This study deals with three interrelated aspects of the economics of nursing home care in Connecticut: financing, cost and efficiency.

The dissertation starts with a discussion of the different sources of nursing home financing. As public payments are primarily made through Medicaid, a review of different Medicaid reimbursement systems is provided along with a discussion of provider incentives for cost containment, access, and quality of care under each system. The financing formulas for Medicaid and self-pay patients in Connecticut are considered in detail to highlight their dependence on costs incurred by a facility. Next, a generalized translog multi-product cost function is estimated for a sample of Connecticut nursing homes to generate information on overall and product specific economies of scale, the complementarity or substitutibility relations between inputs, and complementarity or anti-complementarity between outputs. The estimates of marginal costs for Medicaid and private patient days are used to examine the cross-subsidization hypothesis involving these two outputs.

Any inference about efficiency of a particular home based on an econometric estimate of the cost function uses the average performance of the industry as a benchmark for comparison. Moreover, these efficiency estimates are sensitive to the choice of the specific parametric form of the cost function. The present study uses instead the Data Envelopment Analysis (DEA) developed by Charnes, Cooper and Rhodes to measure efficiency of each nursing home in the sample. Efficiency of a home is computed on the basis of observed physical quantities of outputs and inputs and then compared against the extrema--minimum and maximum--within the peer group rather than with reference to some arbitrary average frontier which might seldom exist in practice. Initially, efficiency is computed within a one-output, seven-input framework. The study is then extended to include four outputs, also taking into account the activity of daily living (ADL) index to see how patient-mix and severity of case-mix variables affect efficiency status achieved by a home.

Finally, an attempt is made to explain the differences in these computed efficiencies in terms of facility specific and market area characteristics. (Abstract shortened with permission of author.).