The editors began this work after some chance observations on theapparent similarities between service areas in the problems faced inassessing ‘what works?’, and the seeming differences between areason how such vital questions are tackled. Some years later we nowknow a lot more about both these similarities and differences. Inthe presentation and exploration of these we hope to enable the manyand diverse parts of the public sector to achieve greater insightinto their own areas through a reflection on the strategies used byothers.

Along the way we have been helped by many generous people. InSeptember 1998 we held a small cross-discipline seminar at theUniversity of St Andrews. We thank all those participants who gaveso freely of their time and ideas, enriching our understanding andproviding added momentum to this project. This St Andrews seminarwas financed in part by the Russell Trust in Fife, whose financialsupport we gratefully acknowledge. We also thank our manycontributors to this text, for sharing their knowledge, and fortheir forbearance under the many demands we placed on them for moreinformation or clearer explications. Early versions of some of theseassembled papers appeared in a special edition of the journalPublic Money and Management (1999,vol 19, issue 1). Hence we are very grateful to the editor of thisjournal, Francis Terry, for encouraging us to build on this earlyeffort.

Over the period September 1998 to August 1999, one of us (Huw Davies)was a Harkness Fellow in Health Care Policy based at the Institutefor Health Policy Studies at the University of California, SanFrancisco. Huw is very grateful for the financial support of theCommonwealth Fund, New York City, which allowed him the freedom todevelop some of the ideas contained in this volume.

Finally, we acknowledge our deep debt to the many people with whom wehave engaged in scholarly or frivolous debate about evidence-basedeverything over recent years. The evidence-based agenda is enormous,and every conversation reveals new pearls to be squirrel led away,which then re-emerge as part of larger arguments later on.