mriginalresear:c6 Is hospital performance related to expenditure on management? Andrew Street, Roy Carr-HiII*, John Posnett York Health Economics Consortium and *Centre for Health Economics, University of York, York, UK Objectives: Reducing the costs of management appears an easy target for those seeking to generate savings or to promote better spending in the National Health Service (NHS). However, an assessment of the appropriate amount of spending on management requires an evaluation of how much management contributes to organizational performance. Methods: Using routinely available NHS acute hospital data, an econometric analysis was undertaken to test the hypothesis that there is a relationship between the proportion of a hospital's income spent on management and the performance of the hospital measured along three dimensions: the achievement of financial targets; performance against waiting time standards defined in the Patient's Charter; and costs of service provision. Results: No general relationship was found between management costs and hospital performance. However, there was some evidence of a quadratic relationship between management spending and the amount of operating surplus generated and performance against the three-month waiting time standard for an inpatient admission specified in the Patient's Charter. These results suggest that performance returns reach an optimum when management expenditure is around 5-6% of hospital income. Conclusions: The evidence is not yet strong enough to draw a general conclusion that management costs in NHS acute hospitals are too high or that an undiscriminating reduction in management costs would have no detrimental effects on hospital performance. However, the findings should prompt managers to identify ways in which their activities are productive and how these can be measured, and what distinguishes effective from ineffective management. Journal of Health Services Research & Policy Vol. 4 No.1, 1999: 16--23 © The Royal Society of Medicine Press Ltd 1999 Introduction There has been an increase in the number of managers in the National Health Service (NHS) since the 1983 Griffiths Report! recommended the creation of general management posts. These replaced the consensus management practised since 1974 among clinicians, nurses and administrative staff. The creation of manage- ment positions was given further impetus by the 1990 NHS and Community Care Act reforms which allowed hospital and community health service providers (known as NHS Trusts) greater autonomy over manage- rial expenditure. Between 1985 and 1995 the number of staff formally classified as managers by the Department of Health rose from 300 to 23000. 2 In principle, effective management should improve the overall efficiency of the health service. However, management costs have recently been targeted by those Andrew Street MSc, Senior Research Fellow, York Health Economics Consortium, Roy Carr-Hill DPhil, Reader, Centre for Health Economics, and John Posnett DPhil, Director, York Health Economics Consortium, University of York, Heslington, York YOlO 5DD, UK. Correspondence to: AS. seeking to generate savings in the NHS: hospitals were set a real reduction of 8% in management costs for 1996/97. In the absence of any body of theory defining the appropriate level of spending, an assessment requires a consistent definition of management across hospitals and an evaluation of how much management contributes to their overall performance. The objective of this paper is to test the hypothesis that there is a relationship between the proportion of hospital income spent on management and its perfor- mance, measured along three dimensions: the achieve- ment of financial targets; performance against waiting time standards defined in the Patient's Charter;3.4 and costs of service provision. Data sources The analysis is based on routine financial and other data for 1994/95 and 1995/96 for acute hospitals in England, defined as those in which more than 30% of expendi- ture is on acute specialties. Data are derived from a number of sources, including the Trust Annual Accounts (TAC) , the Trust Financial Return by programme and 16 J Health Serv Res Policy Volume 4 Number 1January 1999