Efficiency of Hospitals in Victoria under Casemix Funding: A Stochastic Frontier Approach

This paper estimates the cost frontier for large Victorian public hospitals using the econometric technique of stochastic frontier regression. An estimate of the mean level of hospital cost inefficiency in 1994/95 is obtained, in the period following the introduction of casemix funding. One of the major goals of Victoria's casemix funding system introduced in 1993 was to promote efficiency, hence hospital efficiency under the new system is of interest. We might expect to find a low level of cost inefficiency under a casemix funding system. The cost frontier is estimated by the econometric technique of stochastic frontier regression, which is the preferred efficiency measurement technique as it allows for data errors. The application of frontier models allowed us to estimate the degree of random and non-random efficiency variation across hospitals. Inpatient output is found to explain the majority of inter-hospital variation in total cost. The average level of cost inefficiency in total operating expenditure is estimated to be around 3 percent, using an exponential distribution for the inefficiency component. Where there are differences in costs between hospitals, they appear to be related to administrative, medical support and hotel labour inputs