Economies of scale and optimal size of hospitals: Empirical results for Danish public hospitals

Context and aim: The Danish hospital sector is facing a significant rebuilding programme, driven by a political desire to concentrate activity in fewer and larger hospitals. Our aim is to analyse whether the current configuration of Danish hospitals is subject to scale economies that may justify such plans and to estimate an optimal hospital size. Methods: We estimate cost functions using panel data on total costs, DRG-weighted casemix, and number of beds for three years from 2004-2006. A short-run cost function is used to derive estimates of long-run scale economies by applying the envelope condition. Results: We identify moderate to significant long-run economies of scale when applying two alternative translog cost functions. However, using a quadratic functional form we identify constant economies of scale for the medium-sized sub-groups and decreasing economies of scale for the largest sub-groups. The optimal number of beds per hospital is estimated to be 275 beds per site. Sensitivity analysis to partial changes in model parameters yields a joint 95% confidence interval in the range 130 – 585 beds per site. Conclusions: The results indicate that it may be appropriate to consolidate the production of small hospitals (<230 beds) on fewer and larger units.