Efficiency Effects from a Move Towards Activity-Based Payments to Hospitals
Posted: 22 Jun 2007
Abstract: In 2003 the Portuguese Government transformed the legal nature a set of 34 hospitals from the pure state-owned, NHS, status to new statutes, bringing them closer to 'trust hospitals'. The main changes were more flexible statutes regarding labour relations and acquisitions of consumables, on the management freedom side, and a focus on activity-based payments (negotiated with the hospitals). The activity-based payments were to be more explicit than the previous budgeting system (which had already some elements of activity-based payments). Under the new statutes, the hospitals start with a capital endowment. Financial results will add or subtract from that endowment.
These changes were promoted in order to increase efficiency of hospital operations. We provide here an assessment of this change in terms of hospital efficiency. Two previous analyses have been carried out, finding mixed results. One, done by a commission appointed by the Ministry of Health, has found cost savings in the range of 8%, with no adverse effects on quality care. A particular group of hospitals was defined as the control group. It has used a difference-in-difference approach. Another one, using DEA estimates for efficiency, finds essentially no difference in efficiency across groups. We contribute to this discussion on two groups. First, we allow for efficiency effects to differ among hospitals, in particular, to differ according to size. Second, we use also a matching estimator, aimed at a better definition of the 'control' for each hospital. We find that using conventional cost frontier estimates no statistically significant effect seems to be associated with the change of statutes. However, a different story is told by the matching estimates, where a significant difference emerges, favouring the move to more activity-based payments and more flexible management rules. This also reveals the crucial relevance of defining the proper 'control' group to assess the change.
Keywords: hospitals, matching estimators, DRG
JEL Classification: I18
Suggested Citation: Suggested Citation