Resumé

The article assesses the prospects for building a comprehensive theoretical
framework for explaining modern democracy with the aid of the concept of corporate
pluralism. Three district patterns in the literature are detected and reviewed.
None seems entirely satisfactory, though there are grounds for optimism
as regards future progress.

1. Introduction: Departures

In the mid-1960s two political scientists - one a student of British politics (Beer 1966), the other from Norway (Rokkan 1966) - tacitly criticized prevailing theories for failing to account properly for the roles of organized interests in the political process. They and two writers on American politics (McConnell 1966; Lowi 1969) found - for different reasons - assumptions about 'interest groups' in the dominant structural-functional system models and in the theory of democratic pluralism inadequate and inappropriate for understanding the structured, regularized participation of organized interests in policy making.

Though some of these contributions were hardly more than suggestive of anomalies and none was theoretically comprehensive, they were nonetheless seminal. Directly or indirectly, they stimulated theoretical cricitism and a modest amount of empirical research in various settings.

---

* This essay was written while I was Visiting Fulbright Professor at the Institute of Political Science, University of Aarhus. The extraordinary collegiality, support, and intellectual stimulus given me by all at the Institute are acknowledged with pleasure and gratitude. Much of the work on this subject was inspired by the writings of Professor Stein Rokkan, who passed away after this manuscript was submitted. This essay is dedicated to his memory.

Such work indicated a substantial dissatisfaction with existing theories of democratic politics, interest groups, and the policy process. The most important notion that emerged from these writings by the mid-1970s - a common denominator-was that, organizationally and politically, interest organizations were more appropriately conceptualized as continuous, structured participants in policymaking than as seekers of intermittent access and influence (the dominant pluralist view), and that their more or less assured access carried profound implications for the way policies were made, for the substance of policies or outputs - and, less explicitly (except notably for Lowi), for the norms of democratic politics (see especially Kvavik 1974; Heisler, with Kvavik, 1974; Ruin 1974).

From the point of view of theory, the import of this set of studies might be summarized in terms of the suggestion that modern, especially democratic, politics is much more complex than the largely electorally focused theories of the day led us to expect. The attention drawn to the more prominent, institutionalized roles of organized sectoral (and/or territorial or segmental) interests and their intricate and thorough intertwining with the decision-making and administrative aspects of governments was the major corrective provided in this literature. The phenomena on which these observations were based were labelled 'neo-corporatist' or corporate pluralist' by many of the writers (among whom, painful though it is to admit it, I was one); and much of this work came to be identified with an emerging theme of corporate pluralism in comparative political studies.

In retrospect, that was a great mistake. For, what should have served as a corrective - a set of variables to be added to an already extensive list, in order to amend already complex conceptions of politics so that they might more accurately capture even more complex realities - came to be viewed by some as central explanatory variables. In other words, what was intended as counterpoint became, in the works of many - particularly scholars in the Nordic countries, who in the mid-1970s undertook extensive research on interest organizations in politics - the theme.

The purpose of this brief essay is to assess the prospects for building genuine, empirically testable political theory through (or with the aid) of the concept of corporate pluralism. It is emphatically not my aim either to review the very extensive and multifarious literature that has developed around the concept in its very short modern career or to provide a more satisfactory or comprehensive framework within which to cast it. The first of these tasks would require much more space than is available here; and in any case, a meticulous, comprehensive and theoretically conscious review has already been prepared (Olsen 1978). The second is beyond my

ken; and it may be premature to attempt it, considering the accelerating
pace and increasing range of both empirical and theoretical studies on the
subject.

In section 2 a brief sketch of the development of what I see as the three major patterns in the literature is provided. A caveat is in order here: any simple taxonomy of this young but already complex literature is bound to suffer from artificial distinctions, forced combination of works and ideas that differ, and excessive simplicity. But, if pretentions to comprehensiveness and profundity are avoided, such a classification can serve the useful purpose of rendering the literature manageable and tractable for a theoretically motivated assessment.

In section 3 the three patterns are considered from the point of view of their potential for a 'theory of corporate pluralism.' In the brief conclusion, I have permitted myself a few lines to revisit my own pages on the subject. Finally, while my concern is with the theoretical utilities and limitations of the concept in general, given the setting for this article and its primary readership, emphasis will be placed throughout on the Nordic context.

2. Parting of the Ways

Interestingly, the most important stimuli for empirical work on corporate pluralism in the Nordic countries - with the partial exception of some Norwegian studies discussed in section 3.3, below - were not the theoryamending suggestions of Beer, Rokkan and others whose referents were Northern European, highly industrialized democracies. Instead, much of such research was cast into theoretical terms provided by the first comprehensive, modern model of corporatist politics - a lengthy, largely derivative and substantially normative article by Phillippe C. Schmitter (1974).

It is difficult to understand the reasons for the intellectual importance of Schmitter's article for those working on and in Northern European polities (cf. Olsen 1978:34). His primary model of'state corporatism' was derived from Iberian and Latin American referents and was manifestly inappropriate for Northern European settings. That it was empirically inappropriate is shown in large part by the negative or disconfirming stance taken by Nordic researchers from the outset (e.g., Rokkan 1975:219 ff.) and, eventually, by their findings (see, e.g. Damgaard and Eliassen 1978; Buksti and Johansen 1979). This was also acknowledged at least in part by

Schmitter in subsequent assays (1977a, 1977b). Its theoretical inappropriateness derives in the main from its ideological and logical stance, and especially from its focus on the State rather than on government. This focus pervades the 'societal corporatist' variant of the original model, i.e.. the variant intended to be applicable to settings such as the Nordic. (This last point is briefly elaborated in 3.1 below.)

Perhaps the best explanation is that while the earlier work focusing on Northern Europe purposely avoided projecting a single, comprehensive corporatist model and emphasized increasing complexity, Schmitter did project such a framework in relatively parsimonious and, initially at least, seemingly broadly applicable form. Thus, even though his model was found inappropriate or even 'wrong' on the basis of Nordic case studies, it had the virtues of explicitness and theoretical comprehensiveness - qualities eschewed by those more familiar with and more closely attuned to Northern European politics. Though in section 3 I shall discuss briefly some of the theory-related implications of this development, its full exploration would require delving into individual sociologies of knowledge, the impacts of institutional milieux, and the exigencies of research funding - factors clearly outside the scope of my present concerns.

2.1. Three Patterns in the Literature

It should suffice to note here that from these beginnings three distinct patterns have emerged in recent years. One has followed at least implicitly a path reasonably close to that marked by Beer, Rokkan and a few others. It is travelled especially by students of Norwegian politics (e.g., Moren et al. 1974; Olsen 1978; Olsen et al. 1978; Lægreid ånd Olsen 1978; Hernes et al. 1978; and the two reports on Norwegian research in this issue, by Christensen arid Egeberg and by Hernes and Selvik). It is characterized by the subsumption of the concern with organized interests in politics with much broader, often even comprehensive, theoretical frameworks - usually of a generically organization theory type. While this pattern has potential for theory-building, it has, in my judgement, limitations from the point of view of political theory and (cross-national) comparative political analysis. It is treated very sketchily in section 3.3 below.

A second pattern, that which has received by far the most international attention, derives directly from Schmitter's-1974 essay. Reduced to its essentials (but hopefully not distorted or unfairly treated-and in any case, explicated and cited more fully in some of the articles preceding this one), it consists of a theoretically salient model and a subordinate formulation.

The latter is designed to accommodate the former to the universe of Western, democratic, highly industrialized polities that was manifestly inhospitable to it. The central notion in the former is 'state corporatism', wherein singular (monopolistic) peak associations - organized principally but not exclusively in terms of economic functions or sectors - are coopted or even created by the State, thereby preempting meaningful participation by ordinary citizens. (For a model stressing cooptation and preemption without State-centric assumptions, see Heisler, with Kvavik, 1974.) The subordinate formulation, dubbed 'societal corporatist', allows for multiple, even pluralistic organization; and it does not project the State into an overall dominant position vis-å-vis these 'corporate pluralist' entities.

The chief exponents of the 'state corporatism' model (Schmitter; Lehmbruch 1974, 1977,1978; Panitch 1977, 1978) have cast it in quasi-neo terms, with heavy normative loading. They view state corporatism in a decidedly negative, even hostile light, as a threat or impediment to democracy in general and to participatory democracy in particular.

In a fashion largely determined by the way they have cast the problem, those following this path have begun to build a political theory of corporatism (but, given their preoccupation with the state corporatism model, less one of corporate pluralism). In section 3.11 shall briefly assess the prospects for reaching political theory along this path, as well as its implications for other approaches.

The third pattern is the obverse of the second, or nearly so. It has led - particularly among some Danish scholars and a few elsewhere - to detailed, often meticulous 'normal science' studies (in Thomas Kuhn's sense of the term); but it lacks both a Gestalt and a clear notion of dynamics. Often even approximations of the most common tacit or explicit theoretical dependent variable in such research - policy outputs and/or outcomes (i.e., consequences, results or 'performance') - are missing.

The path chosen by those whose work characterizes this third pattern is essentially that of induction. They have accumulated impressive bodies of data (see, e.g., Buksti and Johansen 1977; Damgaard and Eliassen 1978; Elvanderetal. 1974; Grønnegård 1978; Helander 1979b; and Johansen and Kristensen 1979) that might become the raw material for the systematic, cumulative building of theory through induction. But it is work that, at least until now, has been discouragingly devoid of the overarching concepts needed to link narrowly circumscribed case studies, or dynamics, or the weighting or even rank-ordering of the relative importance or key influence of power related variables, and even of structures.

2.2. Different Paths to Theory?

Those travelling along these three paths presumably expect to develop (or enhance existing) explanations of politics through or with the aid of the notion of corporate pluralism. But formulations of that concept, the nature as well as the content, of course, of the theory that might follow, and even general approaches vary greatly across the three patterns - and sometimes even within them.

The concept of corporate pluralism imparts a modicum of coherence to this literature, but only in its broadest, most general sense. In this sense, it can be said to denote the continuous, structured participation of interest organizations in the policymaking (and sometimes the policy implementation) process. Such organizations convey principally but not exclusively particularistic interests - in the largest number of cases dealing with economic subject-matter. And, reciprocally, they are used by policymakers to enhance the legitimacy of policies and to facilitate their implementation.

There is little disagreement among students of the phenomenon, regardless of the path they have taken, on this general description. But sharp differences exist with regard to its significance, and how, why, and with what consequences processes unfold and structures are organized. As noted, theoretical and research approaches also differ markedly.

Depending on the school of corporate pluralism, it is expected to explain (or help to explain) either the nature of politics in whole systems or in one or more of their important (subsystemic?) aspects. Those following the path charted by Schmitter (especially Lehmbruch 1977, 1978; and Panitch 1977, 1978) see corporatism as superstructure and others (e.g., those cited in the first and last paragraphs of section 2.1 above) as more or less invisible infrastructure. In the works of both of these groups of writers, corporate pluralism is viewed either as the essential defining characteristic of politics in some countries or as an important element in some aspects of the political process. In any case, most writers who focus on it expect it to explain much about the style and the content of politics; and for those who see it as superstructure, it also speaks loudly to the normative nature of the polity.

Those identified earlier with the first pattern range close to the second of these two poles. As noted, most have sought consciously to amend or adapt existing theories or politics in both industrialized and industrializing settings (for the latter, see, e.g., Wiarda, 1974; and Malloy et al. 1977) by stressing the previously ignored or poorly understood roles of organized interests in politics. While this group of writers is the most heterogeneous

among the three in terms of approaches and purposes, and hence the most difficult to characterize with common referents, it can be said that most strive to understand the implications of the integration, in some form and varying degree, of what have been traditionally viewed as 'private actors' (i.e., interest organizations) into the policy making mechanisms of modern governments. While this is a concern for those following other paths as well, it seems salient for the Norwegian scholars cited in section 2.1.

Given these differences of orientation, purpose, and style among students of corporate pluralism, it should not be surprising that their approaches to theory should also differ. Some of the salient differences are examined briefly in section 3.

3. The Long Road to Theory: Is It a Dead End?

My concern in this article, as the subtitle indicates, is with the prospects for building political theory with the aid of the concept of corporate pluralism. It seems neither necessary nor feasible to define the notion of political theory in a paragraph or two; and, in any case, it would be foolish even to attempt to do so. The term is used here as it is generally understood in the social sciences: empirically testable explanations of whole or empirically or at least analytically discrete sets of political phenomena. The three major criteria for such theory, then, are empirical testability, explanatory power and scope, and political substance. (This conception of political theory stops far short of a comprehensive political philosophy; and since I perceive some writers - especially Schmitter and Panitch - to be concerned with such a philosophy in relation to their occupation with corporatism, there is bound to be dissonance between the tone of their work and of this essay.) The questions to be addressed, then, are of the following sort: 'is the concept of corporate pluralism, as used by those following each of the paths sketched above, likely to bring us closer to such theory?' 'how?' and 'with what (comparative) likelihood?'.

3.1. Corporatism as Superstructure: The Second Path

Those who began with the notion of state corporatism have tended to impute a high degree of coherence to corporatist structures; and they impart State-centric rationale and dynamics even to the societal corporatist version. Elite-mass and class divisions animate their concerns. In a manner that is harmonious with (derives from?) what I earlier termed their quasi-neo-Marxist orientation, they tend to assume a Gestalt that purposively serves the interests of elites in general and of the bourgeoisie in control of the State in particular (see especially Schmitter 1974:99ff.,

107ff., and passim; Panitch 1977:74-78 and passim). (Successful) corporatism
is thus depicted as a rigidifying, conservative force, in the service
of those who control the State (e.g., Panitch 1977).

Here, then, is a theory - or at least a set of almost organically linked, mutually reinforcing hypotheses - to be proved, rather less to be tested. Indeed, to date, research in this mode has consisted largely of the arraying of validating illustrations; and these are then analyzed in terms of the premises (see Schmitter 1977b; Panitch 1977, 1978). At this juncture, it is difficult to anticipate any empirical or theoretical findings from this approach that do' not support the premises. Explanation and political content are provided, but systematic empirical testing has not yet been - and it appears problematic.

While this observation - from the point of view of building social scientific theory it borders on an indictment - may seem harsh, the internal logic of the approach and the exemplars of work now available make it inescapable. First, public elites (and sometimes even interest organization leaders) are seen as agents of the State, purposively orchestrating structurally determined and rigidified, abiding, class interests - rather than, for instance, as hapless politicians who are the temporary incumbents of government offices and who walk tightropes of narrow coalitions above deep societal divisions, shallow sectoral demands, and intra-sectoral conflicts and problem-imposing environments. While, to be sure, no conspiratorial interpretations are made (at least for the societal corporatist form of interest here), there is an ineluctable sense of order and comprehensiveness in both the state and the societal corporatist models - as well as an imputation of clearly understandable, diachronically consistent motives to all or most classes of actors.

The contrast between the partial, but by now extensive, empirical findings of other scholars (i.e., those associated with the first and third patterns above) and the models and fragmentary illustrative data of the Schmitter-Lembruch-Panitch school could not be sharper. The former, when their work is viewed as a whole, have shown that corporate pluralism - at least in the Nordic countries, for which large bodies of data have been accumulated - is immensely complex, multifarious and poly centric; its norms are characterized by heterodoxy; the actors' motives are often unarticulated (and sometimes perhaps inarticulable), as well as ad hoc or opportunistic. Sometimes such phenomena, wholly unpredictable in terms of the 'societal corporatism' model, as governmental agencies frequently operating] as representatives of organized interests' (Olsen 1978:16) can be found. (On this last point, see also

Kvavik 1974; 1976; and Heisler, with Kvavik, 1974. On the overall
heterogeneity and complexity of corporate pluralism, see Olsen 1978).

Thus, while it would be an exaggeration to say that empirical studies of corporate pluralism in the aggregate depict a system near chaos, it is difficult to escape the conclusion that the relatively parsimonious model of societal corporatism provided by Schmitter et ah imputes a degree of order far greater than that uncovered at least by those who have intensively studied the Nordic cases in the past few years. It is, of course, possible that the sense of extreme complexity conveyed by the detailed empirical work done to date is more a function of the relative youth of such research than of the actual systems of corporate pluralism-i.e., it may be possible eventually to identify inductively the regularities postulated by those who have begun with comprehensive sketches of the superstructure. But for at least two reasons, I do not consider the prospects for such inductive validation good.

First, as I shall argue in the next sub-section and have already suggested above, the outlook for integrated theory through the inductive approach to this subject-area is not bright; and besides, the empirical work done so far has tended to disconflrm the societal corporatism model advanced by Schmitter. Second, the nature of the model - which faintly resembles a deductive-axiomatic construct, but without the rigor, elegance, and parsimony now commonly associated with such - may render it untestable in a direct manner.

The last-noted difficulty arises in part from the assumptions of the societal corporatism model - such as, for instance, its comprehensiveness and overall rationale, and in part from the related choice of its authors to focus on the State rather than the government as both the principal societal actor and the integrating arena for pluralistic interactions among organized interests, between them and their citizen-members, and among organizations, citizens, and the central authorities.

It may be possible to illustrate the problem with a generalized example distilled from the recent experiences of a number of advanced industrial societies in dealing with incomes policy problems. Labor organizations tend to press for higher real, net incomes (i.e., wages and other benefits, in relation to inflation, taxes, etc.) for their members. Employers are interested in the lowes* possible wage and tax costs and the highest possible prices. Governments seek to mediate in societies with comprehensive and effective peak associations (cf. Panitch's statement that 'corporatism, above all, means highly mediated political rule' [1978:21]) but with a view toward maximizing their own policy goals - e.g., minimal inflation, high

economic growth, a healthy balance of payments, etc.

Societal corporatist theory, albeit much less than the state corporatist version, would tend to impute monistic motives to each actor, based largely on a clear-cut functional division of labor - an assumption concomitant with its focus on the State (rather than on government) - and to fit citizen roles and interests into the matrix dictated by the State-centric theory. With such assumptions, the most important variable for policy analysis and outcome prediction would be the relative power of each actor - but particularly of labor and employers (since in true societal corporatism 'functions of public policy have been taken over by interest organizations' [Lembruch 1977:108]). But in fact, as empirical studies of corporatism in incomes policy show (e.g., Elvander 1974; Helander 1979a; etc.), the process of policymaking in this area is much more complex.

Governments intervene, initiate, set parameters and in general strive to achieve their goals; and these tend to conflict with the goals of employers and labor alike across cases and over time. The political complexion of governments (and the composition of coalitions); their willingness and ability to reach over the heads of interest organizations to publics; external economic conditions; organizations' internal structures and resources; their strategies and tactics and the effectiveness of their leaders; and the dynamics of the bargaining process itself may be more significant variables than the existence of the general corporate structure.

The key is less the complexity of the policymaking arena than the fact that it is governments rather than the State that occupy the apex of the structure. For, interest organizations, governments, and citizens do not simply play out roles assigned them in a functional division of influence axiomatic to the sorts of State constructs provided by Schmitter, Panitch, Lehmbruch and a few others. The evidence indicates that roles and utility schedules shift (and the former are sometimes interchanged); governments are far from content simply to mediate; and the outcomes of the interplay of multiplex role-interest sets are not at all predictable from assumptions derived from the status and comprehensiveness of organizations, the postulated interests of citizens, or a posited division of functions.

Empirical research in the Nordic countries has focused on the interplay of interest organizations with governments, without superimposing comprehensive theories of the State. This has made it possible to avoid imputing roles, interests, and priorities to the actors and to treat these as variables to be tested, instead. In this way, systematic empirical research can have a larger and more meaningful scope; but, as the discussion in

section 3.2 below implies, the trade-off entails a loss of overarching
theory.

But if the alternative of retaining the type of theoretical frameworks provided by Schmitter et al. is the heaviest cost exacted for empirical testability, it may be worth paying. For, as evidenced by the proliferation of sub-categories or the ruling of at least partially relevant and significant cases out of bounds (cf., for instance, Panitch 1978; Schmitter 1977a), the ultimate prospect for finding widely applicable theory along this path is not good.

3.2. Mapping Corporate Pluralist Interactions: Can We Reach Theory through Induction?

There are clues in the work, especially of those who have produced analyses of particular systems that initially identified or stressed corporate pluralist elements in the political process (e.g., Rokkan 1966, 1975; Kvavik 1974; Damgaard and Eliassen 1978; etc.), to the difficulties of theory-building based on this concept. These clues lead me to believe at this time that corporate pluralism is not an actual, distinct, discrete political theory, and is not likely to become one in the foreseeable future. Illustrations of these difficulties can be found in the soundly researched, informative, and generally very valuable empirical studies identified with the inductive pattern in the literature above.

Several interrelated lacunae are apparent; and they seem to be inherent in the nature of the phenomena - as well as in the concept - of corporate pluralism. They are essential limitations to the development of a political theory pivoting on the concept.

In summary terms, the key problem is that conceptual and operational limitations in corporate pluralism do not allow us to determine when and under what circumstances the activities or the interplay of interest organizations will be determining for either policy outputs and outcomes or in salience for citizens concerned with instrumentally affecting the political process. This problem can be analyzed into several components - some at the macro-level, others at the individual level. Limitations of space militate in favor of a simple and relatively general analysis.

First, while careful and detailed empirical studies (e.g., Jarlov, Johansen and Kristensen 1976; Buksti and Johansen 1979; Buksti 1979; Grønnegård 1978; Damgaard and Eliassen 1978; Christensen and Egeberg 1979; Elvander 1966; Fivelsdal 1978; liernes and Selvik 1979; etc.) have measured interest organization activity in particular policy sectors or with regard to certain (governmental) institutional foci or levels of government

and, in some cases, associated it statistically with policy outputs (especially in the form of such formal decisions as legislation), most of the components for even a crude regression equation designed to explain outputs are lacking. We do not know how or with what weight interest organization activity, resources, policy goals, political complexion, and so on are related to the other inputs in the policy process. Some of these other relevant factors are: public opinion, electoral and parliamentary party strength, bureaucratic structure, coalition dynamics, objective and perceived economic conditions, international environments and a system's openness to them (cf. Cameron 1978), transnational cultural and ideological movements, the personalities and interpersonal relationships of leaders, the interplay of issues (including such aspects as rapid shifts in saliency and the ever-present problem of at least partial intransitivity), and so on.

In brief, for the impressive and sound empirical research on corporate pluralism to yield effective theory, it must overcome monumental difficulties in variable specification, weighting and measurement, dynamic modelling (especially in terms of interaction phenomena and flow) and causal modelling. The pessimistic prognosis for building, through induction, political theory or political theories hinging on corporate pluralist notions is due less to problems of measurement than of specification (especially of elite-level variables in general and interpersonal relationships in particular, as well as for the individual citizen) and of precise and sensitive representations of interaction or dynamics in flow models.

These problems are not insoluble, to be sure. But, after approximately five years of such work and the accumulation and analysis of substantial data, there are still no discernible clues to deriving overarching (i.e., case or sector-linking) concepts, variable weights, comprehensive flow models and - in terms of more sweeping theoretical goals, an explanation of what difference the activities of organized interests make for outputs and outcomes of political systems - and how and why.

Second, a subset of these and related problems seems, in fact, to be beyond the reach of presently conceivable operationalization; and this augurs ill for reaching theory through induction. 1 refer here to a central phenomenon associated with agenda-setting and policymaking that seems intuitively and on the basis of strong, if indirect evidence to be a key aspect of politics and policy - especially in complex polities with marked corporate pluralist elements: logrolling, generically conceived.

The notion of logrolling as it is used here is a simple one: issues and, at a

more specific level, decisions are tied together in the policy-process both horizontally and vertically. Horizontal logrolling entails the creation of decision-sets, often comprised of substantively loosely related or even unrelated issues, so that political coalitions are created for the set rather than for the single decision. In this way, neither cost-benefit evaluation nor political division is necessary on single issues. The gain is the management of conflict among divergent interests; the loss is a chance of decreased cost-effectiveness (especially in a narrow, economic or budgetary sense) - together with some degree of loss in political accountability to constituents. (This last follows from the blurring of decision-criteria, of course.) Vertical logrolling involves the tying together of decisions across time; and actors can accumulate credits or incur debits by their behavior at each decision-point.

Logrolling is not, of course, unique to systems with noticeable corporate pluralist aspects. It would seem to be particularly crucial in such systems, however, given the structured representation of more or less formally identified interests through regular access to - and in many cases into - the decision-making subsystem (see Olsen 1978:72-73).

But empirical research on interest organization access and participation in policymaking has succeeded so far in tracing the process up to rather than into the 'black box' in which logrolling takes place. Within that black box the input variables not adequately measured (and perhaps not precisely measurable on scales comparable with the indicators used to date) assume great importance. Even more limiting in this regard is the inability of researchers even to estimate the weights of both presently measurable and unmeasurable variables as they interact in the process of logrolling.

Research on the corporate pluralist aspects of Nordic polities has progressed rapidly during the past few years. It has been more extensive and, in the aggregate, more complete on Denmark and Norway than on Sweden and Finland; but considerable data have been collected and analyzed for the latter two countries as well. Such research has shed much light on such matters as - to use illustrations from the work on Denmark, with which I am most familiar - the growth of activity by interest organizations and their incorporation in the policy process over time (Johansen and Kristensen 1979); external participation by interest organizations in the lawmaking process (Damgaard and Eliassen 1978); the internal characteristics of interest organizations and the relationships between internal qualities and resources on the one hand and organizational participation in policymaking (Buksti and Johansen 1979); as well as the 'receiving end¹ of interest organization activity, bureaucracy (Grønnegård 1978, 1979).

Specific policy areas have been treated in great detail (e.g., for Danish policy vis-å-vis the European community, see Buksti 1979). And one imaginative and theoretically sophisticated study (Damgaard 1977) attempts a broader theoretical tack in terms of sectoral activity.

Such research has provided rich and useful data on the characteristics of organizations and their behavior in the political system; on the relationships between organizations and particular aspects of government; on the structuring of government itself in a corporate pluralist milieu; as well as on the structures and processes of corporate pluralism per se. But to date it has not yielded a configurative picture - i.e., a composite picture of a corporate pluralist regime in operation. As long as it is not able to penetrate the black box of decision-making - particularly at the complex stage of the creation and resolution of decision-sets or logrolling, and as long as it does not overcome the difficulties of specification, weighting and measurement of variables, dynamics, and flow and the like noted above, this inductive path to theory is likely to prove frustrating.

These observations can be summarized in terms of the central problem facing empirical researchers on corporate pluralism. In one of the first pointers to the importance of corporate pluralist channels for participation and for exerting policy-related influence, Stein Rokkan used the phrase 'votes count but resources decide' (1966:105). More recently, he reiterated this theme in a brief and general article carrying that title (Rokkan 1975). Rokkan's notion (originally formulated in 1964) was intended to show that especially in such systems as the Nordic the then-standard parly system, electoral behavior, and traditional interest group approaches did not suffice: the roles of regularized participation in the policy process by interest organizations had to be taken into account as well, and the importance of their resources could be seen impressionistically and anecdotally to be significant factors in determining outputs.

The formulation is as valid today as it was fifteen years ago. But, while empirical research has brought us closer to knowing such elements as what kinds and magnitudes of resources are associated with organizational access patterns, we are not much closer to knowing how, under what circumstances, and why they decide (if they in fact do).

The problem is even more severe. For, as Rokkan indicated, votes - and, doubtless other forms of political activity -do count. But how much do they count, in relation to organizational resources? During the past twenty-five years we have developed many effective calculi for electoral, party, parliamentary and coalition politics, both at the individual and the macro-levels. But we have no calculus for measuring or weighing interest

organizational influence in the making of outputs and outcomes; and,
more critically, we have no common calculus for assessing the crucial
interplay of organizational activity with the voter-party channel.

In summary, then, the inductive path has led to many useful and important insights into politics - particularly in the Nordic countries, where research on corporate pluralist phenomena has been most extensive. Much of value has been learned about the activities, internal organization, and qualities and other dimensions of interest organizations in politics. Further, since the earlier theoretical interests of many of those who turned to this subject often persisted, this research has greatly enriched, in both empirical and theoretical terms, various subject areas identifiable in subsystemic, sectoral or issue-area terms.

But, though such theoretical gains may be quite real and important, this path has not brought us appreciably closer to theories of corporate pluralist regimes. With regard to such theories, research to date has helped to disconfirm the comprehensive models advanced by Schmitter and others. That is indeed a theoretically useful result, but not a constructive one. But a theory of corporate pluralism, or, more to the point, a political theory or theories of politics focused on corporate pluralism, has remained elusive. For the reasons adduced above, it is likely to remain so for the foreseeable future.

3.3. Organization Theory-based Approaches: A Brief Note

Recent work by a number of Norwegian scholars (see Olsen 1978; Olsen et al. 1978; Hernes et al. 1978; and Lægreid and Olsen 1978, for example) seems to be closer to the original points of departure of those who sought to treat the political ramifications of regularized interest organization participation in the policy process as part of a broader - if not systemic - whole than either of the other two orientations discussed above. Conceptually and in part theoretically - if not yet appreciably empirically - it makes some progress in the direction of reintegrating interest organizational activity in the policy process into a more comprehensive view of the political process and structure than that displayed to date by those following an essentially inductive path. This orientation is also markedly empirical. However, from the outset it was guided by theoretical notions - largely derived from organization theory (especially in the work of Johan P. Olsen, but cf. Hernes et ah 1978, as well). Thus, its stance conforms to a substantial degree to the initial suggestions of Rokkan and others that the structured participation of organized interests be accorded consideration together with other elements in the political system.

The comprehensive frameworks center on the somewhat similar concepts of coordination (Olsen) and steering (especially Hernes, but also Olsen and others). They are clearly non-deterministic and empirically open (to a far greater degree than the central concepts in the pattern derived from Schmitter's models); and they may make possible the eventual linkage of a large number of elements bearing on policymaking (including logrolling) and administration - with a potential that seems greater at this time than the narrower inductive studies discussed in section 3.2 have shown to date.

However, even this approach as it stands, is less than fully satisfactory from -Apolitical theory-building perspective. Perhaps the most succinct and explicit statement of it in generic terms can be found in Olsen (1978). He identifies four modes of coordination available to organizations for integrated participation: '(a) by specifying rules of procedure, (b) by specifying substantive rules, (c) by ad hoc interactions, and (d) by autonomous adjustments' (14). These options can be examined in terms of substantial data on actual organizational behavior. O'lsen explicitly chose to stress non-joint decision-making (1978:16), which, as he points out, has been one of the primary emphases in organization theory for some time.

The findings (i.e., those reported in Olsen 1978) are those which would probably have been predicted through a deductive-axiomatic view of rational organizational behavior: the general configurative picture of organizational behavior shows that high degrees of specialization and specificity of issues, decision-rules, information, and so on in a given policy-arena are most conducive to high degrees of integrated organization participation.

But while such a milieu may be the preferred setting for organizations, they must per force operate in more 'hostile' environments as well - i.e., where specificity and the other desiderata do not obtain. Even the selection of a focus for effective participation or making the decision of whether to participate or not are likely to entail much less ideal conditions. And, consequently, the approach sheds much more light on the activities of interest organizations in some settings than in others - i.e., in more rigorously structured settings, with clear decision-rules, sharply defined issues, etc.

But in the political arena interest organizations must interact with such politicized entities as parties and such diffuse ones as public opinion. Here some version of a joint decision model may in fact be more appropriate. Robert Goodin has provided an elegant game-theoretic formulation of 'coordination problems' in the formal sense (Goodin 1976:chaps. 4-6),

which, it seems to me, could usefully complement the narrower organization theory conception of coordination. Goodin's essential point is, if I read him correctly, that no conscious cooperation or high degree of integration is required for deriving mutual benefits: the vectors of other actors¹ strategies need to be estimated for such coordination to obtain; and information is then fed back to the decisionmakers. (Olsen assigns a relatively minor place to such feedback calculations - 1978:72.)

In essence, there is no theoretical limit to the range of actors, decision situations, and structural conditions with which such coordination can be attempted. In practice, all organizations use estimates or 'guesses' - if only in selecting settings and conditions in which to pursue the 'tighter' coordination Olsen depicts.

This broader notion of coordination may well permit at least the beginning of empirical research across narrowly delimited sectors or specialized environments - and therein lies the key for a more broadly integrative, more dynamic and perhaps more political framework, one that is less precise but not inconsistent with those characterizing this pattern in the literature. For, it is a common denominator for all types of political actors, at all levels of aggregation.

The point to be made is that actors participating in policymaking and other political processes create an arena in which all actors in the game influence the calculations and behavior of all other actors - initially perhaps in the form of discouraging or encouraging entry by those with marginal interest. At subsequent and more sharply delineated stages it is likely that more and more controlled and controllable conditions will obtain. While it is most reasonable to begin both theoretical and empirical inquiry with reference to conditions that most nearly approximate wellestablished propositions in organization theory - the generic basis for the frameworks under discussion, though more so for Olsen's than Hernes's, if appreciable expansion of the universe of policy situations and actors is to take place, it will behove scholars to expand their central concepts in the direction of more inclusive, if less controllable and precise, notions regarding conditions.

To what is already a painfully inadequate, superficial, and probably distorting set of remarks on the interesting and important work of Olsen, Hernes and others (since the distinct elements in the approaches would require much more space to depict than is available here), one further observation needs to be added. The enormous complexity of governmental and political activity in societies such as the Nordic can be reduced to analytic manageability probably as well - perhaps better - through the

conceptual tools of organization theory (Olsen) or institutional economics (Hernes) as through any others. However, it is possible that such approaches may underemphasize less readily mapped or measured political aspects of policymaking and other relevant processes as authority relationships, political leadership, subtle ideological shifts, and the like.

In effect, the corporate pluralist dimension of modern polities is in part a function of such rather long-term and diffuse developments as, for instance, the necessity for governments to devolve - informally and tacitly in most cases, unwittingly in many - responsibility and authority, in order to make their increasingly extended domains of responsibility manageable, and not only the result of the actions of interest organizations.

It is not my suggestion that the Norwegian scholars cited have ignored such political considerations. Rather, the analytics involved in organization theory-derived approaches may de-emphasize such 'soft' considerations. Olsen and others have concerned themselves with, for instance, the meaning and extent of the legitimacy of organizational activity in various settings. But this and similar political elements - in a fashion parallelling the 'two channels¹ identified by Rokkan - may not be tractable with one calculus any more than they are for those following the inductive path discussed earlier. The solution to this problem, if one is to be found, does not seem to be closer to this path.

4. Conclusion

A number of years ago Robert Kvavik and I suggested (Heisler, with Kvavik, 1974) that the continuous, structured participation of interest organizations in the policymaking and implementation processes of a number of Western European polities had become an important factor. At that time, we could only rely on very fragmentary secondary source materials to support our impressions. It was our aim simply to call attention to this factor, an important one, but only one among others, in the spirit of Stein Rokkan's and Samuel Beer's earlier suggestions that in fact the activities of interest organizations in some systems were of a different sort and of a different degree of importance than more traditional interest group theory recognized.

It is at once heartening to find some validation of our almost wholly abstract notions in the empirical work since done, and rather discouraging to see the relatively slight progress toward correcting or enhancing the political theories against which we (later and less cogently than several others, to be sure) tilted. (It should go without saying that for those of our guesses that have been invalidated by the data recently accumulated we

might shed a few tears, although recognizing that such a result is an
unavoidable part of the scientific process.)

The very extensive empirical studies noted earlier have been useful not only for informing us about particular facts in particular places but because they have generally made contact with some of the hypotheses and outright guesses made in the 1960s and early 19705. Hopefully, such work will continue; but, also hopefully, in the context of broader, more comprehensive frameworks - perhaps similar to those being developed by the scholars noted in section 3.3. In this way, we may be able to reintegrate political theory at increasingly more inclusive levels. At this juncture in the career of 'corporate pluralism' studies, there is ground for optimism - though the direction for such theoretical progress is not quite clear. Perhaps it is sufficient at this stage that those who have in the past concentrated on developing highly abstract formulations now have or soon will have sufficient data bases to test them and that those who have done the difficult, unglamorous empirical work are impatiently and actively searching for overarching concepts and theory.

REFERENCES

Beer, S. H. 1966. British Politics in the Collectivist Age. New York: A. A. Knopf.

Buksti, J. A. 1979. Corporate Structures in Danish EC Policy: Patterns of Organizational
Participation and Adaptation. Paper ipresented at the ECPR Workshops, Brussels (April).

Buksti, J. A. and Johansen, L. N. 1977. Danske organisationers hvem-hvad-hvor.
Copenhagen: Politikens Forlag.

Buksti, J. A. and Johansen, L. N. 1979. 'Variations in Organizational Participation in
Government'. (This issue of SPS.)

Cameron, D. R. 1978. 'The Expansion of the Public Economy: A Comparative Analysis',
American Political Science Review 72, 4, 1243-61.

Christensen, T. and Egeberg, M. 1979. 'Organized Group-Government Relations in Norway:
On the Structured Selection of Participants, Problems, Solutions and Choice Opportunities'.
(This issue of SPS.)

Damgaard, E. 1977. Folketinget under forandring. Copenhagen: Samfundsvidenskabeligt
Forlag.

Damgaard, E. and Eliassen, K. A. 1978. 'Corporate Pluralism in Danish Law-Making',
Scandinavian Political Studies 1 (New Series), 4, 285-313.

Elvander, N. 1966. lnteresseorganisationerna i dagens Sverige. Lund: CWK Gleerup Bokförlag.

Elvander, N. 1974. 'Interest Groups in Sweden', The Annals of the American Academy of
Political and Social Science 413, pp. 27-43.

Elvander, N. et al. 1974. Näringslivets 900 organisationer. Stockholm: Studieförbundet
Näringsliv och Samhälle.

Fivelsdal, E. 1978. Embedsmandsperspektiver på samspillet mellem de økonomiske interesseorganisationer
og centraladministrationen. København: Handelshøjskolen.

Goodin, R. E. 1976. The Politics of Rational Man. London and New York: John Wiley &
Sons.

Grønnegård Christensen, J. 1978. Normerne for samspillet mellem den danske centraladministrativt
perspektiv. Paper presented at the sth Nordic Political Science Congress,
Bergen (August).

Grønnegård Christensen, J. 1979. Rekrutterings- og karrieremønstre for centraladministrationens
øverste embedsmænd 1935-1977. Paper presented at the Danish Political Association
Conference, Sandbjerg (March).

Heisler, M. 0., ed. 1974. Politics in Europe. New York: David. McKay.

Heisler, M. 0., with Kvavik, R. B. 'Patterns of European Politics: The "European Polity"
Model", pp. 27-89 in Heisler, M. 0., ed., Politics in Europe.

Helander, V. 1979a. A Liberal Corporatist Sub-system in Action: The Incomes Policy
System in Finland. University of Turku, Institute of Sociology and Political Science.
Occasional Papers No. 1/1979.

Helander, V. 1979b. 'Interest Representation in the Finnish Committee System ir the
Post-War Era'. (This issue SPS.)

Hernes, G. el al. 1978. Forhandlingsøkonomi og blandingsadministrasjon, ed. Hernes, G.
Bergen: Universitetsforlaget.

Hernes, G. and Selvik, A. 1979. "Local Corporatism". (This issue of SPS.)

Jarlov, C, Johansen, L. N., and Kristensen, O. P. 1976. The Danish Committee System.
University of Aarhus, Institute of Political Science (mimeo).

Johansen, L. N. and Kristensen, O. P. 1979. 'Corporatist Traits in Denmark 1946-76', in
Lehmbruch, G. and Schmitter, P. C, eds., Consequences of Corporatist Policy Making.
Sage Modern Politics Series Volume 5 (forthcoming).

Kvavik, R. B. 1974. 'Interest Groups in a 'Cooptive' Political System: The Case of Norway'.
pp. 93-116 in Heisler, M. 0., ed., Politics in Europe.

Kvavik, R. B. 1976, Interest Groups in Norwegian Politics. Oslo: Universitetsforlaget.

Lehmbruch, G. 1974. Consociationalism, Class Conflict, and the New Corporatism. Paper
presented at the International Political Science Association Roundtable, Jerusalem.

Lehmbruch. G. 1977. 'Liberal Corporatism and Party Government', Comparative Political
Studies 10, 1, 91-126.

Lehmbruch, G. 1978. Corporatism, Labour, and Public Policy. Paper presented at the 9th
World Congress of Sociology, Uppsala (August).

Lowi, T. J. 1969. The End of Liberalism. New York: W. W. Norton.

Lægreid, P. and Olsen, J. P. 1978. Byråkrati og beslutninger. Bergen: Universitetsforlaget.

Malloy et at. 1977. Authoritarianism and Corporatism in Latin America, ed. Malloy.
Pittsburgh: University of Pittsburgh Press.

McConnell. G. 1966. Private Power and American Democracy. New York: A. A. Knopf.
Mcren, J. et al. 1974. Den kollegiale forvaltning, ed. Moren. Oslo: Universitetsforlaget.

Nedelmann, B. and Meier, K. G. 1977. 'Theories of Contemporary Corporatism' Static or
Dynamic?', Comparative Political Studies 10, 1, 39-60.

Olsen, J. P. 1978. Integrated Organizational Participation in Government. Mimeo, University
of Bergen. To be published in Nyström, P. C. and Starbuck, W. H., eds., Handbook of
Organizational Design.

Olsen, J. P. et al. 1978. Politisk Organisering, ed. Olsen J. P. Bergen: Universitetsforlaget.

Panitch, L. 1977. 'The Development of Corporatism in Liberal Democracies', Comparative
Political Studies 10, 1, 61-90.

Panitch, L. 1978. Corporatism in Canada? Paper presented at the ECPR Workshops, Grenoble
(April).

Rokkan, S. 1966. 'Norway: Numerical Democracy and Corporate Pluralism', pp. 70-117 in
Dahl, R. A., ed., Political Oppositions in Western Democracies. New Haven: Yale
University Press.

Rokkan, S. 1975. 'Votes Count, Resources Decide', pp. 216-24 in Makt og Motiv. Oslo:
Gyldendal Norsk Forlag.

Ruin, O. 1974. 'Towards a Corporate Society? Participatory Democracy and Corporativism:
The Case of Sweden', Scandinavian Political Studies 9, 171-184.

Schmitter, P. C. 1974. 'Still the Century of Corporatism?', pp. 85-131 in Pike, F. B. and
Stritch, T., eds., The New Corporatism. Notre Dame: University of Notre Dame Press.

Schmitter, P. C. 1977a. 'Modes of Interest Intermediation and Models of Social Change in
Western Europe', Comparative Political Studies 10, 1, 7-38.

Schmitter, P. C. 1977b. Interest Intermediation and Regime Governability in Contemporary Western Europe. Paper presented at the Social Science Research Council, Committee on Western Europe, Conference on Interest Politics, Madrid (May); and the Annual Meeting of the American Political Science Association, Washington, D. C. (September).

Wiarda, H. 1974. 'Corporatism and Development in the Iberic-Latin World: Persistent
Strains and New Variations', pp. 3-33 in Pike, F. B. and Stritch, T., eds., The New
Corporatism. Notre Dame: University of Notre Dame Press.