Duke 2008 | ASHE | Equity and Efficiency in Health and Healthcare

There are large differences in Medicaid spending among states. In 2002, per-enrollee spending varied by a factor of 3.3. California spent $2,317 and Rhode Island spent $7,540. One should keep in mind, however, that states have the flexibility to administer the Medicaid program and determine its scope within broad statutory limits. States may also choose to include additional eligibility groups or provide additional benefits. As a result, it can be difficult to generalize about the types of enrollees who are covered and the benefits that are offered. Other differences in spending among states can be explained by the magnitude of managed care enrollment.

Using claims data from 2002, we examine the contribution of demographic characteristics (age, sex, and race), health status, and prices in explaining the variation in Medicaid spending. In particular, we follow the methodology used in the Dartmouth Atlas of Health Care to examine geographic differences in Medicare spending. Preliminary results suggest that among beneficiaries enrolled in fee-for-service and receiving full benefits, the coefficient of variation is 0.48. The coefficient of variation ranges from 0.35 for aged, disabled, and adult beneficiaries to 0.62 for children. In contrast, the unadjusted coefficient of variation among Medicare beneficiaries is 0.22. This is not surprising, however, because the uniformity in benefits and, to some extent, prices (both largely determined by the federal government).

The next steps of the analyses will adjust state spending for differences in prices and compute state predicted spending controlling for the age, sex, and race, and health status. In addition, we will attempt to gauge the contribution of other factors to the variation in Medicaid spending (such as supply of providers). Lastly, we plan to study the relationship between relative spending and health outcomes.