OBJECTIVES: In England, the Department of Health places high priority on reducing the variation in unit costs of National Health Service (NHS) hospitals. Efficiency targets are set for hospitals to create incentives for relatively high cost hospitals to reduce their costs and shift performance closer to that of their lower cost counterparts. We examine empirically the dispersion in unit costs to assess the extent of variation in the productivity of hospitals and trends over time. METHODS: We use econometric panel data techniques on data from 235 NHS acute hospital trusts over a six-year period, 1994/95 to 1999/00, supplemented with information from semi-structured interviews with key individuals in hospitals and purchasing bodies. RESULTS: There appears to have been no reduction in variation during this period. Relative unit costs for individual trusts also appear stable, with little movement from relatively high cost to low cost. Judging from limited quantitative evidence outside health care, the variation in costs between NHS hospitals may be comparatively low. CONCLUSIONS: Given all the other aspects of hospital performance that government is seeking to change, reduction in the dispersion of unit costs per se should not be a major policy objective. It is far more important to examine variation in quality-adjusted unit costs.
OBJECTIVES: In England, the Department of Health places high priority on reducing the variation in unit costs of National Health Service (NHS) hospitals. Efficiency targets are set for hospitals to create incentives for relatively high cost hospitals to reduce their costs and shift performance closer to that of their lower cost counterparts. We examine empirically the dispersion in unit costs to assess the extent of variation in the productivity of hospitals and trends over time. METHODS: We use econometric panel data techniques on data from 235 NHS acute hospital trusts over a six-year period, 1994/95 to 1999/00, supplemented with information from semi-structured interviews with key individuals in hospitals and purchasing bodies. RESULTS: There appears to have been no reduction in variation during this period. Relative unit costs for individual trusts also appear stable, with little movement from relatively high cost to low cost. Judging from limited quantitative evidence outside health care, the variation in costs between NHS hospitals may be comparatively low. CONCLUSIONS: Given all the other aspects of hospital performance that government is seeking to change, reduction in the dispersion of unit costs per se should not be a major policy objective. It is far more important to examine variation in quality-adjusted unit costs.