Traditional economic theory operates with the ideal of optimal allocation of resources - formalized as the criterion of pareto-optimality. It can be demonstrated that, as long as a number of institutional requirements are met, perfect competition will ensure optimal allocation of resources.

Economic welfare theory designates as 'market failures' situations in which state intervention in market allocation of resources is required to obtain optimality. In 'public choice' theory, state allocation of resources is shown to result, occasionally, in 'government failure', thus justifying market allocation in cases where 'government failure' exceeds 'market failure'.

Each of these theories incorporates a normative access to allocation,
based on the ideal of optimality. Further, they all regard autonomous state
decisionmaking as the (sole) alternative to market allocation.

Resumé

In the negotiated economy an essential part of the allocation of resources is by definition conducted through institutionalized negotiations between independent decisionmaking centres in state, organizations and corporations. Institutionalized creation of consensus via campaign institutions and forced compromise via negotiating procedures and decisions are central characteristics of the negotiated economy. The article includes evidence to the fact that the present Danish economy is a negotiated economy. During a long historical process, a multi-centred and pluralist political structure has been formed and simultaneously the discursive and institutional basis for co-ordination of decisions made in autonomous organizations been created. It is shown how today's wages policy, labour market policy, public expenditures policy and industrial policy are characterized by negotiation-based economic processes. It is argued that this historical development has far-reaching theoretical consequences. The institutional conditions presupposed in traditional economic theory for the ideal of optimal allocation of resources is enshrined in the constitutional interpretation of the distinction between the authority of the sovereign state and civil society. As a result of the evolvement of the negotiated economy these institutional conditions no longer exist. This development has made a myth of the ideal of optimal allocation of resources and challenges the interpretation of rationality in traditional economic theory.

---

* Translated by Sean Martin.

An analogous situation prevails in modern theory of the firm (e.g. Williamson 1975, 1985), which regards hierarchy as the (sole) alternative to market allocation. Williamson explicitly rejects other decisionmaking systems, such as democracy, as ineffective. He has developed a theory of institutional choice, either of market, (contract) or hierarchy, as the optimal institutional form, depending on the type of transaction being conducted.

A common feature of the theories referred to above is that they adhere exclusively either to autonomous decisions on the market or to autonomous decisions via hierarchies in public or private organizations. Joint decisionmaking via negotiation is not dealt with as a decisionmaking system.

Our aim is to describe the latter type of decisionmaking system, by
discussing and developing the concept of the 'negotiated economy'.

We shall try to show that, in a country like Denmark, the institutional conditions presupposed both in traditional and modern economic theory do not exist in reality. As a corollary, we shall try to prove that the ideal of optimal allocation of resources is thus equally invalid.

The aim is not to launch yet another attack on the institutional premises on which traditional economic theory is based, but rather to contribute towards the further development of a theory of negotiation-based economic interaction.

We are thus pursuing an earlier line of discussion (cf. Hernes 1978;
Berrefjord 1978, 1984) - and establishing a link with the classical institutional
in economic theory (cf. Nielsen 1986).

We shall try to reach a definition of the negotiated economy, which will not only fit into the framework of classical institutional economy, but also - by applying an interdisciplinary approach - try to achieve integration with theories of economy, sociology, and of political and legal science.

In doing so, we shall make the assertion that the institutional conditions presupposed in traditional economic theory for the ideal of optimal allocation resources is enshrined in the constitutional interpretation of the distinction between the authority of the sovereign state and civil society.

We shall then describe the historic creation of a number of new institutions, essential characteristic is that they have altered the relationship the public and private sectors, thereby also altering the legal interpretation of the boundaries between the two.

In cases where decisions are made following negotiation between mutuallyindependent there is no clearly denned boundary between the authority of the state and civil society. The state contributes to the decisionmaking process, without any authoritative jurisdiction; the scope for autonomous action by the agents is restricted by the demand for compromise imposed by the process of negotiation, and is simultaneously subject to changes arising from the creation of consensus, which is an

inherent feature of the environment of negotiation. Consequently, the
assumption of a set of given preferences of agents engaged in isolated
autonomous optimization is not tenable.

We shall end with an assertion that it is precisely this institutional development that makes a myth of the ideal of optimal allocation oi: resources and challenges other tenets, among them the interpretation of rationality in traditional economic theory.

The Negotiated Economy

The concept of the negotiated economy, based on compromise and consensus 'negotiation-based economy', or 'economy via persuasion', or 'round table economy'), is not widespread and has certainly not gained any international currency. It is not recognized, either as a descriptive term or theoretical concept, in economic theory or in political science.

Apart from a few exceptions (e.g. Moore & Booth 1988), its use is probably restricted to Scandinavian references (e.g. Hernes 1978; Berrefjord 1978, 1984; Sæbø 1980; Lindensjø 1986; Pedersen 1986; Pierre 1988; and partly Johansen 1978; and Skogh 1982). In their work, however, it is used as a metaphor rather than as an analytical term; the intention being to depict a specific type of relationship between collective entities (the state, organizations - employers, unions - and corporations) rather than to develop a systematic, clearly defined theory, backed by adequate empirical evidence.

There is no evidence of consistent application of the term (cf. Berrefjord
1984 and Pedersen 1986), nor any agreement on the use of the term to
depict the same phenomena (cf. Hernes 1978 and Berrefjord 1978).

We are thus dealing with a concept that is applied in an attempt to demarcate a phenomenon for which it can barely provide either a coherent set of premises or a historical explanation. This may explain why the concept we are dealing with is so rarely used by scientists, and why scientists may disagree on its interpretation.

Nevertheless, we have decided to apply the concept because it incorporates important components of the phenomenon we are attempting to analyze: (a) allocation of resources, and (b) the role of institutional negotiations in this allocation.

An added reason is the fact that the last few decades have produced gradually more comprehensive and more convincing evidence, at least in Northern Europe, of the historical development of allocation mechanisms and decisionmaking systems, which the dominant paradigms in economic theory, and in political and legal sciences, fail to detect and encapsulate, let alone describe and analyze.

Evidence of the existence of these mechanisms and systems has been

presented in many theoretical disciplines which are mutually isolated but in essence closely interrelated (industrial relations, economic planning, political economy, implementation theory etc.). In many cases this has led to formation of theoretical schools or streams (pluralism, neocorporatism, network theory, segmentation theory, institutional economics, new institutionalismetc.).

However, no attempt has been made, as far as we know, to interrelate or fuse these various contributions so as to produce a coherent theory of a dominant and decisive pattern of development in several countries. This development is the specific phenomenon to which we intend to apply the concept of negotiated economy!

We take the term 'negotiated economy' to mean an organizational structuring of society, where an essential part of the allocation of resources is conducted through institutionalized negotiations between independent decisionmaking centres in the state, organizations (employer/employee) and corporations.

We have decided not to add a specific definition of 'an essential part', as
it is not our aim to present a total survey of the allocation system, but
rather to focus on some crucial features of the pattern.

We shall accordingly put forward the assertion that Danish society is increasingly assuming the character of a negotiated economy as already defined. In Denmark it is not only an essential part, but also a growing part of the allocation of resources which is determined neither on the market, by individual agents adapting to market changes, nor through autonomous decisionmaking by public (governmental, regional, local) authorities.

Instead, the decisionmaking process is conducted via institutionalized
negotiations between the relevant interested agents, who reach binding
decisions based on discursive, political or moral imperatives.

• As distinct from decisions made by public authorities, negotiation-based economic decisions are reached on the basis of interaction between 'independent' agents, where the relevant public authority is usually one of several participants.

• As distinct from authoritative decisions, negotiation-based economic decisions are usually not subject to the possible use of legal sanctions. They are basedon discursive, political or morally binding - rather than legally sanctioned — agreements.

• As distinct from market decisions, which are reached by individual agents, each acting separately and on the basis of given preferences and resources, negotiation-based economic decisions are reached via institutionalized negotiations, in which the deliberate shaping of preferences a decisive factor.

At this stage we must clarify what is meant by the term 'institutionalized
negotiations', and by 'discursive, political or morally binding agreements'.

Institutionalized Negotiations

The classical example of allocation of resources within the framework of a negotiation-based economy is labour market policy, particularly wage policy. In this century wages, working hours and other conditions of work have been determined neither by the single agents nor via legislation, but through institutionalized negotiations between organizations (employers' associations/unions), corporations and the state.

The interaction at the negotiating table is often perceived as the core of these negotiation-based economic decisionmaking processes. It has thus become normal in various fields of study to examine the participating agents' formal rights and obligations, and to study their actions as rational choices based on optimization of utility in relation to given preferences.

In our opinion, this approach has led to a problematic narrowing of the
perspective. Johansen's article on 'The Bargaining Society' (Johansen 1978)
can be cited as an example.

Negotiations are perceived (in this article) as bargaining processes without (designed) to ensure joint decisions before specified deadlines They are perceived as an interaction between parties whose preferences are assumed to exist from the outset and not as the result of the interaction. The parties involved are thus assumed to have a given set of preferences and resources, to take their places at the negotiating table, and to negotiate until they reach agreement.

We believe that viewing the negotiating process from this perspective is counterproductive. The usual result is that the importance of the institutional for the negotiations and the concomitant discursive environment is ignored, thus making it impossible to describe and analyze some of the most important features of the pattern of development since the 19505, at least in Denmark, but probably in many other countries too.

Negotiation-based economic processes are conducted partially, and to an increasing extent, within a discursive framework created by institutions, whose aims are to achieve systematic accumulation of experience, to shape opinions and to influence attitudes. And they are conducted within an institutional framework, intended to achieve compromise, and actually leading to mutually accepted decisions, while meeting deadlines etc.

The negotiation-based economic relationships arising from these discursive
institutional developments differ significantly from the 'bargaining
society' as traditionally represented (cf. Johansen 1978).

The most significant difference is that the requirement pertaining to a
'given' set of preferences is not satisfied, if a systematic accumulation of

experience, shaping of opinions and influencing of attitudes constitute a fundamental component of relationships in a negotiation-based economy. If this is the case, the constitution and manipulation of preferences must be perceived as a rationale for behaviour. Exertion of influence on the counterpart's preferences and on his perception of the 'negotiating space', the room for manoeuver, in the negotiating situation enters the behavioural function of the individual agent.

This attitude-moulding component upsets a basic premise in economic theory; the assumption of given preferences. Theoretical requirements for endogenization of preferences have been developed (cf. von Weizsåcker 1984). These, however, deal with vastly less complex feedback processes than those earlier referred to. The conscious and systematic measures adopted to mould preferences and to map negotiating space, which exert a fundamental influence on perception of what constitutes rational economic behaviour, are likewise omitted.

Institutionalized creation of consensus via 'campaign institutions', and forced compromises via negotiating procedures and decisions, constitute, in our opinion, central characteristics of today's negotiated economy. One of the central theses of this article is that, in Denmark, and since the 19505, institutions have been established, thereby initiating procedures for mobilization of and production of consensus and compromise. Attitudechanging and forced compromise have become important macroeconomic

Institutional History

It is a long-established tradition in Western political theory that the political
institutions, the political order or - in metaphorical terms - the state, are
the indisputable objects for analysis (March & Olsen 1984; Olsen 1985).

The state has come to be understood, in a constitutional context, as the sovereign, autonomous, hierarchically organized, internally powerbalanced, externally undifferentiated administrator of legalized compulsion & Selznick 1978; Olsen 1978).

As such, the ideal of the rule of law (referred to as 'rechtsstaats-ideal' in German) describes the institutional conditions that traditional economic theory stipulates for the ideal of optimal allocation of resources: the boundaries between the authority of the state and civil society are unequivocal; on liberty are made on a priori defined conditions.

In this article we deal solely with a formal or legal concept of boundaries (Maus 1986), such as that typically formulated in Danish constitutional and public administrative law. In constitutional law particular emphasis is laid on parliamentary government (Olsen 1978).

There are four links in parliamentary government: the people above the

representatives of the people; elected representatives above the government;the
above the administration and the administration
above the citizen.

All four links are welded together through the constitutional division of
power among the legislature, the administration and the judiciary (Krarup
& Mathiasson 1976; Haagen Jensen & Nørgaard 1984; Christensen 1980).

The fundamental principle of the rule of law is the sovereignty of the people, guaranteed by the representative system of government. According to the formal concept, there are thus four essential aspects of the rule of law:

(1) parliament derives its power from the sovereignty of the people;

(2) normative and formal rules are laid down for division of power among
the political institutions;

(3) legal limits are imposed on the activities of the administration with
respect to the citizen;

(4) the liberty of the individual citizen or collective groups of citizens is
created, and measures are introduced to protect this liberty in relation
to the state.

This unequivocal separation of the public from the private domain represents important aspect of the rule of law. However, an extensive delegation, over the past 20-30 years, of autonomy and authority from the state to various - also private - institutions invalidates this perception of the state. The boundaries between the public and private domains have become complex, and are no longer unequivocal.

In the following we shall cite examples of how the institutional history of the past 20-30 years in Denmark has altered the boundaries between state and market. Against the background of conclusions drawn from these examples, we shall question the criterion of optimal allocation of resources. We shall describe the development of negotiation-based economic relations as found in wages and labour market policy, public expenditure policy and industrial policy.

The Socio-Economic Ideology

Since the early 19605, there has been a general agreement - at least among the major representative organizations (employers, unions etc.), the central administration, the governments and the major political parties - to consider society as economic, in a cognitive-normative manner. Most of the relationships between the public administration, organizations, corporations households have been considered as part of an overall economic superstructure, referred to as 'the social economy'.

The social economic, or socio-economic conception perceives society as

an economic organism or an aggregate of private and public economic behaviour. However, the link between private and public behaviour is not 'automatically' established. It must, on the contrary, be created, and can be created only if the multitide of individual and collective decisionmakers, motivated by a common incentive, observe a normative injunction to act in the interests of the social economy.

Nor is there any automatic guarantee for the existence of this incentive. It too must be produced, over time. Today it seems as though it is a discursively or politically imperative for governments, organizations, corporations and probably also for households to take socio-economic factors into consideration when formulating demands pertaining to wages policy, public expenditure, investments etc.

The origins of this socio-economic ideology can be traced back to the
19305, although the institutional basis for the model was created at an even
earlier date (Pedersen 1985, 1986).

Even before the turn of the century, the two 'umbrella' (or 'peak') labour market organizations, the Federation of Danish Trade Unions ('Landsorganisation', or 'LO'), and the Confederation of Danish Employers' Associations ('Dansk Arbejdsgiverforening', or 'DA') had won the right to regulate wages and conditions of work, and to do so independently. The 'September Deal', concluded in 1899, formed the nucleus for subsequent development of a number of private and semipublic

Private institutionalization is based on collective agreement between the organizations, including a formalized allocation of competence to institutions, in which the organizations act as representatives of their members.

From 1890 onwards, many private institutions were established to negotiate
on wages and conditions of employment, and to mediate
or arbitrate in case of dispute.

Semi-public institutionalization is based on public law, whereby the organizations, on behalf of their members, are granted autonomy and authority - via legislative rules of procedure and with the assistance of arbitrators nominated by the government - to settle disputes and to issue rulings on conflicts.

These institutions were established primarily during the period 1910-19305.

The entire process of institutionalization of the relationships between
capital and labour is characterized by two distinguishing features:

(1) legal limits are imposed on the central state's discretion to regulate
wages, thus granting to the organizations the independent right to
conclude agreements on wages and working conditions;

(2) there is extensive delegation of public authority to the private organizations
private and semi-public institutionalization.

Institutionalization of the labour market transforms the state form from a monolithic and sovereign structure into a multi-centred and pluralist structure. The labour market establishes not merely a single authoritative decisionmaking centre, but a multi-centred and pluralist structure within which mutually autonomous and authoritative institutions can function.

As far back as the 19305, but with a marked increase in activity from the
late 19505, this state structure laid the foundation for the formulation and
propagation of the socio-economic ideology.

It is thus typical that the socio-economic ideology was formulated in an attempt to impose a comprehensive, coherent structure on a variety of interrelated activities and decisions reached in various independent institutions. Since the 19505, Danish socio-economic problems have been translated into the issue of co-ordinating decisions reached in formally independent institutions.

It is also typical that the socio-economic conception was formulated for the first time in an attempt to co-ordinate the collective wage agreements concluded by the labour market organizations with socio-economic objectives. And it was, in fact, also through introduction of a wages policy and an active labour market policy that the socio-economic ideology was transformed, over time, into a rationale, i.e. into a systematically structured and coherent discourse on society as such.

Since the end of the 19505, three distinct co-ordination problems have
been regarded as the cornerstones of Danish economic policy:

(1) How to co-ordinate the collective wage agreements concluded by the
labour market organizations with overall socio-economic objectives?

(2) How to co-ordinate public expenditures with overall socio-economic
objectives?

(3) How to co-ordinate the development of the industrial structure with
overall socio-economic objectives?

The institutional response to these three socio-economic co-ordination problems - as exemplified in wages and labour market policy, public expenditure policy and industrial policy - develops over time into the pivotal force of successive governments' economic policy, whether social democrat or liberal/conservative.

The need for co-ordination of wage agreements with socio-economic objectives was first described in Green Paper ('Betænkning') No. 154, 'Om samarbejdsproblemer i Danmarks økonomiske politik', issued in 1956. This document also contains the first official proposal for the introduction of an incomes policy (pp. 18, 59).

The notion of incomes policy was propagated institutionally and the Economic Council ('Det Økonomiske Råd', similar to the German Sachverståndigen was set up in 1962 to deal specifically with this issue. Since publishing its first report, in 1962, until today, the Council has given a high priority to incomes policy in its recommendations.

The Council's urgings have been heeded. The participants in negotiations have usually taken the Council's guidelines as a point of departure, and since 1979 wages policy has come to be accepted as an economic political instrument to be applied in the longterm redistribution of income.

The need for co-ordination of public expenditure with socio-economic objectives was localized, from 1957 up to 1973, in a series of green papers, commission reports and medium/longterm ('Perspektivplaner', or blueprints'), 1971 and 'PPII' 1973). This co-ordination was emphasized in connection with the then widespread decentralization of authority and responsibility, from the state to the municipalities ('primærkommuner').

The solution adopted to deal with the problem was a public expenditure policy, which was formulated and promulgated by commissions, and which led to integration of municipal ('kommunale') and county level ('amtskommunale') and the government administration in the socalled Co-operation (Denmark has 273 municipalities, and 14 counties).

Co-ordination of the industrial structure with socio-economic objectives can be traced back to the setting up, in 1973, of the Technology Council ('Teknologirådet'), but was probably not formally recorded as a co-ordination until 1986 (Finansministeriet 1986; Industriministeriet 1986; Arbejdsministeriet et al. 1986).

Here again a solution is presented, one which is apparently gaining an increasing level of acceptance among the most important socio-economic agents. This solution is designated 'technological development', which has been assigned a major role in the attempt to reshape Danish industrial exports - by revising the range of export items and the list of potential import countries - via a goal-oriented raising of standards, in technology, in education and in research.

The socio-economic ideology has thus been constantly formulated and propagated since the 19505, and this process still continues. There is a continuous compilation of and reporting on the experience accumulated in operating wages and labour market policy, public expenditure policy and, most recently, industrial policy.

Implementation of these four policies reveals a common characteristic: a continuous reformulation of the socio-economic ideology. The ideological conception of society as an economic organism is also being subjected to semantic systematization (cf. Pedersen 1987).

However, not all areas of policy have been subjected to the socioeconomicideology,

economicideology,and institutionalization of the negotiating 'game' has not reached the same stage of development in all areas of potential application. Nonetheless, we assert that there is justification for referring to a central trend, that macro-economic policy is increasingly assuming these contours, and that in a number of other areas of policy similar embryonic tendencies can be shown to exist (cf. Nielsen & Pedersen 1986).

The compilation and reporting of experience gained (in operating the four policies) has been made possible precisely through formation of the numerous 'campaign' institutions (e.g. the Economic Council, Budgetdepartmentet', Technology Council, economic secretariats in representative banks, savings banks etc.). An institutional learning process has thus been implemented.

Continuous 'campaigning' is conducted not merely by entering a particular of the current economic situation as an item on the political agenda, nor solely by recording the experience gained from application of economic policy tools. Campaigns are also conducted by creating the semantic frame of reference and the discursive environment.

The very act of recording the experience gained necessarily involves interpretation and reformulation. Forecasts are falsified. Project calculations modified. The correlation between phenomena is confirmed or denied. Earlier approaches to problems or solutions already tried are labelled 'invalid', 'obsolete'. Alternative forecasts and calculations are produced.

As a result, new problems are formulated, alternative solutions are discovered, and new terms and conceptual contexts are created. The socioeconomic is constantly being reformulated as a rationale for the understanding of an explanation of society's economic relationships, problems and patterns of development. The model is subjected to a continuous of rationalization (Habermas 1982).

The socio-economic ideological conception cannot be held to be 'false' yesterday and 'true' today, nor to be useful today and useless tomorrow. Rather, it is subject to three legitimacy requirements, which constantly force policy and campaign institutions to instigate learning processes.

The three requirements are (Pedersen 1986):

(1) that the socio-economic ideology must be normatively correct, viewed

in relation to existing value perceptions of future developments of

society;

(2) it must be technically credible, in relation to prevailing professional

standards and norms;

(3) it must be cognitively true, in relation to the cognitive conception of

the reality it theoretically seeks to reflect.

Since the 19505, these three requirements have been imposed on all policy

and campaign institutions which produce formulations pertaining to the socio-economic conception. If they are to legitimize their own delineation of problems, declarations of aims and proposed solutions, the institutions' assertions must be normatively correct, technically credible and cognitively true.

This continuous compilation and interpretation of the experience gained, the formulation of and popularization of problems, goals and means has, obviously, contributed to development of the socio-economic conception. The development is not an instant phenomenon, nor is it engineered by a single individual or a single representative organization. It has been going on for nearly fifteen years, and through 'campaign battles' waged between numerous political parties, organizations (employers, unions etc.) and institutions (ministries, central bank, councils, boards, commissions and secretariats).

Between these organs a propaganda war is constantly waged, each organ vying with the other in formulating the most compelling account of the lesson to be learned from experience, and in preaching the obvious 'truth' in plain language code. This joint approach to a common corpus of ideas has produced a more or less common language, a discursive environment, which the most important participants (or players) today employ when working towards formulation of and understanding of topical issues and future goals.

The most recent example of this gradual coalescence of terminology is
seen in the field of technology policy, with the Technology Council as a
major supplier.

As stated earlier, the Council was set up in 1973. It was corporatively manned, collegially structured and granted the authority to formulate and propagate a policy for the role of technology in solving socio-economic problems. The Council has succeeded in creating a technology policy and in forcing other campaign institutions, especially 'Budgetdepartmentet' and 'Det Industripolitiske Sekretariat', to adopt and accept the terminology and the related perceptions. Today the socio-economic conception encompasses the issue of technological competitiveness, an issue not considered in this context five years ago. And it also deals with the issue of investment steering, also ignored as recently as five years ago. One of the significant effects is a challenge to the 'classical' autonomy of the corporation. Where the socio-economic ideology was formerly developed in the context of an understanding of the problems of coordination state and the organizations, and between centralized and decentralized organs in the public sector, the relationship between state and corporation is today approaching the stage at which it may be regarded as the most important current and future co-ordination problem arising from socio-economic factors.

The result is further changes in the relationship between state and market, public and private. The conditions on which the corporation makes its economic calculations for its operations have been changed. The resulting phenomenon is an interplay, but also a struggle between socio-economic and corporate considerations.

Institutional Innovation

Formulation and propagation of the socio-economic ideology, from the 1950s until today, has been conducted parallel with sweeping institutional changes in the public administration, and in the latter's links with the private sector (Pedersen 1986). The institutional changes are strikingly apparent on the labour market and in the relationship between state and the municipal and county organizations. The public administration has thus effected a comprehensive integration of private representative organizations.

Institutional integration between the administration and the organizations
taken three forms (Pedersen 1987b):

• co-optation,

• segmentation, and

• formation of networks.

The main difference in the three forms of integration is the degree of
independence granted to the organizations affiliated with the institutions
incorporated in segments or in institutional networks.

Organizations are co-opted when they are granted the right to be heard, or are under an obligation to make a statement as a party involved in a more or less precisely delineated area of interest, or to issue a recommendation deciding on a single case.

There are numerous examples of such involvement. Involvement of private organizations - conducted more or less formally - has become, since the 19505, a common element in the decisionmaking processes of the political institutions (Valentin 1980; Buksti 1980).

Segmentation is applied to create a number of corporatively composed, collegially structured, legally competent, but otherwise, in formal terms, mutually independent institutions. The institutions enter the jurisdiction of one or more ministries, and are thus directly subject to the influence of the latter.

There are numerous examples of segmentation processes (Olsen 1978; Egeberg et al. 1978; Pedersen 1987b). Among the segments created are: the labour market policy segment, from the 19505; the social policy segment, from the 19705; and the industrial policy segment, through the 1980s. The

examples show that a segment emerges over time; that the boundaries
between segments may be floating; that only a few policy fields form
segments, and that not all segments are equally institutionalized.

Institutional networks are not subject to the direct influence of any
ministry. Networks admit both legally competent organs and privately
established or mutually composed institutions.

There are very few examples of an institutional network. Descriptions of these networks (Pedersen 1985,1986,1987a, 1987b) show that a network is created over time; that it has a complex structure, is constantly unstable, that it is rarely found and that it is shortlived. A network is built up slowly, is constantly changing and can rapidly disintegrate.

The three forms of integration describe a development curve. Integration
of organizations and administrations has moved:

• from co-opting single organizations through limited delegation of auth-

ority to single organizations within specific areas of interest or compe-

tence,

• over extended delegation to a large number of different institutions

within loosely demarcated segments,

• and on to unspecified delegation within undefined institutional networks.

However, each step does not replace, or cancel out, the preceding one. Instead, there is an accumulation of phases, in which each form of integration simultaneously and develops along a parallel path. Private, semi-public and public institutions form more and more complex combinations distinct decisionmaking processes.

For each of the four policy fields earlier referred to - wages, labour market, public expenditure and industrial policy - the institutions formed are characterized by a high degree of autonomy in relation to central state organs and by the creation of a type of institution that allows nonauthoritative, communicative regulation of the numerous, varied and formally independent institutions.

Since the end of the 1950s four new types of institution have developed:

1. Policy institutions. Public commissions and 'study' committees - usually
composed along corporative lines - are set up to formulate socio-economic
issues, to analyze and to issue recommendations on steps to be taken.

It is through such commissions that an understanding is formulated of (a) the relationship between the labour market and social economy (e.g. labour market commissions in the 1950s and 19605), (b) the relationship between the growth of the public sector - especially the municipalities - and the social economy ('Perspektivplaner' 1 and 2), and (c) the relationship between the industrial structure and the social economy (the Technology Council 1983).

2. Campaign institutions. These attempt to formulate, systematize

and

propagate the socio-economic ideology. Their constant aim is to have socioeconomic placed on the political agenda, while at the same time transforming the socio-economic ideology into a rationale capable of achieving acceptance throughout the political system and among the entire population, so as to produce 'universal' understanding of and explanations of socio-economic problems.

The Economic Council, established in 1962, conducts analyses of the socio-economic situation and indicates alternative courses of action. In the late 1970s the government set up its own campaign institution, in the Ministry of Finance's Budget Department ('Budgetdepartmentet'). The Technology Council, created in 1973, took the initiative in setting up regional technology information centres. During this period the organizations and strengthened their economic secretariats. The socioeconomic was expressed in mathematical terms, applying macro-econometric models (e.g. 'SMEC and 'ADAM').

3. Negotiation institutions. These can be either ad hoc or permanent bodies, usually consisting of two or three parties. They draw up timetables and outline frameworks for forthcoming negotiations, and possibly for negotiations subordinate organs.

Negotiation institutions were established on the labour market at the turn of the century. But the central negotiation committees with joint representation from the peak organizations were not formed until the late 19505, and bilateral or tripartite organs (for organizations and government) did not emerge until the early 19705. The so-called Budgetary Co-operation ('Budgetsamarbejdet') between the state and the municipal and county organizations was formally established in the 19705, along with a comprehensive planning system. The Technological Development Programme launched in the early 1980s, to co-ordinate the activities of state and corporations, involving negotiation-based co-operation on research and development projects.

4. Arbitration and sanctions institutions. While these are typically labour
market institutions, they are also being developed to deal with disputes
between the state, municipalities and counties.

On the labour market there are private mediation and arbitration institutions, semi-private 'conciliation' and sanctions institutions (the Conciliation i.e. 'Forligsinstitution', and the Labour Court). To deal with disputes between the state, municipalities and counties, ad hoc committees formed, representing, on one hand, the ministries of finance and of the interior, and, on the other hand, the municipal and county organizations. The state can also impose sanctions on the municipalities and counties for various infringements (e.g. overspending and overtaxing).

The most important features common to all these institutional innovations

• the formation of new institutions or sub-organs makes it possible to
formulate and propagate ideological conceptions, and to systematize and
develop them to constitute a rationale,

• at the same time institutions are created and institutional procedures
are introduced to ensure that a compulsory compromise is reached at
negotiations between mutually independent participants,

• within the normative framework dictated by the socio-economic ideology
and the institutional restrictions imposed by procedures and rules, a
complex form of 'negotiating game' is set in motion (Pedersen 1986).

• the negotiating game is basedon communicative rather than authoritative resolutions, and typically results in politically and morally binding agreements, distinct from agreements binding under the provisions of private or public law.

Many types of institution participate in the negotiating game (government, parliament, administration, policy, campaign, negotiation and sanctions institutions - organizations' internal committees, organs, assemblies etc.). These are manned by many types of representative (ministers, civil servants, experts, executives, organizations' representatives etc.). Acting in unison, they regulate and plan wage formation and developments in public expenditure in the industrial structure, in co-operation between the state and corporations.

This regulation and planning is not conducted solely through an organizational a centrally procured and legally binding plan or through a price system. It is conducted through institutionalized procedures for communicative selection of the problems to be tackled and negotiated - wages, labour market issues, public expenditure, matters of technology policy - to define objectives, and to recommend and implement solutions, within the framework of the socio-economic conception.

The numerous institutions and representatives are - typically - mutually independent. They do not always meet as superior and subordinate, nor as isolated entities on a liberal market. A functional allocation of tasks has been effected among the participants - whether dealing with the labour market, public expenditure or technology policy - so that they remain mutually independent of one another and must mutually co-ordinate their decisions, and do so in such a manner that this co-ordination cannot be achieved solely through imposition of rulings or legally binding decisions. The decisions made must be based on political processes that functionally co-ordinate differentiated decisions through communicative rather than authoritative resolutions:

The communicative co-ordination is effected via mobilization and production
a consensus on socio-economic problems, and on the goals and

means which are currently dominant and believed to be most effective.
• Policy and campaign institutions formulate, propagate and transform

perceptions of socio-economic situations into normatively binding

conceptions.

• The negotiation institutions stipulate frameworks, draw up timetables,

conclude agreements, revise them and comply with them.

• The 'reconciliation' and sanctions institutions reconcile the conflicting

parties in case of a breakdown in negotiations, mediate in conflicts on

interpretation, and impose sanctions when agreements are breached.

In all the policy fields already listed, the requisite structure and procedures for the negotiating game are already in place, instantly available to the players. In other areas (e.g. the capital market), such facilities are currently being established. There are, however, major differences in the stage of development reached. As yet, the only area of policy with a fully developed institutional network is wages policy, while other areas are typified by coopting segmentation.

The negotiating game can be played in an institutional network solely through a communicatively elicited consensus or through procedures for compulsory compromise, mediation or arbitration. Within segments the negotiating game is sometimes authoritatively conducted, by virtue of the placing of the segment under the jurisdiction of one or more superior bodies, but typically the game is also communicatively conducted, via creation of a given conceptual context (cf. Chapter 5).

Conclusion

Since the 19505, the organizational structure of Danish society has undergone changes. Formulation, propagation and institutionalization of the socio-economic ideological conception have led to an institutional integration of the public administration, private organizations and, more recently, corporations. The boundaries between public and private, state and society have become very complex.

We have attempted to demonstrate this development in two ways: First, we have made an assertion. We have asserted that the ideal of optimal allocation of resources presupposes compliance with a number of institutional We have further maintained that these requirements are described in the constitutional ideal delineation of boundaries between the authority of the state and civil society.

We have subsequently submitted a historical description. We have

described how, since the 19505, institutions and inter-institutional dependencieshave
created in Denmark, which in a number of policy areas
have in turn established negotiation-based economic relationships.

We had the aim of showing how and to what extent negotiation-based economic relationships have been established in Denmark. We have attempted to define the term - negotiated economy - by identifying its institutional requirements (segments and institutional networks) and the consequences for forms of agreement (discursive, political and morally binding agreements). We regard institutional networks, negotiating games and agreements which are not legally enforceable as ineluctable concomitants, negotiation-based economic relationships are established.

In the following we shall try to amplify this conclusion and indicate some
new approaches to problems suggested by our conclusion.

According to the ideal of rule of law, the state is the autonomous, hierarchically organized, internally power-balanced, but externally undifferentiated of legalized compulsion. Limits are imposed on state intervention in civil society. These limits can be exceeded only on defined, specific and a priori given conditions. A space for individual and collective freedom of behaviour is thus created.

The rule of law has not been unequivocally invalidated by the institutional history we have just presented. Between the ideal and the organizational structure actually existing there is in some cases a state of tension, and in others a rupture.

The ideal of free competition and optimal allocation of resources has, however, been invalidated. The institutional requirements for the ideal have undergone change - to such a degree and on such a scale that the ideal has become a myth.

According to the ideal of rule of law, the administration constitutes an entity. The administration's status as an entity is based on the minister's dual status as political leader and administrative head. The status as an entity is ensured through the hierarchical (superior/subordinate) structure - i.e. successive levels of competence and responsibility. In formal terms the administration is at the same time subject to the requirement that the decisions it makes are sanctioned by legislation adopted by parliament.

Involvement or co-optation of the organizations and segmentation of the administration have diluted both the ideal of the integrity of the administration and the principle of administration grounded on positive law - the so-called principle of legality.

In this context two features of the observed trend are significant:

1. The number of law-based institutions outside the minister's - and thus
parliament's - direct sphere of influence is increasing, as is the importance
of such institutions.

2. Less emphasis is being placed on direct legal authorization for decisions
made in the public administration.

This trend has resulted in the formation of a large number of administrative organs which are more or less independent of the relevant minister(s). At the same time the political order has become multi-centred, and the administrative structure less hierarchical. Another consequence is that the administration is forced to find authorizations - often indirectly - for its decisions at some point along an ever lengthening chain of authorization (Pedersen 1987b). The boundaries between state authority and civil society are no longer distinct; encroachments on liberty are no longer based on defined, specific, and a priori given conditions.

However, this does not mean that the integrity of the administration has been destroyed, nor that administration has become horizontal. Nor does it mean that the principle of legality has been abolished. The minister is still responsible for most administrative decisions (Sørensen 1977). The essential validity of the legality principle remains intact (Christensen 1980).

The political system, however, has become much more differentiated and complex. The formal entity is combined with newly formed segments. A state of tension has arisen between the law as applied and the rule of law ideal.

Simultaneously, institutional networks are formed, destroying the integrity the administration and the principle of legality. As networks are not subject to any jurisdiction, they do not have to play any subordinate role. A network, being a combination of public, semi-public and private institutions, makes it impossible to distinguish the boundaries separating the public from the private sector. Here the legal limits are laid down in extremely complex combinations of public and private law. The boundaries between the state hierarchy and civil society do not exist. The rule of law has been shattered.

Instead, institutional arrangements have been constituted, consisting of numerous, functionally differentiated and mutually independent institutions, co-ordinate actions through networks, reach decisions via the negotiating games and extract commitments from one another through politically and morally binding agreements.

These institutions formulate perceptions of reality, lay down policies, and enter morally or politically binding agreements. The conscious and systematic influence exerted on the negotiating environment and on preferences the cornerstone of negotiation-based economic relations.

The result of this development is invalidation of the ideal of optimal
allocation of resources.

The institutional requirements stipulated for this ideal and the negotiation-basedeconomic

tiation-basedeconomicrelationships actually existing differ radically. The socio-economic conception is not limited to formulation of new ideals for what is to be regarded as the proper course of action. Nor is decisionmaking via segments and institutional networks limited exclusively to establishing relationships between given goals and means, and between decision and action. Rather the propagation of the socio-economic conception and allocation via segments and institutional networks constitute elements in a process of rationalization.

Taken as a whole, the rationalization process implies formulation of a conceptual context, creation of a number of institutional groupings and procedures, and of new relationships - economic, regulatory and personal. The rationalization process is an expression of dynamic learning processes rather than static optimization processes.

The process involves far more complex (and extensive) communication than assumed in traditional economic theory. The 'outer, objective' world is not the sole area in focus. The individual participants do not restrict themselves to adapting to signals received. They also struggle to influence the perceptions of the other participants ('hearts and minds'). Information material is compiled and distributed with the aim of persuasion. The individual participant's 'inner, subjective' world is thus continuously being modified.

In relation to instrumental making of decisions and allocation of resources, negotiation-based economic relationships involve additional and different factors. Moulding of opinion and interpretation are obviously central elements.

The ideal of optimal allocation of resources presupposes a given set of preferences as part of the basis for assessing efficiency/optimality. The preferences of the agents are treated as parameters, rather than as endogenous

The emergence of negotiation-based economic relationships imposes limits on what formalized negotiation theory, using the games theory framework, can offer as a tool for analysis. In this context, as in economic theory in general, preferences of the agents and the negotiating environment assumed to be given. Negotiation processes are taken to mean processes that are not governed by rules which demand joint decisions within specified deadlines, and as interaction between agents whose preferences seen as exogenous factors rather than as endogenous results of the interaction. The theories do not include consideration of behaviour which is aimed at influencing the negotiation environment and at manipulating

If, on the other hand, emphasis is laid on the environment in which the
negotiations are conducted, in the form of systematic moulding of opinions
and attitudes, this will require a fresh examination of a number of the

well known concepts currently applied in the games theory literature on
bargaining.

First, the significance of information will have to be extended. Information transmitted and received, not only to inform the counterpart of one's own (true or untrue) preferences in relation to a given negotiating situation. Information is also compiled and distributed with the aim of influencing factors that can shape preferences in subsequent negotiating situations.

Second, the concept of rationality assumes a new significance. A number of actions, which immediately appear to be irrational when considered from the traditional viewpoint, can be considered as rational in view of their immediate and 'delayed action' impact on the negotiating environment and on preferences.

Third, an extended significance must be attached to strategic behaviour. Sacrifices and the carrying out of threats are thus traditionally seen as appropriate strategic measures, even though they lead to predictable losses, both for the individual participant and for all concerned, if these actions can succeed in influencing the opposition's perception of one's own preferences its perception of one's capacity for struggle (freedom of manoeuvre), thereby leading to more favourable (ultimate) results in later negotiating situations. However, actions intended to influence the environment which subsequent negotiations are conducted can, in a broader sense, e.g. via development of the socio-economic conception, also be regarded as strategically justified.

Current bargaining theory (derived from games theory) can thus offer only a limited range of tools for analysis of negotiation-based economic relationships. It may be more useful to consider the findings reached in behavioural (Simon 1957), institutional (Hodgson 1988) and evolutionary economic theory (Nelson & Winter 1977), in which the participants are not assumed to optimize, but 'satisfy' via norms and routines, which are adapted via learning processes.

It is impossible, on the basis of the ideal of optimal allocation of resources, to evaluate the negotiation-based economic processes. If the ideal is applied, it often leads to decidedly negative conclusions. Leif Johansen (1978, 23) thus concludes that

bargaining will often be an inefficient decision procedure in the sense that it tends to distort the information basis for decisions, it tends to use or waste resources in the process, particularly delay decisions for reasons which are not technically necessary, it will more or less frequently lead to breakdown and failure to realize the potential gains, and threats will sometimes be carried out.

Johansen's conclusions strike us as being well founded - if viewed in
the context of the assumptions made and the theoretical basis applied.

Undoubtedly, such bargaining processes will rarely lead to pareto-optimal
allocation results.

If, however, we abandon the requirement for given preferences, it is not possible to draw such conclusions. Bearing this in mind, we agree with March & Olsen's criticism (1984) of the dominant theories in political science since the 19505. We merely wish to subject the neo-classically oriented economic theory to the same criticism.

March & Olsen (1984, 735) criticize these theories of political science for
being

. . . instrumentalist, inclined to define decisionmaking and the allocation of resources as
the central concern of political life, less attentive to the ways in which political life is
organized around the development of meaning through symbols, rituals and ceremonies.

In line with March & Olsen, we lay emphasis on the fact that politics
in general and negotiation-based economic decisionmaking systems in
particular must primarily be regarded as

. . . education, a place for discovering, elaborating and expressing meaning, establishing shared (or opposing) conceptions of experience, values, and the nature of existence. It is symbolic, not in the recent sense of symbols as devices of the powerful for confusing the weak, but more in the sense of symbols as the instruments of interpretive order (March & Olsen 1984, 741).

REFERENCES

Arbejdsministeriet m.fl. 1986. Debatoplæg om vækst og omstilling. Kbhv.

Berrefjord, O. 1978. 'Fra embetsmannsstat til embetsmannsstat?' in Hernes, G. K., ed.,
Forhandlingsøkonomi og blandingsadministrasjon, Oslo: Universitetsforlaget.

Berrefjord, O. 1984. Rafnes Saken. Forhandlingsøkonomien i aksjon. Oslo.
Betænkning 154. 1986. Samarbejdsproblemer i Danmarks Økonomiske Politik. Kbhv.
Buksti, J. 1980. 'Organisationerne, den politiske proces og samspillet med samfund-

sudviklingen', in Buksti, J., red., Organisationerne under forandring. Århus.
Christensen, B. 1980. Forvaltningsret. Hjemmelspørgsmål. Kbhv.
Egeberg, M., Olsen, J. P. & Sætren, H. 1979. 'Organisasjonssamfunnet og den segmenterte

Finansministeriet. 1986. Budgetredegørelse 1986, Kbhv.

Haagen Jensen, C. & Nørgaard, O. A. 1984. Administration og borger. Kbhv.
Habermas, J. 1982. Theorie des Kommunikativen Handelns. Frankfurt am M.
Hernes, G. 1978. 'Makt, blandingsøkonomi og blandingsadministrasjon', in Hernes, G., ed.,

Forhandlingsøkonomi og blandingsadministrasjon. Oslo: Universitetsforlaget. Hodgson, G. 1988. Economics and Institutions. Cambridge: Polity Press. Industriministeriet. 1986. Den Industripolitiske Redegørelse 1986. Kbhv. Johansen, L. 1978. 'The Bargaining Society and the Inefficiency of Bargaining', Memorandum,

Socialøkonomisk Institut, Oslo.

Krarup, O. & Mathiassen, J. 1976. Elementær Forvaltningsret. Kbhv.
Lindensjø, B. 1984. 'Den Dolda Kompromissen', Skeptron, no. 1.

March, J. G. & Olsen, J. P. 1984. 'The New Institutionalism: Organizational Factors in
Political Life', American Political Science Review, no. 3.

Maus, I. 1986. Rechtstheorie und Politisene Theori im Industrikapitalismus. Miinchen.
Moore, C. & Booth, S. 1988. Beyond Corporatism - Toward Negotiated Economy? Markets,

Networks and the State (to be published).

Nelson, R. & Winter, S. 1977. An Evolutionary Theory of Economic Change. Cambridge,
Mass.

Nielsen, K. 1986. 'Teori om samspillet mellem institutionel og teknologisk udvikling'..
Tidsskrift for Politisk Økonomi, no. 3.

Nielsen, K. & Pedersen, O. K. 1986. 'Fremtidens samfund - forhandlingsøkonomi og kamp
om samfundsopfattelser', in Årbog 1986, Institut for Samfundsøkonomi og Planlægning,
Roskilde Universitetscenter.

Nonet, P. & Selznick, P. 1978. Law and Society in Transition. Toward Responsive Law. New
York.

Olsen, J. P. 1978. 'Folkestyre, byråkrati og korporativisme', in Olsen, J. P., ed., Politisk
Organisering. Organisasjonsteoretiske synspunkt på folkestyre og politisk ulikhet, Oslo:
Universitetsforlaget.

Olsen, J. P. 1985. 'Nyinstitutionalismen og statsvitenskapen', Statsvetenskaplig Tidskrift,
no. 1.

Pedersen, O. K. 1985. Institutionel historie. Om kampagnestyret forhandlingsøkonomi og
konsensusplanlagt løndannelse, Roskilde: Forlaget for Samfundsøkonomi og Planlægning.

Pedersen, O. K. 1986. 'Samfundsøkonomi og forhandlingsøkonomi. Om konsensusplanlægning
netværksstyring i Danmark', Statsvetenskaplig Tidskrift, no. 3.

Pedersen, O. K. 1987a. 'Kampagnekonflikter og forhandlingsspiF, in Amoroso, B. & Jespersen,
ed., Socialøkonomisk Rapport 1987, Kbhv.

Pedersen, O. K. 1987b. 'Den udvandede retsstat'. Memo, (to be published).

Pierre, J. 1988. 'Towards Negotiated Economy?: Political Economy Beyond Meso-Corporatism',
Moore C. & Booth S. Beyond Corporatism - Toward Negotiated Economy?
Markets, Networks and the State (to be published).

PPL 1971. Perspektivplanlægning 1970-1985. Kbhv.

Skogh, G. 1982. 'Forhandlingssamhållet 1990-varkan marknadsøkonomi, blandekonomieller
planhushålling', in Svenska Arbetsgivareforeningen. Facklig foretagsledning? Stockholm.

Sørensen, M. 1977. Statsforfatningsret, Kbhv.

Sæbø, H. V. 1980. 'Olje-økonomi og forhandlingsøkonomi', in Inkjøp 65, 1.

Valentin, F. 1980 Fordelingen af påvirkningsmuligheder. Lavindkomstkommissionens
sekretariat. Kbhv.

Weizsåcker, C. C. von 1984. 'The Influence of Property Rights on Tastes', Zeitschrift fur die
Gesamte Staatswissenschaft, mårz.

Williamson, O. 1975. Market and Hierarchies. New York.

Williamson, O. 1985. The Economic Institutions of Capitalism. New York: Free Press.