Pay-for-performance in primary health care: A comparative study of health policymaking in England and New Zealand
England and New Zealand introduced pay-for-performance schemes in their primary health care systems, with incentives for general practitioners to achieve improved population-based health outcomes, between 2001 and 2007. These schemes were part of health reforms to change the relationship between the state and the medical profession, giving the state increased influence over the quality and allocation of publicly funded health care. Two schemes of differing size, scope and impact were implemented. This research takes a comparative approach to exploring each policymaking process, utilising quasi-natural experimental conditions in these two Westminster governing systems to test the relevance of Kingdon’s multi-theoretic Multiple Streams Framework and other theoretical approaches to explain policy variation and change. The research documented and analysed the agenda-setting, alternative selection and implementation phases in the two policymaking processes and identified the key drivers of policymaking in each case study. A qualitative methodology, based upon documentary analysis and semi-structured interviews with 26 decision-makers, leaders and participants, was used to develop the two case studies, providing rich descriptive details and rare insights into closed policymaking approaches as seen by the participants. From this case study evidence, themes were drawn out and reviewed for consistency with Kingdon’s Multiple Streams Framework as it has been interpreted and adapted by Zahariadis. The case study evidence and themes were considered in a framework of comparative analysis where patterns of similarity and difference were established. The utility of Kingdon’s Multiple Streams Framework in interpreting the case study evidence was assessed. This analysis demonstrated that Kingdon’s Framework, as interpreted by Zahariadis, had high descriptive power for both case studies but failed to predict the patterns of non-incremental change observed or the importance of institutional factors such as ownership and governance arrangements for public services, interest group structure and historical antecedents seen in the two policymaking processes. The research finds that the use of bargaining in England and not in New Zealand is the reason for major differences in speed, scope and outcomes of the two pay-for-performance schemes. Institutional structures in the general practice sub-system are therefore the primary driver of policy change and variation. These acted as enablers of non-incremental change in the English case study, providing incentives for actors individually and collectively to design and rapidly to implement a large-scale pay-for-performance scheme. The institutional features of the general practice sub-system in New Zealand acted as a constraint to the development of a large-scale scheme although non-incremental change was achieved. Phased approaches to implementation in New Zealand were necessary and slowed the delivery of outcomes from the scheme. With respect to other drivers of policy change and variation, the role of individual actors as policy and institutional entrepreneurs was important in facilitating policy design in each country, with different types of entrepreneurs with different skills being observed at different stages of the process. These entrepreneurs were appointed and working within the bureaucracy to the direction of decision-makers in both countries. England and New Zealand shared ideas about the benefits of New Public Management approaches to public policymaking, including support for pay-for-performance approaches, and there was a shared positive socio-economic climate for increased investment in health services. The research provides evidence that Westminster governing systems are capable of purposeful and orderly non-incremental health policy change and that Kingdon’s Multiple Streams Framework, which theorises policy formation in conditions of ambiguity, needs to be enhanced to improve its relevance for such jurisdictions. Recommendations for its enhancement are made.