| While in the past decade or so there has been tremendous growth in Medicare managed care, this growth has been uneven across the country. The limited availability of HMO coverage in certain areas, particularly in rural America, has been a source of on-going concern. Understanding the geographic differences in the Medicare HMO market becomes an important task, not only for guiding the on-going process to improve the Medicare+Choice program and thus bring equity back to the program, but also for evaluating “managed care competition” proposals for future Medicare reform.; This dissertation investigates differences in Medicare managed care development across urban and rural areas between 1993 and 2001. I use a unique and nationally representative dataset combining nine years of enrollment data with other county-level information to examine the impact of federal payment policy in affecting Medicare HMO market. This analysis extends the extant literature in three ways. First, it includes important explanatory variables that are previously missing, notably a measure of payment volatility. Second, it examines not one, but three outcomes associated with the availability of Medicare HMOs, that is, whether a county has access to at least one Medicare HMO plan, the number of available plans, and the Medicare HMO penetration rate in a county. Finally, it improves model specifications by taking the unobserved county-level heterogeneity into consideration and allowing the effect of key variables to differ across urban and rural counties.; The results indicate a positive correlation between the capitation payment rate and the development of Medicare HMO plans. However, this effect is not as strong as suggested by the previous research and proved sensitive to model specifications. Some long overlooked factors, such as the payment volatility, the employers' influence, the Medigap premium and neighborhood dependence, are found significantly related to the outcomes. The evidence also suggests that influences from many explanatory variables differ significantly across urban and rural areas, and even between those rural counties with cities, and those without cities. Simulations show that using payment as the sole policy lever for improving the access for rural areas could prove either ineffective or too expensive. |
