Abstract
Central policy issues in health care systems are the optimal hospital size of hospitals and the capacity utilization of existing structures. If there are economies of scale, the optimal size of hospitals will be larger and long run marginal costs will be below average costs. In a national health service with public planning and financing this has obvious policy implications for structural decisions related to both mergers and/or closing of existing hospitals and to the planning and building of new hospitals. We estimate a multiple output cost function from 133 observations of 19 Norwegian hospital trusts 2013–2019 with data on operating and capital costs as well as the number of somatic patient admissions aggregated by the weights of the diagnosis related groups (DRGs). The cost function is estimated using a) the panel data Stochastic Frontier Analysis (SFA) method with a translog functional form, and b) the non-parametric bootstrapped Data Envelopment Analysis (DEA) method with the Førsund-Hjalmarsson formula for scale elasticities. Based on tests of the variable specification we use three outputs: medical inpatients, surgical inpatients and outpatients. The long-run scale elasticities are estimated to be in the order of 1.07–1.09 for the average hospital and cost efficiencies are in the order of 85%, both significantly different from unity. Both cost efficiencies and scale elasticities are lower in university hospitals.
Similar content being viewed by