The study is a systematic empirical analysis to test some hypotheses. The objectives were (1) to analyze the sufficiency of government subsidy for poor people to cover the cost of hospital care for the hospital to reach cost recovery, and (2) to test whether economic theories can be used to analyze factors affecting the sales volume of product/service ofpublic general hospital.
Population studied were 35 general hospitals owned by district governments in Central Java. From each hospitalfour year time series data were taken (pooling method), so that there were 140 observations (at the end there were only 136 observations, since four observations could not be done due to incompleteness of the hospital records).
Analysis of government subsidy for poor community showed that the subsidy could only cover 5% of the hospital's yearly deficit (total yearly deficit is Rp. 1.5 billions in average). Meanwhile, correlation analysis revealed that only two variables having significance and strong correlation with sales volume of hospital's product/service, i.e. incidence of disease and spending for social marketing. Other variables had no correlation with sales volume of product/service, or the correlations were very weak. Multivariate analysis showed that only incidence of disease, consumer's income, geographic condition, and coverage of insurance, represented in a regression model, were affecting sales volume of hospital 's product/service. Results of the study confirmed that neoclassic economics theories are not suitable to analyze problems faced by government owned general hospitals. Sales volume ofproduct/service ofpublic general hospital was more determined by incidence of disease and consumer 'smotivation created by providers, especially physicians, not by variables regarded as dominant by neoclassic economics theories. The motivation was likely created through abuse of agency relationship which produces supplier-induced demand