Resumé

The purpose of the article is to review four major social science research projects currently under way in the Nordic countries, and to discuss the preconditions for, and the significance of, the manner in which the projects are organised. Four models for organising social science research are outlined, and applied to the projects. The main conclusion of the study, albeit of a tentative nature, is that whether a research project is organised according to one or other model is dependent upon the infrastructure of social science research, and the organisation of the wider society in the country concerned.

Introduction

It has been almost exactly 30 years since C. Wright Mills's The Sociological Imagination (1959) was first published. The book is a penetrating critique of certain aspects of the development of US social science research. According Mills, social scientists had become incorporated into the dominant power structure, and adapted themselves to carrying out the particular type of research demanded by those in power. Briefly put, this had resulted in the bureaucratisation, technocratisation, and passivisation of social science research. Mills observed the emergence of a new type of research organisation on large-scale projects with a hierarchy of administrators and a staff of researchers comprising a form of bureaucratisation. A large project was the basis for technocratisating intellectual work, becoming simply empirical research carried out with the purpose of providing the basis for decisions aimed at solving 'social problems'. Research was reduced to mere social engineering and its theoretical content eroded. As a result, social science research had abandoned its purpose and mission of critically and independently studying the relationship between the individual and society. In the American Social Science Nightmare, research lost its vision, and with that also its proper function, Mills concluded.

Some aspects of the developments that Mills attributed to US social
science are clearly applicable to Swedish social research, and perhaps also

to some extent to social research in the other Nordic countries. This is particularly so with respect to the relationship between elite groups in society and social scientists where the problems of the former tend to become the research topics of the latter (cf. Eliassen & Pedersen 1984, 85 f.; Fridjonsdottir 1987). The epitomy of the type of research C. Wright Mills criticised, the large-scale project, was notably absent for a long time on the Nordic academic scene. This, however, has gradually been changing since the beginning of the 19605, with large social science research projects becoming increasingly common in Scandinavia.

The purpose of this article is to review four such major projects, one each from Denmark, Finland, Norway and Sweden, and to discuss the preconditions for, and the significance of, the manner in which the projects are organised. By 'project' hereafter will be meant 'the coordination of large-scale research resources with the purpose of analysing a phenomenon in an organisational form'.

Within the framework for this general definition of a project a distinction will be made between four models for the organisation of research, namely the Centre, Contract, Network, and Department models. These are introduced order to provide foci for the discussion of the projects under review. The classification is made in relation to the projects' administrative dependence on the established research community, that is to say universities (sector) research institutes, as well as the manner in which research is carried out. The primary characteristic of the Centre model is that it is administratively independent in relation to the established research community, and that the research is carried out in an integrated research milieu. The Contract model is also independent of the established research community, but lacking the resources for developing its own research milieu it contracts research from established research institutions. The Network model is, like the Centre and Contract models, administratively independent the established research community. However, contrary to the latter, the Network model does not command the resources necessary for initiating and carrying out larger research studies. It functions to coordinate existing research by providing a meeting place for researchers; acts as a channel for the diffusion of research ideas, and as a catalyst for the interaction between research that is carried out in different settings. Finally, the Department model is on the whole a part of the traditional academic research community both administratively and in terms of research.

The Centre model can be seen as the final phase in the organisation of research projects which begins with the Department model and develops through the Network and the Contract models. It will not be argued here that individual research-projects pass through these organisational models in a definite order, even if there are examples of this. It will, however, be

argued that the organising of large-scale research projects is influenced by the infrastructure of social science research, and the organisation of society in its broadest sense in such a manner as to make the above described developmental process discernible at the level of social science research as a whole. Described in these terms, the manner in which the Danish, Finnish, Norwegian, and Swedish projects are organised is quite instructive, and they will be reviewed in terms of increasing complexity.

Organising Social Science Research Projects: The Nordic Models

The four Nordic research projects are presented in summary form in Table 1. The projects have several characteristics in common. These are large projects, at least when judged from a Nordic social science point of view, and they are financed externally - that is to say outside the ordinary funding of established research institutions. However, the size of the budget varies considerably from one project to another. Thus, for example, the Finnish project 'Limits of Law' has an annual budget of approximately $100,000, while the budget for the Norwegian project 'Organization and Management' amounts to four million dollars per year. The annual budget for the Danish project 'Public Organization and Management' and for the Swedish project 'Power and Democracy in Sweden' average around half a million dollars and one million dollars respectively (see Table 1). All four projects were launched in the latter part of the 1980s, and three of them, the Danish, Finnish, and Swedish projects, are intended to be concluded around 1990. They focus on what can be called the regulative principles, or mechanisms for government and control, of society or a particular segment of society; that is to say power in its broadest sense. The similarities between the projects mean that, in some cases, the studies carried out in one project have an almost identical equivalent in some of the other projects. In a couple of cases the same researchers appear in two projects carrying out the same kind of studies, underlining the relatively small-scale and integrated nature of Nordic academia.

With the exception of the Finnish project it is possible to argue that the projects are organised in an innovative manner: that is, the Danish, Norwegian, and Swedish projects represent models for the organisation of social science research which have not previously been adopted in these countries. The choice of organisational form can be seen as an attempt to make use of and build further on the resources for research which have been invested in universities and sector research institutes. However, the Nordic countries do differ in respect to the latter. Hence, it is reasonable to except the kind of research projects under discussion, i.e. 'large' projects to be organised in different v/ays among the countries concerned.

The projects will be reviewed below. The presentations do not claim to be comprehensive or exhaustive. Rather, they focus on the purpose, background, and administration of the projects, and on the nature of the research carried out.¹

Finland: The Department Model

Purpose of the Project. The Finnish project 'Limits of Law' (LoL) focuses upon the role of law in social development. The LoL project's overall aim can be very simply expressed as being the study of how the 'rationality' and 'legitimacy' of the legal system influences, and is influenced by, reformist ambitions underlying the modern welfare state.

Background of the Project. The LoL project is to some extent the result of an explicit political interest in the role of the law and the state of the legislature in Finnish society, questions which not least for constitutional reasons have a central place in Finnish societal debate. In its budget for 1986 the Finnish Government proposed the institution of a research professor with emphasis upon these questions. It may have been coincidental the proposal was closely compatible with the theoretical debate within legal studies. Irrespective of this, the appointed professor chose to relate his research very closely to this highly international debate. The LoL project can therefore be seen as a Nordic version of an international within legal studies and legal sociology on the development law and its character in advanced industrial society.

Organisation. The LoL project fits fairly well with the Department model. The project, which is financed by the Academy of Finland, is clearly discipline-based. It comprises two researchers and an assistant at the Department of Public Law at Helsinki University. The idea is that the project researchers will take an active part in the Department's research and teaching, mainly at the doctoral level. No fewer than nine doctoral students have been connected with the project. The project has developed certain characteristics of the Network model with elements of the Contract model. The Network dimension is expressed particularly through the project's attempt to take up and publish research results from other universitiesand departments. As a result, around 30 researchers are either directly or indirectly contributing to the project. This is at least one reason why in a relatively short period of time the project has resulted in the publication of four books within the area of 'law and morality' and 'legal theories' (all in Finnish). The Network aims of the project are also apparent from intensive symposium activity with participation by researchers from several countries, mainly Scandinavia and Germany. The project also has resources, albeit limited, for pieces of consultancy research.

Research. The Finnish project has as a starting point a theoretical tradition represented by names such as Max Weber and Jiirgen Habermas. The project has an explicit theory-testing and theory-refining aim. Within the general aims of the project, research has been organised under three main themes, namely 'law and morality', 'law and societal planning', and 'law and the use of power at different societal levels'. These themes have in their turn been divided up into nine subprojects. Together they cover a broad empirical field. They include studies of, for example, the state budget and planning instruments, the corporatist nature of norm-defining, the development of law and relations between public and private spheres, law and micro-power, and law and social relations.

According to the original programme the LoL project is intended to be concluded in 1989. However, the Academy of Finland is at present considering an application from the project for a further five years of funding. It remains to be seen whether the project will be developed in the direction of the organisation models for research similar to the other Nordic projects.

Denmark: The Network Model

Purpose of the project. The project is called 'Center for offentlig organisation styring' (Centre for Public Organisation and Management), abbreviated as COS in Danish. The main purpose of the project is to initiate and support studies principally of public organisations, their structure and external management functions, and relations to the wider society.

Background of the Project. The project's origins appear to have been stimulated in particular by two circumstances. In the middle of the 1980s two studies of the public sector were initiated at the Copenhagen Business School. The studies had a great deal in common, and so it appeared suitable to co-ordinate them in certain respects. At about the same time discussions were under way within the Danish Social Science Research Council (Statens Samfundsvidenskabelige Forskningsråd) concerning the setting up of a Centre for the study of public economy and organisation. The discussions were very much in line with the ideas behind the studies developed at the Copenhagen Business School. The coming together of the two studies at the Business School and the discussions within the Research Council resulted in the establishment of a special centre for research on public organisation. It may be noted that these ideas took form and were implemented virtually parallel to similar discussions in Norway which resulted in a comparable centre being set up there, though on a significantly larger scale (see below).

Organisation. COS is a Research Centre established by the Danish Social

Science Research Council for an initial period of five years. The Centre is attached to the Copenhagen Business School, and run by a Board staffed by researchers. Available to the Board is an advisory committee - a Centre Council - comprising researchers and practitioners with special knowledge within the Centre's field of research. Hence, the project is administratively independent from traditional research institutions. However, in terms of research the Centre depends very much on the latter. The larger studies connected to the Centre are in general based at different universities, in particular Copenhagen University, Copenhagen Business School, and the University Centre at Roskilde. The COS project does not possess the resources necessary to intiate and implement larger studies and subprojects. Of the approximately 30 researchers who are or have been active in the Centre's research, only a few, mainly doctoral students, have been exclusively funded from the Centre's budget. Hence, it seems fair to state that the COS project resembles the Network model, rather than the Contract or the Centre model. The impression of a network is strengthened by its considerable seminar activity with participation by both researchers and practitioners. The project has a clear ambition to bridge the gap between the research community and the world of practical politics and administration. However, the image of COS as a network should not be exaggerated. The project is still undergoing development and its network activities must be understood as a part of a process of consolidating the COS Centre by means of tying researchers and financiers to the project.

Research. The original nucleus of the COS project's research comprises the studies 'Public Organisation and Management' (Offentlig organisation of ledelse) and 'Public Management Behaviour and Information Utilization' (Offentlige institutioners og lederes styrningsadfærd og informationsanvendelse).Both these are case studies of public organisations displayingvarying of professionalism, political control, and relationship to the market. The main difference between them is that the former deals with basic pre-conditions for the provision of public services, while the latter is an analysis of organisational behaviour as a function of different managment technologies. Over and above these two studies the COS project currently has three larger studies under way. Two of these focus on renewal processes in the public sector. The first is the study 'Modernizing the Public Sector' (Modernisering af den offentlige sektor) which takes as its starting-point the goals for public sector development formulated by the Danish right-of-centre coalition government when it came into power in 1982 (see Bentzon 1988). The second study, 'Behaviour in Public OrganisationsUndergoing (Adfærden i offentlige organisationerunder studies the reaction of employees to different types of organisational change. The decision to study 'the renewal' of the

public sector is significant. As already indicated the COS project strives to produce 'policy relevant' research, including issues relating to current tendencies in practical politics - an amibition the project shares with its Norwegian counterpart (see below). This does not, however, mean that these other studies are non-theoretical. This is most clearly evident in the study 'Relationship Between State and Market' (Stat-marked) which is run in co-operation with researchers at Roskilde University Centre. The study aims to analyse theories concerning the relationship between the state and the market.

The manner in which the COS project is constructed is such that it is hardly possible to talk about a common theoretical framework. The project is characterised by 'theoretical pluralism', and is held together by the participant's common interest in a particular object - the public sector, its organisation, management, and leadership.

Sweden: The Contract Model

Purpose of the Project. The Swedish project 'Makt och demokrati i Sverige', i.e. Power and Democracy in Sweden (PDS), has as its function the description and analysis of the structure and division of power within Swedish society, and the sketching of the nature of power and democracy in tomorrow's Sweden. Special attention is paid to power and influence in and over the business sector, interest organisations, the public sector, and public opinion.

Background of the Project. Three of the four projects in this presentation can be said to have been initiated from 'above', that is to say they have been entirely or partly brought about on the initiative of politicians. The Swedish project is an example of such a politically initiated piece of research. The government has laid down guidelines for what the project will study, and has appointed project leaders. However, the idea of the project did not originate from politicians, despite the fact that during the early 1980s several proposals were made in the Swedish Parliament for the establishment of an investigation into power in Sweden. These proposals were inspired by the Norwegian 'Study of Power' (see below). It is hardly an exaggeration to claim that the Norwegian project paved the way for the Swedish study, and that without the former the latter would probably not have been set up.

Organisation. In certain respects the Swedish project displays the features of the Contract model. The project, which is financed by a governmental committee fund, has a well-developed management function staffed by four researchers. These researchers, one of whom was formerly attached to the Norwegian 'Study of Power', are responsible for determining the

organisation and general direction of the research. The daily administration is handled by a small staff. The project also has an advisory group comprised of social scientists from Europe and the USA, with a degree of overrepresentationfor from prestigious US East Coast universities. The project has resources to initiate and support research within its chosen field of investigation. However, it lacks the resources necessary to create its own research milieu, so the project largely depends upon the universities and sector research institutes for its research. At present there are one or more departments in all the Swedish universities involved in the PDS project, and a couple of studies are being carried out in co-operation with researchers at Norwegian universities. Researchers among the project's leadership group are actively taking part in some of these subprojects. The total number of researchers engaged in the project is not easily determined, but is probably around around 40.

The manner in which the PDS project has been organised is quite clearly influenced by the experiences of its predecessor, the Norwegian 'Study of Power'. From this point of view, the Swedish project not only makes use of the resources for research invested in Swedish academia, but also draws upon the experience and skills of the Norwegian social science research community.

Research. The PDS project comprises at present some 17 larger studies or subprojects (see Petersson 1988). Some of these studies would hardly have come about without the financial and other support of the project. This applies for example to the study 'Social Change from a Local Community Perspective' which is being carried out at the Department of Sociology at Umeå University. However, not all the subprojects have come about on the initiative of the PDS project, nor are they always financed by the project. The study 'Organised Interests and the Public Good' conducted at the Department of Government at Uppsala University exemplifies this, being initiated by the Department and mainly research council financed. In these cases the PDS project has chosen to tie in existing research instead of initiating and financing special studies. Furthermore, it is possible to distinguish a number of subprojects which lie in between these two forms in terms of initiating and financing research. An example of this is 'Power over Companies' which the Ministry of Industry set up and at least provide starting finance for, as well as the study 'Power and Organisation' tied to the Stockholm Business School and which is part of a larger study of organisation and leadership. The subprojects are premissed on the assumptionthat society is characterised by a definite number of 'techniques of social co-ordination', and that these appear in certain combinations in different social areas (see the purpose of the project, and Petersson 1987a, b; 1988). Finally, the guidelines laid down by the government urge the

project to take advantage of certain kind of research, for example, studies financed by the 'Committee for Research on the Public Sector' (Delegationenför om den offentliga sektorn) connected to the Ministry of the Interior. The latter type of studies are not, however, formally a part of the PDS project.

The PDS project should not be seen as merely a sum of these subprojects. The project, or more specifically its leading representative, is not only interested in bringing together the result of the subprojects in order to supply a more or less comprehensive description of power and democracy in contemporary Sweden, but also to take issue with and critically redevelop conclusions put forward by the subprojects.

Against this background it can be asked whether the PDS project really conforms to the Contract model. Perhaps the project is better described as an 'umbrella project' which covers a large number of studies, some of which are only loosely connected with the investigation. In this perspective the project is primarily a meeting ground for researchers from a number of academic disciplines for the exchange of ideas and thoughts on power and democracy in Swedish society. This image of the project is in line with expectations which have been voiced in relation to the investigation, namely that the project should contribute to the improvement and modernisation of the self-understanding of Swedish society (see Petersson 1988). This, of course, invites the question of what kind of society it is that calls on a public investigation to create an updated version of social self-understanding?

Norway: The Centre Model

Purpose of the Project. Interaction between 'economy' and 'politics' is a central theme in the Norwegian project 'Ledelse, organisasjon og styring' (LOS), i.e. Organisation and Management. The project's overall aim is to contribute towards increasing our understanding of, and solving the problems different leadership, organisation, and management systems primarily the public sector. The overall goal has been specified under three main headings. The aim of these is that the project should conduct cross-disciplinary research of high quality, perform a network function, and spread research results to interested groups. Of these goals the first, to carry out high-quality research, has been prioritised.

Background of the Project. The idea of instituting a research programme on the leadership and organisation of the public sector, and on publicprivateinteraction, form in the middle of the 1980s in connection with the Norwegian Government's work on a research policy proposal. Preparing the proposal revealed a significant interest within both the private and public sectors for a special research programme in this area. The Governmentadopted idea, and the proposal won broad support in Parliament.

This party-political unity appears to have been promoted by the positive experiences obtained by an earlier project, namely the 'Study of Power'. The latter study, composed of researchers drawn from a number of social science disciplines, was set up by the Government in 1972 with the task of studying power and democracy in Norwegian society, and was completed in 1982. The way in which the investigation was carried out appears to have strengthened the position of social science research in leading political circles. This without doubt eased the establishment of the LOS programme.

Organisation. The LOS programme is administered by a national committee under the Norwegian Council for Applied Social Science Research (NORAS). The Council makes decisions on the aims of the programme and allocates economic resources. Of the money available to the programme totalling about four million dollars a year, approximately half is channelled into university research on the subject of organisation and management. The remainder goes to a special research institute, the LOS Centre set up in the city of Bergen, which in many respects comprises the core of the LOS programme.

The LOS programme is essentially organised in a manner resembling the Centre model. It has well-developed leadership and staff functions, and employs researchers who work in an integrated research milieu. Within the framework drawn up by the national LOS Committee, the Centre's Council is responsible for its operations. The Council's composition, with representatives the higher echelons of the civil service and the universities, mirrors to some extent the project's 'practical' problem-orientation. The Centre has ten permanent research posts each of which runs from three to five years. Recruitment to these posts takes place in two stages. First the Centre formulates overall research themes (see below). Then contact is made with researchers reputed to be prominent in the chosen fields, who may, if interested in the proposed themes, present specific research proposals. Besides these posts there are a number of additional externally funded researchers working at the Centre. These two categories together, totalling approximately 15 researchers, comprise about a third of the total number of researchers active in the LOS programme. Over and above these posts there are a number of assistantships which are intended to be granted to doctoral students to be recruited mainly from the University of Bergen and the Norwegian School of Business in Bergen. The aim is that the Centre should have a close working relationship with these institutions, especially with the disciplines of economics, business, and political science.

Research. This is currently divided into three blocks - 'Public Administration'(Forvaltningspolitikk),
and Management'
(Desentralising og ledelse), and 'Incentives and Management Problems in

the Welfare State' (Incitamenter og styringsproblemer i velferdsstaten). Within these areas there are a number of studies or subprojects currently under way. The large number of subprojects and their collective breadth make it impossible to present a detailed description here. Thus, research on Public Administration covers at present ten subprojects focused, for example, upon 'pre-conditions of institutional management', 'change and renewal in local government', 'voluntary organisations and the boundary between public and private', 'oil and polities', and so on. The core area of Decentralisation and Management comprises seven subprojects of clearly varying character. These include studies of 'economic organisational theory', 'change and renewal in the business sector', 'gender and leadership',and and control in service production'. The third main area, Incentives and Management Problems in the Welfare State, is, at least in terms of numbers of subprojects, the smallest area. The area is covered by three studies which handle the efficiency of the public sector, company-taxation reforms, and the salary system of the public sector. None of the subprojects have yet presented final reports. However, a larger number of research reports and working papers has been published in the LOS Centre's special publication series.

The organisation of the LOS project has some characteristics in common with Swedish sector research institutes. Like these the LOS project is multidisciplinary and has a rather practical problem-solving orientation. Furthermore, groups with a particular interest in the research of the LOS Centre are represented in senior administration. The main difference between the LOS project and the Swedish type of sectoral research institute is primarily that the LOS project is organised with the intention of flexibility both in terms of personnel and research foci. This is especially evident in the fact that the project contracts researchers for shorter periods of time. Furthermore, researchers not paid by the project but working within the area of the progamme can be invited to join the Centre, thereby both contributing to and drawing upon the intellectual milieu of the Centre. The empirical focus of the Centre has been defined in a significantly broader way than is usually the case in sector research institutes. From this point of view the LOS Centre is better described as a Centre for advanced studies than a sector research institute.

Summing Up: Reflections on Social Science Research and Society

The above descriptions have focused upon the organisation of the four
Nordic social science research projects. The differences between the projectsin
respect are such that they can be said to represent four

types of organisational models for research, namely the Centre model (the Norwegian project), the Contract model (the Swedish project), the Network model (the Danish project), and the Department model (the Finnish project). Naturlly, this labelling of the projects is open to discussion. Generally, the projects display features drawn from all four models of organisation, and there is of course a certain extent of overlap between them. However, on the whole, the classification does capture the most prominent features of each project.

The question is what explains these differences between the projects? It is, of course, impossible to discuss this problem exhaustively in the present paper. All that can be done here is to point to what can be seen as important preconditions for and restrictions on the choice of the way in which projects are organised, and in particular the nature of the infrastructure of social science research. By 'infrastructure' is meant those resources commanded by social science research at universities and in sector research institutes in the form of researchers, research post, libraries, data archives, and so on (ef. Allardt et al. 1988, 15).

The significance of infrastructure for the direction of social science research at the universities has been discussed in varying contexts (cf. ibidem; Eliassen & Pedersen 1984; Kemeny 1988). However, it is clear that the significance of infrastructure is apparent also in the choice of the way in which 'large' research projects are organised. In the introduction it was maintained that the choice of project organisation can be seen as an attempt to make use of, and build further upon, the resources for research invested in universities and sector research institutes. In other words it can be maintained that the infrastructure of social science research sets limits to which project models can be applied. This means that the number of applicable project models increases with an increased level of infrastructural development.

This thesis is well illustrated by the Norwegian and Danish projects. It is hardly a coincidence that the Norwegian project is organised in terms of the Centre model, while the Danish project, with the same general purpose, is organised in terms of the Network model. An important precondition for the availability of the Centre model is that there exists a well-developed social science infrastructure upon which the project can build. Compared with other Nordic countries, Norwegian social science research has a very broad and complex infrastructure.² Danish social science research is still in a stage of basic growth and expansion. This implies that Danish social scientists are hardly able to choose freely among the four models for organising research. This is a particularly clear example of the way in which the infrastructure of social science research impinges upon the organisation of research projects. It is not unreasonable to argue that Finnish and Swedish social scientists face similar problems to those of the Danes. In

short, the manner in which the Nordic research projects are organised
reflects to a large extent the infrastructure of social science research in the
respective countries.

All the projects besides the Finnish are organised as multidisciplinary. Social development in the Nordic countries has, in common with development other advanced industrial countries, resulted in increasing interconnection various social spheres and different types of organisation, management, and control systems. The nature of the research object therefore motivates the organisation of projects along multidisciplinary lines. However, the combination of disciplines that have been chosen in the various projects is not at all obvious. Most noticeable in this sense is the prominent position of the economic disciplines in the Scandinavian projects. To some extent this reflects how the infrastructure of social science has changed in terms of disciplines in recent decades. Business economics exemplifies this. The subject has expanded greatly since it was established at universities at the end of the 19505, and clearly has a strong position in the three projects. Another, and possibly more important, reason for the position of the economic disciplines in this context concerns social organisation in its broadest sense, particularly the size of the public sector and social and economic development during the 1970s and the early 1980s. Budget deficits, and demands for increased government involvement from various interest groups, together with growing resistance to tax increases among the general public, have resulted in considerable political interest in public-sector organisational and management problems analysed from an economic perspective. Surpluses and deficits in the public budget give rise to varying types of problems, thereby affecting the utility of different social science disciplines as perceived by politicans and administrators. As a consequence, some social science disciplines gain ground in the study of societal organisation and the organisation and management of the public sector, while others lose ground due to the nature of problems that according political decision-makers are brought to the fore by developments in the economy (cf. Heidenheimer 1985). It is also significant that Scandinavian science research is harnessed to undertake a number of largescale on the organisation of society and the public sector at a time when the 'Great Society Programme' of the Scandinavian countries appears to be past its zenith.

The Nordic projects exemplify, each in their own way, the continuing process of integration which, it will be argued here, is a prominent characteristicof societies. What this means for social science research is that it becomes progressively integrated into issues of practical social engineering. This development has been observed and hailed by Harold Wilensky (1983) as an expression of the superiority of Scandinavian corporatist'as Anglo-American 'pluralist' democracies in creating a

'dialectic' between political power and intellectual knowledge. Disregarding the lyrical overtones of Wilensky's analysis, the corporatist nature of Nordic societies may provide a partial explanation of why research is understood by Scandinavian decision-makers to be 'a stage in rational policy making', to quote the Swedish government's most recent proposal on research policy (see Prop. 1986/87:80, p. 21). Among the projects under review, the Danish and Norwegian projects are probably the clearest examples of this. The projects have varying kinds of connections to practical policy-making and administration, and have expressed aims of producing practical applied, or policy-relevant, research. In the Norwegian project the aims extend as far as contributing towards the solution of problems connected with different management and organisational systems. Even the Swedish project gives the impression of filling a similar, if less practical, function. In this respect the Finnish project represents a more traditional academic view of the role of research in society since it takes as its starting-point an internal disciplinarydebate a predefined theoretical framework. However, in terms of sheer scale, the Finnish project is in many ways more comparable to a subproject in the other Nordic projects. The conclusion to be drawn from such a comparison may well be that there is no great difference between the Finnish and other projects in terms of the extent to which theory lies at the centre of focus. It cannot be ruled out, however, that the way in which the project is organised, its disciplinary basis, and relationships to financiers, clients, and interest-group representatives influence the theoreticalbasis research and the level of its theoretical ambitiousness (cf. Wagner & Wollmann 1986). There can be little doubt that the disciplinary basis of research is of major significance in respect to one particular aspect of research: it can hardly be a coincidence that the Finnish project, with its foundation in legal studies, has opened a dialogue with primarily German social scientists. Other projects, dominated by economists, political scientists,and economists, appear significantly more influenced by Anglo-American social science research.

On the basis of four 'observations' it is obviously not possible to formulate any definite conclusions about the character of the relationship between the organisation of a research project and its theoretical ambition. The observations indicated above only really apply themselves to positing very preliminary hypotheses. Further, these observations do not in all respects support C. Wright Mills's 'American Nightmare' in its Nordic context. It is clear that social science research projects of the larger type have become more common in Scandinavia. They also tend to be organised in a way that Mills was critical of, that is to say with a 'hierarchy' of administrators and a 'staff of researchers. However, it is not at all apparent that this inevitably leads to atheoretical and one-sidedly empirical research consonant with the interests that elite groups in society represent. How each separate piece of

research is carried out is decided in the final analysis by those researchers who are most directly concerned with it. The central problem therefore becomes how the organisation of a research project affects the autonomy of research in relation to financiers, clients, and interest-group representatives.In respect the four organisational models from the Departmentmodel the Centre model represent increasing degrees of separation from universities, both administratively and in terms of research. This can mean that the protection in terms of integrity which these institutions have traditionally offered to research and the individual researcher become weakened. The result can be that clients, financiers, and/or interest-group representatives increase their influence over research (see further Løland & Mathisen 1987; Olsen 1987). Irrespective of this, one thing is clear; the manner in which a research project is organised will inevitably come to influence what the project focuses upon and how it is analysed. The choice of project organisation is very likely to be of much greater significance for which empirical results and theoretical conclusions the project produces than the nature of the programme formulation, the directives, and goals by which the project is formally determined. In addition, the decision to organise a research project according to one or other model not only reflects the immediately involved researchers' views upon what is suitable or desirable, but also the infrastructure of social science research, and the organisation of society in its broadest sense.

---

NOTES

1. The description is based upon information material as well as a number of published project reports. The material has been complemented with short interviews with persons involved with the projects. Readers wishing further information on the projects are referred to the addresses below. Denmark: Center for offentlig organisation og styring Blågårdsgade 16A, DK-2200 København N Finland: 'Limits of Law' Institutionen for offentlig rått Regeringsgatan 11, SF-00100 Helsingfors Norway: Norsk senter for forskning i ledelse, organisasjon og styring Rosenbergsgate 39, N-5015 Bergen Sweden: Maktutredningen Kungsångsgatan 17-19, S-753 22 Uppsala

2. Which does not, of course, mean that the Norwegians themselves are quite satisfied with the current state of the art; see, for example, Nytt Norsk Tidsskrift, vol. 4 (2-4) (1987).

ACKNOWLEDGEMENTS

The author wishes to acknowledge the advice of J. Kemeny and L. Nord (both at the National
Swedish Institute for Building Research) in preparing this article. The author, of course, takes
full responsibility for the remaining flaws in the article.

REFERENCES

Allardt, E., Lysgaard, S. & Bøttger Sørensen, A. 1988. Sociologin i Sverige (Sociology in
Sweden). Uppsala: Swedish Science Press.

Benzon, K.-H., ed., 1988. Fra vækst til omstilling (From Growth to Re-organisation).
København: Nyt fra Samfundsvidenskaberne.

Eliassen, K. A. & Pedersen, M. N. 1984. Svensk samhällsorganisation och förvaltning
(Swedish Social Organisation and Administration). Uppsala: Forskningsrådens
Förlagstjänst).

Fridjonsdottir, K., ed., 1987. Om svensk sociologi (On Swedish Sociology). Stockholm:
Carlssons.

Heidenheimer, A. 1985. 'Comparative Public Policy at the Crossroads', Journal of Public
Policy 5, 441-465.

Kemeny, J. 1988. 'Defining Housing Reality', Housing Studies 3, 205-218.

Løland, T. & Mathisen, W. C. 1987. 'Det avhengige universitet' (The Dependent University),
Nytt Norsk Tidsskrift 4 (1), 60-76.

Mills, C. Wright 1959. The Sociological Imagination. London: Oxford University Press.

Petersson, O. 1988. 'The Study of Power and Democracy in Sweden', Scandinavian Political
Studies 11, 145-158.

Petersson, O. 1987a. Metaforernas märkt (Power of Metaphors). Stockholm: Carlssons.

Wagner, P. & Wollmann, H. 1986. 'Fluctuations in the Development of Evaluation Research',
International Social Science Journal 108, 205-218.

Wilensky, H. 1983. 'Political Legitimacy and Consensus', Chapter 3 in Spiro, S. E. &
Yuchtman-Yaar, E., eds., Evaluating the Welfare State. New York: Academic Press.