Read Do policy networks matter or what is their output legitimacy text version


Wanted: A Good Network Theory of Policy Making1 Patrick Kenis2 and Jörg Raab3 Paper prepared for the 7th National Public Management Conference, Washington D.C. October 9-10, 2003

Draft ­ Please do not quote without permission


Department of Policy and Organization Studies Tilburg University the Netherlands ([email protected]) Department of Public Policy and Management University of Konstanz Germany ([email protected]) __________________________________________________________________



The title was inspired by Salancik (1995) who in a short review essay ("Wanted: A Good Network Theory of Organization") formulated the need for a network theory of organizations, i.e. a theory that explains the characteristics of organizations by the characteristics of the networks they are embedded in.


Abstract In the policy literature, networks have now for some time been named as an important governance mechanism. From the level of global governance, European integration, and sectoral policy networks to regional arrangements, policy networks are seen as a way to integrate differentiated actor systems and to adjust to problems that cannot be tackled by existing formal institutional configurations. Numerous empirical cases of policy networks have been produced and the development and structures of policy networks have been explained, but what seems missing is theory building on the role of networks in policy making. Therefore, in the paper we will first introduce Wallace's model for theory development. Secondly, we will apply this model and evaluate the literature according to the question what has been achieved so far. We will then suggest what steps should be taken next in order to make some progress towards the development of a network theory of policy making.


1. Introduction In the policy literature, networks have now for some time been named as an important governance mechanism. From the level of global governance, European integration, and sectoral policy networks to regional arrangements, policy networks are seen as a way to integrate differentiated actor systems and to adjust to problems that cannot be tackled by existing formal institutional configurations. Numerous empirical cases of policy networks have been produced and the development and structures of policy networks have been explained, but what is missing is theory building on the role of networks in policy making. There exists no explicit theory that would lead us to predictive claims about how particular network conditions result in particular kinds of policy making. Different authors have recently regretted the absence of this type of theory development (e.g Meyer and O'Tool 2003; Peterson, 2003). We completely agree with this observation since we consider the formation of generalized explanatory principles (i.e. theories) to be the core business of our trade. The availability of theoretical propositions does not only inform us about explanatory principles (and thus equally enriches the practice of policy making) but also helps to specify the factors to measure before engaging in new research and serves as a common language into which the results might be translated after the research is done. Consequently, it seems important to contribute to theory development on policy making (and thus to improve subsequent theorizing). The question is, however, how to arrive at such a theory. The present paper wants to make a contribution to answering this question. This, not by actually trying to develop a substantive network theory (which seems some bridges too far at the moment) but by assessing the degree to which the literature actually develops towards a network theory of policy making and by addressing the question whether and how such theory development could be advanced.


In order to assess the degree to which the policy network literature develops towards a network theory of policy making we will apply a classical model of theory development to the available literature. A confrontation of this model with the existing literature indicates the degree to which components and processes of theory development exist in the present literature. At the same time such a confrontation gives clear indications about the type of scientific work we have to develop in order to arrive at a network theory of policy making. Before turning to an assessment of the policy network literature we will introduce the model by which we will assess the theoretical potential of the policy network literature.

2. Components and Process of Theory Development In his excellent overview of Sociological Theory published in 1969, Wallace presented a model called "Components and Process of Scientific Sociology". The model distinguishes 5 components of scientific enquiry: theories, hypotheses, observations, empirical generalizations and methods (see Figure 1).


Figure 1:

The Components and Process of Scientific Sociology



Source: Wallace (1969: ix)

Wallace uses the model to demonstrate the status of theory in the process of scientific sociology. According to Wallace, in order to examine the uses of theory, or of any other single part, its interrelationship with the others must be shown. The diagram represents scientific work as a succession of manipulations of information (moving clockwise in the diagram) and each of which is controlled by a particular kind of method. Whereas methods are seen here as the principal controls over the way in which scientific enquiry is pursued, theories are the most important informational product of this pursuit. Wallace points out that individual observations contain only very small amounts of information about a given phenomena, and that empirical generalizations and hypotheses are limited to moderate amounts of information but that theories (insofar as each theory is synthesized from several different generalizations, and each empirical generalization is synthesized from several different observations) can contain the maximum amount of information (Wallace


1969: x). This does not imply, of course, that all theories contain the same amount of information since a theory from two empirical generalizations will contain less information than one induced from three empirical generalizations. From such a perspective, theory is not just a storehouse of information but theory itself actively performs two crucial roles in generating the information that is stored within it: first, by specifying the factors one should be able to measure before doing research and theory serves after the research is done, as a common language into which the results may be translated for purposes of comparison and logical integration with the result of other researchers. Arguing why this model is the most valid way to assess the present policy network literature on its theoretical potential (and eventually contribute to its further development) is not straightforward. A number of arguments can, however, be presented, why we consider this model helpful. First, the model was developed in 1969 but it (or variants of it) are still widely used in theory and methodology handbooks. Secondly, the model depicts clearly what the elements are in theory development and how they are linked to each other. Consequently, it provides a clear framework for assessing the literature on policy making. This is particularly important in order to assess a research area like policy studies where theory development is characterized by combining observations form previous literature, common sense and experience. Thirdly, using the model protects us from more recent tendencies in which many things are labeled theory, although they are actually not: references, data, lists of variables, constructs, diagrams, hypotheses (see the informative paper by Sutton and Staw [1995] in which they argue why these are not theory and the many examples of papers in which these are, nevertheless, presented as theories). Fourthly, although the model successfully excludes certain things, which are not theory, it is open to incorporate more recent discussions on theory development.


Most notably, the discussion on the causal logic or causal reconstruction or mechanisms in the analysis of macro-social phenomena (or in other words the question why something occurs) (see e.g. Mayntz 2003) is already an integral part of the model. In the model it figures under "logical induction" and is indeed much less specific and developed compared to what was formulated later by others (in particular by Coleman [1990] ) but does not contradict the model. Finally, the model is not biased towards a certain type or tradition in social science research. It can accommodate voluntaristic as well as structuralistic approaches and qualitative as well as quantitative approaches. For example, the model is consistent with Eisenhardt's excellent conceptual elaborations on "Building Theories from Case Study Research" (Eisenhardt 1989). In what follows we will explain how the model will be used to assess the potential of the policy network literature in developing theoretical statements on how network attributes are related to attributes of policy making. The literature on policy networks will be contrasted with the model presented above. Consequently, we will assess to which extend the different parts of the model as well as their presumed interactions can be identified in the literature. This assessment strategy is based on the assumption that in order to arrive at a network theory of policy making a "theory storehouse" (Wallace) can only develop by recurrently going through the process as described above (see Figure 1). Consequently, the policy network literature will be assessed on the basis of the following questions: Since we are interested in the amount of information in the literature on how and why networks influence the structure, process and outcomes of policy making we will first have to identify that part of the policy network literature, which falls within the definition of "network theory of policy making". Only the literature in which network is considered the explanans and policy is


considered the explanandum, can logically contribute to the development of a network theory of policy making. As we will see, the so-called policy network literature is a miscellany of analytical quite different approaches. Consequently, the literature identified in the first step can, following the principal logic of the model presented, be assessed on how and whether information is produced which can be added to the "network theory of policy making"-storehouse. Such a contribution can be made, again according to the logic of the model, by any of the four "method"-steps presented: developing theory from empirical generalizations, developing empirical generalizations from observations, developing observations from hypothesis, or developing hypothesis from theory. Developing theory from empirical generalizations implies that answers can be found in the literature on the question whether the phenomena-that-explains (attributes of networks) and the phenomenon-to-be-explained (attributes of policy making) are inductively generalized beyond their original formulation, and consequently, the amount of scientific information can be increased. The transformation yielded by answers to this question may be expressed in a theoretical statement like "policy network outcome rates vary with the degree of centralized integration" (in which both the outcome of a policy network ­ the phenomenon-to-be-explained ­ and the phenomenon-that-explains are generalized). Another step through which the literature can contribute to theory development is by making direct observations and transforming them into empirical generalizations. For example, direct observations can be made on several instances of policy making with different degrees of outcomes. Consequently, the observations might then (through counting and computing


rates for various sub-categories of such instance of policy making2) be transformed into an empirical generalization like "policy making at the national level produces solution that are less effective compared to solutions created at the local level". This kind of finding can consequently be used as information input to empirical generalizations. Another type of research, which could equally contribute to theory development is the formulation of hypotheses on how attributes of networks influence attributes of policy making. These hypotheses could be logically deduced from existing network theoretical propositions about policy making and consequently could trigger another scientific cycle and eventually contribute to the information store of the theory. Consequently, the question arises which type of hypothesis have been formulated in the literature on policy networks, so far. In the following part of the paper we will assess the policy network literature on the basis of these questions.


But, of course, also in other ways such as through case studies (see e.g. Eisenhardt 1989).


3. A Short Assessment of the Policy Network Literature 3.1 General Remarks The role of networks in policy making became an important issue on the research agenda in the late 1980s. Social science researchers began to focus both theoretically and empirically on how networks between public, private and non-profit actors shape processes of policy-making and governance. Originally, scholars were interested in understanding and improving policy planning and implementation processes through the state. But the high hopes were profoundly disappointed. The state as the central actor lost more and more of its strong and independent position and had to face the claims of the ever stronger societal actors, which made it impossible to hierarchically implement policy decisions especially in complex policy fields. As a consequence, scholars not only started to describe these more horizontal forms of governance, which developed out of a changed distribution of power, but also tried to argue normatively why these forms of governance were the most effective and efficient for certain types of policy and organizational problems. The term was claimed to becoming the new paradigm for the `architecture of complexity' (Simon) (Kenis and Schneider 1991:26) or the major device to re-integrate differentiated systems of actors in modern societies (Mayntz 1993). On the global level, networks are seen as one important possibility to coordinate activities in an institutional age where a world government is neither realistic for the near future nor desirable for a lot of people at all and where supranational agencies, private and nonprofit organizations are increasingly gaining importance alongside the established nation states (Nölke 1994; Reinicke 1999; Reinicke 1998; Zürn 1998). The studies which deal with actor coordination, policy production and implementation in national policy fields describing and analyzing network-like arrangements whether they are called policy networks, policy communities, issue networks or sub-governments are


meanwhile countless (among others: Laumann and Knoke 1987; Pappi, König and Knoke 1995; Schneider 1988; Schneider and Werle 1991), for the English version see (Marsh 1998; Marsh and Rhodes 1992; Rhodes 1990; Rhodes 1991; Rhodes and Marsh 1992). In the field of European integration the studies are numerous (among others: Bomberg 1998; Bretherton and Sperling 1996, Falkner 2000; Héritier 1993, Nölke 2003) describing how activities and interests are coordinated between the different interest groups, nation states, parliamentary committees and the European commission in a network-like fashion. On the regional or transregional level, cooperation in network-like structure between a variety of actors is seen as crucial to deal with environmental and traffic problems (Beck 1997, Jann 1993, Cappelin and Batey 1993, Church and Reid 1996), coordinate transportation systems (Chisholm 1989), organize industrial restructuration (Best 1990) and are considered to promote innovation to improve the competitive position of regions through industrial districts (Brusco 1982; Herrigel 1996; Pyke, Becattini and Sengenberger 1990; Sabel 1994; Sabel and Herrigel 1987). In the transformation of the former Commecon economies networks are seen as vital institutional arrangements in the privatization and restructuration of the former state owned companies (Grabher and Stark 1997; Stark and Bruszt 1998; Raab 2002a, 2002b). Although the use and the understanding of the network concept varies greatly, a lowest common denominator in the understanding of a policy network can be seen in the definition of a network as a set of actors who are linked by relatively stable relationships of a non-hierarchical and interdependent nature. These actors share common interests with regard to a policy and exchange resources to pursue these shared interests acknowledging that co-operation is the best way to achieve common goals (Börzel 1998:254). Furthermore, one of the basic features, which repeatedly appears in almost all studies is the statement that networks are often the only


governance form that is able to deal with today's complex problems that are not mirrored in the established formal (hierarchic) structure of state bureaucracies and territorial boundaries. In this perspective, it is claimed that the solution should rather be to bring together state agencies, parties, interest groups, companies, civil movements, etc. to negotiate and implement viable solutions. The different perspectives, know-how and resources ought to be brought together and interests could be reconciled. Thus the downside of state bureaucracies (limited perspectives and shrinking resources, inflexibilities, limited and outdated knowledge) as well as of markets (inability to produce collective goods, creation of external effects and social inequalities) could be avoided. Meanwhile, there is some literature on network management (e.g. Agranoff 2001; Kickert 1993, 1997; Kickert, Klijn and Koppenjan 1997; Kooiman 1993; Leach and Percy-Smith 2001) and especially American colleagues try to link governance structures in public and non-profit management to performance (Agranoff 2003; Heinrich and Lynn 2001; Lynn, Heinrich and Hill 2001; Meier and O'Tool 2001,2002; Provan and Milward 1995), although only Agranoff (2003), as well as Provan and Milward (1995) focus on complete interorganizational networks and their impact on outputs/outcomes. In the literature on policy networks ample attention has been given to the conditions for the development of policy making. Some of the case studies mentioned before tried to explain why a specific policy network has developed. Others set out to explain why policy networks as a specific form of governance developed since the 1970's (e.g. Kenis and Schneider 1991, Mayntz 1993, Powell 1990, Messner 1994). If we look at publications in the second half of the 90s and in the last couple of years we can observe that scholars try to bring together at least some of the different perspectives on governance (the German and American version), inter-organizational


networks, interest intermediation and studies in political sociology using quantitative methods of network analysis beginning with the community power studies in the 1970s (see among others Börzel 1998, Klijn 1997, Raab 2002a,b). But as mentioned before, we feel that we are not really making enough progress in creating empirical generalizations especially when it comes to explaining policy with network characteristics. Therefore, in the following, we will try to evaluate, using the Wallace model, what kind of pieces we have and what is still missing. If we look at the literature the vast majority describes empirical observations. This is no surprise, because usually after a phenomenon is discovered the natural first step is

to describe it as thoroughly as possible. In the process, differences, similarities and variations of the empirical phenomenon are discovered which often lead to the construction of different categories. We will therefore start with "observations". 3.2 Observations Numerous quantitative and qualitative studies describing networks were published in the 1980s until the mid 1990s. Important publications in this regard were the "Organizational State" (1987) and subsequent work by Pappi, Laumann, Knoke and colleagues (Knoke and Pappi 1991; Knoke et al. 1996; Pappi and Knoke 1991; Pappi 1995; Pappi and Henning 1999; Pappi and König 1995; Pappi, König and Knoke 1995) in which the importance of large formal organizations for policy making in modern industrialized countries was demonstrated and analyzed what kind of structures had developed in certain policy fields like energy, health and labor. Other empirical articles that were influenced by the "Organizational state but came from an European tradition were published in the edited volume "Policy Networks" by Marin and Mayntz (1991), i.e. Schneider/Wehrle 1991, Jansen 1991, Döhler 1991 and Kenis 1991 (see also Kenis 1992).


Further empirical work was done among others by Bulkeley (2000) on the Australian climate change network, Daguerre (2000) on child care policy in England and France, Daugbjerg (1998) on nitrate policy making in Denmark and Sweden, Forrest (2000) on drought policy in post-apartheid Namibia, Grote (1998) on the economic governance structures of nine European regions, Nunan (1999) on the implementation of EU environmental policy in Britain, Schneider (1988,1989, 1992) on the chemicals control and telecommunication policy, Sciarini (1996) on the Swiss agricultural policy and the GATT negotiations, and Raab (1992) on education policy in Britain. In a rare although theoretically and methodologically not very systematic and stringent comparison Bressers, O'Tool and Richardson (1994) edited a special issue of environmental politics on networks in water policy in among others the Netherlands (Bressers et al.), Germany (Rüdig and Kraemer), the United States (Heilman et al.), the UK (Maloney and Richardson) and the EU (Richardson). Further and more recent work encompasses studies on policy networks by Pemberton (2000) and Greer (2002) in the English, Sager at al. (2001) in the Swiss, Montpetit (2002) and Carpenter, Lazer, Esterling (forthcoming) in the North American3 and Raab (2002a;b) in the German context. Although or because we meanwhile have a myriad of empirical studies using network as a concept to describe structures in policy making it is very hard to come up with some robust and consistent findings (see below) that go beyond some superficial descriptions. However, the following patterns do emerge: 1. In the first roughly two decades of research and discussion it was demonstrated

that policy making structures indeed existed in which corporate actors negotiated solutions for certain policy fields. The strong involvement of private and societal actors led to a wider perspective from a state centered to a more encompassing

In the paper, they take a fresh look at parts of the data collected by Laumann and Knoke in their study "The Organizational State" (1987).



perspective of public policy making. Many studies show that horizontal, decentral and

informal arrangements in policy making that do not resemble the "old" decision making structures of the parliamentary process, corporatism and pluralism do indeed exist. 2. Despite the very different usages of the concept "network", almost all studies see organizations or parts of them as the main actors and conceptualize them as corporate actors and more technically as nodes in the network. 3. However, in many studies, "network" is used only metaphorically, i.e. the studies do not go beyond a merely descriptive approach using organizations as nodes and describe mostly verbally in a more or less systematic way the linkages between these organizations. Moreover, "network" is often used as an empirical tool to qualitatively or quantitatively describe social and political structures, as a special form of social structure (e.g. decentral with horizontal relationships), and as a form of governance almost interchangeably. 4. There are only few studies in which "network" is used as the independent variable, i.e. in which policy outputs or outcomes are explained with certain structural features of the policy making arrangement. Moreover, there are hardly any studies, in which the outputs or outcomes are evaluated in terms of effectiveness whatever the criteria might be (Kenis and Raab 2003). 5. There are multiple observations in many different policy fields and ­ in the EU context on all four levels of government in the EU. Although a lot of studies have been conducted in the field of environmental policy, there does not seem to be an exclusive concentration on specific policy fields. In terms of the level of government, however, the clear majority of studies concentrate on the national and ­ in the EU context - the European level. 6. The overwhelming number of studies has been conducted about policy making in one or several of the Western European states or the EU. There are only very few studies which look at policy making with a network perspective in the US context. This, according to Peters (1998:32) is due to the fact that the American political system consists of a


multitude of political groups all seeking individual dominance in their particular policy areas, rather than working cooperatively within network structures. 7. However, we see some promising empirical work in public management in the US, i.e. more in the area of implementing than formulating policies. This is due to the fact that deregulation and privatization in the provision of public services has progressed much further in the US in recent years and the "Hollow State" (Milward 1993;2000) seems, at first hand, very much a US phenomenon especially if compared with most Western European countries. This relatively new development to provide public services through multi party systems of public, private and non-profit organizations literally calls for a network approach. Interestingly, Western European countries have known these system for decades, although they are based much more on longstanding cooperative relationships and do not know the constant bidding and rebidding process that is currently applied in the US. Astonishingly, compared to the multitude of studies in the area of policy formulation, there are hardly any network studies in the area of implementation and public management.

3.3 Empirical Generalizations According to the model proposed by Wallace, it is possible to arrive at a general theory through logical induction by first making empirical generalizations over a multitude of observations. Unfortunately, to our knowledge, no encompassing metaanalysis has yet been conducted, that tries to come up with more general and more abstract categories and indicators for the dependent as well as independent variables, count the observations and make statements about the effects (consequences, outputs, outcomes, results) for example in terms of the number and types of actors, the types of relationships, clique structure, density, centralization, etc.


In general, we have not gone beyond some very preliminary attempts to summarize some results by constructing descriptive categories, as van Waarden (1992) and Jordan and Schubert (1992) among others suggested which have resonated in the field only to a limited extent, however.

What we would need is a study of the type that has recently been conducted by Hill and Lynn (2003) on empirical results in studies looking at governance and its effects in a wide variety of policy areas. In such a study the following questions should be addressed (taken from Hill and Lynn 2003 and applied accordingly, it is assumed that the basic question, do networks matter has already been answered affirmatively): 1. How do networks influence outputs/outcomes (performance)? 2. How do particular aspects of networks matter? However, as it was stated before, the number of studies reporting specific effects of policy networks is rather limited so far. Therefore, it is doubtful whether a comprehensive meta-analysis with a limited number of cases will lead to meaningful results.

3.4 Theories In our view, which seems to be the general opinion in the field (among others König 1998; Peters 1998:26; Peterson 2003:8) , a consistent body of hypotheses does not exist that could be called a network theory. To be precise: since a theory is defined by its dependent variable it is likely, if there is progress in that direction at all, that different theories will be developed that might share some important traits but would be considerably different from each other. They could be called a network theory of policy, of effectiveness, of organizations, etc. depending what the main phenomenonto-be-explained is.


The decisive question in this endeavor, however, is whether it is possible to come up with more general categories both for the dependent and independent variables compared to the ones constructed for the step "empirical generalizations". The question to be asked at this point would be "of what factors are the structural features of a network a special case?" and "of what categories are the different policies a special case?" (see Wallace 1969:ix). The three categories usually used to categorize policies, i.e. distributive, redistributive and regulative, might be a good starting point in this respect). This should then allow us to formulate causal relationships of a more general kind between structural features of the policy making arrangements and policy outputs/outcomes.

3.5 Hypotheses If we start with making empirical observations and through empirical generalizations manage to formulate a theory the next step is to come up with hypotheses derived from the theory that can subsequently be tested by making empirical observations. The circle would then be closed. Hypotheses in this respect are concise about what is expected to occur, not why it is expected to occur (Sutton and Staw 1995:377). That the Wallace model is suited for theory development from empirical research is demonstrated if we look at the study by Provan and Milward (1995) on mental health networks in the US. In that study they make empirical observations (with an n=4), come up with empirical generalizations and inductively construct a theoretical model from which they derive four propositions about the relationship between network features and effectiveness. Since there is no unified theoretical body it is logical that we hardly find specific hypotheses making statements about what to expect in terms of policy outcomes given certain structural features of the network in the literature. There are, however,


bits and pieces in the literature that formulate propositions based on a range of theories, using network as a dependent or an intervening variable. In their study "The Organizational State", Laumann and Knoke (1987:272) use the prominence position of actors in the communication and the resource exchange network as intervening variables to explain the participation of actors in policy events (decision making). The positions are in turn explained by the variables "issue interest", "monitoring capacity", and "influence reputation". König and Breuninger (1998) try to explain the question why actors form policy networks, i.e. they look at policy networks as a dependent variable. Based on rational choice theory, they formulate hypotheses based on the relationship between policy preferences and tie formation as well as sector membership and tie formation. A sober evaluation of the research and findings in terms of developing a network theory of policy so far reveals that the situation seems to be rather bleak. Despite the multitude of observations we have not been able to really go beyond the merely metaphorical use of the network concept. From this perspective the strategy to start at empirical observations and arrive at a theory through logical induction from empirical generalizations seems highly doubtful. Also, most of the studies reporting empirical observations do this without any hypothesis. Therefore it is no surprise that the results seem to lack the necessary consistency in terms of time, institutional frame, number of actors, importance of the policy, etc. in order to come to empirical generalizations. Given the effort especially necessary in conducting comparative quantitative network studies it does not seem very promising to engage in studies that simply add empirical observations without any hypotheses based on a network theory of policy. Therefore, we will evaluate in the next section, how such a theory could be developed as a starting point for further empirical research.


4. Where to go from here? From the analysis above we can safely conclude that the "network theory of policymaking" storehouse is rather empty. We seem to have hardly any knowledge on the effects of network attributes on the characteristics of policy making. In order to hold the claim that networks are relevant in policy making we have to demonstrate that the presence and absence of relationships between actors in a policy-making setting makes a difference to the policy making (i.e. its structure, processes and outcomes). Only if we can demonstrate the types of effects of interactions or links between actors on policy-making we have created a genuine and exclusive basis for a network theoretical approach to policy-making. Rather than claiming again and again that relations are important, we should arrive at theoretical propositions about whether interactions make a difference with respect to policy making. Instead of using the network terminology just as a tool or a terminology4 we should use it for understanding policy making per se. How do interactions support policy-making? How does policy-making function in terms of interactions? How does adding or subtracting an interaction in a policy setting change the policy-making? How should interactions be structured ideally to increase, for example, innovation in policy making. An information rich theoretical storehouse should be able to answer these types of questions.


Peterson is particularly critical on this point: "Sometimes, debates in the public policy literature between advocates of competing models ­and especially between network `theorists' and their detractors ­ seem increasingly unproductive. They often focus on rather trivial questions of terminology, and can be embarrassingly selfabsorbed (Rhodes 1997; Richardson 2000; Marsh and Smith 2000;2001; Dowding 2001)." He adds a footnote in which he continues: ,,As much as the authors cited here may be diligently seeking to advance or critique policy network analysis, their frequent resort to self-citation and tendencies to try to rewrite their and other authors' places in the literature mean that a diligent postgraduate student could be forgiven for concluding that recent debates seem to have become so petty and personal that the approach itself is best avoided." (Peterson 2003: 15).


But what happens in the absence of such a storehouse or better, how could we fill the storehouse. Here again, the classical model of Wallace might be useful. On the basis of this model we can identify what has to be achieved in order to develop theory. Since network theory building on policy-making is in such a premature state, the most logical place to start at in the Wallace model is "observations". This might on the first sight contradict the earlier remark, which claimed that we have an abundance of observations but a deficit of theories. The issue seems, however, that we need different types of observations. Since we are primarily interested in the effects of a specific independent variable (i.e. the absence and presence of relations) and not just in the dependent variable (i.e. traits of policy making) we cannot just start collecting information about policy-making cases, which vary on a specific trait. The reason being, that we are at this point not so much interested in explaining differences in certain traits of policy making (e.g. its degree of innovation) but are interested in the effect of a specific explanans (i.e. the absence and presence of relations). Therefore, we should not select cases of policy making with different degrees of innovation (and, consequently produce empirical generalizations and eventually theories about the factors explaining these differences). In order to develop a network theory of policy-making we need to select observations in function of the operationalization of hypotheses. These hypotheses on their part should be logically deduced from some kind of theoretical statement, which proposes that the presence or absence of interactions between policy actors affects a specific characteristic of the policy making. For example, a hypothesis could be that "The higher the density of a policy field the higher the chance that it produces an innovative outcome". This hypothesis does obviously not result from a strong or sophisticated network theory of policy making but it is at least consistent with the


general theoretical claim that networks play an important role in policy making. Moreover, such a hypothesis can be operationalized and thus guide our empirical observations. On the basis of these observations the hypothesis can be falsified or rejected and thus leads to an empirical generalization like "Policy-fields with high density rates have higher innovation rates compared to policy-fields with low density rates". In order to transform this type of information to the level of "theory", the "why" question has to be answered. Why is it that the degree of density in a policy field increases its level of innovation? This question can be answered by finding an answer to the question what other distinctive characteristics policy-fields do have in common because they have a high or low level of density (and which can explain differences in the level of innovation) as well as the question of what the degree of innovation is a special case of. In other words, in which way can the phenomenon-that-explains (i.e. the density of the policy field) and the phenomenon-to-be-explained (i.e. innovation in policy-fields) be inductively generalized beyond its original formulation and thus increasing the scientific information.

5. Conclusion: Are We Expecting Too Much? The answer is yes and no. We are not expecting too much for at least two reasons. On the one hand there is a constant and increasing stream of claims from the academic and practice literature that relations and networks are important in the functioning, process and outcome of policy making. As our review of the literature has demonstrated, however, the evidence for this claim is hardly present. On the other hand, network analysis develops more and more sophisticated and widely used methods for describing and analyzing relational structures (also in the field of policy making). But in order to be productive in understanding policy making, network


analysts will also need to become more theoretical about the study of policy making. As convincingly argued by Salancik (1995), network analysts tend to use other theories (such as resource dependency theory or diffusion theory) to explain phenomena but often do not ask how their perspective addresses a theoretical problem. In line with Salancik, it seems that we are not expecting too much that if network analysis is used, its theoretical advantages should go beyond other welldeveloped theories. Consequently, in case networks are considered important in policy making and network analysis wants to be productive in describing policy making arrangements and in explaining policy outputs or outcomes we do expect two things: first, the formulation of propositions about how adding or subtracting a particular interaction in a policy network will change the policy coordination among the actors; and secondly, the formulation of propositions on how a network structure enables or disenables the interactions between two parties in a policy setting (see e.g. Kenis and Knoke 2002). But at the same time we could also say that at least at this time we are indeed expecting too much. The main reason being that developing a theory as a rich "information storehouse" in the sense of the Wallace model is an extremely complicated journey. It assumes that we study a substantive number of policy making cases, agree on the most important independent variables, use comparable operationalizations and measurements, concentrate on comparable traits of policy making, develop causal reconstructions which will ultimately have to be based on a theory of action, etc.

We have to admit that this paper has not contributed much to this journey, except, may be, by demonstrating how such a journey could look like (and, as we all know, real journeys always look different form those in travel catalogues).



Agranoff, Robert. 2001. "Big Questions in Public Network Management Research" Journal of Public Administration Research and Theory 11(July): 295-326. Agranoff, Robert I. 2003. "A New Look at the Value-Adding Functions of Intergovernmental Networks." Paper presented at the Seventh National Public Management Research Conference. Washington, D.C. Beck, Joachim 1997. Netzwerke in der transnationalen Regionalpolitik. Rahmenbedingungen, Funktionsweisen, Folgen. Baden-Baden: Nomos Verlag. Best, Micheal H. 1990. The New Competition: Institutions of Industrial Restructuring. Cambridge, UK.: Polity Pr. Bomberg, Elizabeth. 1998. "Issue networks and the environment: explaining European Union environmental policy." Pp. 167-184 in Comparing Policy Networks, edited by David Marsh. Buckingham/Philadelphia: Open University Press. Bressers, Hans, Dave Huitema, and Stefan M. M. Kuks. 1994. "Policy Networks in Dutch Water Policy." Environmental Politics, Special Issue: Networks for Water Policy: A Comparative Perspective 3:25-51. Bressers, Hans, Laurence J. O'Tool, and Jeremy Richardson (Eds.). 1994. Environmental Politics, Special Issue: Networks for Water Policy: A Comparative Perspective. Bulkeley, Harriet. 2000. "Discourse coalitions and the Australian climate change policy network." Environment and Planning C: Government and Policy 18:727-748. Börzel, Tanja A. 1998. "Organizing Babylon - On the Different Conceptions of Policy Networks." Public Administration 76:253-274. Bretherton,C., and L. Sperling. 1996. Women's Networks and the European Union: Towards an Inclusive Approach? Journal of Common Market Studies 34 (4): p.487-508 Brusco, S. 1982. "The Emilian Model: Productive Decentralization and Social Integration." Cambridge Journal of Economics 6:167-184. Cappelin, R., and P.W.J. Batey (Eds.). 1993. Regional Networks, Border Regions and European Integration. European Research in Regional Science Series Nr. 3. London:Pion. Carpenter, David, David Lazer, and Kevin Esterling. "The Strength of Strong Ties: A Model of Contact Making in Policy Networks with Evidence from U.S. Health Politics," Rationality and Society, forthcoming. Chisholm, Donald 1989. Coordination Without Hierarchy. Berkeley: University of California Press. Church, A., and P. Reid. 1996. Urban Power, International Networks and Competition: The Example of Cross-border Cooperation. In: Urban Studies,33 (8):1297-1318. Coleman, James S.1990. Foundations of Social Theory. Cambridge: HarvardUniversity Press. Coleman, William D., and Grace Skogstad. 1990. "Policy Communities and Policy networks: a structural Approach." Pp. 14-33 in Policy Communities and public policy in canada, edited by William D. Coleman and Grace Skogstad. Mississauga, Ontario: Copp Clark Pitman Ltd. Daguerre, Anne. 2000. "Policy networks in England and France: the case of child care policy 1980-1989." Journal of European Public Policy 7:244-60. Daugbjerg, Carsten. 1998. "Linking Policy Networks and Environmental Policies: Nitrate Policy Making in Denmark and Sweden 1970-1995." Public Administration 76:275-294. Döhler, Marian. 1991. "Policy Networks, Opportunity Structures and Neo-Conservative Reform Strategies in Health Policy." Pp. 235-297 in Policy Networks. Empirical Evidence and Theoretical Considerations, edited by Bernd Marin and Renate Mayntz. Boulder/Frankfurt: Westview Press/Campus Verlag.


Eisenhardt, E. 1989. "Building Theories from Case Study Research". Academy of Management Review 14(4): 532-550. Falkner, Gerda. 2000. "Policy Networks in a Multi-Level System: Convergence Towards Moderate Diversity?" West European Politics 23:94-120. Forrest, Joshua Bernard. 2000. "The Drought Policy: Bureaucracy, Decentralization, and Policy Networks in Post-Arpartheid Namibia." American Review of Public Administration 30:307-333. Greer, Alan. 2002. "Policy Networks and Policy Change in Organic Agriculture: a Comparative Analysis of the UK and Ireland." Public Administration 80:453-474. Grabher, Gernot, and David Stark. 1997. Restructuring Networks in Post-Socialism. Oxford: Oxford University Press. Grote, Jürgen R. 1998. "The Political Ecology of Regionalism. State-Society Relations in Nine European Regions." PhD Thesis. Department of Political and Social Sciences. Florence: European University Institute. Heilman, John G. et al. 1994. "Water Policy Networks in the United States." Environmental Politics, Special Issue: Networks for Water Policy: A Comparative Perspective 3:80-107. Héritier, Adrienne. 1993. "Policy Netzwerkanalyse als Untersuchungsinstrument im europäischen Kontext: Folgerungen aus einer empirischen Studie regulativer Politik." Pp. 432-447 in Policy-Analyse. Kritik und Neurorientierung (PVS Sonderheft), edited by Adrienne Héritier. Opladen: Westdeutscher Verlag. Herrigel, Gary. 1996. Industrial Constructions: The Sources of German Industrial Power. Cambridge: University Press. Hill, Carolyn J., Laurence E. Lynn, Jr. 2003. "Is Hierarchical Governance History? An Analysis of Evidence from Empirical Research." Washington DC/College Station/TX, unpublished manuscript. Jann, Werner. 1993. "Regieren im Netzwerk der Regionen - Das Beispiel Ostseeregion." in Regieren im 21. Jahrhundert - zwischen Globalisierung und Regionalisierung., edited by Carl Böhret and Göttrik Wewer. Opladen: Leske + Budrich. Jansen, Dorothea. 1991. "Policy Networks and Change: The Case of High-Tc Superconductors." in Policy Networks. Empirical Evidence and Theoretical Consideration, edited by Bernd Marin and Renate Mayntz. Frankfurt/Boulder: Campus/Westview. Jordan, Grant, and Klaus Schubert. 1992. "A preliminary ordering of policy network labels." European Journal of Political Research 21:7-27. Kenis, Patrick. 1991. "The Preconditions for Policy Networks: Some Findings from a Three-Country Study on Industrial Restructuring." Pp. 297-330 in Policy Networks. Empirical Evidence and Theoretical Considerations, edited by Bernd Marin and Renate Mayntz. Boulder/Colorado, Frankfurt: Campus Verlag/Westview Press. --. 1992. The Social Construction of an Industry. Frankfurt/Bolder: Campus Verlag/Westview Press. Kenis, Patrick, and Jörg Raab. 2003. "What Do Policy Networks Do?" Paper presented at the Conference on Democratic Network Governance. Copenhagen, Denmark. Kenis, Patrick, and Volker Schneider. 1991. "Policy Networks and Policy Analysis: Scrutinizing a New Analytical Toolbox." Pp. 25-62 in Policy Networks. Empirical Evidence and Theoretical Considerations, edited by Bernd Marin and Renate Mayntz. Boulder/Colorado, Frankfurt: Campus Verlag/Westview Press.

Kenis, Patrick and David Knoke. 2002. "How Organizational Field Networks shape Interorganizational Tie-formation Rates". Academy of Management Review 27(2): 275-293.


Kickert, W.J.M. 1993. ,,Complexity, governance and dynamics: conceptual explorations of public network management", pp. 191-204 in J. Kooiman (ed.) Modern Governance: New Government­Society Interactions. Sage. --. 1997. ,,Public Governance in the Netherlands: An Alternative to Anglo-American Managerialism", Public Administration 75: 731-52. Kickert, Walter J.M., Erik-Hans Klijn, and Joop F.M. Koppenjan (Eds.). 1997. Managing Complex Networks. Strategies for the Public Sector. London: Sage. Klijn, Erik-Hans. 1997. "Policy Networks: An Overview." in Managing Complex Networks. Strategies for the Public Sector, edited by Walter J.M. Kickert, Erik-Hans Klijn, and Joop F. M. Koppenjan. London: Sage Publications. Knoke, David, and Franz U. Pappi. 1991. "Organizational Action Sets in the U.S. and German Labor Policy Domains." American Sociological Review 56:509-523. Knoke, David, Franz U. Pappi, Broadbent J., and Y. Tsujinaka. 1996. Comparing Policy Networks: Labor Politics in the U.S., Germany, and Japan. New York: Cambridge University Press. König, Thomas. 1998. "Introduction." Journal of Theoretical Politics: Modeling Policy Networks (Special Issue) 10:387-8. König, Thomas, and Thomas Bräuninger. 1998. "The formation of policy networks. preferences, institutions and actors' choice of information and exchange relations." Journal of Theoretical Politics 10:445-471. Kooiman, Jan (Ed.). 1993. Modern Governance. London: Sage. Laumann, Edward O., and David Knoke. 1987. The Organizational State. Social Choice in National Policy Domains. Madison/Wisconsin: University of Wisconsin Press. Leach, Robert, and Janie Percy-Smith. 2001. Local governance in Britain. Basingstoke: Palgrave. Lynn, Laurence E., Carolyn J. Heinrich, and Carolyn J. Hill. 2001. Improving Governance : A New Logic for Empirical Research. Washington D.C.: Georgetown University Press. Maloney, William A., and Jeremy Richardson. 1994. "Water-Policy Making in England and Wales: Policy Communities under Pressure?" Environmental Politics, Special Issue: Networks for Water Policy: A Comparative Perspective 3:110-167. Marin, Bernd, and Renate Mayntz (Eds.). 1991. Policy Networks. Empirical Evidence and Theoretical Considerations. Boulder/Colorado, Frankfurt: Campus Verlag/Westview Press. Marsh, David (Ed.). 1998. Comparing Policy Networks. Buckingham/Philadelphia: Open University Press. Marsh, David, and Rod A.W. Rhodes. 1992. Policy Networks in British Government. Oxford: Clarendon Press. Mayntz, Renate. 1993. "Modernization and the Logic of Interorganizational Networks." Pp. 3-18 in Societal Change Between Market and Organization, edited by J. Child, M. Crozier, and R. Mayntz. Aldershot: Avebury. --. 2003. Mechanisms in the Analysis of Macro-Social Phenomena. MPIfG Working Paper 03/3, April 2003. Meier, Kenneth J., and Laurence J. O'Tool. 2001. "Managerial Strategies and Behavior in Networks: A Model with Evidence from US Public Education." Journal of Public Administration Research and Theory 11:271-93. --. 2002. "Public Management and Organizational Performance: The Impact of Managerial Quality." Journal of Policy Analysis and Management 21. Messner, Dirk. 1994. "Fallstricke und Grenzen der Netzwerksteuerung." PROKLA: Zeitschrift für kritische Sozialwissenschaft 24:563-596. Milward, Brinton H. 1993. "The Hollow State: Private Provision of Public Services." Pp. 222-237 in Public Policy for Democracy, edited by Helen Ingram and Rathgeb Steven Smith. Washington, D.C.: Brookings Institution. --. 2000. "Governing the Hollow State" Journal of Public Administration Research and Theory 10 (2): 359-379.


Montpetit, Éric. 2002. "Policy Networks, Federal Arrangements, and the Development of Environmental Regulations: A Comparison of the Canadian and American Agricultural Sectors." Governance 15:1-20. Nunan, Fiona. 1999. "Policy network transformation: the implementation of the EC Directive on packaging and packaging waste." Public Administration 77:621-638. Nölke, Andreas. 1994. "Internationale Interorganisations-Netzwerke. Das Beispiel der Geberkoordination für Afrika südlich der Sahara." Zeitschrift für Internationale Beziehungen 1. --. 2003. "Multi-Level Governance in the European Union: The Evolution of Transnational Policy Networks." in Israel and Europe, A Complex Relationship, edited by Klaus Boehnke. Wiesbaden: Deutscher Universitäts-Verlag. O'Tool, Laurence J., Kenneth J. Meier, and Sean Nicholson-Crotty. 2003. "Managing Upward, Downward, and Outward: Networks, Hierarchical Relationships and Performace." in Annual Meeting of the American Political Science Association. Philadelphia. Pappi, Franz U., and David Knoke. 1991. "Political Exchange in the German and American Labor Policy Domain." in Policy Networks, edited by Bernd Marin and Renate Mayntz. Frankfurt/Boulder: Campus Verlag/Westview Press. Pappi, Franz Urban. 1995. "Macht in Politikfeld-Netzen: Die Beziehungen zwischen arbeits- und sozialpolitischen Akteuren der Bundesrepublik im internationalen Vergleich." Pp. 101-136 in Die deutsche Gesellschaft in vergleichender Perspektive: Festschrift für Erwin K. Scheuch zum 65. Geburtstag, edited by KarlHeinz Reuband. Opladen: Westdeutscher Verlag. Pappi, Franz Urban, and Christian H.C.A. Henning. 1999. "The organization of influence on the EC's common agricultural policy: A network approach." European Journal of Political Research 36:257-281. Pappi, Franz Urban, and Thomas König. 1995. "Informationsaustausch in politischen Netzwerken." in Netzwerke und Politikproduktion. Konzepte, Methoden, Perspektiven, edited by Dorothea Jansen and Klaus Schubert. Marburg: Schüren. Pappi, Franz Urban, Thomas König, and David Knoke. 1995. Entscheidungsprozesse in der Arbeits- und Sozialpolitik: Der Zugang der Interessengruppen zum Regierungssystem über Politikfeldnetze; ein deutsch-amerikanischer Vergleich. Frankfurt/New York: Campus Verlag. Pemberton, Hugh. 2000. "Policy Networks and Policy Learning: UK Economic Policy in the 1960s and 1970s." Public Administration 78:771-792. Peters, Guy. 1998. "Policy networks: myth, metaphor and reality." Pp. 21-32 in Comparing Policy Networks, edited by David Marsh. Buckingham/Philadelphia: Open University Press. Peterson, John. 2003. "Policy Networks." in Reihe Politikwissenschaft des Instituts für Höhere Studien. Wien. Powell, Walter W. 1990. "Neither Market Nor Hierarchy: Network Forms of Organization." Research in Organizational Behavior 12:295-336. Provan, Keith G., and H. Brinton Milward. 1995. "A Preliminary Theory of Interorganizational Network Effectiveness: A Comparative Study of Four Community Mental Health Systems." Administrative Science Quarterly 40:1-33. Pyke, F., G. Becattini, and W. Sengenberger. 1990. Industrial Districts and Inter-Firm CoOperation in Italy. Geneva: International Institute for Labour Studies. Raab, C.D. 1992. "Taking Networks seriously: Education Policy in Britain." European Journal of Political Research 21:69-91. Raab, Jörg. 2002a. Steuerung von Privatisierung. Eine Analyse der Steuerungsstrukturen der Privatisierung der ostdeutschen Werft- und Stahlindustrie 1990 - 1994. Wiesbaden: Westdeutscher Verlag. --. 2002b. "Where Do Policy Networks Come From?" Journal of Public Administration Research and Theory 12:581-622.


Reinicke, Wolfgang. 1999. "Globale Wirtschaft, territoriale Politik? Überlegungen zu einem "Regieren ohne Regierungen"." FAZ 44:9. Reinicke, Wolfgang H. 1998. Global Public Policy. Governing without Government? Washington D.C.: Brooking Institution Press. Rhodes, Rod A.W. 1990. "Policy Networks: A British Perspective." Journal of Theoretical Politics 2:293-317. --. 1991. "Policy networks and sub-central government." Pp. 203-214 in Markets, Hierarchies and Networks. The Coordination of Social Life, edited by Grahame et al Thompson. London/Newbury Park, Calif.: Sage Publications. Rhodes, Rod A.W., and David Marsh. 1992. "New directions in the study of policy networks." European Journal of Political Research 21:181-205. Richardson, Jeremy. 1994. "EU Water Policy: Uncertain Agendas, Shifting Networks and Complex Coalitions." Environmental Politics, Special Issue: Networks for Water Policy: A Comparative Perspective 3:139-167. Rüdig, Wolfgang, and Andreas Kraemer. 1994. "Networks of Cooperation: Water Policy in Germany." Environmental Politics, Special Issue: Networks for Water Policy: A Comparative Perspective 3:52-79. Sabel, Charles. 1994. "Flexible Specialisation and the Re-emergence of Regional Economies." in Post-Fordism. A Reader, edited by Ash Amin. Oxford: Blackwell. Sabel, Charles, and Gary et al. Herrigel. 1987. "How to Keep Mature Industries Innovative." Technology Review 90:26-35. Salancik, Gerald R. 1995. "Wanted: A Good Network Theory of Organization." Administrative Science Quarterly 40:345-349. Schneider, Volker. 1988. Politiknetzwerke in der Chemikalienkontrolle - Eine Analyse transnationaler Politikentwicklung. Berlin/New York: de Gruyter. --. 1989. Technikentwicklung zwischen Politik und Markt. Der Fall Bildschirmtext. Frankfurt/M.: Campus. --. 1992. "The Structure of Policy Networks: A Comparison of the Chemicals Control and Telecommunications Policy Domains in Germany." European Journal of Political Research 21:109-129. Schneider, Volker, and Raymund Werle. 1991. "Policy Networks in the German Telecommunications Domain." Pp. 97-137 in Policy Networks. Empirical Evidence and Theoretical Considerations, edited by Bernd Marin and Renate Mayntz. Boulder/Frankfurt: Westview Press/Campus Verlag. Sciarini, Pascal. 1996. "Elaboration of the Swiss Agricultural Policy for the Gatt Negotiations: A Network Analysis." Scheizer Zeitschrift für Soziologie 22:85-115. Stark, David, and Laszlo Bruszt. 1998. Postsocialist Pathways. Transforming Politics and Property in East Central Europe. Cambridge: Cambridge University Press. Sutton, R.I. and Staw, B.M. 1995. "What a Theory is Not", Administrative Science Quarterly 40 (3): 371-384. Waarden, van Frans. 1992. "Dimensions and types of policy network labels." European Journal of Political Research 21:29-52. Wallace, Walter L. (Ed.). 1969. Sociological Theory. An Introduction. Chicago: Aldine Publishing Company Zürn, Michael. 1998. Regieren jenseits des Nationalstaates. Globalisierung und Denationalisierung als Chance. Frankfurt a. Main: Suhrkamp Verlag.




Do policy networks matter or what is their output legitimacy

29 pages

Report File (DMCA)

Our content is added by our users. We aim to remove reported files within 1 working day. Please use this link to notify us:

Report this file as copyright or inappropriate


You might also be interested in

Friedman Flat Earth Apr 17 2006.doc
Microsoft Word - Hopkins, Monaghan, Hansman - Who Has Access_MWR2P 06 Article_.doc
Do policy networks matter or what is their output legitimacy
Microsoft Word - Belwal and Teshome pdf