Marija Kaštelan Mrak The economic role of governments is attracting more and more attention day by day. Many argue that, compared to bureaucratic centralized governments, decentralized governments would be more effective and more efficient in supplying services that help create and promote the competitive advantages of their local communities.1 In any case, increasing capacity and responsibility of decentralized government units turns out to be the first step that should contribute towards making the government more capable and efficient.
The most recently established Croatian central government officials proclaim public administration reforms to be among their first priorities. In a way, this process will continue the line of major social and economic transformations that Croatia has been over during the past decade, including a program of democratization and decentralization of government functions started by the previous government.2
The Croatian transitional effort cannot be seen just as an individual matter in a historically and geographically isolated space. Apart from being strongly marked by the process of political and economic transition from a more or less centrally planned system to a more market oriented one, the intended decentralization of the public sector is also a part of a broader, world wide, phenomenon of government decentralization going on around the world.
The shifts that will occur, both in the public and in the private sector, will have a long-lasting effect on the economy.
Here, we will dispense with politics and confine to exploring economic, organizational and managerial issues arising from public sector decentralization in Croatia with the focus on costs and effectiveness of the maintaining a body of public administrators. We will assume that in studying government decentralization, a parallel can be traced to the process through which the private sector has gone under while decentralizing its management and operations during the second half of the XX century. So, after establishing the general settings for government decentralization, the paper provides a short historical account of the considered of some issues considered important in the evolution of the organization in the private enterprise sector. The paper concludes by drawing analogies and establishing expectations for the Croatian public sector.
I.Global settings and the restructuring of governments
Globally, the tightening of competitive struggles between firms, regions, nations and continents has led into increased interests for the efficiency of the public sector. A major issue for increasing the economic competitiveness of local and regional communities would be related to the model of government organization and its capability for perceiving, building (developing) and maintaining the competitive potentials of the community it serves.
The interest for the role and organization of the public sector has been revived in Europe, but also all around the world, for over a decade. Expectations on the ways in which the public sector delivers its services are being broadened and reformulated.3 And, along with proposals for reform, came the need to establish expectations for efficiency gains and attempts at formalizing targets so that improvement achieved can be monitored and managed.4
The involvement of the public sector into economic and business considerations derives from two divergent sets of facts: On one side, the government, and the public sector in general, use up resources that might otherwise have had alternative usages in the private sector. Also, the public sector comes by these resources through non-market mechanisms.5 On the other hand, the government performs services that contribute to the general efficiency of an economy. Many of these services would not be provided at all (or would be of lower quality) if not performed by governments. Such are the actions and investments in sustaining the production and preservation of resources such as education and natural resources, the efforts to build and maintain the legal frameworks, establishments and agencies for directing, monitoring and correcting spontaneous (possibly privately profitable but socially disruptive) market practices.
Economically, it is thus acceptable for governments to use up resources if their internal rate of efficiency is higher than that achievable by other sectors, and, if the overall efficiency is raised when resources are allocated through government.6 So, the question concerning government spending is not so much a question of government share but of efficient incurred by government spending. A related question is how can alternative efficiencies be determined and/or measured.
The measures of effectiveness applied to the private (business) sector are usually related to financial or market performance. In case of government spending, measures such as market share and profits do not apply. Therefore, what we do have as evaluation instruments are mainly intuitive reasoning by some form of analogy and international comparisons.
Another set of questions deals with the correlation between government efficiency and the chosen organizational model. For Croatia, the question is whether and how will the intended decentralization contribute to various aspects of internal efficiency. For, the same amount of capital can produce a higher or lower output depending on the mode of organization and on the applied technology. Economists call this effect of achieving the best possible economic usage of the available pool of resources efficiency.
Efficiency is usually classified as external and internal efficiency. External efficiency arises from a better allocation of resources among economic units (sectors, industries and firms). Internal efficiency depends on achieving the best possible mix of resources inside an individual economic unit.
Internal efficiency can be further divided into operational efficiency and governance efficiency. Operational (technical) efficiency derives from the adequate use of the available technology.7 The efficiency of the governance system designates the efficiency of the system of control. It relies primarily on the quality of information and on adequate motivation. Operational efficiency and governance efficiency are interrelated. Relevant information and motivation aligned to performance are the two ingredients needed for efficient decision making so that operational processes may run smoothly and so that the right changes may be introduced at the right moment. A comparatively higher operational efficiency (achieved, for example, because one firm possesses superior new technology), reflects on the productivity of resources and eventually higher profits that might be allocated to further developing the governance system (for example, by adding new professionals to the staff).
II. Defining government decentralization as an efficiency problem
The problem of government decentralization is an example of a typical organizational phenomenon. We assume that theoretical concepts developed under organization theory are suitable for analyzing the problem of government decentralization. Decentralization will therefore be understood as a change of organizational model. Organization theory also poses that it is possible to achieve efficiency gains by changing organizational forms (organizational models, design or architecture and organizational routines).
Theproblem encountered here is to define efficiency shifts likeable to occur if the government is decentralized.
Efficiency is multidimensional economic phenomenon (many services are performed by a same administrative body)8
Efficiency is path-dependant.
Efficiency may be established only in comparative relations.
Efficiency can be modified through organizational design and managerial practices
The business literature has traditionally associated decentralization with less hierarchy and higher autonomy at lower managerial levels. Decentralization can also be accessed in relative terms implying that one system relies more heavily on vertical control (central authority) while another relies more on horizontal control (self-management and lateral adjustments between agents).
Usually, centralization was considered to be economically more efficient when environments were steady and predictable. Then economies of standardizations, specialization, planning, scale, came into action further enforcing the centralized organizational model. Turbulent environments, on the other hand, called for more initiative and autonomy of individuals at the operational level. It was taken for granted those individuals in direct interactions with problems would be informed enough, or at least faster in providing responses to threats and opportunities. Compared to the intricate process of collecting and evaluating information by some specialized body, decisions by individuals would be quicker, would cost less and, in many cases, would even be more pertinent.
On the theoretical level, the decentralization problem appears simple. At first sight, economists will assume governmental functions where hierarchy and scale economies are more intense should be centralized, while those where local expertise and flexibility are needed will be decentralized.
However, some additional questions come to mind concerning not only the final result but also the process and the costs of change:
In Croatia, if we assume that every reorganization resulting from major shifts in government and governance systems combines centralizing some functions and decentralizing other (especially when territorial changes occur) we might be breaking up some standards, routines, bodies of knowledge and loosing efficiency faster than we are able to create new efficiencies (a reason for the so called "transition lag"), citizens might wonder is there a well defined program of change.9
If efficiency is a moving target, a new question arises of whether we are capable of moving at the pace of comparative economies, and what obstacles to speed and effectiveness of change can most easily be surpassed,
To what point are the present decentralization processes in Croatia consistent with developments in the EU to which Croatia aspires? Would there be another period of readjustment once we enter the EU?10
Up to what point should the system be modeled after EU plans at the sacrifice of present efficiency, and at the sacrifice of aspects of efficiency more valued by Croatian citizens. Time frame considerations, also lead to conclusions that some aspects of the emerging organization, for instance the choice of functions, personnel size and expertise, might be efficient only after joining the EU but temporarily be inefficient.
How precise and predictable are scale economies? Should we, as a small country of four million people, expect to achieve an effective organizational and managerial structure in specific tasks? How do individual tasks add or subtract from aggregate efficiency levels, and how do individual tasks compete for limited resources? Would the problem of scale be even more aggravated by decentralization?
How to achieve complementarities and synergies leading to superior economic performance, what amount of resources should be dedicated to specific purposes, for example e-government, policy analysis and drafting, or to existing and new services?
Are available resources consistent with the new functions, especially in terms of personnel, can there be comparable specialization of knowledge and skills?
How do rules and procedures contribute to increasing resource efficiency, and where rules block efficiency?
Up to what point is efficiency predictable and measurable, to what precision can we calculate some potential "marginal rate of return" for new resources being added or for resources being redistributed among functions?
How reliable are techniques to value non-material resources (knowledge, skill, personal networks) inherent in the present system? So that, rather than destroy them, we might successfully re-deploy them in the emerging organization.
III. Overview of organizational developments in the private sector during the second half of the XX century
In the business world the decentralization trend came about as a response to intensifying competitive pressure forcing firms to lower operational costs and managerial costs. The search for competitiveness also manifested itself first in the restructuring of firms, then of industries, and finally of the world economy. When technology and markets change, so do processes, and so do organizations. The attributes used to describe the directions of organizational change were: decentralization, delegation, empowerment... while the resulting organizational models were called organic, non-hierarchical, horizontal organizations, implying that central organizational authorities have restrained their powers.
Factors speaking for and against of decentralization in the real private sector were sensed and explored very early the business literature. The concept of decentralization became popular in the managerial literature already during the 1940's, while in the 1960's the advocates of decentralization represented the main stream of managerial-organizational thought. Most authors advocated employee involvement and self-control. In the literature, self-control was interpreted as a mean for increasing motivation and effort on the part of the individual employee and so increase individual productivity.11 Decentralization has also allowed favored entrepreneurial behavior at lower organizational levels.
Concerning operational efficiency, decentralization led to the redefinition of internal control mechanisms. Control systems once dependent on line structures and personalized control were now being depersonalized through the process of creating alternative governance structures, this time based primarily on formalized, mainly financial controls.12
Formalized, standard controls also bear some risks. The risks appear through new possibilities of "fooling" formal controls. By simply by adjusting "the numbers", formal systems could be eluded for extended periods of time. Among the new terms were coined in the process of recognizing the intensifying problem of reliability of control measures, the most important for understanding the issues of governance costs are: "information economics", "incomplete information" and "opportunistic behavior".
Motivational-psychological considerations combined with information economics are to be found in agency theory. Organization theory defines every authority relation as an agency situation, or a situation where one person is given control over resources belonging to another person.13 The economic effect of agency relations reflects itself in three forms of agency costs: monitoring costs, bonding costs and residual loss. All three types of costs are subtracted from the possible maximum efficiency, which might have been achieved by the original pool of resources. The higher the agency costs, the lower the efficiency of the system of governance.
Another contemporary line of theory dealing with incomplete information and opportunism is transactions costs theory. Transaction costs theory revealed how the problem monitoring the behavior of suppliers (and business partners in general) reflects on organizational structure, including the problem the choosing the most adequate modes of control. Transaction cost theory related opportunistic behavior to two new concepts, the concepts of asset specificity and the concept of institutional safeguards.
During the XX century, many industries, especially the technologically most intensive and most progressive ones, in terms of new and innovative products, have undergone through waves of concentration. Eventually, world-scale, dominant business systems rose to action in order to beat competitors in the race to acquire control over most valuable resources. The new organizational model that consequently appeared was a network of firms connected in partnerships of various levels of stability and formalization. The inside structures of those systems evolved through forms of networks of semi-dependent entities. Very often these networks consisted of a dominating firm and its satellites, giving the rise to the same problems once spotted in controlling semi-independent groups and business units inside a single firm. The issues of efficient control now became more aggravated for the fact that often there was no ownership authority links and coordinating the operations of formally distinct business units. Agency costs and transaction costs became relevant for shaping the economic structure of dominating firms while dependant firms wondered about lock-ins and strategic options for nesting in some world-scale value chain.
The very central issue in achieving efficiently functioning enterprise networks relies on the problem of ensuring responsibility, or at least accountability, without being able to rely on traditional governance tools, and so extend the scope and speed the process of acquiring control.
In an enterprise network, control tools still depend on financial controls as standard efficiency indicators. In case of contract relations, additional performance measures define quality, delivery times, responsiveness, etc. Through contracts, dominating firms also achieve less formal and less capital-intensive market mechanisms of control. The treat of being "substituted" by another supplier gives a decent degree of control to the dominating firm. The dominant firm is in the position to direct the behavior of the dependant firm even though there may be no direct hierarchic links, no ownership control, in fact, even no regular personal contact among the formally distinct business entities that compose the network.
One might wonder whether these "extended enterprises" present a temporary compromise, or will they become the dominant organizational pattern in the business world. For one consideration, the rather sudden growth of these business systems left no time, or better, no economically sustainable possibility to establish extricate hierarchies and procedures, especially if the pressures for maintaining the flexibility of operational chains is taken into account.
Compared to the breakthrough in the sixties, the issues that predominated in designing organizations in the last decades were preoccupied mostly with inter-firm relations and only secondarily with the consequences on in-firm arrangements. There were other issues, such a shedding excess staff, attaining flat structures, restructuring along process lines, but always directed towards reducing costs, gaining competitive advantages and nesting inside a value-chain as the most cost-efficient supplier of products and services. Economically, the result of networking was a dynamically efficient business system, broader in size and more compatible with scale economies. Control systems for individual firms inside the network, for a firm technologically and strategically more compact, became easier to administer and more cost effective.
Still, during the eighties and nineties, with more enterprises joining the networks the issues of controlling behavior in the big network showed the tendency of becoming ever more complicated. In its size, defined by employed resources and manpower, and by the complexity of functions performed, the problems in enterprise networks compare to those expected to be encountered at decentralizing public administration. Enforcing desired behavior has become a dominant economic problem both for the business sector and for the public sector.
In big, networked business systems, whit geographically dispersed and technological diverse, dominantly "self-controlled" business entities, the systemof control became more complex (and so more expensive) for several reasons:
Additional standardized performance measures had to be developed to reduce the risks of contracting out,14
Free riding and opportunistic behavior had more chance of passing unnoticed over longer periods of time once the system became more complex,
Switching costs (hold-up problems) had to be taken into consideration. Those are generally higher when associated with bringing in another associate business to the network compared to the cost of hiring new employees.
The same issues of effectiveness and costs of governance are to be evaluated when contemplating the economic consequences of public sector decentralization.
IV. Analogies and comments – Decentralized models of government organization and efficiency implications for Croatia
The system of public administration is presently becoming more aware of the possibilities of decentralization and de-bureaucratization. Two factors to be considered wile looking at the economic effects of decentralization are the costs of governance and the operational costs. Another set of problems relates to the relative share of resources used up public administration, and its comparative efficiency. Yet a third set of problems deals with the costs of changing the organizational model. With no intension to give a complete and final prediction on the overall effect of government decentralization on national competitiveness, the proceeding part of this paper will just note few observations.
A. Internal efficiency
As some geographic areas sense the treat for its products of becoming obsolete in world markets, their governments came under more social pressure to start dealing more actively with national economic efficiency and with national competitive rankings.
However, this manifestation of "competitive pressure" is supposed to be attenuated as we move from central government levels towards lower levels of government.15 In an theoretical line of reasoning, this would mean that once informal, market-like "measures" of efficiency dim, and they do since the market for government services is definitely less than perfect,16 a certain level of efficiency and reliability must be enforced by alternative (bureaucratic control mechanisms, namely by more standards and rules.
The drawbacks of government decentralization lie in this very problem of attenuated exposure to the reactions (cross-evaluation, eventual discontent) by the consumers. Central national authorities are more scrutinized, if for no other reason then because the process of their election is more visible. It is monitored and discussed, both by the political opposition, the press and the regular citizens. On the opposite end, on the level of local authorities, transparency is more questionable.17 However, the advantage of local authorities, especially in small communities is their sense of belonging and tighter social strings with the community (contributing towards self-control and self-motivation), which might and might not make them more motivated to act as to contribute to the well being of their local community. The final effect is that individual performance may vary broadly, from person to person, especially since very few standard performance measures exist at the moment.
Control costs- informed decision making and agency costs
Concerning the cost of maintaining a body of administrators, and considering their efficiency in administering information, the decentralized organizational model is believed to be capable of offering more accuracy in interpreting situations on the "field". Thus the accuracy of information available for decision making should be higher and decisions should be made faster. However, local administrations might be expected to have a better understanding of operational problems, only if there is enough personnel at the local level, with relevant professional training and with enough available (financial) resources to autonomously conceptualize problems and deliver solutions.
Apart from efficiency problem in administering information, the role of local government units, and the role of the civil servant in general, is a genuine agency relationship. Same as higher level managers in business, in order to raise the performance of their LGUs, or even to expend their personal (individual) reach of control, local government administrators, as agency theory suggests, would be inclined to demand more resources than might be optimal for the society as a whole.
The solution of the agency problem, at least the one suggested by textbooks,18 would be to design more elaborate motivation mechanisms that will enforce self-control. Practically, the problem is more complex. For Croatia, since formal controls are underdeveloped, and, since defining and enforcing control standards is costly, public administration are left with professional pride and personal power-status as the dominant modes of controlling opportunistic behavior. For future considerations, self-control based solely on salaries might be costly.19 In a more dispersed and personalized range of salaries, the total amount dedicated for salaries might rise, making public administration costlier, especially is the additional cost is not offset by improvements in productivity.
When non-monetary factors are taken into consideration, some might argue that, on the psychological or motivational level, the simple feeling of being able to make decisions to be achieved by decentralization, would act as a form of individual compensation. The logic behind is that the feeling of personal accomplishment to be provided by empowering local/lower level managers, might to some point, even compensate for the scarcity of financial resources (and small dispersion of salaries).
Still, as and when more functions (decisions) are delegated to lower authorities, control costs are expected to rise as more people will be allowed more (entrepreneurial) discretion.
A good side of decentralization is that the lower the number of hierarchical levels, the lower the costs of salaries for purely hierarchical positions. A positive effect of flattening organizational structure is that the total number of line managers should be decreased, while the number of relatively autonomous professionals (better trained and better informed decision makers) should increase. The correlated effect would also be that with fewer tiers of managers, the range of salary dispersions should fall, thus decreasing the total amount paid in salaries.
At the moment, considering government decentralization in Croatia, the conclusion is that government decentralization will, at least temporarily, demand additional costs. Part of those costs will go for employee compensation. Another part will go for devising more elaborate remuneration schemes able to differentiate salaries and benefits on individual and group performance. But, hopefully, more elaborate compensation packages, especially if based on performance, would also encourage professionalism and raise the quality of public management.
Process restructuring – improving cost-performance and transaction costs
In order to make a reform in the public administration sector successful, the costs of governance and the costs of transforming the system should be offset by gains in operational efficiency.
More autonomy at local government levels might be an impulse for establishing new public-private partnerships. If a new market for public sector supporting services develops, at the rate that more commercially based services appear, they would face competition among themselves producing a favorable effect of market control.
As discussed earlier, more government decentralization should affect the role of public employees and allow them to act with more as entrepreneurs. For, we might define administrators either as decision takers (facilitators) or as performers of public services. In case we see officials in local governments as those responsible for supplying a service to society, and not as those concentrating authority over a specific range of employees (performing the task themselves), we create more space for rationalizing the processes by which public services are delivered. When we see public officers are those concentrating authority over a pool of financial resources, they might be motivated to increase the productivity of these resources if they have the authority to decide among various organizational arrangements.
What we have at hand is the problem of "make or buy" decisions contemplated by transaction costs theory.
In the real sector, the problem of transaction costs is explicitly evident when the restructuring of "business" processes is concerned. So, the issue is whether local governments should have entrepreneurial authorities. Should they be granted the rights to "lead autonomous business strategies" and form the "mix of businesses" that is most appropriate for the moment and the situation.20 A related question is how are these authorities to be regulated by laws and other institutions. Here we have in mind the need to discuss the intensity and form is regulation most desirable considering its implications for efficiency.21
Streamlining processes efficiently and promptly when the opportunities arise, depends on constantly monitoring processes and on depersonalizing the issue of contracting out. When individuals and firms face a treat of competitors that may replace them in case of sub-standard performance, they will behave efficiently even in the absence of perfectly efficient control mechanisms. If we define managers in the public sectors primarily as facilitators of public and not performers, we might have a market control that can complement (or perhaps substitute) for formal administrative control through standard and constant measurements and reports.
If some activities now performed by government employees are to be eliminated and transferred to contractors, either business services or NGO's (education, sports, culture) someone has to identify the opportunity and make the decision of closing down some jobs in one's own department because the outside contractor can perform the job better or at a lower cost. Besides from being an unpopular measure for the public manger to take (the decision on firing someone will not reflect on his salary, but will on his popularity as superior), there is the problem of examining possible future hold-up problems (dependencies on some contractors), and the problem of authority for strategic organizational rearrangements.22
There is also the problem of capacity for searching for new "operational methods", or for new service providers. In order to do so, public officials must be first educated, then sensitize to the problems and finally aware of the market possibilities. Probably smaller constituencies, where there is less personnel, less possibility of employing well trained staff, of sending staff to seminars, but also, there is a smaller work load (lees attractive business to contractors), and dislocation problems (contractors firms or NGOs might not be present in the local market) will be more confined to preserving the "old" functioning models. It is a fact that in the business sector, professionally led BPRs seem to work better for bigger companies with strong emphasis on streamlining processes from the top (the top-down approach).
B. Some notes on the size of government - allocative efficiency
Considering the available data it is not possible to argue for or against allocative efficiency. What can be done is to make some observations on the size of the Croatian government, comparing its share in the available national resources to the share of resources used up by governments in other countries.
An important consideration about the effects of governments displacing resources is that there is no proof that the public sector is making better or worse use of resources than would the private sector.23 So, in fact, in cross-country comparisons in levels of government involvement in the economy (table 1, table 2), we are making assumptions that, if certain countries, perceives as developed (or developing faster than others) had shown some pattern of government spending, other countries, not so developed, should try to adapt their patterns of spending and/or organization to achieve similar efficiencies.
Table 1. Central government expenditures as percent of GDP