Efficiency Measurement in the Public Sector: An Appraisal z g Q MICHAEL BARROW and ADAM WAGSTAFF* zyx QP 1. zyxwvut INTRODUCTION One ofthe consequences of the present Government’s commitment to improving the efficiency of the public sector has been a continuing search fo better ways of monitoring efficiency. This process has resulted in, infer alia, various government agencies developing performance indicator (PI) packages. The Department of Health, for example, has a package of PIS for the National Health Service (NHS) that it uses to ‘identify aspects of services which warrant investigation’ (Department of Health and Social Security (1983)). The PIS developed to date have not, however, been free fro criticism. Indeed, government departments such as the Department of Health acknowledge that, focusing as they do on input usage and throughput the PIS developed to date shed relatively little light on the ques efficiency (see, for example, DHSS (1983)). Dissatisfaction with the current performance indicators zyx ML - and, indeed, with their precursors - has led several researchers to try to improve on the using statistical and other quantitative techniques. Initially this work t to be based on the standard multiple regression model. Though this undoubtedly represented a step forward, it was still far from ideal. Fortunately the recent interest amongst policy-makers in performance indicators has coincided with a renewed interest on the part of academ the problems of efficiency measurement. The resulthas been the development of a large number of new methods for measuring efficiency. All of these have in common the concept of thefronfier: efficient organisa are those operating on the cost or production frontier, whilst ine organisations operate either below the frontier (in the case of the production frontier) or above the frontier (in the case of the cost frontier). Thoug empirical applications of these frontier techniques tended to be confined to the private sector, several of the methods have now been used to analyse the Both authors are at the School of Social Sciences, University of Sussex; Adam Wagstaff is also at the Centre for Health Economics, University of zyxwvu York.