Measuring Hospital Efficiency through Data Envelopment Analysis when Policy-makers' Preferences Matter. An Application to a Sample of Italian NHS Hospitals

In this paper we show how both the choice of specific constraints on output weights (in accordance with health care policy-makers' preferences) and the consideration of exogenous variables outside the control of hospital management (and linked to past policy-makers' decisions) can affect the measurement of hospital technical efficiency using the "Data Envelopment Analysis" (DEA). Considering these issues, the DEA method is applied to measure the efficiency of 85 (public and private) hospitals in Veneto, a Northern region of Italy. The empirical analysis allows us to verify the role of weight restrictions and of demand in measuring the efficiency of hospitals operating within a National Health Service (NHS). We find that the imposition of a lower bound on the virtual weight of acute care discharges weighted by case-mix (in order to consider policy-maker objectives) reduces average hospital efficiency. Moreover, we show that, in many cases, low efficiency scores are attributable to external factors, which are not fully controlled by the hospital management. Finally, we show that accredited private hospitals exhibit a higher level of total inefficiency than public ones: for-profit hospitals are mostly characterised by scale inefficiency, while non-profit hospitals are affected by different sources of inefficiency. Most of the hospitals in Veneto are too small in relation to their output levels (i.e. are characterised by IRS) and this problem of scale inefficiency characterises mainly the accredited private hospitals. This result indicates that private hospitals are considered important within regional health care planning as providers of supplementary services integrating public supply, even though they operate at a sub-optimal scale.