Central and Spending Agencies

Whether an agency is part of the cluster of central agencies responsible for fiscal management or part of the administrative agencies primarily responsible for spending, three management approaches are discernible. The first is strategic management, which seeks to maximize results to the government and the community through the best allocation and efficient use of resources. The second is political management, which recognizes the diversity of interests of voters and the consequent competition for resource allocation. Government is a mosaic of agencies and programs amenable to influence. The third is administrative management, which implies that the state is an autonomous and technocratic administration with its own established procedures and rules of governance that determine the operational procedures and techniques for allocating and utilizing resources. The configuration of these approaches differs between central agencies and administrative agencies in times of normalcy and in times of growing uncertainty. For purposes of analysis here, the factors contributing to that uncertainty are ignored, and for simplicity, central and administrative agencies are considered homogeneous categories representing their tasks and interests.

Although the central agencies are many and each has a different administrative territory, macroeconomic stabilization is their unifying theme, and although their influence has been declining, their primary purpose is to achieve an appropriate economic environment in which both public and private sectors play a constructive role—through organizational and market approaches—and provide services to the community. Their outlook and consequently their relationships with administrative agencies are based on their understanding of the availability of financial resources and the desirable level of resources that can be deployed through the government budget and associated instruments. This abstract role is fulfilled through the various stages of expenditure management in which the central agencies function as financial and economic advisers, as budgeteers, as accountants, and as controllers. In theory, they are endowed with a good deal of power. As power is naturally open to abuse and to ensure the respect due to such power, they are obliged to demonstrate that what they are doing is the best or the most expedient course of action. They have to demonstrate that their advice is right. They try to overinsure against all risks (indeed, they see little need to keep such insurance to the minimum), and they therefore tend to be conservative. In allocating resources, they are obliged to depend on the information provided by the administrative agencies. As this information is furnished in a self-serving manner, the central agencies often compile their own information to forecast, to verify, and finally to allot the resources. This information dependency or the degree of independence may determine the empirical underpinning of the analysis of budget bids from the agencies.

The outlook of the administrative agencies is colored in two shades; one is a recognition that resources are limited and that everybody is competing for them; the second is a desire to pursue policies that reflect the interests of the clients in which primacy is attached to pursuing their needs, rather than to limiting resources. To achieve the second goal, information may in the short term be doctored and other relevant stratagems may in theory be used. From the point of view of public choice theory, the main aim of the agencies is to expand their budgets. (Other studies by those in public administration view this aim as naive and too simplistic. In fact, they argue that the main interest of the agencies is protection of their turf rather than expansion. They also suggest that the agencies engage in rent seeking through the various types of regulations, expenditures, and loans, so that benefits can be targeted to the favored clients.)¹ If such self-interest is carried to its logical conclusions, expenditures and other instruments could expand with adverse effects for all, including the agencies themselves. To that extent, predatory self-interest will yield to a modified or more pragmatic self-interest. Once the budget allotments are made, the administrative agencies view them as property rights or entitlements—unalterable and available for them to spend for a specified period.

This description may seem to indicate that the differences between central and administrative agencies are fundamental and irreconcilable. But rules of governance have to be recognized, which aim to ensure a code of behavior that controls conflicts and converts them into meaningful and accountable goals. But the situation changes because the variables that had been taken into account in formulating the budget have changed and the rate of change itself has been changing, with consequent implications for the direction and speed of remedial action. The central agencies may therefore change their strategy, acquire draconian powers, and resort to new management techniques. The administrative agencies can then no longer adhere to the belief that the budget allotments are inviolate and have to either reduce the needs or incur excess expenditures or arrears or other vague approaches that tide them over without explicitly violating the rules. In all these phases, how can compliance be ensured so that each of the agencies does what is expected of it and so that what it does fits with what others are doing? To achieve this objective without resorting to a “command” situation has been the concern of the economics of organizations. The answers to these issues vary. Broadly, however, they may be considered here under (1) traditional organizational techniques; (2) principal-agent relationships; (3) improved organizational design through information systems; and (4) overall organizational development. The theory, its relevance, and general experience are considered for each area.

Traditional Organizational Techniques

The need to improve an organization’s management to ensure it complies with the goals specified has always been at the forefront of organizational research, although definitive answers have not yet been found. However, the search for the underlying trends and attempts to formulate answers illustrate and clarify how organizations work. These issues have received only sporadic attention in economics and are now being pursued more vigorously.² One such recent attempt is the analysis of Simon (1991), who, while conceding that the term “well” reflecting the functioning of an organization is relative and somewhat ambiguous, raises the fundamental issue of what motivates real people in real organizations to perform better. Simon argues that this aspect can be better understood, not by maximizing the utility, but by analyzing some assumptions derived from organizational phenomena. These are authority, rewards, identification with the organization, and coordination. Authority is not always used to command specific actions but endows the decision maker with the power to take the initiative in advancing organizational objectives by taking appropriate decisions in periods of uncertainty. Rewards try to motivate the employees to accept authority and to fulfill objectives. Simon recognizes that if only rewards were used, organizations would be far less effective than they are. Loyalty and complete identification with the organization are more likely to be the dominant factors. He assumes that employees would be docile or receptive to the inculcation of a spirit of loyalty. Coordination is needed to ensure that each individual acts in unison with other links in the program. This coordination is implicit in the authority mechanism of every organization, and the need for it becomes even more relevant in periods of uncertainty.

Simon’s efforts are geared more to the study of generic organizations than to public expenditure management. In applying his approach to expenditure management, some of its unique features come to light. Authority in public expenditure management is specific in matters like payments, custody of moneys, and reporting. In the observance of economy, efficiency, and effectiveness, they become the goals of the organization on the assumption that if everybody does share (avoiding moral hazard), then somehow the goals will be achieved. But as the government framework is much wider, experience shows that responsibility for the broader goals becomes diffuse and may even lose its identity. The use of rewards requires that certain conditions be met, such as measuring contribution visibly. Where, on the other hand, the considerations for rewards are broad, the basis for incentive may lose its integrity. Loyalty is more personal, abstract, and its impact on the achievement of a goal difficult to ascertain. In public expenditure management, given the differences enumerated above on the approaches of central and administrative agencies, employees in the latter may prefer to be expedient and to comply with what they perceive to be in their department’s interest rather than with the overall goals of government. The failures of expenditure control are usually due not to cupidity but to the pursuit of organizational loyalty as perceived by the individual. Coordination is vital for public expenditure management and requires a high degree of ability in all levels of officials. At issue, however, is the latitude to be given to lower-level officials or program managers in the context of uncertainty when the variables are changing and speedy action needs to be taken. Would uncertainty contribute to greater central control?³ If it did, would it not have an adverse impact on loyalty in the agencies? Simon did not address these issues.

Wilson (1989), while recognizing that a sense of mission can be found in public agencies, says that they differ in managing compliance, mainly because of how they are structured for their operations. He offers two considerations for typology of organizations: Can the activities of the organizations be observed? Can the results of those activities be observed? (Output and outcome, respectively.) The monitoring of outputs and outcomes is not always easy as each agency is unique. From this point of view, Wilson proposes a fourfold typology that explains the diversity of organization in public authorities: (1) “agencies in which both outputs and outcomes can be observed” (production organization); (2) “agencies in which outputs but not outcomes can be observed” (procedural organization); (3) “agencies in which outcomes but not outputs can be observed” (craft organization); and (4) “agencies in which neither outputs nor outcomes can be observed” (coping organization) (p. 159).⁴ He adds that management styles also have an impact on the degree of compliance.

Like Simon’s approach, Wilson’s analysis is general and does not seek to integrate the financial or budgetary considerations to ensure compliance. It illustrates the complexity of organization in government and raises issues about the nature of compliance and the extent to which it may dominate the thinking of different organizations. Public expenditure management is thus doubly complex because it has to ensure that the moneys are utilized properly while progress is achieved in program implementation. This “money” dimension has, however, not been explored. This omission is somewhat inherent in Wilson’s analysis; he seeks to find the factors other than finances that contribute to organizational systems. This is not the place, however, to assess the relative merits of finances and other organizational features in contributing to an organization’s success: finances remain a crucial factor.

Principal-Agent Relationship

How can continuing compliance from the various agencies in a dynamic setting be ensured? How can the agencies’ participation be ensured in a way that would accord with the primary aim of macroeconomic stabilization—an avowed objective of the government? These issues have been explored in economic theory of the firm where one or more persons (the principal) engage another person (the agent) to perform some service on their behalf that in turn involves decision making by the agent. It is sought to extend the results of these efforts, known as principal-agent relationships, to the government sector in general and more specifically to public expenditure management.⁵ This approach assumes that all important decisions affecting the macroeconomic health of the country are being taken by the central agencies. The decisions so taken are to be implemented by the relevant agencies, their managers, line and staff personnel, and the clients in much the same spirit as the decisions are taken, that is, their reactions will be economic rather than political. It also implies that the agencies, given some incentives, will forsake their own utility-maximizing behavior of seeking a higher budget but will accept a discretionary one that represents the difference between the maximum obtainable budget and the minimum cost of producing the required services. The issue does not arise when the actions of the agents can be directly monitored by the principal. Only when there is a distance—in the literary and policy sense—between the principal and the agent does the principal-agent theory have to be applied. It is relevant (as with public expenditure management) where the central agencies determine the policy and institutional environment, which the agent accepts as immutable and seeks to maximize its utility within those parameters.

To ensure that the agent acts in the broader interests of the central agencies or the principal, psychological and monetary incentives are provided to the agent so that its actions are on the desired lines. Translating this goal into action requires a number of special features. First, the beliefs that govern policy actions should be symmetrical. If full information on the various facets of the economy is available, the reactions of most policymakers will be more or less predictable. The differences, in addition to those for ideological reasons, are likely to stem from how each decision maker reacts to uncertainty. As Sappington (1991) states, “implicit in the strong assumption of symmetric beliefs is the presumption that both parties are able to anticipate fully all possible contingencies that might arise during their relationship” (p. 48). Second, it is assumed that the agent’s approach to risk is neutral. Any deviation from the result intended caused by a change in the implicit contract would not be the agent’s liability. Third, it is assumed that the agent can carry out the policies without any cost to him or to his agency. Fourth, the activities of the agent can be monitored by the principal in a transparent manner. The agent may resort to doctoring his estimates, but direct monitoring would in due course prevent a reoccurrence.

The application of this theory to public expenditure management has several limitations. The first is the viability of the discretionary budget. If the central agencies were aware of the inherent slack in budgetary provisions, or of the minimum cost needed to provide a service, the budgetary allocation would be at that minimum level. The second is the notion that a single principal deals with budget making. While this is true in some industrial countries and other countries still practicing central planning, in a number of other countries the responsibility, and consequently the approaches, are still fragmented. Too many principals exist. Third, principals and agents differ on their respective roles. In a revenue shortfall, the principals may expect the agents to adjust to the new realities and to scale down services or to maintain them at reduced levels. The agents, on the other hand, may value stability once the budget is approved and may be totally risk averse because they consider the budget provision a property right. In the end, this partly depends on the relative power of the principals and the agents. Also, the theory does not offer sufficient guidance on the optimal sharing arrangement between the principal and the agents. Although the available literature provides insights into the working of relatively simple organizations, applying them to the government as a whole remains problematic.

An essential first step is to have an organizational format that will facilitate the application of the principal-agent theory. Experience in this regard is limited to the United Kingdom, where, since 1988, several organizations have been formed as agencies to provide them with defined tasks and managerial autonomy to carry them out.⁶ In each case, before an agency was established, the need for its activity was reviewed and alternative steps including privatization were explored. Once it was determined to be a government activity, an agency was created with objectives, performance targets, powers, flexibility, and the accountability of the chief executive as well as the regimes for financial and personnel managers. In addition to providing autonomy, specifying tasks and performance targets are expected to bring the approaches of the principal and the agent closer together.

Incentives, however, cover a wider area and are applicable to at least four different areas—central/state financial relationships; relationships between central and administrative agencies; relationships between administrative agencies and their subordinate organizations; and financial relationships between government and public enterprises. China is one of the few countries that has applied the theory of principal-agent relationships to regulate the financial relations between the Central Government and its provinces.⁷ Since 1988, a system of contract budgets under which the provincial governments collect taxes and retain a specified or increasing percentage over the amounts to be transferred to the central governments has been in force. The revenue-sharing arrangements are intended to provide an incentive to the provincial and local governments. Although the system was to have been continued after 1990, some modifications were introduced because interprovincial disparities were widening, and also because of the growing revenue needs of the center itself.

Two features of the experience of Australia, New Zealand, and the United Kingdom with relationships between central agencies and administrative departments are worthy of note.⁸ The first is the relaxation of traditional input controls in favor of an amalgamated category of running costs as a single item comprising salaries and associated costs. The second is the central determination of a fiscal dividend to take productivity gains into account. Savings over and above the specified efficiency dividend are retained by the management of the agencies to provide an incentive for greater efficiency. Less is known, however, about extending these principles between the administrative agencies and their subordinate organizations. For public enterprises, China offers a unique example of the application of principal-agent relationships through what is known as a contract system under which wide-ranging incentives are provided to the enterprises. Granick (1990) provides a well-documented study of the experience and its major contribution toward a smoother transition to a market economy.⁹

Although experience with incentives is limited, it indicates that there is room for extended application as well as for more detailed study of their impact during volatility and stress. Experience in a number of cases suggests that guarantees cannot be affected in terms of providing a degree of stability in budgetary resources. If such guarantees were to be provided, they would contribute to a higher fiscal deficit that might have more adverse consequences than those accruing from the suspension of any implicit guarantees. No conclusive pattern can yet be discerned from evidence available on the relationships when the agencies cannot be assured of stable budgetary allocations. Another feature integral to the art of budget making is that most problems experienced during implementation are those that were avoided during budget formulation. Principal-agent relationships require these issues to be addressed during budget formulation, and, if they are not, it is hoped that their impact will not exceed the “contingency” amount included in budgetary plans.

Improved Organizational Design: Role of Financial Management Initiative (FMI)

The importance of achieving a budgetary outcome corresponding to budgetary intentions prompted a revival of the role of administrative agencies and the need to strengthen their overall financial management system. The major issue was the need to generate a commitment within the agencies that accorded with the general approaches of the government as a whole, and more specifically with the short-term goals of macroeconomic stabilization as enunciated through the budget. It was to conform to the specified parameters so as to bring out the agencies’ best creative abilities in implementing budgetary policies. Experience indicated that command-oriented controls had not been adequate to deal with the fiscal crises that dominated much of the seventies and eighties and that continue to dominate policy approaches in the nineties. The issue, as posed by Metcalfe and Richards (1987), was how to live with budgetary standards and yet be motivated by them (p. 180). Compliance depended on motivation, and managerial autonomy contributed to the introduction of improved organizational design in the administrative ministries, departments, and agencies through the financial management initiative (FMI) in the United Kingdom. The British Government’s experience with FMI is instructive in several ways.

Introduced in 1982, FMI represented an essential link between the Government’s overall economic policy and the day-to-day pursuit of efficiency in departments and other agencies. The objectives of the initiative were

to promote in each department an organisation and a system (emphasis added) in which managers at all levels have:

(a) a clear view of their objectives; and means to assess, and wherever possible measure, outputs or performance in relation to those objectives;

(b) well-defined responsibility for making the best use of their resources, including a critical scrutiny of output and value for money; and

(c) the information (particularly about costs), the training and the access to expert advice which they need to exercise their responsibilities effectively.¹⁰

The initiative stipulated in addition that budget implementation requires objectives to be specified in detail so that program managers can ascertain how much can be spent, why, and what benefits will accrue. These aspects depended on better information as a basis for action. More important, managers were to have a major role in composing their budgets and greater freedom in managing them.

The introduction of FMI should be seen as the culmination of a series of earlier efforts, and as an illustration of an innovation that drew heavily on practices in the commercial world—and in particular on the earlier experience of the Management Information System for Ministers (MINIS) in the Department of Environment in the early eighties to find out “who does what, why, and what does it cost?”¹¹ MINIS showed that ministers and civil servants would act more as managers of the resources under their control through the close review of developments that the system provided. Commercial experience showed how delegated budgets would contribute to improved performance. In comparison with the past system characterized by excessive fragmentation and a consequent loss of motivation and identification, the new one offered them a well-defined responsibility for resource allocation planning and resource utilization. Another feature—managerial autonomy—was also drawn from the experience of the commercial sector. The agent works within the overall parameters of supervision and monitoring by the top management. If the information was computerized and all the relevant parties had access to the same information, it reduced the hierarchical distance between the top management and its agent. Also, managerial autonomy seeks to build a solid bridge between tasks and responsibilities.

When these aspects are applied to the government, what is being done as part of the budget and how it could be changed has to be identified. In other words, the manager (in the agency) in composing and implementing his budget will have to reckon with the degree of flexibility inherent in the situation. For this purpose, a distinction was drawn between administrative expenditure (direct government consumption) and program expenditure (expenditure largely external to government like transfer payments, grants, or procurement contracts). For the administrative expenditure, budget allocations were divided into blocks, and managers were made accountable for costs within their blocks. Thus, the technique of running costs covering all the administrative expenditure was developed, and central controls moved away from specifying civil service numbers to the overall control of running costs. Similarly, for program expenditure, particularly for transfer payments, FMI sought to clarify the nature and extent of each manager’s responsibility and the results, and the relationships with other agencies where the spending authority is outside the manager’s purview.

FMI contributed additional stimulus to delegation, managerial autonomy, and to the application of a type of management accounting aimed at measuring and monitoring costs. It also spurred the establishment of information systems that facilitated the review of goals and their achievement. It depended on the development of management systems and their translation into public expenditure plans delegated down the line. This approach, on the surface, may remind the former centrally planned economies of their previous systems. Fundamental differences exist, however, because costs computed as a part of FMI are real and not artificial as was so with the standard costs of centrally planned economies. The management information system also differs from the statistical systems in centrally planned economies, which were intended primarily for the central planner rather than for the managers. The monitoring and assessment of targets, outputs, and achievements takes place in FMI in an open and accountable way that had no counterpart in the centrally planned systems. More important, responsibility for setting and reviewing their own budgets is with the managers rather than with an all-powerful central planning agency.¹²

Overall Organizational Development (OOD)

Public expenditure management is an integral part of the overall management of public services. The assets or liabilities of the latter affect the former in its sphere of work. Viewing public expenditure management as separate, in spirit and content, may hinder progress. The umbilical cords between the two should be recognized, and any improvement in expenditure management should be made as part of the overall management of the public sector. This approach was adopted in a series of studies by the office of the Auditor General of Canada during the eighties.¹³ Although the Canadian effort was initially viewed as an exercise in obtaining value for money, it soon broadened in an attempt to understand the constraints affecting the work of government organizations. It symbolized an extension of the normally perceived role of audit as critic to that of constructive ally.

This inquiry covered organizations, systems, and values. In the first stage, it identified three major factors in the work of government organizations:

• the impact of political priorities on productive management;

• the negative impact of the numerous administrative and procedural constraints; and

• the influence of disincentives on productive management.

It was recognized that political priorities overshadow and even displace productive management. The politician’s interest in seeking support for his initiatives may conflict with the pursuit of prudent fiscal management. Efforts to reduce personal costs and operational costs can be incompatible with political priorities. Good government does not necessarily mean good management. Similarly, the numerous administrative and procedural constraints, however well intentioned, may frequently be counterproductive in their impact. Managers in administrative agencies see themselves in a melting pot with conflicting expectations and demands made on them by central agencies and special agencies. Thus, while working within the overall financial regimen specified by the central agencies, they have to comply with the personnel requirements of a civil service board, legal specifications of legal agencies, the environmental considerations of public works, and the procurement specifications of general administrative services, and find a cost-effective formula for providing a service. The issue is not the conflicting demands of the agencies inherent in government, but the persistence of outmoded or long-winded regulations.¹⁴ Disincentives, the Canadian study pointed out, overwhelmed the few incentives and were widespread in expenditure management. Personnel controls were on the number of posts rather than on the salary budget (since addressed in a few countries, as noted earlier). Lapse of funds was a stigma, and there was every incentive to spend regardless of value for money. Across-the-board cuts treated the profligate the same as the thrifty; government-wide instructions and regulations can be disincentives to some departments. These features emphasized the unique tasks that an organization had to contend with in a government.

Identifying the above problems was only a preliminary if essential step. How were they to be avoided? And what can be learned from those organizations considered to be well performing? The 1988 study identified the features of a well-performing organization. It stated that emphasis on people (explicitly recognizing that people are the key to high performance), participation leadership (abandoning command or coercive leadership), innovative work styles (monitoring, feedback, and control systems), strong client orientation, and a mindset that seeks optimum performance were the key factors. But examining the features of an organization or its systems and structures is unlikely to be rewarding if the people, their dominant values, and their role are not identified. An organization spawns an organizational culture that is the net product of the interaction between people and systems. Such values, which are intangibles, may not always be measurable when they contribute to success, but are clearly visible in a failure.

Accordingly, the Auditor General of Canada made a study in 1990 of values, service, and performance in the public sector. While recognizing that values have different meanings, the study defines them in normative terms of what the goals and standards of an organization should be and points out that a balance is needed between managing by control and compliance and managing by caring and commitment. At an agency level, the need for a balance is inherent because it has to interact with its clients as well as with its principals. For a central agency, the need for controls is also viewed as inherent as controls provide accountability. Therefore, the central agencies see a uniform framework of controls as both inevitable and desirable. Both groups are sincere in their views, which are based on what they perceive and how that perception is internalized in the day-to-day working of the organization. But controls may conflict with commitment. The study (1990) points out that “instead of trying to attain the elusive goal of flawless operations through controls, and thereby maintain public support, the suggestion is to explain to the public, that such perfection is unattainable… it is possible to come closer to perfection by managing through understanding” (par. 7.64). In developing this framework for management through understanding, the study notes that conversion of the entire public service would be impossible. Accordingly, it suggests a selective application or what it calls “islands of innovation.”

These studies seek greater understanding of the complexities of public sector management and recognize the difficulty of converting into reality the Glassco Commission’s dictum, “let managers manage.”¹⁵

An Assessment

There is an African proverb to the effect that “until the lions have their own historians, tales of hunting will always glorify the hunter.” While no definitive history is being written from the point of view of the administrative agencies, as the preceding discussion illustrates, the use of command and coercive power has yielded to the use of cooptive power both in literature and to a lesser extent in practice. The growth in government in the last two decades has been such that even the benevolent despot envisaged by Wicksell would have a hard time imagining the numerous kinds of action required at various stages. As the economic and organizational contexts varied, so did the connotation of compliance. From a somewhat narrow adherence to compliance with the budget as formulated, it moved into management with flexibility so that the agreed objectives could be achieved with economical and efficient use of resources. The relationships between central and administrative agencies became less coercive, although greater centralization emerged in some cases, which was viewed as an appropriate response to the uncertainty. Regardless of the merits of centralization, continuing that approach, even after the fiscal crises abated, had been much debated. Providing autonomy to the spending agencies is not necessarily at the expense of central power and control. Such autonomy presumes a change in the nature and content of control from a seat-of-the-pants approach to a more strategic approach based on recognition of the organizational needs. The approaches described above have brought a new awareness of the fiscal responsibilities in the agencies, but that awareness does not necessarily mean greater compliance. The strength of these approaches lie in their capacity to permit a clearer understanding of not only the symptoms but also the substance, so that the latter can be addressed.¹⁶

The themes discussed above have a commonness about them even though their applicability is different. Together they offer a means of converting relevant principles into meaningful practice. Whether applying these principles and packaged approaches would help public expenditure management or whether they in turn would become obsolete can only be determined in the future, which, as Polanyi (1951, p. 199) stated, is by its very nature beyond our comprehension and informed humility.