Evidence from both anecdotal and investigative sources suggests that nursing home staffing and clinical practice varies considerably across geographical area, even after the functional status of nursing home residents is taken into account. This dissertation documents the existence of significant inter-regional variation in nursing home staffing and clinical practice, and identifies regional market characteristics associated with variation.

The conceptual model guiding this analysis assumes that the output of the nursing home production process is measurable along the dimensions of quantity and quality. Severity-adjusted prevalence rates for nursing home clinical practices represent quality. Staffing is modeled as an input into the production of both quality and quantity.

As efficient providers, nursing home providers have the incentive to substitute across and within categories of inputs as relative wages change. Since all facilities within a circumscribed market area face the same wage structure, they have the same incentives for substitution, a potential source of variability between markets. Structural characteristics of the market which determine the supply and demand for quality were identified from industrial organization theory and incorporated into a simultaneous equation model.

The unit of analysis for this study is the county. Information on nursing homes was obtained from a 1987 federal survey. The Area Resource File provided variables describing the structure of the nursing home market. The model was estimated by 2SLS regression. In reduced form, the model was estimated by weighted least squares.

Substitution in response to relative wages or to the relative supply of nursing home personnel is strongly suggested by the results. Evidence for substitution between labor and capital is also obtained from the results of the 2SLS estimations. Competition in the nursing home market was hypothesized to enhance the quality of nursing home services. Results indicate that in markets in which regulation attempts to limit capacity or contain costs, quality is lower. However, the degree of competitive rivalry or the availability of community substitutes for nursing home care does not have an association with nursing home quality.

This study contributes to current knowledge of nursing home markets, and has implications for the understanding of the working of health care markets in general. It demonstrates that rather than resulting in waste and duplication with little or no identifiable contribution to quality, non-price competition may lead to beneficial increases in the quality of care provided in nursing homes. By considering a treatment setting in which physician participation and incentives do not predominate, this research should contribute to our understanding of the determinants of variation in all health care settings.