This thesis addresses the effects of increased costs of access to health care on the demand for care. Four hypotheses concerning the demand for ambulatory care are tested. 1. Increases in the cost of access to physician care will cause a decrease in expenditures on physician care and alter the types of treatment individuals receive. 2. The effect of changes in money prices and access to care is different for non-elderly adults, elderly adults, and children. 3. The cost of access to medical care and money prices will have more pronounced effects on chronic conditions as opposed to acute conditions. 4. Estimation of health care expenditures is sensitive to the model employed. The 1987 National Medical Expenditure Survey (NMES) data base is used to estimate a multi-stage model of health care demand. The NMES data base includes a measure of willingness to pay for a reduction in doctor's office waiting time. This variable is used to measure the opportunity cost of time which has proved difficult to measure in previous studies. The NMES also includes measures of travel times, waiting times, and travel modes. The multi-stage model of health care demand is composed of health expenditures and types of treatment. Health expenditures are estimated with two different models because of the sensitivity of model choice. The models are the two-part model of demand and Ordinary Least Squares (OLS) estimation of transformed expenditures, where expenditures are transformed by the Inverse Hyperbolic Sine (IHS) transformation. The IHS transformation is defined for zero values and works well with extreme- valued distributions - both properties of the distribution of health expenditures. Treatment choices are estimated conditional on positive expenditures using the almost ideal demand system.
