In my research, I investigate self-insured health plans and compare them to conventional insurance plans. Employment-based health insurance plans constitute the primary source of private health insurance coverage in the United States, provided to about 160 million people (Claxton, et al., 2005; Kaiser Commission on Medicaid and the Uninsured, 2007). Continually rising health care costs and the preferential tax treatment of fringe benefits (e.g., the premiums paid by employers are excluded from an employee's taxable income) have largely contributed to their increasing popularity. As a result of this tax subsidy, the relative price of health coverage obtained through the workplace is estimated to be 27% lower (Gruber and Poterba 1996).

Previous economic analyses of health insurance have largely focused on traditional health plans—those plans in which an employer contracts with a health insurance company (Buchmueller et al. 2002; Dranove et al. 2000; Long and Marquis 1999; Marquis and Long 2001). In contrast, little research has been done regarding employers' contributions towards health insurance premiums and the price elasticity of demand in self-insured health plans. In my econometric studies utilizing the 1987 National Medical Expenditure Survey (NMES) data series, specifically the Employment-Related Coverage and Household Survey data sets, I analyzed factors affecting employers' decisions on health insurance plans they offer to employees as well as the degree of responsiveness to changes in price (e.g., the price elasticity of demand) for employment provided health insurance.

My empirical analyses showed that self-insurance predicts higher employer contributions to health insurance premiums, which is consistent with the earlier empirical findings for conventional health insurance plans (Buchmueller et al. 2002; Marquis and Long 2001). Higher premium contributions offered by self-insuring employers may imply the cost advantage of self-insured health plans relative to conventional health insurance, as firms that self-insure do not have to obey various health insurance laws and regulations. Moreover, the results showed that self-insuring employers spend 18–25% more per policy holder on their health benefits (thus, on their medical treatments in general). Hence, for advocates of medical treatments, this effect may be a strong argument for a public policy that reduces regulations on health insurance. Similarly, for those not favoring medical treatments, it may be an argument for increasing regulation on health coverage. Finally, my studies have shown that the price elasticity of demand for self-insured health plans is within an inelastic range (from -0.08 to 0.01 depending on the model specification). This implies that policy holders are insensitive to changes in price of their health coverage.

More detailed description of this work can be found here and in Chapters 2 and 3 of my dissertation.

In another study, I investigated the effects of behavioral differences regarding the ownership form and the certification status on the length of hospice service use. Both of these issues, for-profit versus nonprofit organizational forms and certified versus not certified status of hospice agencies, are highly relevant especially to policymakers, care providers, and patients because they have implications with respect to the access, cost and quality of health care provided.

In recent years, much attention has been focused on the ownership form of hospices, mainly due to the fact that for-profit ownership of hospices has grown rapidly. In fact, the number of for-profit hospice (Medicare-certified) organizations has increased nearly 300% from 1992 to 1999, whereas the number of corresponding nonprofit organizations has risen 43% during the same period (General Accounting Office, 2000). This observation has become a starting point for this study, as I was interested in explaining differences between for-profit and nonprofit forms of hospices and the impact of the hospice certification status on the length of service use.

The existing empirical literature provides contradictory results regarding the differences between for-profit versus nonprofit organizational forms of hospices. Some findings imply the differential behavior across the ownership types (Carlson et al. 2004; Christakis and Escarce 1996; Foliart et al. 2001; Lindrooth and Weisbrod 2007; Ohri 2007) whereas others do not (Hamilton 1994; Lindrooth and Weisbrod 2007; Ohri 2007). My approach builds mostly on the work of Christakis and Escarce (1996) and Ohri (2007) and examines the behavioral differences with respect to the length of service use. Moreover, similarly to Lindrooth and Weisbrod (2007), I hypothesized that based on the U-shaped cost function, hospices may be inclined to maximize the duration of the patient's intermediate days in order to maximize their profits. Specifically, I investigated whether for-profit and nonprofit hospices behave differently with respect to "short-stay" versus "long-stay" patients and whether there is a selection bias towards patients with the longest expected service use.

My analyses have shown a positive impact of the for-profit organizational form with respect to the length of hospice use as compared to their nonprofit counterparts. In particular, the obtained results suggest that among individuals with short expected length of hospice use, patients at nonprofit hospices have lower mortality while using hospice care. In contrast, among those with long expected length of hospice use, patients at for-profit hospices have longer survival times. There is, however, no evidence of systematic selection of long-stay patients by for-profit hospices or of short-stay patients by nonprofit hospices. Furthermore, the results of the impact of hospice certification show that the length of hospice use is shorter at certified hospices as compared to noncertified ones.

More detailed description of this study can be found in Chapters 4 of my dissertation.