The purpose of this introductory article is to set implementation research in a political science setting. We understand it as having a distinctive and central place within the discipline. We argue that implementation research is not public administration research in disguise. Its frame of reference is not formal constitutional organization but the arrangements and procedures of the living constitution. And implementation research is not just public policy analysis by another name. Its ordering principle is not policy problems as defined and addressed by the ‘political system’ but policy problems as defined and addressed by relevant societal actors. The importance of implementation research for political science is that it offers scope for empirical constitutionalism: to ask, first, what are the organizational arrangements which generate policies and their effects, in order to ask, second, which of those various arrangements are or could be made optimal forms from a prescriptive constitutional perspective. Implementation research acknowledges the written constitution-but also the living constitution. It is a mind-set for sceptical enquiry into the structure and functions of policy processes, and to confront these manifestations of the living constitution with the prescriptions of the written.

The case was selected as being (reportedly) one of successful implementation: the interest was in the preconditions of success. To meet a forecast shortage of housing for oil-related construction workers, the Scottish Development Department convened a working party on which it, eight local authorities, a representative of the local industrial lobby and various other relevant central government departments were involved. A joint programme was organized so that the local authority contracts would be ‘pooled’. This was intended to attract a national-scale builder with his own workforce into the area and relieve the skilled labour shortage by phased development. The housing problem was turned round to a housing surplus, but this resulted from the over-optimism of the forecast need and a variety of building responses rather than the straightforward success of the joint operation. The case illustrates that implementation success depends on a range of dimensions; the implementation question dissolves into multiple aspects.

Implementation analyses have increasingly gone beyond the study of the working of well-defined bureaucratic hierarchies. The key role played by informally co-ordinated clusters or complexes of different organizations, so-called implementation structures, has instead been emphasized. This article on a major programme for energy research and development in Sweden agrees with the thrust of this latter approach. However, three substantial reformulations of the basic thesis about non-hierarchical implementation are advanced. First, it is argued that the role of implementation structures is by no means limited to the implementation process proper but may involve a crucial contribution to the definition and setting of a policy problem. Second, the article rejects the thesis that a purposive commitment to some common programme objective or rationale is a prerequisite for the successful implementation of a programme. Third, it underscores the importance of understanding linkages between public and private sectors in different bargaining arenas. But analysis cannot focus only on the uses of existing linkages and arenas. Programmes may well involve policies for affecting distant transactions. They will then have to cope with uncertainties about the future nature of such arenas as well as the conditions for shaping, or even creating, them.

The paper discusses the implementation of manpower policy as viewed from below, and is a partial critique of ‘bottom-up approaches’ to implementation studies. It reports on a Social Science Research Council financed study at the School of Advanced Urban Studies which employed both bottom-up and top-down methods, and explored the hypothesis that the nature of the employment problem tackled by policy affected implementation. Restructuring policy would be easier to implement than policy designed to ‘mop up’ the effects of restructuring, the least easy to implement being policy designed to redistribute rewards in the labour market. The research attempts to include the view from below the ‘bottom’ usually referred to by bottom-up theorists. The paper reports, from the point of view of those affected, the failed implementation of the multiple responses at central and local level to the creation of large-scale redundancies. It shows the presence of ‘garbage can’, ‘symbolic’, ‘crisis-response’, and ‘ambiguity’ elements in the implementation process. It also shows ‘mobilization of bias’, reluctance to assume responsibility for policy decisions, and the power of the market in implementation. It concludes that ‘economic’ is a quality akin to ‘magic’ which makes its manipulation in implementation extremely difficult, and suggests that powerful implementation institutions are needed if those below the ‘bottoms’ are not to suffer from restructuring in the labour market.

The implementation structure approach was developed for research in the area of human services delivery as an alternative to the traditional top-down model of implementation. A preliminary test of the general applicability of this perspective to other policy types and instruments is presented in this article, which describes the design and assumptions of an international comparison of the implementation of regulatory programmes in the area of air quality control. On the basis of field work and preliminary results, some observations are offered with regard to the contribution the implementation structure approach can make to that comparative analysis on which the cumulative knowledge of implementation processes across policy areas and in different institutional contexts depends. While it is suggested that more work must be done on conceptualizing the analytic unit ‘implementation structure’, experience in this project indicates that the level of analysis demarcated by the bottom-up perspective, and the general focus on the pattern of interactions through which regulatory outputs are produced, are useful points of departure for description and analysis of implementation in this policy area.

This paper reports results of a research project dealing with the role of private actors and semi-public organizations in inter-organizational networks of policy implementation. The first part addresses more general and longer-term issues of how and why mixed public and private implementation networks are generated. The central part of the article describes the results of a case study of regional economic development policy in Switzerland by comparing two models of implementation: an administrative one specifically attuned to the interests of the policy's target (firms), and a ‘consociational’ one emphasizing the participation of regional economic interest groups. Using this comparison we try to evaluate the impact of institutional configurations on policy output.

The article describes the design of, and presents some first results from, the West German component of a four-country comparative study of organizational arrangements for providing small firms with assistance to help them overcome their problems of development. In describing the research design, particular attention is given to the unit-of-analysis problem and to a method for inductively reconstructing the unit of analysis. An analytic ‘grid’, based on a simple model of problem-solving, is proposed which would seem well suited to the project's cross-cultural frame of reference. Preliminary results provided in the latter part of the article should serve to illustrate some of the analytic possibilities which may be expected from the project.