University of Bergen

I. Towards a New Perspective on Budgetary Behavior

The work reported here began with an interest in budgetary behavior as a topic for political studies. The point of departure was that budgeting can be viewed as the most important decision-making process in many organizations. Slowly, however, the focus changed toward the importance of rituals and ceremonies in modern, complex organizations, and how they contribute to the formation of legitimacy, compliance and consensus. It is the thrust of this article to connect these two themes. This will be attempted within the framework of a case study of the budgetary process in a Norwegian commune.

There seems to be a widespread consensus about the importance of budgeting today. Organizations and agencies, public and private, draw up budgets, usually within a formalized set of rules governing time-limits for formulating proposals and deciding on them, which units can make which proposals and decisions, who should have an opportunity to comment on them, etc. This consensus on the importance of budgeting is not matched with a similar level of conceptual clarity and order.

The growth of complex organizations has been viewed as one of the characteristics of modern society, distinguishing it from earlier, feudal forms (Gouldner, 1959, p. 400). This view emphasizes the instrumental aspects of organizations. Budgeting in most theories and in conventional wisdom has been viewed as one of the most important organizational instruments for coordination of scarce resources in order to reach desired states of affairs in the future. Burkhead says: "Government budgeting is one of the major processes through which the use of public resources is planned and controlled" (Burkhead, 1968, VII), and similar points of view can be found concerning budgeting in private organizations. The budgetary process is often assumed to be the most important policy-making process and one giving the organization a comprehensive program for a certain time-period in the future. We are left with an image of coordination in time, and also of decision among alternative expenditures, a process through which the participants intervene with ongoing patterns in order to guide the organizations in a desired direction by transforming individual and group values and beliefs into policies.

This conception of budgeting obviously has its roots in classical economic theories of how budgeting ought to be done. We have to remember that these exercise, through which one was maximizing the joint social benefit, reaching theories assumed goal consensus, and looked upon budgeting as an intellectual exercise, through which one was maximizing the joint social benefit, reaching common goals with the lowest possible costs. That is, budgeting is viewed as a means to efficient and rational utilization of resources, a process that disciplines all planning by requiring systematic comparison among alternatives, by taking into account their whole impacts (Mosher, 1954, pp. 48—49).

Lately we have seen a reaction against these abstract, often highly formalized models of budgeting, which never refer to the organization as a social unit and do not take into account social and psychological factors. However, in the few empirical studies of how budgeting is actually carried on, describing it as political processes where participants and subunits may have conflicting goals, where coalition building and bargaining take place, and where one assumes "bounded rationality", (Wildavsky, 1964, 1968, Crecine, 1969), it has still been taken as given that the allocations of resources are the most interesting aspects of budgeting. In a way we have a variant of the general debate between people studying decision-making by means of "economic man"-models and those using "muddling through"-models. No one has asked whether we can better understand budgetary behavior if we do not take it as given that the most important things taking place are policy-making and resource allocation.

This article grew out of a dissatisfaction with using either of these two decision-making models in order to understand the data gathered in a case study of budgeting in a Norwegian commune. Both economic man- and muddling throughmodels emphasize the instrumental aspects of budgetary behavior. They both assume self-conscious people with goals, information and ability to influence the final budget (even if the requirements of goal-clarity, information, and influence are of very different degrees in the two models). Working with the Norwegian data, we found that;

(1) The size of the organizational apparatus activated during the budgetary
process clearly indicated the importance of the decision.

(2) We found, however, few participants with future directed goals, information,
initiative, perceived influence or felt responsibility for the decision.

(3) We seldom or never found situations where competing alternatives were
presented and fought for. The rule was consensus solutions.

(4) However, both the participants themselves and the environment (e.g. the newspapers) seemed to be convinced about the importance of the budgetary process as the most important (or at least one of the most important) decisionmaking processes in the commune.

It slowly became clear that we should not take it as given that we could
understand the patterns of behavior observed by assuming that the people taking
part in the budgetary process were participating in policy-making. In this situation

it seemed necessary to develop a new theory about what was taking place. This could not be done within the conceptual framework of decision-making. On the contrary, the data gathered made more sense if we looked upon the whole process as a ritual or a ceremony.

We are here confronted with an old problem in political research — the importance we will give categories provided by the society (or parts of it) in our attempts at theory building. (By and large our experience with using such concepts, most often at a low level of abstraction, has not been very good.) We may choose to (1) define budgeting as a special class of decisions with some specified characteristics. Even if we produce several alternative definitions, we may find that much of the activity called budgeting by the people participating in it, will fall outside our definitions. Or, (2) we may want, in principle, to study all the processes and activity called budgeting. Here we must be careful. When we decide to study the set of activities called "budgeting", we must not unwittingly accept the meaning given the term in conventional wisdom as well as in more fully articulated prescriptive theories.

If we do not question whether budgeting activity can be better understood as something other than policy- or decision-making, we have at least partly fallen into a trap. Given, then, that we are interested in budgets, and in the activity called budgeting, several questions become interesting. We may be interested in conditions under which a budget will have any consequences at all (that it will be implemented). We may be interested in what consequences it eventually had, consequences for the economy, for the social structure (who-gets-what, relationships between classes etc.), consequences for future behavior of the organization members (a control aspect), or consequences for the next budget. Or we may take as a point of departure a given budget and try to track down when and how each expenditure was "bounded". We may find that the connections between the formal budgetary process and the final product may or may not be of any interest, and at least in some cases it will be difficult to use "process" regarding the way the budget is made.

In this article, however, we will focus on the behavior taking place during the formal budgetary cycle. Following Leach (1968) we will take it as given that most social action under most circumstances will have many aspects other than the instrumental aspects. Here we will especially be concerned with the expressive aspects of budgetary behavior. In the same way as our actions may do things (affecting the choice at stake, and the state of the world), they may also say, or express, things (affecting the values and beliefs of the participants and the on-lookers). We therefore should not take it as given that budgetary behavior is always primarily concerned with transforming individual and group values and beliefs into policies. We may as well ask how the process itself produces, strengthens or weakens, certain values and beliefs. Since we so easily accept (at least in our part of the world) that budgeting has something to do with decisionmaking, future directed action, coordination, goal-directed activity, etc., "budgeting" will carry strong meanings in our culture. The word will have a positiveloaded,

expressive potential, and it should not surprise us if activity called budgeting produces a certain level of acceptance, legitimacy and reverence. That it affects our loyalty, and gives a certain amount of reassurance. By providing an opportunity for expressive behavior (both enthusiasm and discontentment) such a process may give a feeling of participation, provide a Catharsis-function (Lasswell, 1930, p. 195) and socialize the participants into accepting both the procedures and their products. We know that no political system or organization can survive without a minimum of such affective and emotional support, and an analysis of the expressive elements of budgetary behavior may throw some light on the values and beliefs the organization is going to serve and also the general processes of consensus-formation.

We do not believe that every social action belongs unambiguously to one or the other of two readily distinguishable categories (e.g. decision-making, ritual behavior). Rather, we would believe that budgetary behavior under most circumstances will be a certain combination of both. But then we may also think that the expressive aspects under certain circumstances may be dominant, so that the behavior of the participants in a budgetary process can best be understood as expressive activity. Analyzing expressive behavior we have to refer to its effects, and not to the intentions of the actor, or his opinions about what he is doing.

When we introduce the word ritual, however, certain problems arise. On the one hand "ritual" in conventional wisdom (like bureaucracy) often has a negative meaning, and it may be difficult to change this conception in people's minds. On the other hand different scholars give the term different meanings (certainly across disciplines, but also within each of them). Says E. R. Leach, a well known student of ritual activity

Even among those who have specialized in this field there is the widest possible disagreement
as how the word ritual should be used and how performance of ritual should be
understood (Leach, 1968, p. 526).

As a first approach to the concept (we will return to it in section IV) we will talk about budgeting as a ritual act in situations where the instrumental aspects of the process are minor, and the expressive aspects dominant for a majority of the people studied. (Since we, when using expressive act, refer to the effects of behavior, we also have to accept that the combination of instrumental/expressive aspects may differ from person to person.) Decision-making will be looked at as a subset of choices, characterized by a conscious selection between alternatives for action, based on some degree of spelling out their consequences in relation to some (more or less clearly) defined scale of preferences. In this paper budgeting will be referred to as a decision-making act in situations where the instrumental aspects of the behavior are dominant and the expressive aspects are minor. We will in the following primarily devote our interest to these two extremes. (Decision-making act: instrumental aspects dominant, expressive minor. Ritual act: instrumental aspects minor, expressive dominant.) We will seldom discuss situations where both aspects are important, or both of little importance,

or the numerous other combinations we may find under different circumstances. By focusing on the extremes, we will be more able to discuss some organizational factors affecting the probability that a budgetary process will take the character of decision-making or a ritual act.

The problems we will now deal with are:

First, we will inquire further into the way students of politics have looked upon budgeting. It will be argued that: while budgeting for a long time was treated as an a-political, intellectual exercise, it is now described as a political process, but in a limited sense of the word. Obviously the budgetary process, if it can best be described as a ritual, still has consequences which should interest the student of politics. However, he might do well to focus, not so much on budgeting as a vehicle for translating needs and wants into policies, but rather on the process as a mechanism for forming or confirming wants, needs, beliefs, and therefore influencing how people will be "prepared" to accept and support decisions made otherwise. Furthermore, within the conceptual framework of decision-making, political scientists have been more interested in structural aspects than the content of choices made, the who-gets-what-question.

In section 111 we will present the data from the case studied. First we will briefly describe the important place Norwegian communes have in the national polity and economy, and some of the values, beliefs and feelings strongly connected with local self-government in Norway. We will also sketch the setting of the case studied and show how it differs along important dimensions from other cases found in the budgeting literature. Finally, we will focus on the data, and try to show how it is difficult to understand them in the conceptual framework of decision-making.

In section IV we will first briefly discuss some main points in the literature on rituals, and then show how we may understand the data presented if we assume that the budgetary process is primarily expressive behavior for a majority of those participating.

The important question, then, becomes to what degree we can spell out some of the organizational conditions under which the budgetary process tends to gravitate towards being a ritual act or a decision-making act. This will be done in section V by first discussing two important dimensions: (a) the budget as an externally or an internally determined event, and (b) the organization as a "planning" unit which coordinates its problems and resources over time, or a "fire station" organization, a model which assumes that "organizations make decisions by solving a series of problems, each problem is solved as it arises, the organization then waits for another problem to appear" (Cyert and March, 1963, p. 119). This type of organization allocates its resources in order to solve crises in its environment as they arise rather than basing its behavior on long-range planning. After discussing in this way some organizational conditions affecting the degree to which a budgetary process will have any decision-making content, we will approach the question of which factors affect the importance of the expressive aspects.

Finally, in the conclusion we will present the hypothesis that there might be connections between how important the budgetary process is as a vehicle for decision-making and resource allocation, and who-gets-what of the resources available.

Obviously the frame chosen is a very wide one. With only one case study as background, and given the existing level of knowledge or rather ignorance of the problems raised, we have to limit ourselves to asking some questions and only tentatively proposing how we may approach them. Needless to say, we will not be "testing hypotheses", we do not have enough data for that. If this paper is successful it will not be for having answered old questions, but for having made it possible to ask some new ones.

II. Budgeting and Political Science

A surprisingly long time had passed before political scientists realized that budgeting might have something to do with politics, and should not be left to people treating it as an intellectual exercise of maximizing an assumed social welfare function. But still, we will argue, "political" so far has been given too narrow a meaning. We still have not inquired into all the "political" aspects of budgeting.

In 1961 Aaron Wildavsky seemed a little uneasy when telling the readers of a professional social science journal "the obvious truth that the budget is inextricably linked to the political system". But, he said, this fact was not yet recognized in the literature (Wildavsky, 1961, also Crecine, 1969). In his later writings Wildavsky has repeated that "in the most integral sense the budget lies at the heart of the political process", that "budgeting is the life blood of government", etc. (Wildavsky, 1964). Yet, even in 1968 he had to point out that "the intensive study of budgetary behavior has just begun" (Wildavsky, 1968, p. 1968).

For at least four reasons we would have expected budgeting to have been an
important theme for political scientists long ago:

(1) The importance budgeting has been given in the normative literature. We all know the weight of the who-has-control-of-the-purse-string argument, and the important connections seen between who has this control and the state of democracy (e.g. Cleveland, 1919, Smith, 1944).

(2) The gap between existing theories on budgeting and "reality". Traditional economic theories on expenditure-decisions seemed completely sterile. Says the economist Buchanan; "rarely has there existed a wider gap between theory and practice than that between the normative theory of the optimal budget and the way budgets are actually made" ("Public finances", 1965). And thirty years ago V. O. Key stated that "the doctrine of marginal utility . . . has a ring of unreality when applied to public expenditures. The most advantageous utilization of public funds resolves itself into a matter of value preferences between ends lacking a common denominator (Key, 1940).

(3) Budgeting is a process one finds in most organizations and agencies. We

have a recurring event, and a process that in itself produces data more easily quantifiable than most other phenomena political scientists have been interested in. Obviously budgeting would be an ideal focus for a discipline like political science so interested in being more "scientific" via more formalized and quantified models.

(4) From time to time the people in the discipline have been reminded about the connections between budgeting and politics. As mentioned, V. O. Key pointed this out in 1940, but even this prestigious political scientist did not inspire anybody to follow up his ideas.¹ The political aspects were formulated even more clearly by Mabel Walker. Writing in 1930 she stated the problem in this way:

The budget maker faces a tremendous problem. He must balance municipal functions, one against the other. Whether consciously or not, he determines whether he shall permit a few houses to burn down, pickpockets to thrive, babies to die, traffic accidents to result, paupers to suffer, illiterates to remain untaught, or children to miss their chance to play. All of these evils and many more are present. None of them can be entirely stamped out, but within limits all will diminish as appropriations to fight them are increased (Walker, 1930, p. 10).

And about solving the problem:

Fantastic problems never to be answered, the average person might say. But they are answered every year in every city of the country. Blindly and unconsciously it is true, and in every city a different answer. But however confused and crude the methods may be, a decision is reached (Walker, loc. cit.).

But in spite of these statements, and a later one telling us that allocations of resources are made in an almost complete absence of the evidence which would be necessary to validate them (Simon, 1965), the notion of the budget as a valuefree tool without political aspects has had much greater impact on the thinking and writings (and action) on budgeting than the view of, for instance, Mabel Walker. Recently we have seen attempts to present program budgeting as primarily a question of technique and efficiency (Novick, 1965, pp. 258 and 370, Churchman and Schainblatt, 1969, especially pp. 179—181). Mosher comments (1969, pp. 160—161): Many seem to be back in a search for "the one best way" and dreaming about a millenium for rationality and efficiency.

Attacking this latter view and the built-in assumptions about continuous competition at the margin among all alternatives for allocation of available resources, people like Wildavsky and Crecine have wanted to contrast traditional "normative" budget-theory — looking upon budgeting only from a narrow efficiency perspective — with a description of how the numbers in the budget are actually arrived at. That is to create a "'positive" theory of budgeting as a political process, a mechanism for translating financial resources into human purposes (Wildavsky, 1968, p. 192). Wildavsky also argues for his theory as the better one for prescriptive purposes, and in this way joins other writers (e.g.

Simon, March, Cyert, Lindblom, Braybrook, etc.) in their general attack both on the descriptive and prescriptive value of the classical "economic man" model assuming "total rationality". The main question is which model is the most fruitful one (both from a "positive" and "normative" point of view, (a) a model assuming total rational comprehensive and simultaneous evaluation of means and ends, calculation of sets of alternatives and their consequences (at the margin) against a well defined, given scale of preferences, or (b) a "muddling through" model assuming that budgeting is incremental, fragmented and sequential, that decisions are made without clear, common goals, and that reliable information about alternatives and consequences are often lacking.

Wildavsky and Crecine, however, still seem to assume that the most important thing taking place in the budgetary process is decision-making or allocation of money.² This has obviously been true in the contexts they have been studying. But we may also argue that their conception of "political" becomes too narrow if they want to build a more general, political theory of budgeting. Two points should be made:

(1) There is an (apparent) lack of interest in the content of the decisions made. In the writings of Wildavsky we can find a change from an interest in the "who-gets-what" aspect (Wildavsky, 1961) to a concentration on the strategies and calculations of the participants, and on the structure of the choices made (Wildavsky, 1964, Davis, Dempster, Wildavsky, 1966). In many ways Wildavsky is giving us a theory of the politically strong. Allen Schick (1969) says Wildavsky seems to have forgotten that budgets have social consequences, and that there are losers. In Schick's opinion Wildavsky does not evaluate the whogets-what aspect because he is taking the view of the "pluralists" in American social science. That is, he is accepting the content, and that the process is working properly, because he views it as a result of a bargaining-coalition-building process where all interest groups are participating.3

(2) But as proposed in this paper, budgeting under certain circumstances might not primarily be instrumental activity, transforming values and beliefs into policies. If we find that the expressive aspects of budgeting affect which values and beliefs an organization is going to serve, and its basis of legitimacy and consensus, this would not make budgeting of less interest for the student of politics.

In important ways the expressive dimension would be a supplement to the literature on budgeting as decision-making and make the frame of analysis wider. We now turn to the case studied in order to illustrate some of the points made above.

III. A Study of Budgeting in a Norwegian Commune

The puzzle we are going to present in this section is made up of three types of data, which together cannot easily be understood within the conceptual framework of decision-making. First, we will give some indicators of the importance Norwegian national law makers have given local budgeting, requesting a thor

ough treatment of each part of the budget by activating a large number of laymen in committees and boards. The laws also give a detailed set of rules which govern how these units should interact. Secondly, the data from the case study clearly indicate the "difficulties" of finding decision-makers and real choice situations. Thirdly, however, the attitudes and perceptions of the participants themselves (and some on-lookers) again indicate the importance of the budget decision as a means for finding the "right solutions" concerning how available resources should be allocated. Before we go into this puzzle in more detail, we have to give a brief introduction to the important role the communes have in the Norwegian economy and the polity, and give a short presentation of the commune studied.

(a) The Importance of Local Self-Government by the Layman

A firm belief in the value of local self-government is a very important part of the democratic creed in Norway. When the communes got their "Constitution" in 1837 it was seen as a strengthening of the local units (vs. the national government), and the layman (vs. the bureaucracy).

The communes have been looked upon as "schools in democracy", and empirical studies have shown that training in local councils, boards and committees has been one important route to the national political leadership (Valen, 1966). The communes have been very reluctant to create full-time administrations. In 1965 Norway had only 3,796 persons in administrative positions in the 20 counties and 451 communes. The layman, with his part-time activity is strongly assumed to "govern" and administer the communes in Norway. However, the number of full-time employees increased 60% in the period 1951—65, and the prognosis is for an increase of 6 % a year in the coming period (InnstilUng om utdanning av tjenestemenn for administrasjonen i kommunene m. v., Bergen 1966, P- H9).

It is widely believed that Norwegian communes today have more functions and control more resources than in most other countries.⁴ The local units can do everything not explicitly prohibited by law. This gives them a very broad frame of activity, and the tendency has clearly been to take up issues which cannot be defined as purely local. The activity of the communes has traditionally been based on local taxes on income. In the 1960's the communes have received approximately 2/3 of all direct taxes (St.meld. nr. 54, 1960—61, p. 66). In 1968 the net budgets reached 5.9 billion Norwegian kroner, or 1,566 kroner ($ 1 = 7.15 N. kroner) per inhabitant, (St.meld. nr. 58, 1968—69), which was a 16 °/o increase from the preceding year. While the national income in the period 1947— 63 increased 223%, the communal net budgets increased 410% (St.meld. nr. 15, 1963-64).

These figures might give the impression of a vital local government. However, from time to time, we are told that, while it is correct that much money is channeled through the communes, the local units no longer have control over how this money is used,⁵ that today the communes are only administering national

policies, setting minimum levels in most sectors (e.g. schools, general welfare, etc.). Each autumn, at least when the local councils have to make the budget decisions, we find the newspapers overflowing with statements that the commune are "tied up" by the national government and the provinces.

Certainly there is also a tension between, on the one hand, the general picture we get by reading the non-empirical (and mostly prescriptive) literature on Norwegian communes (Nordanger and Engh, 1960, Heltne, 1965, Engh, 1968, Tarjem, 1969) — and on the other hand, the points of view found in empirical studies on local decision-making (e.g. Christie, 1964, Olsen, 1967) and reports from people having central positions in the administrative apparatus.⁶ Without doubt the literature based on deductions from existing rules and laws are the best known, and here we learn (1) that local government is important, (2) that we have local government in Norway, (3) that budgeting is the most important single decision made in the communes (also that this decision, therefore, is taken by the local council, chosen directly by the people), (4) that budget makers should balance all demands for resources against each other in order to find the "right" use of the taxpayers' money. It is not difficult to recognize the impact of the classical theory of economic choice.

(b) The Case Studied

The idea behind the case study was to look upon budgeting as a complex decision, a stream of premises (Simon, 1965), and to examine the potential bargaining and coalition-building taking place. The intent was to place the study within a now well-known tradition: Focusing on a key decision, and trying to understand the content by studying the processes (Cyert and March, 1963). This was done by using all written material, having structured interviews with all the 91 participants in the formal budgetary process, and less structured interviews with the people who seemed to be the most central in the decision.

The commune studied is near to the capital. In 1966 it had 8,000 inhabitants, having doubled in the last 20 years. 46 °/o of the work force was employed in the capital. The local council had a socialist majority of 18, while the nonsocialist parties had 17 representatives. The net budget was 9.2 million kroner, gross budget 14.7 millions. Traditionally the full-time administration had been very weak. In 1939 the commune had only 6 employees, 1950 12, and in 1966 34. Only 15 of these participated directly in the work on the budget.

While the size of the budget was modest, at least compared to those reported on in American studies, the complexity of the choices involved was considerable. The budget makers in most Norwegian communes are faced with an increasing number of alternatives for allocation of resources. Also, each alternative has a growing complexity as one leaves the purely "local" issues. At the same time we find here few persons in full-time positions, very little data, and in general a low degree of knowledge about social, political and economic processes taking place. One of the reasons for choosing the commune studied was that it was an expanding unit with a number of alternatives for the allocation of its resources.

At the same time, we knew that there was some degree of local discretion remaining
(everything was not bounded from outside). In the following sections
the commune will be presented in more detail.

C. The Data (or the presentation of a puzzle)

Cl: The National Law Makers' Point of View: Norwegian communes make one-year budgets (following the Calendar), and basically the handling is the same in all communes. The set of rules governing the process reflects on the one hand the perceived importance of the decision, on the other hand the wish for layman control.

In the preparation phase a full-time manager for each sector of the budget (if the commune has such a person) puts forward a proposal to a layman committee or board. In the commune studied, 25 committees with from 5—20 members are activated each year. It is these laymen committees which make the formal decision about what the proposal should look like, and it is their proposal which is sent to the next level. Here we enter the recommendation phase where the municipal manager (only communes with more than 10,000 inhabitants have to have such a manager)⁷ gathers all proposals, and prepares a recommendation to the executive board, with its 9 laymen (chosen by and among the members of the council). Again it is this layman unit which submits the final recommendation to the formal decision unit, the local council.⁸ All these units are required by the Communal Law, which also states their spheres of competence. Both the law and the normative literature emphasize that since the budget is the most important decision taken in the communes, it cannot be delegated from the Council to other units. This again is closely connected to the fact that the Council is elected directly by the people. Furthermore the Communal Law requires that the budget proposal be made public at least 8 days before final adoption. These 91 people, only 15 of them full-time employees, will in the period from April to October Ist draw up the budget. Many of the laymen will participate in more than one committee. 20 of the 35 members of the Council also participated in the preparation phase, and 5 laymen participated in all three phases. In the following we will call this apparatus activated in order to get a decision, the budget-organization. Both indirectly — by stipulating the number of units activated, the number of meetings arranged, and the number of man-hours devoted to the decision — and by more direct statements, the national law makers clearly indicate the importance of local budgeting.

C2: "Where Have All the Decision-Makers Gone?": We will now try to compare
this picture of the budget decision with some data on what actually took
place (and did not take place) during the work with the budget.

First we analyzed the distributions among the 91 members of the budgetorganization of (a) goals, (b) information, (c) initiative, (d) perceived influence, (e) time devoted to the budget, and (f) who felt responsibility for the product.⁸ Certainly it is not difficult to show that the future directed, decision-maker

model was not a very good description of the plurality of the participants. The
first report from the case study concluded that one could describe three main
types of participants:

(a) "Experts on distribution": Only five persons were included in this category. These people were all focusing on the budget as a whole and as a problem of distribution of resources between sectors. This small group was characterized by devoting much time to the budget, participating in the preparation, recommendation, and formal decision phases. They were well informed, discussed the budget with many people, and often gave information to others. Especially they played an "interpreting" role between the three phases because formalized, written communication was very bad. They also took care of most of the contacts with the environment, and held opinions about how the commune should be developed. The elected representatives had long periods of activity within their parties, had been living in the commune longer than the rest of the participants, and seemed to organize most of their time around political activity. For the employees in this group obviously the formal position in the budget-organization was the most important factor.

(b) The "advocates": This was a somewhat larger group of people, (approx. 20) not interested in the budget problem as a whole, but in obtaining as much as they could for the sector they themselves were working in. The group consisted of employees and elected representatives participating in two phases in the budget work (and 2 or 3 in only one phase). They knew their own sector very well, but had a low degree of knowledge about other sectors. They were active, took initiatives and were often giving information to others working in the same sector. Most of them were also active in their parties. They had goals connected only to their own sector. Even if they devoted less time to political activity than the group above, local politics took a lot of their "free time".

(c) "Democratic accessories": In the work with the budget, this group, which included approximately 75 % of the participants, were onlookers. Most of them participated in only one phase, and had a low degree of information, activity, perceived influence, and responsibility. They seldom had had any office in the parties. They had participated in local politics for a short time period. It should also be noted that their possibilities for using a veto were reduced by the fact that the commune has to make a decision before October Ist, and also that this group got all its information from the persons it eventually should veto.

This picture should be supplemented with the fact that the decision on the budget seems to be an "internal affair" for the budget organization. Only 5 out of the 91 respondents reported that they had discussed the budget in any organization (even if nearly everybody were members of several such organizations). The interest groups did not activate themselves to any extent. This point of view is also supported by the fact that nearly all criticism against the budget came from people already inside the budget organization, and that nobody could remember having got any reaction during the 8 days the budget proposals had been made public. (It is also interesting from our point of view that nobody has

seen this fact as a problem, but follow the rules year after year.) The data so far, however, have only indicated what dozens of community studies tell us, namely that the number of decision makers are not as large as we would like to think they are in our democracies. Since this is not the main point in this connection, we should take a somewhat closer look at those few with goals, information, initiative, etc., who are assumed to make real choices.

The municipal manager is by a large plurality assumed to be the decision maker, and it is therefore of some interest to examine the major changes he actually made (in 1966). The reasons for the major changes are: (a) the commune has to take into account earlier decisions (e.g. concerning new positions), (b) some changes have to be made because the province has taken over some schools, (c) there have been some changes in the taxes to the province, (d) some of the money proposed to be set aside as a reserve for future school buildings, etc., was cut off. (It can be discussed what differences this makes, because all reserves will automatically be cut off at the province level if the budget has not been brought into balance.)

These brief examples should illustrate that the general manager as well as being looked upon as an omnipotent decision maker can be viewed as the central memory of the commune, keeping track of all the very complicated relations between the budget and the national government, the province, earlier decisions made in the commune, etc., and reminding the others about the consequences of these factors. When we look at the smaller changes the manager made, he only once used a future directed argument in his comments on the budget. In 23 cases he gave no reason other than the budget was too tight. When the manager does not find any need for giving his reasons and the other participants accept this, we may (as we will comment below) assume that this is because there is really not much discretion left, even for the manager.

In the same way we could analyze the proposals from the managers in the different sectors. Like the municipal manager they seem to have an important role as "memories" for their sectors. They look upon their proposals as products of last year's budget, changes following new laws or regulations, increases in the number of clients in the sector, and of other decisions throughout the year. In this way they also will know pretty well what it is possible "to get through". Certainly they will ask for a little more, but the cuts made seldom produce any surprise. When the people participating in the preparation phase indicate that they know what it is possible to get, and the people making the recommendations and the formal decision say that they cannot change much because the work in the first phase has been so thorough, this has to be viewed in close connection with the general lack of decision-alternatives in the process.

Most clearly this is indicated by the fact that all decisions in the preparation phase were unanimous. So were the decisions of the executive board, while two representatives voted against some smaller appropriations in the Council. In general we found a low legitimacy of conflict both along party and geographical

lines, (often justified by the participants by saying that the size of the commune
was too small to permit such conflicts).

What do people do, then, in the many, and often long, meetings, if they do
not change or try to change the proposals discussed? Three major patterns
should be noted:

(a) The first pattern has many versions, but common 10 them all is an emphasis on verbal activity connected with important parts of the value and belief system in the communes. Very seldom do debates of this type affect the issue at hand (the budget proposal). Sometimes the actions will be quite the opposite of the expressions. Often the most time-consuming part of budget meetings is the debate about the conditions in general for local self-government today. An important aspect of this debate is the fact that one very seldom makes any attempt to define what one means by local self-government. For instance, one very seldom takes up certain programs that the national government demands, but for which the commune would not otherwise allocate resources. The general fuzziness of the concept of local self-government today makes it possible for both sides on important questions to say that their solution would strengthen local self-government.¹⁰ A second "big" question in the debate is the relationship between the bureaucracy and the layman. While voting for new positions for fulltime employees, the members of the Council especially will tell each other about the dangers involved in hiring more full-time people. This debate is also often connected to the question of revision of wages of the employees. A third version centers on the relationship between the Council and all other bodies. The Council members will state that they have got too little time to work with the budget, and that they — who should make the real decision — can say or do nothing when the budget proposal reaches them.

(b) The second pattern, could be called the floor-carpet syndrome, and is based on the fact that "the big issues" can be discussed but seldom — or never — changed, while some smaller issues can be both discussed and changed. And many representatives have their "pet issues" which they are very fond of. The result is that we often get a strong negative correlation between the amount of money involved in an issue and the time devoted to it. (As C. N. Parkinson has proposed, this seems to be a very general phenomenon.) A very popular (and time consuming) version of this debate is when someone discovers that some communal office has bought a floor carpet which is some kroner more expensive than another one which the representative thinks is good enough.

(c) The third pattern we may name the "not now but later"-syndrome. While most participants seem to accept that they have to proceed without real alternatives regarding how the money should be allocated, we find lots of expressions like: "something should be done about this problem", "when we try to find a solution we have to take into consideration that", "someone (alt. we) should look more closely at this problem", "we should be given an opportunity to take a new look at this issue", etc. The important aspect with statements like these is that while they are not directed straight at the choice at hand (proposing

alternatives, mobilizing support, etc.), they can be viewed as attempts to put
some constraints on future decisions, by signaling support for certain points of
view, values or social groups.

We will, then, try to summarize the argument of this paragraph. We have given several types of data indicating that decision-making does not seem to be a very important aspect of what is going on. It is very difficult to understand the mobilization of all these people and committees each year from a participation-in-policy-making point of view. In section V we will argue more generally that communes with an organization structure like the one studied will be forced into a fire station type of activity. In this way resources will be allocated throughout the year, and the formal budgetary process will be (more or less) empty of decision content. By presenting, now, the data on how the 91 respondents themselves look upon the activity they participate in, it will be possible for us to propose an alternative model of budgetary behavior in communes like the one studied (Section IV).

C3: The Participants' Point of View: The work on the budget was surrounded by certain opinions about the meaning of the activity the participants were taking part in, how they ought to behave and why. We will now see that for a majority these opinions were closer to the lawmakers' point of view than the data we have presented above, but also that there exist some interesting differences in these opinions.

The importance of the budget decision was emphasized by a very strong majority of the respondents. 46 °/o answered that the budget decision was the most important decision made in the commune, 50 °/o said it was one of the most important, only 4 °/o felt it was just a decision like many others, and nobody said it was a decision of little importance.¹¹ This sense of importance is apparently also accepted by Norwegian newspapers, which give in detail reports from the budget meetings.¹² These reports not only tell how the money is allocated, but also give the opinions stated by the different participants.

The respondents' answers on the importance of the budget decision should not be taken at face value. A certain dualism can be illustrated by the fact that while a strong majority emphasizes the importance of the budget decision, the same people also say that the commune is all "tied up" from outside. When, however, we asked to what degree it made any difference which party had the majority in the commune, that is, if any party would have freedom enough to make a policy of its own, a strong majority of the people with an opinion about the question, felt it would make a difference, and that the majority party could make a policy of its own.¹³ One interpretation of these data is simply to recognize the different needs any politician has. On the one hand he needs to rally support and win elections (it makes a difference), on the other hand he wants to avoid criticism, when he has to explain why x and y did not get the money they asked for (— we are all tied up). These different needs very often are reflected in the way the mayor (or the manager) in most Norwegian communes

describes the budget situation when he is interviewed by the local newspaper. Certainly he will state that the commune is all "tied-up", but just as surely he will say that after all, the budget is not that bad, and that all important interests in the commune have been taken care of.

We do not think the use of these different "explanations" are the results of conscious game playing by the respondents, but more a product of what frame of reference our questions are evoking. The feeling of the importance of budgeting is closely connected to the feeling of expertise among the respondents. When asked about the degree of criticism from outside, the most common explanation for the lack of such criticism was that people outside the budget organization did not really understand the problems. Outsiders could ask for more money for their own fields of interest, but they would not have any insight into the problems involved in balancing demands and resources, etc. Even among people on the periphery of the decision, who realized that they were not very important decision-makers themselves, we found a feeling of expertise when these respondents compared themselves to people outside the budget organization.

How, then, should a budget be drawn up? The rational decision-maker planning his future was a surprisingly strong ideal. The elements of classical theory of choice, and the prescriptive literature on Norwegian communes, were reflected in the demands on committee members and the managers of the different sectors. They should follow Pigou's commandment and evaluate what they requested for their own sector in light of the needs in all other sectors. These opinions were strongest among the majority party in the Council, where 76 °/o held these opinions. Only the employees had another decision-model. 73 °/o of the full-time employees felt that a committee member and a manager of a sector had to fight in order to get as much as possible for their sectors. Others had to weigh the different demands.¹⁴ The elected participants' emphasis on "rational", non-conflict solutions seems to be an important part of the socialization in the budget organization. Among the people who had participated for less than five years, 33 % strongly stressed the importance of reaching consensus concerning the budget as a value in itself. Among those who had been involved in local budgeting for five years or more, the percentage was 52. Wildavsky also has argued (1964, p. 6) that participants in a budgetary process after a while are socialized into certain roles. For the plurality of the participants in the case studied, they are socialized into the onlooker's role. They are not very well informed onlookers, but they seem to be content to believe in the ideals described earlier, and that these ideals are realized. Some differences between the participants should, however, be mentioned. Asking the elected members if they wanted any changes in the way the budget was made, 62 °/o (N = 13) of the people most central in the process¹⁵ had proposals for changes, while only 33 °/o of the least involved (N = 63) had any such proposals. Most of them were of a very minor character.

The demand that the Council should have a powerful voice in all local matters
has always been an important part of the democratic creed and local self
government ideology in Norway. The perceptions of the real importance of

this unit today, therefore, can be used as an indicator of the degree to which different participants believe that the ideals found in the prescriptive literature are realized. In general we find a strong negative correlation (—.614) between degrees of "centrality" in the decision (as measured by our index) and belief in the importance of the Council in the budgetary process. The participants giving it the highest degree of importance are the committee members who are not members of any party, and who are participating in the least important committees. We may assume that these people are most like the "private voter" in the commune, and it is therefore interesting to note that this group believes most strongly that the ideals found in the prescriptive literature are actually implemented. This picture is also consistent with the image one gets through the newspapers, where the debates in the Council get most publicity.

The fact that the Council seldom or never makes any changes in the budget proposals was explained by a majority by referring to the thorough preparation the budget gets in the committees.¹⁶ And the same pattern is found in the explanations given by those who said they did not take any initiative. In general the least active participants seem to feel that things are done in a proper way, related to consensus values in the community. When this group "explains" why they do not have any influence, they often emphasize that this is because they have not done their homework (read budget papers etc.). But still the elected representatives feel that they as a group have control. Again — that the ideals in fact are realized — 61 % of the Council members and 80 % of the members of committees and boards say that the elected representatives have most influence over the budget (compared with the full time employees), while only 25 % of the employees share their point of view.

Consider finally some attitudes towards and perceptions of party politics in the work on the budget. 75 °/o (N == 75) of the people participating on the proposing level said that there was seldom or never any party politics in the committees ¹⁷ (Party politics defined as discussions and voting along party lines.) In the Council, however, 69 % (N = 35) said that discussions and voting often followed party lines, 23 °/o said this happened sometimes, and only 8 % answered seldom or never. When checking back, we found that with very few exceptions all budgets had been passed unanimously in the Council. The important thing is that while the respondents say that the Council is the least important unit in the budgetary process, and that very seldom can changes be made when the budget reaches this formal decision level, the participants are left with the image of being involved in politics. Since it is very difficult to relate these perceptions to actual changes in the allocation of money taking place, we will assume that the expressive, verbal activity has some important impacts.18

Summarizing we find that the respondents are convinced that the budgetary process is an important decision, and that they are left with a feeling of expertise. Emphasizing rationality, consensus and a strong role for the layman, a majority seems to be content with the way budgets are drawn up. The more the participant is on the periphery of the decision, the more content he seems to be,

and he believes that the prescribed ideals are in fact realized. While in this way we find a close connection between the importance given, and the ideals prescribed by national law makers on the one hand, and the perceptions of a majority of the respondents on the other, it is very difficult to locate decisionmakers and actual choices in our data. In the next section we will propose an alternative way of looking upon local budgeting, and see how this framework, better than a decision-making model, accommodates the data given above. We feel that the data presented support the hypothesis that while the decisionmaking content of the budgetary process might be a minor one, the activity taking place may have impacts on how some people are thinking about the process.

IV. Budgeting as a Ritual Act

When we try to clarify what we mean by ritual, however, we have to realize that very little is known about the importance of rituals and ceremonies in "modern society", and that the level of conceptual consensus and order is obviously no higher concerning rituals than budgeting. The somewhat strange fact is that we have not really accepted that rituals and ceremonies are important phenomena, worthwhile to study also in modern, industrialized society. We have recognized that the transition from feudal to modern society was followed (in most interpersonal relationships) by a change from use of brutal force to a relatively higher degree of manipulation of the conditions under which other people are choosing. And, we have also realized that while man needs material benefits, he also needs symbols.¹⁹ However, students of politics in modern societies have so far not paid much attention to the way in which rituals and ceremonies affect our thinking and feeling about different phenomena, and the way we act (or do not act) towards them. Existing theory in this field is basically found in social anthropology and psycho-analysis. Here the importance of raising these questions is well recognized, but today one still has to say that "the problems in the social and emotional effectiveness of symbols are immensely complicated ... (and that we) ... do not yet know many answers clearly" (Gluckman, 1965, p. 252).

Let us for a moment engage in a little "name-dropping". Confucius stated that "ceremonies are the bond that holds the multitudes together, and if the bond be removed the multitudes will fall apart" (quoted from Sahlins, 1968, p. 12). Plato said that each system of social control had to be founded on a "royal lie" in order to work (also, see Thompson, 1969, pp. 14—15). Social anthropologists have argued that ritual activity is related to the inherent conflicts within social structure (Gluckman, 1965, p. 258). We have been told that there can be no social order without the "mystification" of symbolism, and that this mystification makes it possible for the social order to survive (Duncan, 1962, Cohen, 1969). At least some argue that such statements should not only be seen as relevant for more primitive societies, but that rituals, ceremonies, and symbols are as important in modern, industrialized societies (e.g. Cohen, op.cit.).

Also among political scientists it would (probably) be accepted that every regime and institution surrounds itself with a body of dogma, self-justification, mystification and glorification, and that no regime or institution can be completely without affective and emotional support. R. A. Dahl says (1961, p. 112):

The use of ritual, ceremonial and pageantry to give legitimacy and propriety to important actions in the life of an association is so widespread in human societies that it should astonish none to discover that modern democratic societies have also created their own ceremonials. Nor should it surprise that a community where the democratic ethos is powerful and traditional the rituals take predominantly democratic aspects.

Edelman and others have proposed that the most conspicuously "democratic" institutions and the most cherished forms of popular participation in government are largely symbolic and expressive. One hypothesis is that while the strong, organized interests get the material benefits, the non-organized gets symbolic reassurance.

Thus the importance of rituals was fully recognized in ancient times. Today students of primitive societies are also working with the ritual problem, and it is not totally unknown to students of politics in industrialized states. However, we do not have any integrated body of literature on the importance of rituals and ceremonies in modern organizations. Assuming that they are an important device for social control, (and production of consensus), and that such activity is located in the unconscious or semiconscious part of the mind, it should not surprise us that we have been more able to identify and analyze the importance of rituals in "primitive" societies than in our own. We are ourselves caught up in the same system of symbols which we are trying to decode (Mannheim, 1968).

The most relevant literature seems to be based on two different (but probably closely connected) sets of assumptions: (1) The importance of rituals for giving the actions of leaders legitimacy, and as a "cover" for the procedures actually used. By evoking and reinforcing attitudes and responses in their audiences leaders can be viewed as intentionally using rituals to prevent others from being rational in political life. But we will assume that rituals may have the same consequences even if not so intended by leaders. (2) The other set of assumptions is based on the cognitive and affective needs of the followers. Here one may view rituals as the only way people who are not in a position to analyze a complex situation can adjust themselves to it, that is through stereotyping, oversimplification and reassurance (Edelman, 1967, p. 40, also Campbell et al., 1964, ch. 9). Man will instinctively try to find meaning and order in a confusing and ambiguous situation, and one "solution" is to "believe" that leaders can deal with uncertainty, the complex and the threatening. We know that other societies "used" the rain-dance, the Oracle, the candle, mass and bell, etc., but we have not asked if the plan, the budget, the committee and certain types of social research could be analyzed from the same perspective, as producing an illusion of order and rationality and therefore giving some kind of security (and preventing a comparison of certain beliefs against reality). We also have to remember that

participating in a decision may fulfill personality functions that are not directly related to the issue ostensibly to be decided (Dror, 1968, p. 79). By participating in a value expressive way an individual may become able to meet his emotional needs. His reward is establishing his self-identity and self-importance, and confirming his notion of the sort of person he sees himself to be (Katz, 1960).

Thus, we will talk about ritual behavior when an activity is calling attention to the participants' relatedness and joint interest through the expression of shared feelings and attitudes. It is value expressive activity, that is, activity focused more on making it clear what kind of persons the participants are, than changing the world. And it is ego-defensiveness, that is, activity protecting the individual from certain kinds of situations that are (real or assumed) threats to him. We are here again referring to the effects of the activity, not to the intentions (or in other ways to the cognitive apparatus of the actor). An "external" characteristic of such ritual behavior is that the participants are handling symbols more than their empirical referents. Ritual activity, however, has to be distinguished from other types of treatment of symbols by the fact that the concepts used are often fuzzier, more difficult to define, and more difficult to connect to the issue at hand, and to the material rewards at stake. Some would say that ritual behavior cannot be justified by a rational means-end type of explanation and that it does not have any technological consequences.²⁰ A ritual act, as characterized here, is in most ways the antithesis of the concept of budgeting described earlier, which emphasizes the instrumental, the rational, the conscious, the strategic, and the need of theories of the world and clear-cut concepts to handle it.

Returning to the data, we may relate them to Edelman's (1967) prediction that in political processes where the ritual aspects are dominating, we will find: (1) Repeated and well publicized attention to significant problems which are never solved, and (2) noisy attacks on trivia. It is of some interest to look at the general debates on local self-government, lay or bureaucratic rule, etc., on the one hand, and the "floor carpet" debates on the other hand, from this perspective. We would propose that these types of debates (the first on a general level and very difficult to relate to the budget allocations, the second too narrow to really affect the values the organization is going to serve), primarily legitimize resource allocations on the basis of yesterday's values — often not made explicit. The third type of debate — ("something should be done — not now, but later") may give certain corrections on how resources are allocated in the long run.

We feel that these debates together may give the participants an opportunity to emphasize their support of important shared values such as local self-government, democracy and lay rule, rationality and order, etc. Especially the many statements about the national government putting some of these values in danger by tying up the communes, can be assumed to support a feeling of relatedness against external forces. But we may also see how these patterns may accommodate the ego-defensive needs. A major function (probably in any) society is to shorthand the functional act or acts symbolized., to evoke previously

socialized memories of important group activities. Ritual symbolizes relevance. The more individuals are removed from the normative structure of budgeting (that is involved in activities which violate normative conceptions of what the process should be), the more important it is, the greater the need, to engange in activities which reinforce the concept of their own relevance. Thus to coin a phrase, if trivia in the budgetary process did not exist, it would have to be invented. Here we have to confront two theses of opposite directions. On the one hand people are assumed to participate in local government because it affects their everyday life. On the other hand people should avoid participation in local government because it provides immediate evidence of the effects of their activity. By giving an opportunity for debate, criticism and a few minor changes in the proposals, the budgetary process may provide an image of decisionmakers protecting the set of values mentioned, the tax-payers' money, and guiding the community towards a desired future in spite of many difficulties.

We will hypothesize that there are basically three different roles involved:
(1) The leaders (earlier called "experts on distribution" and "advocates")

who can be assumed to run the commune on a "fire station" basis, will, through the budgetary process, get legitimacy for the allocations made throughout the year. Drawing up the budgets they also find out what they (and others) really have done, and discover the consequences of different decisions made earlier. In a way they are the high priests in the ritual, and to a certain extent they probably also are reassured about the important role they play, leading the commune.

(2) The secondary leadership, earlier called "democratic accessories", has an opportunity to put some constraints on and propose corrections for the way the commune is governed in the long run. In the budgetary process they are participating in a very meaningful activity, connected to important values in the communities they are recruited from, and they can do it without devoting too much of their time and energy to the job. They feel that things are well done, and use this as an "explanation" for not being more active.

(3) The third groups are inhabitants in the commune (especially those not able to define "fires" throughout the year, and get their problems solved in this way). Wc have seen that this group gives few or no inputs to the budgetary process. Our hypothesis is that in most communes, most of these onlookers have a very vague picture of what is going on, but that the apparatus activated gives them a certain reassurance that the elected people and the employees do the best they can.²¹ We have only been able to illustrate this point indirectly by showing some attitudes among those people mostly on the periphery of the budgetary process (and therefore most like the "private voter").

We believe that this way of interpreting the data from the case study is roughly correct. This framework makes it possible to connect all three types of data mentioned above. But while the national lawmakers demanded a strong lay organization in order to make decisions on how the resources should be allocated, we propose (especially for future students of budgeting in Norwegian communes)

that the major effects of this relatively large budget apparatus may be quite different from those presupposed. Obviously we do not have data which could "prove" our hypothesis. The study was designed for quite different purposes. In the next section, however, we will try to strengthen our arguments by discussing more generally which aspects of the budget organization make the budgetary process (a) empty of decision content, but (b) effective as a means for affecting the way the participants and the onlookers think about the process.

V. The Relevance of the Organizational Context

If one, when talking about building a theory of budgeting, understands it as a theory about the activity called budgeting by the people participating, it will be necessary to inquire into the conditions under which the instrumental and/or the expressive aspects of behavior become important. In this section we will try to make a modest contribution to this goal. We will attempt to generate more general statements regarding: (1) under which organizational conditions a formal budgetary process like the one studied will be of any interest for those who want to study how the organization allocates its money, and (2) under which conditions we by studying this process will primarily find data which will enable us to say something about how the organization is kept together, how consensus, legitimacy and compliance are produced.²² First we will discuss some organizational conditions under which the budgetary process will be important or not important as a decision-making process. The second question, under which conditions the process itself will have a major impact on how people think about the process and its outcome, is more difficult to answer. We will at least try to mention some dimensions which seem relevant for the case studied. And especially we will be interested in how it can be possible for a process to have expressive effects if it has no instrumental importance. Our point of departure will be some aspects of the relationship between the budget organization and its environment.

Theories focusing on budget makers, their goals, strategies, expectations, etc. primarily explain the budget as an internal event. One assumes that internal decision-makers are able and willing and that they have both the power and the knowledge to intervene in ongoing activities, so that they can by means of their intellect, guide the organization towards a more desirable state of affairs in the future (as compared to what would happen if they did not intervene). Everybody does not attribute such an active role to internal "budget makers". On the contrary, several models assume that they are passive instruments of little or no interest, and that budgets can best be explained as "external" events. We may just mention (a) economic theories assuming the effective operation of an external market, making studies of the internal operation of the firm of no interest, (b) most pressure group theories, using their simple force models, and viewing policies simply as results of external pressure groups (some assuming, however, that the agency or organization studied can be treated as just another pressure group), (c) the Floyd Hunter school of community studies, and other theories assuming that decision-makers within an agency are only a cover for the real

decision-makers, the backroom boys, (d) a growing literature on spending in American states where resource allocations are better explained by socio-economic characteristics of the society than by variables connected to the political system (Dawson and Robinson, 1963, Dye, 1966, Hofferbert, 1966). These studies also receive some support from research on spending in Norwegian communes (Nordbø, 1968),²³ (e) Norton Long's "ecology of games" model also assumes that coordination is mainly an ecological phenomenon, and not a product of conscious, rational contriving of any central agency. For many purposes overall organization is weak or non-existent (Long, 1958). From the point of view of all these models, to call budgeting a decision would just be rationalization — a post factum-product.

Approaching first the question of under which conditions a budget may best be explained as an external or internal event, we may start with the most obvious point, asking to what degree other units can, by means of formal rules and laws, constrain the budget organization. Without any doubt, Norwegian communes are "tied up" by decisions made by the national government and the provinces. It has not, however, been easy to obtain a good statistical measurement of the degree of "boundedness". Many of the national policies are initiated by the communes, and the governing units of the provinces are recruited from the communes. Also, most of the laws and rules binding the communes are formulated in such a way that they do not indicate exactly the expenses involved in introducing a given policy. For years, however, a main theme in Norwegian politics has been the allocation of functions and resources between the different levels within the public sector. To some degree there have also been discussions of the degree to which party units take decision power from the local government. We may also mention that the wages of the employees, a very important part of every budget, is decided on in the collective bargaining between the employees' unions and the union of the communes, and that the results have to be taken as given by each single commune. We could, probably, for a plurality of Norwegian communes, summarize all the decisions made outside the commune having a formal binding effect on it, and find that the budget makers were completely tied up. This, however, would leave us with a theoretically uninteresting case. We chose the commune studied because we knew it had some degree of local discretion over the budget. The question then becomes how this degree of discretion is used. Raising this question we have to broaden the statement of the problem of "tying up" a commune. We should not only examine to what degree other formal decision-making units (national government, counties, parties) can constrain the budget organization, but also which actions taken by firms, organizations, private persons and the commune itself have in fact a binding effect. This will be done by focusing on the relationships between the budget organization and the general set of activities and processes going on in its environment. Very few budget makers can start completely from scratch. They will find themselves within a set of ongoing activities, formulated problems, expectations, etc.

Now it is time for two important distinctions. First we will primarily be

interested in budget organizations which are not working in environments that are completely stable, but focus on situations where changes occur, where the environment, and thereby uncertainty, has to be taken into account in the decision model.²⁴ Secondly, it is important to make clear to what degree the participants in budgeting think they should, and can, intervene in events taking place and crises developing throughout the year (also unpredicted consequences of the behavior of the budget organization itself). Pre-Keynesian economics is an example of an orientation within which most crises had to be taken as given, and not to be affected by governmental activity. We have also seen that people working with planning and program budgeting believe that some distance from "pluralistic pressure" is a condition for effective use of their models (Mushkin, 1969, p. 169). We will not, however, be interested in budget organizations which can "close" their relationships with the environment (and their clients) after making a budget, but have an ideology that one should intervene if crises develop ²⁵ If a budget organization takes such an active approach, we will find (even in a small commune like the one studied), that the range of values it may affect is very large. So is the number of issues one has to consider (some of them involving a high degree of technical complexity) and one will have to realize that means-end connections are often surrounded by a high degree of uncertainty.

In organization theory we find the prediction that organizations will try to manipulate their environments in order to avoid uncertainty²⁶ (Cyert and March, 1963, Thompson, 1967). In Norway, the situation, by and large, has been that many communes have not been willing, or have not the resources to use laws enabling them to handle uncertainty produced by actors in the private sector. It should be mentioned, however, that one of the first consequences of a recent enlargement of the full-time administration in the technical sector in the commune studied, was a total stop in all house building outside areas already zoned for such purposes. Also, it was made more difficult for the many people having summer houses in the commune to transform these into year round homes. Certainly in these ways the commune could reduce the number of crises in the educational, general welfare —, (etc.) sectors produced by actions in the private sector. (Because of its character as a suburban commune, actions of firms did not have the same importance.)

In the short run, however, any organization has to choose between two main strategies for living with uncertainty; (1) Forecast problems and plan for their solutions ahead of time, by allocating resources via the budget, (2) solve problems as they arise by finding solutions acceptable to the governing coalition, and thereby allocating existing resources during the year. We may here talk about a "planning" — "fire station" dimension. The nearer the fire station model, the less interesting will the budgetary process be from a decision-making point of

view.

The important point now is that (given the "openness" as discussed above),
the less an organization is able/willing to anticipate problems and allocate resources
for their solutions ahead of time, the more it will be forced towards fire

station activity. That is, it becomes especially important to try to spell out
(some) conditions under which an organization will be able to keep itself at the
"planning" end of the "planning" — "fire station" — continuum.

To sum up the argument so far, and connect it to our definition of decisionmaking, we may say that in order to make the budgetary process interesting from a decision-making point of view, an organization will have to, (1) be able to forecast what will happen if one does not intervene with ongoing activities, (2) understand to some degree means-end connections so that one can have meaningful opinions about how potential interventions will affect different sets of values connected to desired future states of affairs, (3) be able to imagine alternative futures, or alternative sets of values that can be implemented. In the following we will discuss the mentioned factors in relation to which (a) structures, (b) people, and (c) knowledge have to be present in order to make a budgetary process interesting from a decision-making point of view.

Consider first the ability to imagine alternative futures. One critical factor is probably the degree of competing leadership. We know that conflict can be organized into or out of politics, (Schattsneider, 1960) and that "followers" will have few possibilities for influencing decisions if we do not have competing groups of leaders providing alternatives for them. At least three sets of conditions seem to work against the development of such alternative leadership in communes of the type studied. First, the Norwegian principle that the executive board has representatives from all parties in the Council in relation to their numbers of representatives there, can be assumed to provide less competition than a parliamentarian system with one governing group and one opposition. We have seen a tendency towards a "professionalization" of local government leaders in Norway, a leadership consisting of a small group of elected laymen and employees in higher formal positions (Kjellberg, 1965, Olsen, 1967, concerning this phenomenon in general, see Dror, 1968, p. 91) Further the tendency seems to be that leaders from different parties are more like each other (concerning values and beliefs) than their respective followers (Strand, 1968). A second factor working in the same direction is the tradition that the bureaucracy (with very few exceptions) always proposes only one alternative for action. And thirdly, we will propose that the general tendency towards low legitimacy of conflict will also work to reduce the development of competing leadership.

Production of alternative futures following Gresham's lay of planning (March and Simon, 1958), will furthermore be assumed to depend on freedom from routine work. Only very few Norwegian communes have special units for planning. On the other hand, we know that a great amount of time is consumed working on routine matters. The fact that most of the participants are lay people working in their free time, and that the numbers of professionals is very small,²⁷ again supports the conclusion that the conditions for developing alternative futures seem to be unfavorable. Finally we will mention the time-horizon used. The communes are using one year budgets, and only very recently have there been experiments connecting these budgets to longer-range planning. The connections

between the "general plans" developed in some communes and the one
year budgets produced are often very difficult to see. The case studied would be
one example (also Handegård, 1969).

In sum, none of the three sets of factors mentioned seems to make it easy to imagine alternative futures, as demanded by a decision-making model. Now we will argue that the situation is no better concerning ability to forecast future problems, and to develop instruments for guiding the development.

As we know from the planning literature, ability to describe ongoing activities,
the connections between them, their relations to potential alternatives for
actions, and connections to the relevant preferences and goals, are highly dependent
on the development of theories and models able to catch the most important
aspects. This includes operational concepts which as far as possible make
goals and relations quantifiable, but one also needs data which can be fed into
these models. Clearly expertise of many kinds is required. In general, the less
successful we are in developing models which enable us to predict the consequences
of our intervention with ongoing patterns, the more unexpected the
consequences which will demand resources and tie up budget makers in the futno

ure.

In most Norwegian communes we do not find any special units focusing on such problems. On the contrary, those communes interested in working out plans for their development, have in most cases hired private consultants. They have thereby prevented themselves from developing the necessary expertise, both manpower and knowledge, and often also made them unfit to implement the plans provided by the consultants. We may also mention that the profession playing the most important role in national planning, the economists, has never taken any important part in the local bureaucracies. Neither have sociologists and political scientists, who potentially could contribute an analysis of how the system is really working today, spelling out social consequences of potential alternatives for actions. Finally, we may mention the way the budget is structured, focusing on inputs, not on outputs and programs. The importance of this factor has been discussed in great detail by the PPBS people attacking "traditional" budgeting, and need not be supplemented here.

If we try to understand this state of affairs we have to mention that the strong emphasis on the layman and the great trust in "good common sense" is still very strong in local government in Norway, even though the layman's ability to comprehend everything that is going on in the commune was much higher some years ago, especially in the small and static communes, than it is today in the bigger communes with their growing complexities. Equally important is the fact that local government has not been, and is not, able to compete with the national government and the private sector, for highly-qualified manpower and equipment. This is partly due to wages, etc., but we would also assume that the attractiveness of working in the communes is dependent on (a) the fact that many of the "great" problems today are not congruent with the borders of the communes, but have to be solved in cooperation with several communes.29 That is,

working within a commune one cannot have a direct effect on these problems. We may here mention that a new planning apparatus is growing up outside the traditional structure of local government (Mydske, 1968). (b) The "nearness" to the clients, making the number of potential decision-makers very high, and therefore the number of issues and values that should be taken into account are also high, which again produces goal ambiguity. The national planners have been able to build their models around a few consensus values like economic growth and avoidance of unemployment.

Summarizing the arguments so far, many Norwegian communes, at least of the type studied, seem to lack most of the structural manpower and knowledge requirements needed in order to keep them away from the "fire station" end of the continuum. This, combined with the heavy general constraints via national laws, certainly reduces the probability of getting a budgetary process of great interest from a decision-making point of view. But how can we, then, defend the point of view that the activities engaged in each year by the members of the budget organization have any effect on the way they (and others) are thinking about the process and its outcome? Obviously we would need other types of data and other research techniques in order to say something directly about the psychological processes involved. Here we will only try to indicate how aspects of the organization itself make our conclusions a likely interpretation.

We should remember that the budget organization is surrounded by rather clear expectations about how participants should behave, and we have indicated that the organization established cannot fulfill these expectations. In this way the participants will have a common interest in "improving" (not necessarily in any conscious way) the picture of what is going on. We may see reasons both why participants do not "want" to see and cannot see the discrepencies between reality and the ideal. An important part of the picture given is that the organization "satisfies" in important ways the needs of the participants. The leaders are given legitimacy and freedom to keep the "fire station" going. This point of view gives us the prediction that a potential "revolt" against the ideal picture of budgeting will come from the leaders, if they are too much tied up by outside forces. The secondary leadership can be assumed to satisfy their needs for a feeling of importance and for community through the debates, which nearly always end in consensus. This would indicate that the motivation for participation in this group is not primarily power or a desire for certain policy outcomes, but rather ego-defensive and value expressive needs.³⁰ The low issue-orientation (goals, information, initiative focused on the choice at hand) also supports this view. Furthermore, this low issue-orientation prohibits a real comparison of beliefs and attitudes against reality. A high degree of rotation among the secondary leadership works in the same direction. Among the people participating in only one phase of the budgetary process, nearly 60 °/o had participated in local government for less than five years. Among those participating in all three phases, nearly 80 °/o of the elected members had more than 10 years practice in local government. This picture of, on the one hand, a core of people with a long spell

of duty behind them, and on the other hand a secondary leadership which changes frequently, is also supported by other case studies. For the third group — more distant onlookers — we have assumed that the (vague) picture of a representative budget organization gives them a necessary reassurance, a picture of a rational and fair distribution of the resources available.

But even if we assume that the participants and onlookers "get what they want", why are they satisfied with these pay-offs? Why do they not activate themselves to a higher degree in order to affect the choice at stake and the material benefits available? Why do we not get alternative leaders? We think that it is correct to say that given the degrees of freedom left to most Norwegian communes today and the complexity of the budget decision (this is no simple yes-no choice but one demanding a heavy investment of time and energy in order to find out what is going on, how the commune is affected by the complicated set of rules governing the relations between local-regional and national government, etc.), the leadership positions are not very attractive.³¹ We would also assume that the openness of the organization is fairly large. That is, people desiring strongly to be part of the leadership have a high probability of doing so.³² Finally we will mention the general feeling of consensus,³³ that the alternatives for choice are not too distant from each other, and the low degree of criticism. This gives a general picture of middle of the road solutions.³⁴

In short, we think that the organizational characteristics important for the ritual effect of the budgetary process are: the strong element of part-time participants with a majority of people not taking the decision too seriously and being unwilling to invest too much time and energy on it, together with a feeling of consensus and a somewhat high probability for those who want to be leaders to reach their goals.

VI. Conclusion — and Some Further Questions

In this article we have primarily tried to raise certain questions concerning the way we may think about the activity called budgeting by the people participating in it. Doing this we have proposed a new dimension which should be taken into account by students of budgetary behavior. This again has made it necessary to give a fairly broad review of the relevant literature.

Confronted with some data on budgetary behavior which could not be explained within a decision-making framework, we had to try to develop a new perspective on what was taking place. Important parts of the data from the case study of budgeting in a Norwegian commune could be understood if we looked upon the budgetary process as a ritual or a ceremony rather than decisionmaking. The success of proposed comparative studies of budgeting (Wildavsky, 1968) will depend heavily on the ability to clarify the concepts used. At least we have to avoid the trap caused by accepting a priori the meaning given budgeting in prescriptive theories and in conventional wisdom. Probably we should also ask if budgeting has any value at all as a theoretical concept.³⁵ By proposing that budgeting, in some contexts, can be better understood as rituals, affecting

the way people think about and feel towards ongoing processes, rather than as vehicles for policy-making, we have entered unexplored territory. We have, however, tried to discuss some organizational conditions under which the ritualistic vs. the decision-making aspects of a budgetary process will be the most important ones.

The important question now is, what difference does it make if an organization (e.g. a commune) is working as a "fire station" and has a budgetary process with primarily ritualistic functions, rather than a real coordination of its use of resources through budgeting. Consider the concept of quasi resolution of conflict as used by Cyert and March (1963). Here they argue that organizations do not solve their conflicts, but have several procedures which make it possible to live with them. Three such procedures are (1) local rationality, different parts of the organization are working with different problems, (2) sequential attention to goals, the organization attends to different problems and goals over time, and (3) acceptable level decision-making rules.

Working, then, as a fire station, a budget organization will be able to benefit from these mechanisms for quasi resolution of conflict. In Lowi's (1964) vocabulary, most issues can be solved as questions of distributions (that is, one is approaching Pareto optimality — some group may get something, and nobody perceives this as a loss for himself). If, however, an organization tries to allocate its resources through one budgeting process, it will not "benefit" from the quasi resolution mechanisms, and with Lowi, the issue becomes distributive. The participants will realize that in most situations no one can get something without someone losing, and we would assume these expectations about potential outcomes to have certain effects.

Now, we may very well predict who will be able to define crises, and thereby get resources, given a "fire station" organization. The probability that the strong will get stronger, is very high (Jacobsen, 1965). Most issues will be solved by the people able to define their problems, together with the fire-station keepers. On the other side we cannot take as given that the weaker parts (e.g. unorganized groups) will automatically benefit from a redistributive situation. But we will expect that a higher visibility of what is at stake, and a clearer recognition of the competitive element, will produce a higher level of conflict and the activation of new groups. This again can be assumed to have a potential effect on the definitions of what a given choice is all about, and how tangible resources should be allocated.

---

NOTES

I wish to thank the following people for constructive comments on earlier drafts of this paper: James G. March, Stanford University; Michael Cohen and Deane Neubauer, University of California, Irvine; Aaron Wildavsky, University of California, Berkeley; Søren Christensen, Copenhagen; Francesco Kjellberg, Henry Valen and Tone Schou Wetlesen, Oslo, and Fredrik Barth, Knut D. Jacobsen and Kåre Rommetveit, Bergen. However, none of these good people should be taken as guarantors for any part of this paper. I also want to thank Michael Cohen and Mairi MacNeil-Lande for helping me to make my English more understandable, and Eva Sivertsen for doing the typing. 1 However, see Lewis (1952) who followed up the normative problem of determining on what basis it should be decided to allocate X dollars to activity A instead of allocating them to activity B, or letting the taxpayers keep their money.

---

I wish to thank the following people for constructive comments on earlier drafts of this paper: James G. March, Stanford University; Michael Cohen and Deane Neubauer, University of California, Irvine; Aaron Wildavsky, University of California, Berkeley; Søren Christensen, Copenhagen; Francesco Kjellberg, Henry Valen and Tone Schou Wetlesen, Oslo, and Fredrik Barth, Knut D. Jacobsen and Kåre Rommetveit, Bergen. However, none of these good people should be taken as guarantors for any part of this paper. I also want to thank Michael Cohen and Mairi MacNeil-Lande for helping me to make my English more understandable, and Eva Sivertsen for doing the typing. 1 However, see Lewis (1952) who followed up the normative problem of determining on what basis it should be decided to allocate X dollars to activity A instead of allocating them to activity B, or letting the taxpayers keep their money.

2 This is no longer an entirely accurate description. Wildavsky after making a study of budgeting in Oakland concludes that resource allocation is either not feasible or not done, and starts to look for other purposes of budgeting (personal communication, January 13, 1969). Reporting on the research, the conclusion is that budgetary activity does serve to control bureaucratic behavior and facilitate tax policy. (Meltsner and Wildavsky, 1969.) The last point is consistent with our emphasis on budgeting as a compliance and consensus forming activity. The control aspect cannot be discussed in detail in this paper. Somewhat large discrepancies between the budget and the accounting document, and the general lack of information among the participants lead us to conclude that the control function is not very important in the case studied here. At least it does not explain why the large apparatus of lay men are activated in the process.

3 In his now famous "The Science of Muddling Through" (Lindblom, 1959, p. 85) Lindblom says: "Almost every interest has its watchdog. Without claiming that every interest has a sufficiently powerful watchdog, it can be argued that our system often can assure a more comprehensive regard for the values of the whole society than any attempt at intellectual comprehensiveness".

4 Even if comparable statistics are not easily found, this conclusion is supported by a publication from the International Information Center of Local Credit for the period 1956^—60. See, "Local Government in the 20th Century". Proceedings of the lULA Jubilee Congress, Brussel, June 17th—25th, 1963, M. Nijhoff, Haag, 1964.

5 The Norwegian laws in 1969 stipulated that local taxes on income had to be between 15% and 19%. 441 out of 451 communes, with 90% of the population, were using 19% (St. meld. 58, 1968/69), which has been used as an indicator for the freedom of local government.

9 Many reports of this type can be found in the journal, Kommunalt Tidsskrift.

7 Smaller communes often have a treasurer. In bigger < board preparing the budget-proposal for the executive board.

8 The timing is: starting in April, the managers of the different sectors will have their proposals ready for the committees in May_—June. In June^—July these will be gathered by the municipal manager. His proposal will then be discussed in the different parties, before the executive board takes over in the middle of August. The council has to make a decision before October Ist. The events taking place in this period are what we have called the formal budgetary process.

9 Where perception data had to be used, we compared the respondents' answers concerning his own behavior with other participants' perceptions of his activity. Some of these distibutions are analysed in Olsen (1968).

10 The best example is the discussion concerning the boundaries of the local units. Both people advocating larger communes and those who wanted to keep smaller units said their solution would strengthen the local self-government. (Om en revisjon av den kommunale inndeling m. v. Prinsipielle retningslinjer. Innstilling II fra Kommuneinndelingskomitéen. Oslo, T952. Arbeidernes Aktietrykkeri.)

11 Q: Compared to the other issues the commune is working on, would you say that the budget is the most important decision, one of the most important, an issue like many others, or an issue of little importance?

12 The meetings of the Council are open to the public (but very seldom does a large audience gather). The meetings in the Executive Board are closed. It is of some interest to note that the School Board which is the only one in the preparation phase which attracts an audience, is the only one having a special subcommittee preparing a budget proposal. The newspapers, however, will give reports about what happens in the closed meetings, but not in the same detail (what each person said) as from the Council-meetings. In general we may suggest that the environment (e.g. newspapers) tends to account as important those things which are most accessible and available. A budget is a very visible thing, it has numbers, and is easy to report. (The parallel to elections seems fairly close).

---

12 The meetings of the Council are open to the public (but very seldom does a large audience gather). The meetings in the Executive Board are closed. It is of some interest to note that the School Board which is the only one in the preparation phase which attracts an audience, is the only one having a special subcommittee preparing a budget proposal. The newspapers, however, will give reports about what happens in the closed meetings, but not in the same detail (what each person said) as from the Council-meetings. In general we may suggest that the environment (e.g. newspapers) tends to account as important those things which are most accessible and available. A budget is a very visible thing, it has numbers, and is easy to report. (The parallel to elections seems fairly close).

13 Using a scale: very free, somewhat free, somewhat tied up and very tied up, only three persons used the category "very tied up". 66 °/o of the Council members answered very free/ somewhat free. Only among the employees do we find the opposite pattern with a majority of those who had any opinion answering "somewhat bounded".

14 The findings of Wildavsky (1964, p. 18) are much nearer the view found among the employees, and indicate some important differences between full-time and part-time participants.

15 This index is a very crude one, basedon, (a) who had operational goals, (b) who were information centers, (c) who took initiative and succeeded, (d) who were perceiving themselves as influential, and (e) who felt responsibility for the budget decision. Since the distributions all followed the same pattern, even such a simple index should indicate some degree of centrality in the decision.

16 Using an open question with no alternatives given, 64 % among those who had an opinion gave this answer. The rest of the answers were scattered over a large number of other alternatives.

17 63 °/o of these again (N = 57) explained this with common goals and values, but again we find clear differences among the respondents: from 53% of the majority fraction in the Council to 80 % of the committee members not members of any parties, used this explanation.

18 These differences in perceptions of party politics between the Council and the committees could be explained by the fact that the meetings in the Council are more outwardly "open", but here the perceptions of the Executive Board do not fit in. The members of the Executive Board report that party politics frequently takes place. Probably perception of party politics is connected with to what degree one perceives oneself as a leader or not.

19 It is important to make a distinction between, (1) referential symbols, which are economic ways of referring to the objective elements in objects or situations. Here the elements will be identified in the same way by different people (at least those who understand the special set of symbols involved). An example would be mathematics. And on the other side (2) we have condensation symbols, which evoke emotions associated with a situation (pride, humiliation, anxieties, etc.), and condense them into one symbolic event. (Edelman, 1967, p. 6). We will primarily be interested in condensation symbols.

20 Other "external" characteristics often mentioned are that the type of behavior is predictable, and that it has a cyclical aspect, that is it will be performed on appropriate occasions, a certain time of the year, or when a special event takes place (death, birth, war, etc.).

21 Contrary to some of my colleagues I do not think that we can activate a budget organization like the one we have studied once a year without consequences. We should remember that the majority of decisions in most large organizations most of the time are taken by relatively few people. For the majority of those affected it is not possible to check the content "directly", one has to rely on indirect indicators, for instance who participates (and other aspects of the procedure). Thus, "who participates" will in itself be an important signal.

22 Obviously these two aspects are closely interconnected in the long run.

23 For a more balanced view on this literature, see Salisbury, (1968) and Sharkansky and Hofferbert (1969).

24 This means that many small rural communes, dominated by the primary occupations, will be excluded from the following analysis.

25 We have to make a distinction between this "open"-"closed" dimension, and the question of who defines a "crisis". The argument should be relevant both in environments where one thinks the public sector should only solve problems the private sector cannot, or will not, solve, and in environments where the public sector is assumed to have a stronger impact on the

---

private sector. However, this does not mean that it is not important to work with the question of who can define crises. Obviously this is an important aspect of the problem of who can "tie up" a commune, but also, the more it is up to the environment to define crises, the more difficult is the problem of forecasting for the organization.

26 One way to handle uncertainty is to use large unspecified grants. This obviously takes "decision-making" content out of the budgetary process and makes the decisions on how these funds are really going to be used more important. Norwegian communes in general, however, use very specified appropriations.

27 A connection is often seen between the numbers of professionals and the degree of innovation. One important factor is that the professionals have better contact with their colleagues in other organizations, producing a scale for comparison and a possibility for defining situations as not "satisfying".

28 The former Czeckoslovakian minister of finance gives some very interesting illustrations of unexpected consequences as a result of inadequate theories of how the system was really working (Sik, 1969).

29 This statement is least true for some communes, where geographical conditions on the one hand reduce the interaction with other communes, and on the other hand divide the commune in such a way that we get a clear conflict dimension along geographical lines.

30 Even if we do not have systematic data for the country as a whole, several case studies indicate this (Valen and Katz, 1964), and studies reported in Kjellberg, 1970 (forthcoming).

31 A certain support for this is shown by the tendency we have observed that groups usually involved in political leadership, do not seem to be very interested in positions in local government, especially not in the communes where other alternatives (national leadership, leadership in interest groups and in private firms etc.) are available (Kjellberg, 1970).

32 The debates of the "something should be done"-type may be a way of indicating one's interest in investing the time and energy demanded by those who desire to be leaders, and, one may here also indicate ability to comprehend a complex issue like the budget.

33 It is no tautology to propose consensus both as a cause and an effect of ritual. This is completely consistent with the fact that the state of a system at a certain point in time (Ti) in most cases is primarily a product of the state of the same system at an earlier point in time (To).

34 We have already mentioned that the fire station-keepers — eve.n if they come from different parties — as a product of close interaction over many years (especially in the executive board) show a tendency to accept the same values and beliefs.

35 The concept of budgeting as first drafting a document and then its implementation, is clearly a parallel to older definitions of planning (where one emphasized the distinction between a thinking-phase and a doing-phase). Newer literature, however, has left this static view, and is looking more upon planning as continual activity basedon clarifying goals and wanted future states of affairs, knowledge about how the system is working, and development of instruments for guidance of the organization (Jacobsen, 1969). What we need then is a general theory of resource allocation, which again opens the possibility of classifying organizations according to the way they make their allocations.

REFERENCES

Burkhead, Jesse. Government Budgeting, N.Y., John Wiley & Sons Inc., 1968 (Bth printing).

Campbell, A. (et al.). The American Voter, N.Y., John Wiley & Sons Inc., 1964, (An Abridgement,
paperback).

Christie, Nils. "Edruelighetsnemnder. Analyse av en velferdslov", Nordisk tidsskrift for kriminalvidenskab,
1964, pp. 89—118.

Churchman, C. W. and A. H. Scheinblatt. "BPB: How Can It Be Implemented", Public Administration
Review, vol. XXIX., no. 2, 1969.

Cleveland, Frederick. "Popular Control of Government. Three Schools of Opinion in the United
States with Respect to Budget Making", Political Science Quarterly, vol. XXXIV, 1919,
pp. 237—261.

Cohen, Abner. "Political Anthropology: The Analysis of the Symbolism of Power Relations",
Man, vol. IV, no. 2, 1969.

Crecine, J. P. Governmental Problem-Solving: A Computer Simulation of Municipal Budgeting,
Chicago, Rand McNally, 1969.

Cyert, R. M. and J. G. March. A Behavioral Theory of the Firm, Englewood Cliffs, N.J.,
Prentice-Hall, 1963.

Dahl, R. A. Who Governs, New Haven, Yale University Press, 1964 (sth printing).

Davis, O. A., M. A. H. Dempster and Aa. Wildavsky. "A Theory of the Budgetary Process",
The American Political Science Review, vol. LX, no. 3, 1966, pp. 524—547.

Dawson, R. E. and J. R. Robinson. "Interparty Competition, Economic Variables and Welfare
Policies in The American States", Journal of Politics, vol. XXV, 1963.

Dror, Y. Public Policy Reexamined, San Francisco, Chandler Publishing Company, 1968.

Dye, Th. R. Politics, Economics, and the Public Policy Outcomes in the American States, Chicago,
Rand McNally, 1966.

Edelman, M. The Symbolic Uses of Politics, Urbana, Univ. of Illinois Press, 1967, (paperback).

Gluckman, M. Politics. Law and Ritual in Tribal Society, Chicago, Aldine Publishing Comp.,
1965.

Gouldner, A. W. "Organizational analysis" in Robert Merton. Sociology Today, N.Y., 1959.

Hofferbert, R. I. "The Relation Between Public Policy and Some Structural and Environmental
Variables in the American States", The American Political Science Review, vol. LX, no. 1,
1966.

Jacobsen, K. D. "Informasjonstilgang og likebehandling i den offentlige virksomhet", Tidsskrift
for samfunnsforskning, no. 2, 1965.

Jacobsen, K. D. "Institusjonelle betingelser for planlegging", Plan og Arbeid, no. 2, 1969.

Katz, D. "The Functional Approach to the Study of Attitudes", Public Opinion Quarterly,
vol. XIV, Summer 1960, pp. 173—175.

Key, V. O. "The Lack of a Budgetary Theory", The American Political Science Review, vol.
XXXIV, Dec. 1940.

Kjellberg, F. "Politisk lederskap i en utkantkommune", Tidsskrift for samjunnsjorskning, no. l,
1965, pp. 74—90.

Kjellberg, F. (ed.). Kommunalpolitiske studier, Universitetsforlaget, Oslo, 1970 (forthcoming).

Leach, E. R. "Ritual", International Encyclopedia of the Social Sciences, 13, N.Y., Macmillan
and Free Press, 1968.

Lewis, Verne B. "Toward a Theory of Budgeting", Public Administration Review, vol. XII,
Winter 1952.

Lindblom, Ch. E. "The Science of Muddling Through", Public Administration Review, vol.
XIX, Spring 1959.

Long, N. "The Local Community as an Ecology of Games", American Journal of Sociology, vol.
XLIV, Nov. 1958.

Lewi, Th. J. "American Business, Public Policy, Case Studies, and Political Theory", World Politics,
vol. XVI, July 1964.

Mannheim, Karl. Ideology and Utopia, N.Y., Harcourt, Brace & World, 1968.

Meltsner, A. J. and Aa. Wildavsky. Leave City Budgeting Alone!, Graduate School of Public
Affairs, University of California, Berkeley, 1969, (mimeo).

Mosher, F. C. Program Budgeting, Chicago, Public Administration Service, 1954.

Mosher, F. C. "Limitations and Problems of PPBS in the States", Public Administration Re-view,
vol. XXIX, no. 2, 1969.

Mushkin, S. "PPB in Cities", Public Administration Review, vol. XXIX, no. 2, 1969.

Mydske, Per. Problemer og strukturer — En organisasjonsanalyse av norsk regionplanlegging,
Oslo, 1967, (mimeo).

Nordanger, K. M. and A. Engh. Kommunalkunnskap, Oslo, Tiden, 1960, (sth printing).

Olsen, J. P. Kommunal budsjettering — En beslutningsprosess, Oslo, 1967 (mimeo).

Olsen, J. P. Public Finances, Needs, Sources and Utilization, (A Report of the National Bureau
of Economic Research) N.Y., Princeton Univ. Press, 1965 (3rd printing).

Sahlins, M. D. Tribesmen, New Jersey, Prentice-Hall, 1968.

Salisbury, R. H. "The Analysis of Public Policy: A Search for Theories and Roles" in Ranney,
A. (ed.). Political Science and Public Policy, Chicago, Markham, 1968.

Schatt&neider, E. E. The Semi-Sovereign People, N.Y., Holt, Rinehart and Winston, 1960.

Schick, Allen. "System Politics and System Budgeting", Public Administration Review, vol.
XXIX, no. 2, 1969.

Sharkansky, Ira and R. I. Hofferbert. "Dimensions of State Politics, Economics and Public
Policy", The American Political Science Review, vol. LXIII, no. 3, 1969.

Sik, Ota. Plan og marked, Oslo, Aschehoug, 1969.

Simon, H. A. Administrative Behavior. N.Y., The Macmillan Company, 1965, (paperback).

Smith, H. D. "The Budget as an Instrument of Legislative Control and Executive Management",
Public Administration Review, vol. IV, no. 3, 1944.

Strand, Arne. Det folkevalgte politiske lederskap i Oslo, Oslo 1968, (mimeo).

Tarjem, Kåre. "Kommuner og fylkeskommuner" in Johnsrud and Sandberg. Politisk ABC,
Oslo, Gyldendal, 1969.

Thompson, James. Organizations in Action, N.Y., McGraw Hill, 1967.

Thompson, Victor. Bureaucracy and Innovation, University of Alabama Press, 1969.

Valen, Henry. "Den sosiale og politiske bakgrunn for rekrutteringen av det politiske lederskap",
Tidsskrift for samfunnsforskning, no. 2—3, 1966.

Valen, Henry og D. Katz. Political Parties in Norway, Oslo, Universitetsforlaget, 1964.

Wildavsky, Aaron. "Political Implications of Budgetary Reform", Public Administration Review,
vol. XXI, Autumn, 1961.

Wildavsky, Aaron. The Politics of the Budgetary Process, Boston, Little, Brown & Co., 1964.

Wildavsky, Aaron. "Budgeting as a Political Process", International Encyclopedia of the Social
Sciences, vol. 2, pp. 192—199.

Walker, Mabel. Municipal Expenditures, Baltimore, The Johns Hopkins Press, 1930.