New Government-Society Interactions Edited by Jan Kooiman SAGEPubhcations London • Thousand Oaks ■ New Delhi PUBLIC MANAGEMENT: FROM IMITATION TO INNOVATION LES METCALFE Public Management in the 1990s During the 1980s public management became a growth industry (Kapil and Kekkonen, 1990). There is now a world-wide movement to upgrade public administration by using management concepts, tools and techniques, many of which were originally developed in business. But public management wiii not flourish as a growth industry in the 1990s merely by extrapolating from the approach on which its initial successes were built. New and different problems of public management are emerging for which there are no ready-made answers. There is already a worrying tendency for solutions to be more predictable than the problems they are supposed to solve. Privatization, contracting out, formation of agencies, creation of internal markets and the use of a whole battery of specific management techniques are almost reflex responses to the problems of governmental underperformance. A fundamental reappraisal is urgently needed if reform is not to narrow into this self-limiting process of relying upon conventional business solutions. Before investigating alternative solutions we need to be sure that the key questions are correctly formulated. The agenda for reform must be revised to take account of distinctive features of public management. Public management reform is not just an apolitical quest for technical efficiency, within a given institutional framework. Public management in the 1990s will have to develop new concepts and models of governance which alter policy processes, redefine the criteria of policy effectiveness and transform the patterns and processes of public accountability as well as deploy more familiar management techniques. The theme of this chapter is that we are reaching the limits of useful application of business management practices in government but we have hardly scratched the surface of public management proper. Public management cannot meet the needs of government if it remains little more than a collection of second-hand business management methods. The real challenge is a different one. Public management needs to be recognized as the main area for creative developments in the theory and practice of management. Governments, by accident or design, have far more difficult management problems than any business would attempt to tackle. If the basis of public management reform until now has been imitation of business management, the challenge for the 1990s is to move on to innovation; the development of new methods of management appropriate to the distinctive needs of government. Les Metcalfe The challenges facing public management stem from two interrelated features of contemporary government; the pace and type of change with which government must cope and the fact that government operates tiirough networks of interdependent organizations rather than through independent organizations which simply pursue their own objectives. If there is a single contrast between the traditional concerns of public administration and public management modernization, it lies in the shift of emphasis from stability to change; from keeping the ship of state on an even keel, to ensuring that flotillas of organizations are equipped to steer a common course through rough and sometimes uncharted waters. Public managers must not only manage the work of their organizations efficiently, they must also participate in a more strategic and political process of managing large-scale structural transformations in the networks of organizations through which public policies are developed and delivered. It is a matter of personal preference whether one regards this as politicizing management or managing politics. Public management and politics are inextricably intertwined in the process of managing change at the interorganizational level. Dror (1990) coined the term 'reformcraft' to highlight the importance to government of developing new skills and capacities for governance appropriate to the difficult and demanding task of steering strategic changes. This chapter considers what reformcraft involves in contemporary government where the critical public management problems are interorganizational in scope, while many of the standard business management solutions are pitched at the level of the single organization. The basis on which management modernization was launched in the 1980s is not obsolete. It still has some value. But it is inadequate to answer the crucial contemporary questions. The core values around which reform evolved in the 1980s are usually summarized as the 'three Es' -Economy, Efficiency, Effectiveness - denoting the eradication of waste, the streamlining of services and specification of objectives to ensure that resources are targeted on problems. Within this mind-set, imitation of business methods seems the rational management response. But in the 1990s the pace of change, and the complexity of the tasks facing governments mean that the 'three Es' are no longer a sufficient basis for reform. They presume a degree of constitutional and institutional stability that cannot be taken for granted. In the turbulent contexts of Eastern Europe and the Soviet Union, this is obvious. Less dramatically, but no less importantly, in the European Community and Canada where federalism is an issue for markedly different reasons, problems of large-scale institutional change are also prominent. But even where the basic framework of government is not an issue, the challenges of governance are severe, in many policy fields; the economy, the environment, education, health, welfare, transport, urban planning; the problems are outrunning the capacities of the organizations charged with dealing with them. The Fublic management: from imitation to innovation 175 common underlying theme in these different situations is the need fornew capacities to handle structural change and manage reforms at the interorganizational level. Bui 'reformcraft', at this level, is in short supply. Just when it is vital that organizations work together, crisis conditions undermine the trust and mutual confidence on which coordinated action depends. Paradoxically, as the discussion below will show, the business management ethos may make things worse radier than better, by reinforcing go-it-alone attitudes and weakening incentives to coordinate. However, this is not an argument for throwing the management baby out with the bathwater. Rather, it points to theoretical shortcomings which it has been possible to gloss over until now, but which become an increasingly powerful constraint on reforms. Having borrowed heavily from business, public management lacks an adequate theoretical underpinning. It lacks a logic of its own. Filling this gap involves going back to basics and addressing a set of conceptual and theoretical issues which, for a variety of reasons, have been mishandled. These include some of the public choice assumptions on which advocacy of business management solutions ostensibly depends. Paradoxically, their reliance on market-oriented prescriptions has diverted the attention of management reformers away from the governance problems which public choice theory was originally developed to resolve. Public and Private: Locus or Focus? A pivotal issue in the public management debate is the meaning of the public-private distinction. Generally, this is regarded as a matter of specifying, on efficiency grounds, where the boundary should be between public and private sectors. While this has the appearance of objectivity, such distinctions are not mere products of empirical observation and technical analysis, they are socially constructed means of organizing experience and, in this case, guiding policy (Malkin and Wildavsky, 1991). They provide us with ways of perceiving the world and interpreting experience. They exert a powerful, but subüe, influence on how we see the world and seek to shape it. The reason for laying such stress on this is that imitation of business management practices rests on a particular way of making the public-private distinction which, on closer inspection, turns out to be seriously flawed. The flaws have been brushed aside, not without some justification, by reformers in their haste to implement ready-made solutions. If the private sector has solutions to public sector problems, why not imitate them? Government presents such a large target for management reform that almost random attacks can quickly produce some positive results. In operational management and policy implementation, there is ample scope for a piecemeal approach and little reason to delay in taking action. However, the political attractions of early successes mask the danger of compromising longer term reform objectives. Sustained progress depends 176 Les Metcalfe on something more than a succession of ad hoc interventions. At a minimum, public management requires a framework to guide and consolidate different contributions. Yet, this minimum condition has been regularly, if inadvertently, disregarded because of a widespread tendency to confuse and conflate two different ways of drawing the public-private distinction which are part of a broader debate in the public administration tradition. Henry (1975) distinguished between 'locus' and 'focus' starting points for reform and research in public administration; the one institutional, the other analytical. 'Locus is the institutional 'where' of the field. A recurring locus of public administration is the government bureaucracy, but this has not always been the case and often this traditional locus has been blurred. Focus is the specialized 'what' of the field. One focus of public administration has been the study of certain 'principles of administration' but again, the foci of the discipline have altered with the changing paradigms of public administration' (Henry, 1975: 318). The distinction between locus and focus is useful for analyzing different ways of conceptualizing management responses to public policy problems. For clarity of exposition, the term public sector management will be used for a locus approach and public management for a focus approach. The former is concerned with government in the structural sense of a grouping of institutions and the latter with governance in the functional sense of a process of steering and conducting policy. Reformers and researchers, have created confusion by assuming a locus approach embodies a consistent theoretical focus. Reformers, in particular, dogmatically assume that public sector management problems are sufficiently similar to those of business to apply the principles of private sector management. Researchers often adopt the more sceptical and, on the face of it, open-minded attitude that empirical studies will uncover the theoretically distinctive features of public sector management if there are any. In practice, investigations with a locus starting point are inconclusive. Efforts to identify empirical similarities and differences between management in the public and private sectors lead nowhere. The answers are not 'out there' waiting to be found. Any assumption of two distinct and homogeneous sectors rapidly evaporates. The boundary between public and private sectors, which the locus approach assumes, is blurred and ambiguous. Different criteria, such as ownership, legal status, use of public funds and profit motivation produce overlapping definitions of the scope of the public and private sectors, introducing a voluntary sector, non-business and non-government adds to the confusion. To accommodate these disparities the initial sharp dichotomy between public sector and private sector management is often watered down into a spectrum or continuum embodying degrees of publicness. For some authors, 'all organizations are public' (Bozeman, 1987). The task of making comparisons is further confounded because many organizations do not fit neaüy into one or other 'sector'. Marginal Public management: from imitation to innovation ill cases, intermediate types and hybrid organizations variously labelled as quangos or para-state organizations, proliferate haphazardly. Just to add to the confusion, there is clear evidence that the organizational diversity of the public 'sector' is as great, if not greater, than that of the private 'sector'. Not surprisingly perhaps, in view the lack of coherence in public sector management, eclecticism rules. Surveying the literature, Gunn (1987) identified six contending attitudes which he summarised as: • Public administration is unique. • Public and business management are alike in unimportant respects. • Public management is the integrative paradigm. • Convergence between public and business management. • Management is generic. • Public management is a less efficient form of business management Advocacy of these different interpretations of the extent to which public and private management are alike or unlike has created a debating society atmosphere. Well-worn argument and counter-argument are arrayed against each other in ritual combat, without a basis for determining the importance of similarities and differences. The protagonists see little need to adjust to each other's point of view. It is symptomatic of the disordered state of the field that Perry and Kraemer (1983) entitled their compendium 'Public Management: Private and Public Perspectives'. 'Those who advocate that public management and private management are similar and their counterparts who argue that they are different approach this issue from vastly different perspectives or biases.' (Perry and Kraemer, Í983: 55). Some may regard this as healthy academic pluralism. My own view is that it is an unnecessarily confused mess. Progress is blocked because basic concepts remain ambiguous. The same terms are used in diverse and inconsistent ways, providing no basis for cumulative theoretical progress. This leaves government dangerously vulnerable to ideological fads and business management fashions - and the public badly served. In short, the belief that there is a public sector in which public organizations have a clearly distinguishable set of management problems is undermined by those who seek to use it. The 'locus' approach to drawing the public/ private distinction fails because the world does not fracture along a clear institutional fault-line, dividing the public sector from the private sector. Fortunately, the alternative starting point; treating the public/private distinction as an analytical 'focus' rather than an institutional fact; provides a much more fruitful way forward. Most of the remainder of this paper is devoted to outlining the set of problems and processes of public management which provide this theoretical focus. Some of the analysis is familiar in the sense that it starts from the problems of collective action or public choice which have had such an influence on the direction of 178 Les Metcalfe administrative reform. Where it takes a new direction is in developing management solutions from those tiiat have become conventional wisdom. Management; Taking Responsi busty for the Performance of a System There is a prior question to be addressed before defining the differences between public and private management. We need to consider what they have in common and how, in turn, die common characteristics of management distinguish it from other images of (he public policy process. This is no more than an application of the basic logic of comparison (Sartori, 1970; 1991). Several definitions are possible but, to begin with, management will be defined as taking responsibility for the performance of a system (Metcalfe and Richards, 1987). Responsibility is an essential attribute of management. Getting things done through others, to quote another definition of management, means that some individual or group must be invested with the right and duty to ensure that group efforts are coordinated. This definition distinguishes a managerial conception of the public policy process from the incremental and rational images which dominate theoretical and practical discourse about governance. Neither gives central importance to responsibility. On one side, in the incremental ist image of the policy process no one is effectively responsible for the overall performance of systems governed by the interplay of different interests. Groups and organizations pursue their own objectives in an organizational ecology. The parts are managed; the whole is not. The performance of the system is an unintended consequence of the interactions of the parts. Markets and many political systems rely upon the disjointed incrementalism underlying Lindblom's (1959) 'science of muddling through' to coordinate their activities. On the other side, control rather than responsibility underlies rationalist images of the policy process. A unitary model of governance underpins rational decision-making with performance standards set at the top and sub-units acting merely as agencies carrying out policies determined above. Rationalist prescriptions usually assume the feasibility of hierarchical control as the precondition of shouldering responsibility. Although management and control are often used as synonyms, there is a vital distinction between them: to manage is not to control (Landau and Stout, 1979). Control in a strict sense presumes an ability to determine outcomes; control is possible if objectives are well-defined, well-ordered and stable, and the techniques needed to achieve them are proven and reliable. Predictability is the condition of control. Routine is the servant of control. It is when non-routine responses are needed that control breaks down and management qualities; ingenuity, creativity, risk-taking, conflict-resolution; come into their own. Management is an intermediate category. It involves acceptance of responsibility for steering a system when control is not possible but Public management: from imitation to innovation 179 unguided ecological processes produce substandard results. Management is an adaptive process which proceeds by experimentation, learning and innovation in a changing environment. It is not the enactment of a preconceived programme or the enforcement of rules. Management is a more widely applicable concept than control in government precisely because public policies require intensive and sustained cooperation among many organizations. By the same token, it is difficult, because it requires common action where there are diverse and possibly divergent organized interests. These comparisons suggest a first approximation to the distinctive character of public management. If management in general is getting things done through other people, public management is getting things done through other organizations. Public policies typically engage the efforts of many organizations. Public management usually involves the complex and delicate task of taking responsibility for steering an interorganizational network. Coordination among organizations which are formally autonomous but functionally interdependent is one of the keys to ensuring effectiveness in public management. The next section considers how the public-private distinction should be formulated to illuminate the problems of managing in a multi-organizational context, where goal consensus and unitary authority cannot be assumed. The Public-Private Distinction: Macro and Micro Management The challenge for public management is to bridge the gap created by the emergence of interorganizational tasks and public policy problems where central control is impossible but completely decentralized adjustment processes among organizations are too slow and reactive to achieve satisfactory performance. The problematique of public management is most starkly revealed in the logic of collective action or public choice, of which Olson's (1965) analysis is probably the best known. While the theoretical focus of this problem is quite familiar, the form of an effective solution is not. Reduced to essentials, the problem is that in certain situations private interests override public interests and individually rational actions produce collectively irrational outcomes. Rational, self-interested individuals will choose not to cooperate; first, because they do not expect others similarly motivated to cooperate and second, because if others do display cooperative behaviour, non-cooperative free riders will benefit without contributing. The pessimistic conclusion of the logic of collective action is the lack of collective action. All are worse off than they would have been if they had cooperated. Yet, given the circumstances, each does what seems the best for themselves by pursuing their private interests. The interests of each diverge from the interests of all. 180 Les Metcalfe Public Goods and Private Interests The source of the problem of collective action resides in the nature of public goods. Public goods are often, wrongly, reified as goods and services produced by government. This reification is founded on, and r sustains, a locus approach and therefore assumes the validity of the traditional distinction between public and private sectors which was discarded earlier. As Malkin and Wildavsky observed; 'if public goods should be provided by the government, then the answer to the question, 'what is a public good?' is the equivalent of deciding, 'what should the government do?' When to classify is to decide, definitions have powerful implications for public policy'. (Malkin and Wiidavsky, 1991: 357-58). Bearing these strictures in mind a clear theoretical focus is needed. At least in this chapter, what distinguishes public goods is not who produces them. It is the peculiarly difficult problems of producing them and the harsh consequences of failing to do so. Two characteristics distinguish public goods from private goods; joinmess of supply and the impossibility of excluding others from benefitting from them if they are provided (Mueller, 1979). Joint action is necessary to produce public goods. Individuals cannot provide them for themselves. But, once produced, no-one can prevent free riders from benefitting. Stripped of all substantive content, this is the heart of the problem of governance; the delicate balance between shared public interests and individual private interests. Public choice theorists, having posed this problem, show little real interest in finding constructive solutions to it. More often than not, they evade the issues it raises. Having established that individualistic pursuit of private interests does not satisfy the requirements for producing public goods, they use the intractability of public choice problems to justify their own preferences for reducing the scope of the public sector, rolling back the frontiers of the state and increasing the sphere in which competitive market forces and commercial incentives operate free of government regulation. They seek to absolve government of responsibility for dealing with these problems by shifting them from the public sector to the private sector. But in doing so they also shift from their original theoretical focus, which is clear, to a locus approach which is both unclear and misleading. Familiar 'trust the market' reform proposals like privatization, contracting out, deregulation and imitation of business practice in government are too narrowly conceived to provide solutions to the urgent and important problems of collective action such as environmental protection, economic reconstruction, regional development or the reform of major services of the welfare state like education or health. However, disregarding these very difficult problems does not make them disappear. The popularity of business management solutions of these kinds depends far more on ideological preconceptions than on logical inference. The conclusion that the remedies to collective action failures lie in more competition is quite at odds with the fact that the pathologies of public choice stem from unbridled competition in the first place. Simply assuming that a complex Public management: from imitation to innovation 181 mixed-motive situation can be transformed into a straightforward zero-sum game in which purely competitive behaviour is socially beneficial, is as unsatisfactory as naively assuming that cooperation will emerge spontaneously in interorganizational networks where there are severe competitive pressures and acute political constraints. The real question is whether collective action problems can be managed to further vital common interests. An affirmative answer to this question depends on formulating the public-private distinction in a new way. The most fruitful way of doing so is as a sharp analytical distinction between macro and micro levels of management rather than a fuzzy institutional distinction between ill-defined sectors. In these terms, public management at the macro level is concerned with the performance of a whole system and private management at the micro level with the performance of the parts. The analogy with the macro-micro distinction in economics is obvious but incomplete. Macroeconomic management, whatever its limitations, is basically a stabilization process. Public management as a macro-process is mainly concerned with change and especially with structural change. It is in phases of major reorganization that the balance between public and private interests is most precarious and a macro management process is most needed to steer the transition to a new institutional framework. For clarity, it is useful to regard macro-management as concerned with steering interorganizational networks and micro-management as steering individual organizations. The more interdependent the functions of organizations become, the greater the probability that collective action problems will arise which are irresolvable at the micro level. When they do, a macro management process is required to ensure coordination. This does not mean that macro management should displace or supplant micro management. To a considerable degree, individual organizations can retain autonomy in dealing with other organizations within the existing institutional framework. The distinctive focus of public management is on achieving cooperation among organizations in circumstances where the basic framework of public policy and interorganizational cooperation comes into question. These circumstances must be defined more precisely. The acute problems of achieving cooperation under adverse circumstances have been explored in several social science analyses conceptually equivalent to the logic of collective action. They highlight different facets of the problem and, indirectly, provide pointers towards their resolution. The tragedy of the commons (Hardin, 1968) is directly relevant to the over-exploitation of the physical and biological environment. The congestion of the social environment in cities and transport systems through the mal-adaptive ecology of micro-motives (Schelling, 1978) and the social limits to growth (Hirsch, 1977) centre on the problems which arise when there are macro problems but no macro competencies for dealing with them. The well known difficulties of establishing trust in a context of uncertainty and mutual dependence such as the prisoners' dilemma game and Mintz's (1951) analysis of the social psychology of panics and crises 182 Les Metcalfe reveal how suspicion of others' intentions (whether or not it is well-founded) triggers a vicious circle of devil-take-the-hindmost competition and destructive conflict Finally, and most directly relevant to public management, there is the problem of turbulent organizational environments (Emery and Trist, 1965; Metcalfe 1974, 1978). In contests of high interdependence the performance and effectiveness of any organization depends to a large extent on how its activities mesh with those of other organizations. Go-it-alone strategies disrupt the framework of mutual expectations and undermine the institutional ground rules on which continuing cooperation depends. The more complex the causal texture of the organizational environment the greater the risk that individualistic actions will precipitate destructive conflict and disintegration. These variations on the theme of individually rational actions combining to produce collectively irrational outcomes display important continuities which support the idea of a macro-micro distinction. First they make it clear that public goods are not merely private goods produced by governmental means; they are qualitatively different. They are of a higher logical type (Bateson, 1972; Watzlawick, Weakland and Fisch, 1974) than private goods. Second, more concretely, public goods provide the macro context of governance within which private goods are produced. They are necessary complements rather than competing alternatives. The distinction is akin to the relationship between the rules of a game and the strategies of individual players within the game. Third, changing the rules of the game is a different order of change from the mutual adaptation of individual organizations to each other in the normal run of business. It is structural change rather than incremental change and instead of contributing to the maintenance of an existing pattern of interorganizational relations precipitates a process of redefining the roles and redesigning the relationships among the participating organizations. Incremental and Structural Change This interorganizational perspective fixes the theoretical focus of public management innovation on the issues of structural change that the public choice approach tries to evade by reducing structural problems to incremental problems. These different orders of change must be clearly understood, because the consequences of failing to distinguish them in practice and manage them appropriately are extremely serious. Incremental change fits the 'muddling-through' model of the public policy process. It both assumes and maintains macro-stability. Structural change is change of a. system rather than change within a system. It involves the modification of the rules of the game that frame organizational interaction, define organizational roles, responsibilities and relationships and regulate interorganizational competition and conflict. In government-business relations, macro management involves the design and operation of what are usually called regulatory regimes. The failure to construct and manage these Public management; from imitation to innovation 183 regimes effectively underlies, for example, the US Savings and Loans catastrophe and the collapse of BCCÍ and the Maxwell empire. In each case the cumulative impact of commercial exploitation of a weak institutional framework precipitated an enormously costly and unmanageable crisis. It is only at the macro level that it is possible to gain a full picture of structural problems and manage the transition to a new institutional framework. But who should take responsibility for the performance of the whole system? It is tempting, but mistaken, to see the solution as some form of central direction. But, structural change in pluralistic systems is not amenable to central control. Ascribing to public management the task of dealing with macro organizational problems is not a return to a master-minding synoptic decision-maker. Since individual organizations cannot resolve these problems by acting independently and solutions cannot be effectively imposed by outside interventions, ways must be found of sharing responsibility for managing stnictura! change. An effective public management process must embody a capacity for joint decision-making and action specifically dedicated to the task of steering structural change at the macro level. Public choice theorists will resist this conclusion, which certainly implies more govemance,though not necessarily more government. But it is artificial and arbitrary to exclude the possibility that the participants in the system can make collaborative efforts to resolve structural problems by remodelling the macro-environment in which micro-decisions are made (Buckley, Burns and Meeker, 1974). Public Management as a Macro Process The difference between imitation and innovation in public management is now clear. The imitative task of management reform in government is to adopt and adapt business or other management ideas to upgrade micro-organizational capacities. Individual organizations are then better able to cope with incremental change. The innovative task of public management as a macro process is to develop new and quite distinctive macro-organizational capacities to deal with structural change at the interorganizational level. The requisite macro-management capacities for steering large-scale reorganization and major reform have no counterpart in business. Public management as a macro process requires an interorganizational management process in which the various organizations and interests involved in a public policy system share responsibility for managing structural change while retaining their autonomy in managing incremental change. What is required to deal effectively with structural change is a participative process in which different organizations come together in a joint problem-solving, collective-decision process to redesign the rules of the game and redefine their mutual roles and responsibilities around an agreed definition of public interests. 184 Les Metcalfe Public management at Üie macro level is not a natural outgrowth of the relationsliips that evolve among organizations seeking to protect their own prerogatives and advance their separate interests. It will not emerge spontaneously. Managing structural change requires interorganizational dynamics which are the reverse of those that evolve from the politics of managing incremental change. Instead of working from micro to macro it starts from the macro level Instead of aggregating predefined private interests and seeking allocative compromises within established structural parameters, public management begins by diagnosing and formulating systemic problems at the macro level and designing integrative solutions based on common interests and collective goals. After a new macro framework has been established individual organizations can once more focus on the pursuit of private interests within a new structure of governance. The difficulty, as with all public goods problems, is that the public interest in the resolution of structural problems is everyone's concern and no one's responsibility. Part of the task of public management is to develop institutions and processes which ensure priority for formulating structural problems and managing structural change. Solutions cannot be imposed from outside. Attempts to do so are liable to trigger the disintegrative dynamics they are intended to avert. The willing cooperation, commitment and dedication which are essential to resolving structural problems must be developed from within. Macro-management solutions are needed which are, paradoxically, participative and top-down. They require extensive active involvement in diagnosing problems and designing solutions by those who will be involved in implementation. Their participation must be re-framed in terms of joint problem-solving rather than special pleading. Psychologically and politically this involves a risky shift from the defensive 'what-we-have-we-hold* postures of distributive bargaining to the more innovative reforming attitudes of integrative bargaining. An important example of this process is the '1992' programme to complete the internal market of the EC, although, for political reasons it has been portrayed as aprocess of deregulation and removal of micro level obstacles to the operation of free markets. But the internal market will not operate in apolitical vacuum. Positive integration requires collective action by the Member States to restructure the economic environment of the EC. The European Commission has a key facilitating role in formulating common interests and integrating the efforts of the administrations in the Member States as participants in a macro-management process. But it can not undertake the macro management function alone. To be effective in managing structural change it should create the context in which the macro issues are collaboratively identified and a new basis of integration designed and developed. Public management: from imitation to innovation 185 Effectiveness; From Individual Optimising to Pluralistic Learning The discussion of public management as a macro process raises important general issues about policy effectiveness and the contribution of reform to improving performance. Public management as the guiding theme of reform raises strong expectations of better performance. The primary motivation for administrative reform is improved performance. Improvements in performance may take different forms: better service, greater value for money, lower costs and so on. However, in practice, reform efforts which venture into large-scale structural reorganization often fail to produce the desired results. Reformers tend to concentrate on localised improvements which promise a quick payoff. Ministries are divisionatized and agencies are formed with sharply defined missions and presumed autonomy. Reform relies mainly on micro-management prescriptions, constrained by the conventional management-by-objectives model. The assumption underlying this 'look after the pennies' approach is that better management involves targeting resources at specific problems. However commonsensical this may seem, it is far too restrictive. It falls well short of providing a model of effectiveness for public management at the macro level. Reformers have been reluctant to go further into the issues surrounding the meaning and measurement of effectiveness. This is because of the theoretical complexities of managing with multiple conflicting objectives and ill-defined shifting priorities. But to sustain the claims of public management as a macro process a broader concept of effectiveness than the attainment of predefined objectives is indispensable. Targeting resources at problems assumes the validity of a model of management based on individual rationality which equates effectiveness with optimising within constraints. But public management operates in a pluralistic context in which goal consensus cannot be assumed, in which authority is dispersed, in which conflict is legitimate and in which, nevertheless, the different constituents are interdependent and have common interests -however dimly perceived. The rational model provides no clear prescriptions when objectives uncertain and constraints alter - as they do in phases of structural change. Being based on 'the logic of individual rationality', it presupposes the existence of a unitary mind which is capable of ordering its preferences, of formulating non-contradictory goals and of evaluating alternative means and actual performance in the light of its integrated goal system' (Scharpf, 1978: 346). These conditions accord better with amodel of hierarchical control than with the inter-organizational realities of governance. Even in business, the rational model is being superseded as an ideal or standard of excellence. The stresses experienced by business in the 1970s and 1980s forced a reformulation of ideas about the meaning of managerial effectiveness. Ouchi's (1981) contrasts between Japanese and 186 Les Metcalfe American management practices is instructive because tine qualities that he criticizes in American management closely resemble those that are being introduced into government. Flexibility and innovative capacities acquired much greater importance than they had had in a more stable environment. Ideas about management effectiveness are in flux (Goodman and Pennings, 1977; Cameron and Whetten, 1983) and criticisms of the rational model have been moving for some time towards a more general and inclusive concept of effectiveness. Effectiveness criteria are not givens. March (1962,1978) observed that values and preferences are often formed in the course of policy development and implementation rather than being optimising criteria derived from a predetermined objective function. Vickers (1973) pointed out that, in the appreciative systems of policy makers, goals are subsidiary to norms and values. The orthodox model of rationality disregards the extent to which management involves problem-formulation as well as problem-solving as leaders confront the creative political task of developing new visions and generating new values from which future goals can be derived (Dror, 1988). The public management function of bank supervision depends on evaluating the private management of commercial banks against prudential norms and values defined within a macro level framework (Metcalfe, 1982). The BCCI disaster illustrates the consequences failing to establish an international regulatory framework for banking. Landau (1973) criticized the conventional management wisdom that optimal effectiveness is achieved by eradicating organizational overlap and duplication in the execution of policies. Contrary to many popular critiques of public bureaucracy, overlap and duplication are not always wasted effort. They can increase reliability by providing a measure of insurance against the risks of human error and organizational failure. The interorganizational context of public management multiplies the sources of error and underscores the importance of strengthening the interfaces among organizations. An interorganizational network is only as strong as its weakest link, but often interdependence is undermanaged. A common characteristic of a succession of recent disasters in and around Britain which have taken their toll of human life - a passenger ship capsizing; trains crashing; an explosion on a North Sea oil rig; a fire at one football stadium and fatal overcrowding at another - has been the failure or breakdown of coordination among various authorities so that no-one saw the whole picture until too late. Safety, as a public good, was everyone's concern and no-one's responsibility. These specific criticisms of the goal-seeking model of management have been accompanied by more thoroughgoing challenges. Weick (1977) advanced what may appear to be merely flippant proposals that; effective organizations are garrulous, clumsy, superstitious, hypocritical, monstrous, octopoid, wandering and grouchy. However, they have a definite theoretical underpinning; centred on a concern with adaptability, flexibility and resilience to meet unforeseen situations, as Pondy (1977) showed. Weick sought to shift the emphasis from conventional models that place the Public management: from imitation to innovation 187 emphasis on specialization and optimal efficiency in dealing with a specific task in a given situation; models which inform current developments such as the formation of agencies in the UK and similar developments in the EC, Spain and other countries. Instead he underlined the need for generalized learning capacities to cope with an evolving set of problems in a changing environment. In the short run this may seem less efficient, but in the long run it is more effective. Nystrom and Starbuck (1983) presented a positive case for an evolutionary rather than an optimizing view of organizational effectiveness. 'Judgements about organizational effectiveness generally undervalue information, discovery and learning. Effectiveness that falls below the best performance attainable is nevertheless desirable if it includes information about better criteria or better methods. Conversely, supe/ficially optimal effectiveness is actually undesirable if it forecloses learning. The managers of successful organizations quite often inflict crises on their organizations by deciding that they have already attained optimal effectiveness or that they know how to attain it' (Starbuck and Nystrom 1983: 153). These contributions set a new agenda. They call for theory of effectiveness in public management as a pluralistic process of interorganizational learning. While few people would dispute the desirability of organizational learning as a useful adjunct to public management, the argument here is that organizational learning and interorganizational learning capacities are central to effectiveness in public management. This is not as vague and ephemeral as it might seem. A rigorous and well-defined general concept of learning is contained in Ashby's (1960) exposition of the adaptive processes of ultrastable systems. Ultrastability is sometimes misunderstood as implying extreme rigidity. The opposite is true, it denotes an enhanced capacity for flexibility, along the lines of the discussion of incremental and structural change above. The distinguishing feature of an ultrastable system is a dual capacity for gradual, step-by-step adjustment in a given environment, backed up by a reserve capacity for major self- reorganization to match environmental discontinuity. An effective learning system displays this self-designing capacity for responding to major changes in policy objectives or environmental circumstances. Learning Processes in Public Management In the political environment of public management learning processes are especially difficult to create and maintain. Individual learning is a psychological process. Organizational. learning is also a political process (Metcalfe, 1981). A critical task of public management is to build institutional learning capacities at the macro level to manage the environment in which private management operates. But, conventional political pro- 188 Les Metcalfe cesses often block learning because ideology overrides evidence or vested interests resist policy evaluation and change. The practical difficulties of building learning capacities at the interorganizational level are enormous. But at least in concept, ultrastability provides a model which can guide the development of organizational learning capacities at the macro level. Managing structural change depends on the constituent organizations pooling their adaptive capacities and acting in concert as multistable system (Metcalfe, 1974) with what Dror (1984) has termed a central mind and what Deutsch (1963) called a learning network. At this level, public management reforms are better regarded as management by design rather than by direction. It should be concerned with designing adaptable systems rather than producing blueprints for specific reforms. In a world of rapid and discontinuous change, reform is not a single isolated event, but a permanent responsibility for public learning. There is no need to try to improve upon a classic statement. 'If government is to learn to solve new public problems, it must also learn to create the systems for doing so and discard the structures and mechanisms grown up around old problems. The need is not merely to cope with a particular set of new problems, or even to discard the organizational vestiges of a particular form of governmental activity which happen at present to be particularly cumbersome. It is to design and bring into being the institutional process through which new problems can be continually be confronted and old structures continually discarded.' (Schon, 1971). While the design of individual organizations as adaptive, learning systems can draw on a growing body of work in organization theory, this is less true of the design of interorganizational networks which have so much importance in public management. The agenda of future research should focus on the problems of designing interorganizational networks which have macro-level capacities to learn and manage structural change as well as micro-level capacities to deliver services efficiently. In democratic systems of government this includes setting the framework of accountability. Private management operates within whatever accountability framework exists and exploits whatever opportunities and loopholes it offers. Public management should create incentives and accountability systems which align private interests with public purposes. In the past public accountability has been associated almost exclusively with the negative function of preventing abuse of power. In a democratic society, this is, of course, one of the essential bulwarks of constitutionalism. However, in addition to preventing the abuse of power, well-designed accountability processes can serve the positive function of promoting the effective use of power, by creating incentives for effective performance (Metcalfe, 1989). This substantially enlarges the scope of public management beyond the narrow limits of applying business methods in government, but it responds to strategic problems of governance at the interface of politics and administration where strategic decisions about forms of Public management: from imitation to innovation 189 organization and accountability are made and broad issues about the design of regulatory regimes are decided. Conclusions If necessity is the mother of invention, public management reform will see the development of new and original theories of management in the near future. The problems facing government are such that it will become the forcing-house of management innovation. But to make the shift from imitation to innovation, public management needs new theoretical foundations. This article has proposed a theoretical focus on the distinctive problems of public management at the macro level. The function of public management is defined as dealing with structural problems at the macro level in contradistinction to private management concerned with managing incremental change at the micro level. The crucial task of public management is to recognize the emergence of these problems and initiate action at the macro level to remedy them and forestall the turbulence that otherwise occurs. Some of the most important implications that flow from this analysis have to do with the meaning of effectiveness and the requirements for achieving it. The conventional goal seeking model of effectiveness must be superseded by models based on a logic of learning which takes account of the pluralistic context of public management and accepts the integrative challenge of building interorganizational cooperation in the midst of structural change. This concept of public management implies a very significant expansion of the role of management in government beyond the sphere where existing business management methods can be applied more or less directly. It proposes the extension of management into areas of strategic policy-making and large-scale reform. Creating the macro-management capacities necessary to handle structural change will require significant investments in human resources and organizational development; investments on a much greater scale than governments have made in the past. But the price is small compared with the costs of failing to resolve structural problems or the benefits that flow from overcoming them. In any case, wherever governments are facing the challenge of managing structural change, there is no viable option to developing macro-management capacities.