An Instrumental Perspective

1. Organisations as instruments
Public organisations carry out tasks on behalf of society. In higher education this could mean, for example, preparing study reforms through a government ministry, and the implementation of new study programmes through public universities and colleges. Organisations can thus be understood as tools or instruments for achieving certain goals seen as important in society, such as raising the standard of higher education. This can come to expression partly by public organisations and their members acting with instrumental rationality in fulfilling tasks, and by achieving the desired results. It entails that members of an organisation assess available alternatives or tools according to their consequences, in relation to the selected goals, make wilful choices between alternatives, and achieve the effects desired through those choices. Yet instrumentality can also come to expression through an organisations structure being designed in accord with means-ends assessments, and that the organisational structure, in turn, determines the members behaviour while carrying out tasks. Instrumental rationality can thus involve the effects of organisational structure and the process through which it is determined and formed. In the study of organisations, there are many tone setting traditions that view organisations as instruments. Among the classic social scientists, Max Weber has particularly influenced organisational literature, through his analysis of bureaucracy as an organisational form. About the same time in the USA, a scientific view of management in productionorganisations developed, particularly through the work of engineer Frederick Taylor. In this Scientific Management tradition, often called Taylorism, great emphasis is placed on finding effective organisational forms and work techniques. In the inter-war years several attempts were made to develop a set of general administration principles, among others, by the American political scientist Luther Gulick, in connection with the development of public organisations in the USA (Scientific Administration). As a critique of the contributions of Taylor, Gulick and others, the American political scientist Herbert Simon and his colleagues, in the aftermath of the Second World War, developed a theory about administrative decision-making behaviour. Simon

doubted the realism of the assumptions Taylor and Gulick built their work upon, but he nevertheless emphasised that members of organisations try to act in instrumentally rational ways, and that an organisational structure has great significance for what members actually do or can do. Explanations from the vantage point of organisations as instruments, are concerned to clarify goals and means-ends conceptions of organisations and individual actors, which choice of action they follow, and if and how the actions result harmonises with what was desired. We shall therefore first sketch which elements are included in actions built upon such a logic of consequences. By formal organisational structure, often merely called organisational structure, we mean a structure that consists of positions and rules for who shall or can do what, and how various tasks should be executed. Organisations are composed of a set of positions and subordinated units, and can themselves fall in under larger units. Organisational units can be divided up and coordinated in different ways. In the following account of organisational structure, we take our point of departure in bureaucratic organisational forms with strong elements of hierarchy, division of labour and routines. Thereafter we will sketch how structural features of organisations can influence what organisations do and how their members think and act. Instances of several sub-units and divisions of labour also invite a view of organisations as heterogeneous, with coalitions that make room for disparate goals or interests, and diverse resources for the articulation of that interest. The individual sub-units and their members can act in an instrumentally rational way, but the results here also will depend on which resources others have and what they do. Public organisations and their sub-units can also enter into coalitions with other public and private organisations, and in a similar way be dependent upon what these do to achieve their objectives. Other such organisations constitute a part of the organisations environment, and we will address how these and other aspects of the environment can influence structural features of the organisation. From an instrumental perspective of organisations, steering can occur, in part, through designing the organisations structure and relation to the environment, and, in part, through instrumentally rational actions within these boundaries. Organisational leaders can, to varying degrees, influence their own and others contexts of action. As we shall return to at the end of this chapter, an instrumental perspective entails that leaders are assumed to have the ability for

rational calculation, and for political and social control, but that there can be some limitations to these abilities.

1.1.1.

The logic of consequences instrumentally rational actions

As we shall account for more fully in chapter 5, organisations have goals ideas about what they would like to achieve or realize in the future. Within this context, a problem for an organisation can be defined as a perceived distance between a desired and an actual state of affairs. For public organisations, this can involve problems defined from within public administration, or from outside. Thus problem solving implies actions that aim to reduce or eliminate this distance. Another way of describing this is through the concept of instrumentally rational actions. The implementation of such actions consists of four elements: Goal or problem: What does one want to achieve, and what is the distance between that and the current state of affairs? Alternatives: What actions are possible? (Expectations about) consequences: What future consequences in relation to the goals can follow from each alternative, and how likely are these consequences assuming that the alternative is chosen? Decision-making rules: How shall the choice between alternatives be made? In some cases, the point of departure is an organisations goal or perception of a problem. It assesses possible alternatives based upon their consequences, and a choice is made accordingly. E.g., a university can have apprehensions about its forms of leadership not providing the basis for achieving professional quality goals, and it assesses various alternatives that are expected to give a greater degree of goal-achievement. In other cases, the point of departure is an organisations alternatives for action. It assesses how valuable the consequences following from each alternative are, and a choice is made based upon this. A university can be faced with several possible forms of leadership, and it chooses amongst these according to the expected consequences related to goal-achievement. The actions required to implement tasks are marked, in any case, by their relation to a logic of consequences, where the organisation chooses between alternatives, based upon rational calculations of possible consequences.

The concept of full instrumental rationality refers to an organisation having clear and consistent goals, it has full overview of all alternatives, and full insight into which consequences these alternatives render in relation to its goals. From this it often follows that the organisation chooses the alternative that gives maximal or the greatest degree of goal achievement. Even so, many empirical studies of how organisations act show that only to a certain degree is this realistic, particularly for complex public organisations where many considerations come to bear. This state of affairs is crystallised in the concept of bounded rationality. As we will return to in chapter 5, the concept implies that an organisations goals are diffuse, inconsistent or unstable, and that the problems it faces are complex. The concept also includes the idea that an organisation has incomplete information about alternatives and consequences. It knows of only a limited set of alternatives because it has limited capacities; it must select information and premises for decision-making, but probably has unsure knowledge of means-ends-relations. It requires time and resources to acquire a better knowledge base, and complete insight will be impossible to achieve. From this it follows that the organisation chooses an alternative that gives good enough, or an acceptable degree of goal achievement. In other words, the organisation will have a decision-making rule built upon satisfisation, rather than maximisation, where satisfactory, but not necessarily optimal solutions are chosen. It must be underscored, nevertheless, that also actions based upon bounded rationality are marked by a logic of consequences. In his analysis of decision-making behaviour, Herbert Simon takes as his point of departure the individuals actions within an organisation. As the antithesis of economic man, motivated by self interest and with full knowledge of all alternatives and consequences, Simon sketches the administrative man, who acts within a determinate structural framework, but who has incomplete knowledge of alternatives and consequences. A specialised organisational structure renders to individuals a relatively narrow range of understanding and purview in their roles; this can ease understanding and capacity problems, but can also cause knowledge problems and difficulty in seeing ones own activity and role in a wider perspective. As we shall return to below, designing an organisational structure could entail that rationality can be somewhat reduced at the organisational level, in relation to the individual level. We shall also return to how possible conflicts of interest between individual members and their organisation can be dealt with through organisational structure.

2. Formal organisational structure

Who shall or can do something on behalf of an organisation by carrying out tasks is determined by which formal roles or positions organisational members have, which sub-unit they are in and which larger units the organisation is a part of. Here we are concerned with the formal structure of organisations. This entails that expectations for those holding positions are impersonal. Norms for practices thus exist independently from the personal characteristics of those who at any certain time hold positions. These formal norms can come to expression through, e.g., organisational charts, manuals, rules and regulations. Organising through designing the formal organisational structure happens, to varying degrees, through specialisation and coordination. How an organisation is specialised and coordinated can pertain to the relations between positions and sub-units the organisation is composed of, the relations to larger units where the organisation is a part, and its relations to other organisations it has dealings with. A bureaucratic organisational form, as Weber describes it, is marked by hierarchy, division of labour and routines. Hierarchy entails superior and subordinate positions and various vertical levels in an organisation. In a government ministry, various sections or agencies can, e.g., be part of a division, and all divisions will be subordinate to the political and administrative leadership. Someone in a superior position has the right to command and instruct subordinates, and this is achieved through downward channels of communication. In addition, information travels upwards through various forms of case proposals and reporting systems. The hierarchy in a bureaucratic organisation is often tied to a career system, where members endeavour to rise to higher positions, and are promoted on the basis of qualifications and performance. Division of labour entails that an organisations tasks are grouped into different units and tied to concrete positions; in other words, it is a matter of horizontal specialisation. As we shall return to shortly, various principles can lie at the root of this division of labour within and between organisations. What characterises a bureaucratic organisational form is not the specific principle, but rather that there is a large degree of division of labour. Such an organisational form is also characterised by many routines, i.e., rules and procedures for who shall carry out tasks

and how they should be accomplished. These routines can have different contents, but will often be codified in written documents such as regulations, memos and manuals. Through the superior-subordinate relation between different levels in a hierarchy, there will be a great degree of vertical coordination within and between organisations. In the central governments of many countries, vertical coordination comes to expression, among other ways, through principles pertaining to ministerial responsibility, i.e., the cabinet minister is responsible to the parliament for activities in his or her own ministry and subordinate agencies. At the same time, the minister, through participation in the cabinet, is responsible for shaping the cabinets general policy, and this entails elements of horizontal coordination at this level. Hierarchy can also involve vertical specialisation, in that different types of tasks are assigned to different levels in the organisation, or to organisations at different levels. For instance, tasks related to advising political leaders could be assigned to a ministry, while more professional tasks could be assigned to a subordinate agency. Routines can constitute a form of coordination both vertically and horizontally. Procedural rules can be used as tools within an organisation, but can also be intended to coordinate activities in a why that cuts across organisations. For civil service organisations, this can apply to, e.g., general rules for casework, financial management and human resource management. Horizontal specialisation expresses how different tasks are thought to be allocated on a certain level by means of organisational structure. It can, for example, be decided that one and the same unit will deal with transportation and environmental issues, or these can be allocated to different units on the same level. Another example is that one can choose to have ownership tasks and supervisory tasks in the same agency, or allocate these to two agencies, such as has become more common now. Through the influence of administration theorists such as Luther Gulick, it has become common to distinguish between four different principles for horizontal specialisation. The purpose principle, or sector principle, distributes cases according to purpose or sector. At the ministry level, the principle may come to expression through there being distinct ministries for defence, education, health and various industries such as agriculture and fisheries. The process principle distributes cases according to the procedural method or type of process used in order to achieve a purpose. In a ministry or a municipality, there can, e.g., be distinct units for cases concerning financial management, human resource management and judicial issues. These cases are not goals in themselves, but steps in the process of advancing towards a

purpose, such as, e.g., education. The third horizontal divisional method for classification is the client principle. Here all the cases concerning a particular section of the population are gathered into one organisational unit, such that through collaborative expertise, a comprehensive perspective on them can be achieved. In some municipalities there can be, e.g., individual sections for the elderly or for children or youth. The geographical principle implies that an organisations structure reflects a territorial division of the society. The principle is exemplarily expressed through the three levels of government the national, the regional and the local and, not least, the ever-increasing international level. It may come to expression within a municipality through the existence of several neighbourhood councils that each take on tasks such as schools or care for the elderly. In an organisational hierarchy, one specialisation principle may lie at the root of one level and another principle at a lower or higher level. This can mean, for instance, that a government ministry with a purpose such as education can have its own sub-units for various processes or courses of action. By horizontally allocating issues of concern, the responsibility for coordination will move upwards to a higher level in the organisation, or to a larger unit the organisation is a part of. For example, a ministrys leadership will have responsibility for coordinating activities for several divisions, and the cabinet will have responsibility for the activities of several ministries. Many types of organisational forms can be thought of as alternative or supplementary to a bureaucratic organisational form. At one extreme is a flat structure, i.e., an organisation without superior and subordinate hierarchical ordering, but with several positions and sub-units at the same level. In a flat structure, members of an organisation can either always be tied to the same position or sub-unit, or rotate between these. In a collegial structure, a board of directors or an advisory council can be set up instead of, or in addition to, the top leadership in the hierarchy. For example, cabinet ministers deal with many issues as a collegium, and so also do city government commissioners. Many public organisations have a permanent board of directors devoted to leadership, e.g., state-owned companies and public universities. A collegial structure can also cut across organisations, e.g., by establishing coordinating committees that include members from various organisations, for horizontal coordination between them. Pertinent Norwegian examples can be the coordinating committee a government uses in relation to large labour market and economic organisations, and the so-called Coordinating Committee for

European Union/European Economic Area Agreement (EU/EEA) issues, where all the ministries are represented. Another alternative is a matrix structure. Here a position or sub-unit is subordinated to several superior units simultaneously. These superior units usually operate according to different principles of specialisation. An example could be a unit for elderly care in a neighbourhood under the jurisdiction of a municipal service for elderly care and a wider neighbourhood administration, in other words, the client principle and the geographical principle would apply for the same unit. In addition to arrangements such as these without time limitations, public organisations have many temporary arrangements that extend beyond the bounds of bureaucratic organisational forms. This is the case for commissions or task forces appointed by a government ministry in order to determine alternatives for accomplishing specific tasks, such as a commission that reviews new or revised laws. Here persons often take on such duties in addition to their regular positions, but someone may also have it as a fulltime job for a limited period. Among pertinent examples within Norwegian higher education, we mention the Mjs Commission that considered study reforms, and the Ryssdal Commission that considered organisational forms for universities and colleges. This is also the case for project-organisations that are established in order to review and implement specific tasks, such as the building of Gardermoen Airport Train Line. Here persons usually are removed from their regular positions and placed in a temporary unit that disbands when the task is accomplished. This method is frequently used for reorganisations in public administration. Instances of boards of directors, advisory councils, coordinating committees, matrix structures, commissions and task forces are all various forms of network structures, which supplement the bureaucratic organisational form. This is not merely the case within individual public organisations, but also applies to the interrelation between public organisations, and to organisations in the private sector, so-called public-private partnerships. Public organisations can also be distinguished from each other according to whether they have a simple or complex structure. While an organisation with a simple structure has few positions and sub-units, and has few connections between these, a complex structure will have many levels, many units on each level, and many connections vertically and horizontally. Moreover, connections can be differentiated from each other according to how tight or loose they are, that is, whether they are tightly integrated or consist of more or less independent units. The

larger an organisation is, measured in number of members, the greater chance there is for having a complex structure. A large organisation with a bureaucratic organisational form can nevertheless be relatively simple. Meanwhile, complexity can increase by dividing up public organisations, such as we have seen tendencies for in recent years, through structural devolution, divisionalization and separating independent result-units, all of which make public organisations increasingly appear as conglomerates. Public organisations can also be described according to the degree to which they are centralised or decentralised. This normally concerns at which level decisions are made in and between organisations. In a centralised organisation or group of organisations, the final or important decisions almost always are taken at superior levels, and it is expected that these decisions be followed up at lower levels. In contrast, such decisions in a decentralised organisation or group of organisations will be delegated to a lower level. The concept of decentralisation often is used to describe such a condition, but will more precisely concern processes where decisional authority is delegated to a lower level. Reforming public organisations through re-structuring does not necessarily lead to either centralisation or decentralisation, but can involve both simultaneously. The Norwegian hospital reform illustrates this well: On the one hand, it centralised power by having the central government take over ownership of hospitals from county governments; at the same time, the hospitals were transformed from being administrative organisations into health enterprises, which entailed decentralisation. Structural features and concrete actions Although the formal organisational structure does not necessarily indicate anything about the actual behaviour of members in an organisation, it will constrain how tasks are carried out. Among other ways, this comes to expression through the use of various specialisation principles. Through geographical specialisation, the likelihood will increase that different sectors will be coordinated within each individual geographical area, e.g., neighbourhoods. The opposite would be a specialisation built upon the purpose principle, implying that focus is placed upon lines of demarcation between sectors rather than between territories, and such specialisation will probably further the standardisation of public policies across geographical areas. Specialisation built upon the client principle will, to a greater degree, imply policies directed towards sectors

being seen in relation to each other. Specialisation based upon the process principle can support the development of professional expertise in an organisation, e.g., expertise related to financial management and juridical issues. On the other hand, such specialisations can direct attention away from the actual purpose of the organisation. Choosing a specialisation principle could thus have great significance for how problems are dealt with, and for the contents of public policy. To a certain degree, this can also be influenced by different forms of coordination, vertically through hierarchy, or horizontally through structures such as coordinating committees. An example of different forms of coordination within a specific policy domain can be Norwegian governmental youth policy, which, up until the early 1990s was placed together with sports, in a unit that fell under a division (later a ministry) for culture, but where there also were several coordinating committees that cut across ministries. In the last decade, however, issues related to youth have been placed in a division together with issues related to children, in a ministry focused on family policies. Important youth issues, such as education and work, nevertheless now fall under other ministries. Another example of coordination through formal organisational structure can be the Norwegian governments national security and emergency planning policy. While such issues formerly had been allocated to units under several ministries (among others, the Ministry of Justice and the Ministry of Defence), with certain elements of vertical coordination through ministers and horizontal coordination through coordinating committees, they have recently been collected under one agency the Directorate for Civil Defence and Emergency Planning under the Ministry of Justice. For problems concerning a specific group within society (e.g., youth), or a specific geographical area (e.g., a valley prone to flood), it can nevertheless be difficult to find good ways to deal with such issues, if, as is the case in Norway, central government is largely based on the purpose principle. How tasks are accomplished also comes to expression through established routines in an organisation. In part, rules and procedures can relate to participation, i.e., who has the duty or permission to carry out tasks. Given that time and attention are limited, an organisation and its members cannot participate in all issues that are available to them at any given time. In some instances, matters will be handled at a low level in the department, while in other and more important instances the entire hierarchy will participate. What is more, rules and procedures can concern contents, e.g., which goals and problems are viewed as pertinent, which alternatives of

action and consequences shall be assessed, and how the choice between alternatives is made. Through existing routines, known problems can be linked with specific alternatives of action, and the choice of actions will then be based on assessing the problem definition and relevant alternatives. On the other hand, new and unknown problems, and problems involving uncertainty over which consequences will follow from each alternative, cannot, without further ado, be handled through established rules and procedures, for such can result in ad hoc routines and organisational solutions. Elements of hierarchy in an organisation will, as has been mentioned, entail that a superior can command and instruct subordinates. What then causes the subordinate to follow injunctions? Firstly, the emphasis on impersonal relationships in formal organisations can cause members to distinguish between private goals and interests on one hand, and the organisations goals on the other, and they can use the latter as the foundation for carrying out tasks. The tasks are seen as a public obligation or duty. Secondly, it can be in a members own interest to comply with organisational norms, since future promotion in the hierarchy usually is based on performance in present and former positions within the organisation. An organisational structure divided into different levels can therefore function in a disciplinary manner, i.e., it will help members to distinguish their individual decision-making behaviour from their personal opinions. Degrees and forms of dividing up labour, along with routines and hierarchy, also clearly delimit individual decision-making behaviour in public organisations. Some of these limitations are related to the capacity for action, which means that organisational members are unable to participate in all the decisions they have the right to participate in. Other limitations are related to the capacity for analysis, i.e., an organisational members ability to calculate rationally. The organisational structure provides its members with limited goals and commitments they must emphasise, it also filters which information about alternatives and consequences the members have or are able to acquire. Engrained routines for accomplishing tasks can also cause individual decision-making behaviour to be more characterised by rule following than problem solving. Even though the formal organisational structure poses clear limitation on an individual organisational members choice of actions, it simultaneously creates the possibility for the organisation to realize specific goals. Put differently, rationality at the organisational level can be strengthened through structural features: They both constrain and enable the organisations instrumental rational actions. Thus organisational theorists such as W. Richard Scott on the

basis of Herbert Simons ideas about bounded rationality linked to individual decision-making behaviour and the significance of formal organisational structure develop what they call a rational organisational perspective, where organisations are instruments for goal achievement. Coalitions and interest articulation Thus far we have viewed public organisations as unified actors, and emphasised how they make wilful choices on the basis of agreed upon or common goals. Different individuals or groups within an organisation can nevertheless be committed to different goals and interests, and the organisation or its individual parts must relate to other organisations that can have other goals and interests. We call this heterogeneity or plurality. From a negotiation-based instrumental perspective, organisations can be understood as coalitions, where each actor acts in an instrumentally rational way, is motivated by self-interest, and can also enter into coalitions with actors outside the organisation, who, according to their interests, act in similar ways. Interest distribution may be rooted in formal structures within and between organisations, e.g., it may be related to actors carrying out specialised tasks. Their resources for articulating their own interests can also to some extent be rooted in the formal structure, e.g., by elements of super- or subordination of actors and different forms of horizontal coordination. If we view an organisation, or a part of one, as an instrumental rational actor, other actors will in many respects be able to influence its possibilities for making wilful choices based upon self-interest. Firstly, other actors can set limitations on which alternatives of action are relevant. This can concern knowledge about alternatives, but also which alternatives can be chosen in concert. Secondly, other actors choice of actions can be significant for which consequences the rational actors own actions bring to bear. The consequences in relation to an actors goals or interests will thus depend on what other actors do. Furthermore, these actors can contribute with knowledge about means-ends-relations. One can clearly identify both situations in cases where governments deal with complex issues. Issues concerning asylum and immigration policies, labour market policies, environmental policies and education policies engage and affect different government ministries, although such policies, in most central governments, are primarily anchored in one ministry. Thirdly, actors cannot of their own accord lay down decision-making rules, regardless of whether they are based on maximisation or satisfisation. These decisionmaking rules will also be influenced by the distribution of resources between actors.

How then can conflicts of interests between organisations be dealt with? As we shall return to in chapter 5, this can be done in four different ways, all related to formulating and developing goals. Firstly, a dominant coalition can, out from rational calculations, choose between relevant alternatives of action and push through its own goals and interests. Secondly, the actors can negotiate a compromise between different interests, which in turn provides the basis for an instrumental rational choice based on knowledge about alternatives and consequences. This is quite normal within central governments. Thirdly, the competing goals can be addressed one at a time, so as not to be confronted with each other. Such sequential attention to goals is an example of what is called quasi-resolution of conflict, for it tones down the consideration for consistent solutions. Fourthly, goals in different parts of an organisation, or in different organisations, do not always need to be viewed vis vis each other. A great degree of specialisation and decentralisation can in fact entail that each unit has more limited goals and problems that can be addressed independent of what other units do. This way of handling conflicts of interest is often called local rationality, and is another example of a quasi-resolution to a conflict. Conflicts of interest can also be dealt with by actors who come to agreement on means. If an alternative is acceptable for everyone, there is no need to come to a decision on how potentially conflicting goals should be evaluated in relation to each other. Organisational structure, environment and uncertainty In the early theoretical contributions concerning organisations as instruments, which we mentioned introductorily, organisations are often understood as closed systems. For instance, Weber emphasised internal relations in public organisations, such as hierarchical elements, divisions of labour and routines. For him, the external relations mainly were limited to how such bureaucratic organisations were subordinated to political leadership and regulated by laws. In the post-war era it has been more common to view organisations as open systems. An organisations environment will be important, both for how its structure is formed, and for how tasks can be accomplished through instrumental rational actions. In organisational literature, a distinction is often made between environments in the sense of the inter-organisational network an organisation is a part of, the general environment, and the international/global environment. The American organisational theorist Mary Jo Hatch also divides the general environment into different types: social, cultural, legal, political, economic, technological and physical. From an

instrumental perspective of public organisations, the most important parts of the environment are those that influence an organisations possibility for achieving its goals by accomplishing its tasks. These are called the technical environment, and may concern parts of an interorganisational network, some types of general environments and, in certain cases, perhaps also parts of the international/global environment. The task environment refers to the actors in an inter-organisational network, whom an organisation particularly depends on for supplying resources, or for getting a return on the results of accomplished tasks. As the American organisational theorist James D. Thompson suggests, the characteristics of the task environment hold great significance for how an organisational structure should be designed in line with instrumental rationality. He sketches two dimensions for the environment of action: whether it is stable or shifting, and whether it is homogeneous or heterogeneous. These parameters, Thompson maintains, can provide the basis for four different types of organisational structure:
Stable environment 1. Simple structure 2. Several sub-units based on specialisation according to purpose or process principle, which each correspond to homogeneous parts of the task environment Shifting environment 3. Decentralisation to regional subunits 4. Decentralisation to sub-units based on specialisation according to purpose or process principle

Homogeneous environment Heterogeneous environment

The first category craves few resources and often makes it simple to act in accord with the environment because the environment represents predictability and often stands for similar types of expertise, tasks and interests. An example of this can be the cooperation between a ministry/ government department and a subordinate agency, or between a ministry/government department and a specialised interest organisation. The second category craves more of an organisation because it needs extra resources, both in order to partition its own organisation and to coordinate contact with the environment. A recent Norwegian example of this is when a heterogeneous department, such as the Ministry of Local Government and Regional Development, had to dialogue with many different actors in the environment, e.g., with regard to asylum- and refugee

policies, housing policies, regional- and district policies, and policies for indigenous groups, regardless of whether these other actors were public or private interest organisations. Because of the dynamic environment, the third category represents potentially greater uncertainly and even more use of resources than the first two. This can, e.g., be the cooperation between a ministry or agency at a central level and regional sub-units for education, health care or agriculture. The fourth category represents the greatest amount of uncertainty and resources, because it requires splitting up the regional sub-units. Norwegian examples of this can be the offices of county councillors or regional units of the transportation sector. The distinguishing characteristics of actors in the task environment sketched above are significant because they create differing degrees and forms of uncertainty for organisations. This uncertainty can be related more directly to properties of tasks and how they are accomplished. Based upon central elements in a logic of consequences, we can distinguish between whether or not there is agreement about goals, and whether or not there is agreement/certainty about the understanding of means-ends-relations. Such reasoning can provide the basis for sketching four forms of action with their accompanying structural features:
Goals Disagreement 3. Bargaining in a

Agreement Disagreement

Agreement 1. Computation in a

bureaucratic structure representative structure 2. Consensus in a collegial 4. Inspiration in a network structure structure

The first category represents a stable form that is relatively unchallenging and needs few resources. It can reflect homogeneity and agreement in a political and administrative leadership. As to dealing with decisions in a ministry, this can mean that a division, a section or an individual member has a clear goal, knows the means for achieving it and its consequences, thus the decision-making process is relatively simple. The second category reflects heterogeneity and divergent organisational positions. There may be agreement on a relatively general goal, e.g., reducing unemployment, while various cabinet ministers, out from different ministerial commitments, can disagree about the means for achieving that goal. The third category craves more resources, holds greater uncertainty, and potentially involves more conflicts. Take for example a cabinet and large special interest organisations; they

can disagree over which political goals are the most important but yet agree on how these should be reached. Thus there is disagreement about political priorities, which may reflect different ideological commitments, different positions and different fields of expertise. The fourth and last category represents the most uncertainty and conflict, because there is great disagreement about goals, and established knowledge is debateable and uncertain. This is often what one sees in dealings between public authorities and more ad hoc, action-oriented environmental protection organisations. An organisation nevertheless is not only influenced by its environment, but also to some degree can influence its environment. Among other things, the organisation can constitute part of the technical environment and task environment of other organisations, and thus can have significance for how these are designed. What is more, many public organisations will be included in what, using Hatchs terminology, are called different types of general environments for other organisations, such as legal and political environments. Steering through designing and exploiting leeway/discretion for action From an instrumental perspective of organisations, steering will involve influencing the relationships that are significant for achieving goals. Seen from the vantage point of an organisations leaders, this can happen partly through designing structural features, and partly through instrumental rational actions within these parameters. Through the formal organisational structure, leaders can secure for themselves the capacity for analysis and the right to participate, and in turn exploit the leeway/discretion for action these parameters provide. Also through this organisational structure leaders can regulate other actors participation and determine which alternatives of action will be relevant for them. As to the design of structural features, these can include features of bureaucratic organisational forms, i.e., degrees and forms of hierarchy, divisions of labour and routines. They can also include features of other organisational forms, such as boards of directors, coordinating committees and commissions. For instance, when reviewing a new or revised law through a commission, merely stipulating who has the right to participate in this phase of work may be done through the commissions composition, and through the commissions mandate it can be specified which goals, alternatives and consequences are deemed relevant. Also, the choice between reviewing the law through a commission or of allowing the work to be done within a

department will indicate something about who has the right to participate during this phase. Similarly, in the next instance, when the proposed law is sent for a hearing to affected parties, it is determined who shall be allowed to articulate their views, which problem definitions and which alternative solutions are particularly relevant. As to actions within the structural framework, we have mentioned that an organisations leadership does not have the capacity to participate in all decisions where it has the right to participate, and this can result in problems of control. Where the leadership does participate directly, it can build upon its means-ends insight and rational calculations to make decisions, or it can build upon its bargaining strength for advancing its interests in relation to other actors who have other interests and resources. The organisational leadership can nevertheless also influence decisions without participating directly, for other actors can base their actions on insight into what leaders would do without them needing to be physically present. These actors can, among other things, refrain from putting forward an alternative solution based upon their insight that leaders will disapprove of it. Instances of such anticipated reactions and autonomous adaptation are widespread within and between public organisations, and can be rooted in experiences of what has happened in the past. Rational calculation and political/social control We have started from a perspective of organisations as tools for achieving given goals determined by leaders or groups in a society. The possibilities for achieving these goals through carrying out tasks will thus be limited, partly by an organisations understanding of means and ends, and partly by its power to influence situations in relation to other actors who have interests connected to the same tasks. The instrumental rational ideal is a situation in which organisational leaders possesses great abilities for rational calculation and, through hierarchical authority, great control over other actors. As we noted above, situations can nevertheless be found where leaders have incomplete means-ends-understanding, and where they must negotiate with other actors who have different interests and resources. Such a view of organisations can also be related to more general characteristics of models of governance and autonomy in public administration. In what the Norwegian political scientist Johan P. Olsen calls the sovereign, rationality-bounded state, the foundation for knowledge is problematized through ideas about limitations on rationality, but he emphasises how

organisational design can turn public administration into an instrument for political leadership. This sort of governance model is marked by instrumental rational actions and hierarchical authority. In what Olsen calls the corporate-pluralist state, political and social control are problematized. Specialisation in public administration provides the basis for developing a range of interests in different organisational units, and organized groups of society are enabled to participate in shaping and implementing public policy. In such a model of governance, actors act in an instrumentally rational way, but the power base of other actors limits the political leadership. The challenges facing a public leadership vary according to combinations of rational calculation and political control. To begin with, the ideal of achieving high values for both of these variables puts pressure on leaders to expend great resources in order to develop expertise in organisations, and to often be directly involved and have a high degree of control over other participants. Secondly, the combination of placing high value on rational calculation and medium or low value on political control can create significant problems for implementing desired changes; in other words, proposed changes risk remaining on the drawing board. Thirdly, the combination of high control and medium or little insight can result in solutions that are not well reflected or which change almost nothing a situation that immediately calls for new reform processes. Fourthly, a mid to low value for rational calculation and political control can easily lead to chaotic or anarchic processes, where few problems in a society are solved and conflicts between actors increase.