The starting point for this study is a paper presented by Sandiford and Rossmiller at the Agricultural Economic Society's meeting at Newcastle-upon-Tyne, England in March 1996 - 'Many a Slip: Studying Policy Delivery Systems'. The paper is appended as Annex I. In their paper Sandiford and Rossmiller discuss the change in the theoretical approach to the analysis of public policy in the last two decades as a result of a shift towards public choice theory which stresses the self-interest of groups who forge alliances to further their own ends. This has become an alternative to the traditional normative welfare approach. Sandiford and Rossmiller's examples are concerned with the fact that advice and analysis usually concentrate on the policies themselves and their apparent results. There is in the authors' view a growing tendency for the "distancing" of policy making and advice from policy implementation with the increasing tendency to employ non-governmental agencies in "the policy delivery system" (PDS). The system of policy implementation is in the authors' view a neglected area and their concern is to contribute "towards a framework for structured thinking about policy delivery systems" to improve the relevance and practicability of advice.
Sandiford and Rossmiller suggest the adaptation of the so-called "structure-conduct-performance paradigm" (S-C-P) from the industrial organisation literature. Starting with a description of the policy-making process itself, there should then be an analysis of the stated policy objectives and the manner of their agreement as "an exogenous reference point". There needs to be clarity about what it is intended to achieve and the degree of commitment of those who wish to achieve it. Analysis should then move to the policy delivery system (PDS) and the functions that agents in the structure of the system are intended to perform in implementing the policy. This should be followed by analysis of the actual conduct of policy by agents in the PDS to pinpoint any differences between intended and actual behaviour. Finally, there is the assessment of the performance of the policy in meeting its original objectives and delivering benefits to the "targeted recipients", which, it is argued, should be done by reference to four criteria: effectiveness, efficiency, equity and enforceability - the four Es. The authors stress the complexity of the task of assessing performance and some of the difficulties of interpreting their criteria. There is overlap in these criteria, and in the real world bureaucrats and politicians (Adam Smith's "crafty and insidious animals", Smith [1976. IV.2.39]) are forced to trade off one against the other. It will be argued in this study of European dairy policy that particularly has this been the case as the policy has developed over the years.
Sandiford and Rossmiller's paper deals mainly with the theory of the new general approach to policy analysis, but it also attempts to apply the method briefly to three examples. These relate to 1) agricultural statistics collection in Korea, 2) agricultural input supply in Korea and 3) export crops in Sierra Leone. It has been suggested that adaptation of the 'new institutional approach' to policy making and its extension to policy delivery provides a structured framework for studying the European dairy policy and its delivery across countries. An attempt will accordingly be made in this paper to apply the framework first in a general overview of the development of the dairy policy and its delivery across what has now become 15 member states. Policy agreed at the European level after reconciliation of conflicting interests may be implemented differently with differing results in member states. To study such matters in more detail this paper will consider the policy delivery system in three countries as sub-case studies. The three countries chosen are Italy, the United Kingdom and the Netherlands. The interest in these countries lies in the fact that the Netherlands is one of the original Six with the least exceptions from Community norms, whilst the other two countries have market and processing industry structures that are different from the rest of the European Union members and where the policy itself has had to be sometimes cumbersomely adapted to cope with these exceptions.
Milk policy in the EU and Sandiford and Rossmiller's paper both have an intrinsic interest which might be reason enough for looking at one through the other. There is, however, more than methodological interest in bringing the two together. It will be argued at the end of this study that during the first half of the 21st century, with the likelihood of further GATT rounds concluding along the lines of the 1992 agreement, dairy policy is likely to become "globalised" rather than merely turned into a laissez-faire system. Not only therefore will the numerous new regional international organisations, free trade areas and preferential groupings that have come into existence in recent years face problems similar to the EU but so will the new World Trade Organisation (WTO). Sandiford and Rossmiller's intention was to provide an adaptable methodological framework for development assistance projects, and such like. Application of their approach in the complex area of milk policy and co-ordination of national policies may have wider benefits.