Swedish organization research has got a vitalizing injection through the studies of local government that have been conducted over the last 10 years. From the beginning the studies were focussed on the budgetaryprocess in austerity (Jonsson 1982, Brunsson-Rombach 1982), but gradually it was discovered that more fundamental management problemsthan budgetary ones are at hand. It might for example be asked whether a municipality can be looked at as one organization — isn't it rather a number of organizations governed by professional norms where politicians often in vain try to find out what is really going on? A fundamental difference from most ouher organizations is that a municipalityhas a top management where leaders are in constant conflict. It even recruits decision makers because they are good at quarreling with the other members of the management team. And there are ritual debates (when everybody has made up their minds there is a public debate where nobody expects to influence anyone of his fher collegues) after which a wooden club is thumped on the table. In many decision situations about half of the management team (the minority) states publicly that this was a very stupid decision that should not be

Summary

Against a backdrop of current organization el research on local government in Sweden this paper deals with the management problems faced by municipalties. Characteristics of these organizations are printed out and the necessity for pratical leadership in decentralized municipal structures is discussed.

---

*) Professor, dr., University of Gothenburg.

carried out. Furthermore these strange organizations choose to do things where it is extremely difficult to measure outcomes. This howeveris of no great significance since decision makers keep talking about the future all the time. In the meantime the different professions in the large administrative apparatus try to uphold high a professional standard. These organizations have been very successful by most standards;they enjoy high legitimacy, they have grown rapidly over a long period, and they keep growing in spite of the faet the state is in severe financial trouble.

From an interest in financial management, research has moved its locus of interest to the general management issues where democracy and effectiveness goals are to be satisfied at the same time, a problem that faces industrial organizations that want to implement industrial democracy, as well.

This paper deals with some of the above mentioned issues and tries to
point out in what directions solutions might be found.

The choice between basic control approaches

When the new interest in organizational studies of local government was put on the agenda in the early 70ies we found an organization where planning as the method of solving current problems dominated among officers as well as politicians. Plans for the future, long term and comprehensive, were believed to create an order that would guarantee concerted action of all organizational members (it would also help legitimize a high level of taxation). This approach had been very successful, in the post war establishment of the welfare state, which is to a large extent managed by local government, and in building away the shortage of housing during the late 60ies. With the arrival of the oil crisis with its financial consequences and the intensification of polities the premises for organizational control had changed. During the earlier period there had been virtually no disagreement on goals and methods, now there was; politics had become a reality and the plan ning situation was turbulent. The standard reaction was »more of the same« - centralize and provide the center with more sophisticated instruments like e.g. program budgeting.

Lacking the necessary environmental stability planning as a control
method was seen as failing. Frustrated local government managers

saw power as the key to better control. Instead of compiling detail into comprehensive plans at central planning offices directed by central politicians, the solution was seen in decentralization. Detail should be pushed out to operating units and operative plans made there. The center should govern by directives and resource limits, and decentralize responsibility to local decision making bodies (i.e. keep the power to discipline units centrally). Experiments with decentralization are being conducted in a number of Swedish communities today. Expectations are very high; deeper democracy and higher efficiency.

A third approach to control, that may come as a corrective outcome of the decentralization experiments, i.e learning. Eearning means simply that you do something and see what happens. ]f outcomes are not the desired ones you change the behavioural rules. This control approach is myopic, sub-optimizing and backwards-looking, but it has survival capability if targets are based on experience. The learning organization will tend to stay in benevolent environments, and opt out of threatening

Planning, power and learning has different implications for organizational
design and for political leadership.

Openness and closure

The problem with local government is that it is supposed to be democratic as well as efficient. To the former end the decision making body is designed to be in conflict with itself. We want it to have an inbuilt opposition that works tirelessly to point out that there are better ways to conduct business and more desirable ends to strive for. By arguing for alternatives and pointing out deficiences in current policies it serves democracy, it makes politics out of current problems and undermines the authority of the majority to direct operations. The majority therefore has to pay constant attention to the legitimacy of its policies, not only in election campaigns but in all decision making situations. The establishment of criteria of choice is an open process that is easily influenced by the environment. The stability of preferences over time and the exogenous nature of goals in relation to the choice of alternatives of action, presupposed in the rational planning model that we cherish, are seldom present. Ambiguity and frustration is a necessary consequence.

Efficiency presupposes closure. Alternatives should be evaluated and carried out according to fixed and consistent criteria. They should also be evaluated on their own merits and not in terms of how they effect future political situations. Compromises should be avoided since they are, by definition, suboptimal choices. But they are not. The control process is to a large extent open.

The method to achieve closure is to rely on formalistic and ritualistic decision events (like pounding tables with wooden clubs). Implementation of decisions is the responsibility of a hierarchy inhabited by neutral/loyal competent professionals. Although these bureaucrats have superior know how, specialize in on problem area, and devote very much more time to issues than politicians, they are supposed to let themselves be guided by the outcomes of the »irrational« political process. Techniques like programme budgeting have been introduced to increase the rationality of the joint functioning of the political and the administrative parts of the organization. But politics is driven by the will to achieve a better world; it cannot accept being defined out of the planning process. Politics permeates everything; party politics, grass roots, special interests, users' influence, and professional norms are legitimate parts of democracy. Political leadership and rational action have to be achieved in an organization that oscillates between openness and closure — goal reformulations as well as efficient adaption of service operations to changing client needs. A formidable task indeed! It might be ventured that political leadership is the management of openness and closure — debate and action.

A two-tier system

Organizationally this formidable task is treated on two levels; one where the output almost exclusively is talk and statements about what should be done and another where action is taken. It is essential for democracyas well as for efficient operation that the link between the two is intact. Politicians and officers are mutually dependent and the relationis built on trust. Problem solving requires that both parties contribute.The ideal image we have of the public sector tells us that officers are unpolitical experts that only deal with facts while politicians deal with values and goals. Usually we do not reflect very much about whether it is possible to set relevant goals without knowing very much

about what can be achieved (i.e. without causal knowledge) and whether it is possible to provide relevant causal information without knowing the specific intentions of the top. Temporary closure is achievedthrough decisions that are surrounded by formal rules making it a duty for officers to carry them out. Decisions are very visible in local government; debate on pros and cons culminate and is then cut offby the decision that turns it into law^T for the administrative hierarchy. Or der and action is built on the power vested in the elected bodies.

Technical competence and a superior information handling capacity are properties of the administrative hierarchy. Furthermore officers can specialize and devote much more time to each issue than politicians can. And officers are more numerous! Viewed against this background it is not at all self-evident which oi the two parts of the organization that has the greatest influence on the course of events (Brunsson &J6nsson 1979). One could maintain that officers have the lead in formulating alternatives due to their know-how, while politicians figure as foremost in terms of taking responsibility because of their leading role in decision making. This asymmetry between influence and responsibility is non-problematic when there is basic agreement between all parties on the goals of the welfare state. There might be disagreement on the timing and rate of the expansion of services but the general direction is fwas not in question.

When resources become scarce, how^rever, and politics intensify, conflicts of interests become more visible. Latent power issues turn manifest. In stead of sharing praise for success the two parts of the organization will have to share blame for failures. Trust relations may break when either party attributes failure to the other (Jonsson 1982). Observers tend to attribute failure to inner causes (e.g. incompetence) while actors tend to attribute to external factors (e.g. lack of resources). Newleadership and organizational issues are raised by austerity.

Is decentralization the solution?

Lack of resources has an inherent centralizing effect — priorities has to be set centrally and budgetary discipline maintained — but adaptation to a turbulent environment necessitates decision making close to the sources of disturbance. In this situation the solution a la mode, at least in Sweden, seems to be decentralization — in the name of democracy as

well as efficiency. If the legitimacy of a large public sector is to be upheldit has to adapt services to local needs within a tight budget. To make budgetary allocations stick the local government needs the disciplineol the hierarchical structure for financial control. Central monitoringol performance against an annual budget divided into a number of shorter periods does not require any new inventions whatever the local organization. When it comes to controlling and adapting services to local needs and deepening democracy however innovations seem needed.

Obviously we have here a »classic« case of contingency theory application. There are problems with differentiation as well as with integration (Lawrence & Lorsch 1967). Differentiation in terms of adaptations to local service needs as well as mobilization of citizens to engage in local common worries is difficult in an environment used to uniform practices governed by corporatist and for professional norms that are firmly institutionalized. Integration is difficult to achieve in an environment of intensified politics and with a much publicized ideology of decentralization and encouragement of local initiatives. It is most likely under those conditions that decentralization will generate a number of uniform rule-fettered local bureaucracies. (Meyer & Scott 1983) i.e. the problem will not be solved but magnified.

The solution will, firstly, have to include strict financial controls so that resource limitations are viewed as preconditions for operations, not as something that can be disregarded and referred to somebody else to solve should resources seem inadequate. Given this, the next step is to achieve differentiation. The municipality should establish a pattern of local service units that are as far as possible adapted to local conditions and designed to further local citizen participation (legitimacy). These units must be big enough to ensure professional competence and small enough to be responsive to environmental change. Services depending on »technical infrastructure« like water and energy are more difficult to decentralize than »social« services.

To realize the potential benefits of coordinated services over sectorial borders the local unit should be organized across sectors on a geographicalbasis with the principal aim to focus and coordinate problem — solving activities. The school sector, the social sector, libraries, recreationetc. should join forces in development and intelligence operations. To achieve this in the face of strong professional and institutional

norms, creative leadership on the local level is needed. The task is quite demanding; to promote cooperative attitudes and risk takirig outsideconventional propfessional boundries. Grass roots democracy also has different characteristics than the representarive democracy practicedcentrally. The latter is a formal decision making mode while the formeris a matter of attitude, process and participation.

Integration over local units — to ensure that the will of the people as expressed by the central elected body guides local operations - is less problematic as to the financial contyrols than concerning operations. To clearify this we have to return to the approaches mentioned previously. The planning approach is built on the accunulation of knowledge centrally and deductive reasoning around forecast, alternatives and stable criteria of choice. It can be applied to the financial field but not to local operations. The power approach is built on keeping central authrity to establish rules of behaviour and for reward and punish on the basis of performance on central criteria. It might be acceptable from an efficiency point of view but it is probably insufficient if emancipatory aspects are important and if the environment is to turbulent for the centre to be able to adapt rules and criteria adequately. Remains the learning approach which is built on experiment and direct observation of results. This kind of cybernetic approach can be applied locally but it has the shortcoming of being myopic and particularistic. It will not seek grand optima. It has to be complemented with a central device that directs local attention and develops existing criteria of goodness. Such a strategic device could be called political leadership in the local government environment. It works, not on operations or the application of resources to operations, but the interpretative schema that guide how local operations managers define the situation. It works through interaction and communication. Ree (1985) points out three dimensions of political leadership: political credibility, administrative effectiveness and political semantic ability. A balance between the three is necessary.

Credibility is reached by being seen as consistent and goal-directed, effectiveness by being able to coordinate political programs and being able to steer them through the decision making process (which sometimesmakes compromises necessary, which might be detrimental to credibility). The most important dimension here is however the third one; the ability to communicate. Communication takes place on

several levels e.g. using a technical program language, or using a more contextual organizational language, or a quite open societal language. There are also shifts between attention directing mass communication, attitude forming personal communication and combinations between the two in legitimation processes (Zaltman et al 197 3, Hussein 1981). It seems reasonable to conclude that the integreation of operations in a decentralized organization could be provided by political leadership. (The nature of political leadership from an organizational point of view deserves more research attention!)

An organizational structure adapted to these prescriptions would have financial controls through the hierarchical structure and attention directing, standard setting communication through other channels, maybe through the environment (e.g. the political parties). The structure should not be designed on the assumption that operative and financial information can meaningfully be integrated on other levels than the operative one.

Learning will chiefly take place on the operative level where services are produced. Occasionally the conceptual filter through which learning takes place may be changed by political signals but responsibility for the adequancy of services rests with the local units.

When operations have been established and routinized the they can be
managed through either of the power modes; behavioural or output
oriented control (Ouchi 1978).

References:

Brunsson, N. Scjonsson, S. Beslut och Handling (Decision and Action) Li her. Stockholm, 1979.

Bi unsson. N. & Rombach, B. Går det atl spara.J (Is economizing possible.') Doxa. Bodaiors, 1982.

Hussein, M.K. Ihe Innovative Process m Financial Accounting .Standard Setting. Accounting,
Organi/.ations and Society. 1981, p. 27-39.

|on.sson, S. A City Administration Facing Stagnation. Political organization and action in
CJothfiiburg. Studentlitteratui. Lund, 1982.

Mevci. J.VV. Scott, W.R. Organizational Environments. Ritual and Rationaliiv. Sagf. Be\erl\
Mills. 198-J.

l.awicncc, F.R. & Lorsch, |.W. Organization and Environment. Harvard Univt'isity Press. Boston,
191.7.

Onchi, \V. Ihe Relationship between Organizational Structue and Organi/.iiii)nal Control,
Administrative Science Quarterly, p. 95 (I. 1978.

Rec, S. Reflections on the nature of political leadership dupl. Copenhagen Kconomics and
Business Administration, 1985.

/.ah iiuii, (>., Du man, R. & Holbcck, J. Innovations in Organizations. W'ilev. New York, 197^.