry, but no significant effects on exit from self-employment or the likelihood of being self-employed.5,6 It is important to note that the responsiveness of self-employment behavior to changes in the after-tax price of health insurance could take several forms. An employee of a firm that offers health insurance may decide to leave their wage and salary employment altogether to start an independent business when the increased deductibility of premiums lowers the cost of health insurance. A taxpayer that is partially self-employed and receives health insurance from their employer may decide to decrease or eliminate their wage employment, becoming primarily or exclusively self-employed with insurance purchased on their own. In addition, the lowered cost of health insurance may induce individuals to experiment with self-employment. A part-time worker who already purchases insurance in the individual market may decide to start up a side business, since any profits up to the cost of their health insurance policy are effectively exempt from income taxes. Or, a worker who is not covered by insurance may decide to start a small business, and also purchase health insurance because of the increased deduction.7 This study, then, examines self-employment behavior using a panel of tax returns from over 236,000 taxpayers that spans from 1999 to 2004. It advances the literature in at least three ways. First, because it uses a panel in which the self-employed are followed for up to six years, it is possible to include fixed effects to account for unobserved characteristics that are correlated with the regressor of interest and outcome variables. Second, the paper examines whether the impact of the policy differs depending on the intensity of self-employment under examination (whether the taxpayer has at least a minimal amount of self-employment income, whether the majority of their income flows come from self-employment, or whether all of the taxpayer's earned income comes self-employment). Third, since the tax data reports whether the self-employed health insurance deduction was claimed, it is possible to test whether any responsiveness to the policy change is centered among those who actually claim the deduction, which would be expected if any estimated effect is indeed due to the policy change. Three limitations to this study should be noted. First, unlike the Current Population Survey, which has often been used in studies of self-employment, tax data does not contain information on whether taxpayers identify themselves as self-employed. Thus, we must resort to using information on the amount of reported self-employment income at the tax unit level to create measures that attempt to ascertain whether the taxpayer is self-employed. Second, because eligibility to claim the self-employed health insurance deduction depends in part on whether the taxpayer is offered health insurance by an employer, which is not observed in tax data, we cannot account for this in our key independent variable, the after-tax price of self-employed health insurance. Because this is more likely to be a problem for married taxpayers (for whom an offer of family health insurance for one spouse can make the other spouse ineligible to claim the deduction), we report some results separately by marital status. Third, the effect of the deduction may differ across taxpayers depending on a number of characteristics, some of which (such as the desire to have children or health status) are unobserved in tax data. We are able, however, to examine whether the estimated effect differs along observable dimensions (such as marital status and the age of the taxpayer). Using fixed effect instrumental variable models, we estimate the extent to which the level of self-employment is related to the after-tax price of health insurance. Our results suggest that the decline in the tax price of health insurance while self-employed between 1999 and 2004 increased the probability of being self-employed by about 1.5 percentage points. We also find significant effects on different intensities of self-employment, with an increase in the probability that a taxpayer would be primarily or exclusively self-employed by 1.1 and 0.35 percentage points respectively. These effects explain about a third to a half of the total increase in self-employment by these definitions over the sample period. As would be expected, we find that this increase in the propensity to be self-employed is centered among those who claimed the self-employed health insurance deduction. We also find this decrease in price increased the probability of entering self-employment by 0.8 percentage points, and find suggestive evidence that the decrease in price decreased the probability of exiting self-employment by 2.8 percentage points. The paper proceeds as follows. Section 2 describes tax policy toward health insurance for the self-employed. In Section 3, the estimation strategy is outlined, and Section 4 describes the panel of tax returns that is used in this study. Section 5 presents the estimation results, and Section 6 concludes.
نتیجه گیری انگلیسی
The Omnibus Consolidated and Emergency Supplemental Appropriations Act of 1998 (OCESA) increased the deductibility of health insurance premiums and in effect reduced the price of health insurance while self-employed. This exogenous change in the cost of health insurance provides a natural experiment to assess the sensitivity of the self-employment decision to health insurance prices. Using a panel of tax returns from 1999 to 2004, we estimate the extent to which the level of self-employment, the entry into self-employment and exit from self-employment are related to the after-tax price of health insurance. Our results suggest that the decline in the tax price of health insurance while self-employed between 1999 and 2004 increased the probability of being self-employed by about 1.7 percentage points. We are able to show that tax units responded along numerous margins, both in switching between no self-employment and self-employment and in increasing the intensity of self-employment (including up to the point of being exclusively self-employed), with the estimated coefficients implying that the rate of self-employment increased by between 9.1% and 14.9% due to the change in policy during these years. For the margin between no and some self-employment, we also find that the decrease in the tax price of health insurance for the self-employed increased the probability of entering self-employment by 0.8 percentage points, and find suggestive evidence that the probability of exit decreased by 2.8 percentage points. Overall, the results in this paper imply substantial effects of the increase in the deductibility of self-employed health insurance on the propensity of taxpayers to be self-employed. Since a number of recent papers (including Gruber and Poterba, 1994, Selden, 2009 and Heim and Lurie, 2009) found significant effects of the same or earlier increases in the deduction on health insurance purchasing behavior, it is perhaps not surprising that this paper finds significant effects on the decision to be self-employed. In addition, the results are consistent with Velamuri (2008), who found that the TRA86 introduction of the self-employed health insurance deduction increased the rate of self-employment among woman by between 14% and 25%, though they are inconsistent with Gumus and Regan (2009), who find small or no effects. Finally, the fact that we find a significant effect of health insurance considerations on the decision to be self-employed is also consistent with results in Madrian and Lefgren, 1998, Wellington, 2001 and Gumus and Regan, 2009, and Fairlie et al. (2008), who find that having some other source of health insurance coverage has a significant effect on the decision to be self-employed. To put the size of our estimated effects in context, it is important to note that, during the time period under analysis, the rate of self-employment increased for each of our three definitions of self-employment by between 21.5% and 27.4%. Thus, although our estimated effects are somewhat large, the policy change still only explains around a third to a half of the actual increase in self-employment during this period. Nevertheless, our results suggest that making health insurance for the self-employed cheaper by providing a larger tax deduction had a large and significant effect on the propensity to be self-employed, the numbers of individuals entering self-employment and the likelihood individuals will remain self-employed. Consequently, these results imply that policies which further reduce the cost of health insurance for the self-employed are likely to increase the rate of self-employment.