Abstract
The World Bank has recently released its Public Sector Management (PSM) Approach for 2011–20. This commentary reviews the core messages of this document and then indicates how it embodies a convergence between academic research and practice in its approach and analytic framework. It then presents a ‘thought experiment’ about how practitioners might bring scholarship and practice together as suggested by the PSM approach. Nevertheless, effective implementation of the approach will depend on the convergence between the path it lays out for practice and the incentives that World Bank officials face in efforts to improve public sector management in real-world situations.
Keywords
The World Bank’s Public Sector Management Approach signals the importance of improved public sector performance in developing countries. Yet it also indicates how difficult this process is and how frustrating it can be for those working in the field because there really is no ‘science’ of what is most important in reform; it reminds us that reform initiatives are often episodic. In addition, the document reiterates the always unwelcome news that there are no easy solutions and certainly no universal solutions or certainty that selected interventions will regularly produce expected results. It explains this bad news in terms of the impact that context, history, tradition, culture, and a variety of other factors have on the definition of problems in specific countries and on the ‘fit’ of a solution to those realities.
Altogether, then, the document directs our attention to the importance of deeper knowledge of specific countries and the critical need for finding appropriate diagnostic tools to identify which among a variety of intervention options best fits a given situation. Thus, it frames the issue of improved public sector performance in terms of focused problem-solving rather than in terms of major sectoral reforms. Importantly, it reminds us that there is often a difference between de jure and de facto when it comes to performance and behavior. At the same time, it values the possibility of learning from the Bank’s long experience in public sector management. Finally, it advises flexible and creative approaches to problem solving and the assessment of ideas as they are put into practice.
These points very effectively reflect recent trends in academic research. Clearly, among academics in the field of development, context and specificity are capturing important attention; increasing numbers of researchers focus more on how countries differ in their development histories and how institutions vary in structures and functions across countries or regions. They are more interested in specific problems and the range of options for addressing them.
In addition, much current scholarship in development accepts that some ideal state of performance does not exist anywhere in the real world and that a process of development is just that – a process that unfolds over time with great variation and many surprises along the way. As a consequence of these perspectives, researchers are less concerned with big, overarching analyses and universal prescriptions and more focused on identifying analytic tools that respond quite specifically to the process of development in specific places, not just some imagined end-point of it. The most valued metric is not ‘perfect’ – or even ‘good’ – it is just ‘better than before’.
If I think of what is happening among scholars of development at my own institution, the Kennedy School, for example, I can point to many ways in which more contextually specific analysis is encouraging new ways of thinking about the challenges of development. Thus, for example, I have written quite extensively about ‘good enough governance’, why some local governments perform better than others, and most recently, about the distinct historical experiences of countries in introducing Weberian civil services. 1 Ricardo Hausmann has developed a ‘growth diagnostics’ approach that provides a general framework for teasing out the specific economic opportunities in particular countries. 2 Dani Rodrik has argued that while there may be general principles in development, constraints on growth differ across countries. 3 Matt Andrews, has captured our imaginations with the idea of ‘hippos in the Sahara’, which points to what Lant Pritchett and others have referred to as ‘isomorphic mimicry’ – the tendency to adopt similar institutional forms even in the face of great evidence that institutions perform differently in different contexts. 4
We are certainly not the only ones thinking along these lines. ‘Next steps’, ‘good enough’, ‘best fit’, contextualized diagnosis, deep knowledge of the historical, political, and social roots of current conditions, and the identification of binding constraints and bottlenecks are concepts that provide an important thread in recent scholarship in development. It is important that these concerns span economics, political science, and management sciences. They are leading all of us to appreciate our colleagues in history departments more – in large part because they have long appreciated how history is constructed uniquely in each village, region, and country in the world.
The central idea across this research is that development can be advanced through situationally determined responses to specific problems. This new generation of thinking is clear: what works in India may not work in China, Rwanda, Argentina, or Albania. The new Public Sector Management Approach document does a masterful job of translating this new research focus into knowledge and guidance that are relevant to practitioners.
Yet, this problem- and context-oriented approach in the academic world puts practitioners in a very difficult place. It directs them to be flexible, to be creative, and to learn along the way, but without providing many clues as to how to do this. It provides a few frameworks for those in the field, but no specific advice about appropriate actions to recommend or put in place. It exposes practitioners to the demands of clients for ‘answers’, but then indicates that the appropriate response to these demands is that there are no general recipes. It insists that problem solving in public sector management, and development more generally, has to be discovered organically and incrementally. It suggests frankness about the fallacy of universal best practices when responding to clients. These admonitions can certainly be experienced as very disempowering for those whose jobs it is to recommend changes.
In the real world, it is indeed difficult to hold the attention of those eager for progress and clear answers with responses that amount to ‘Well, I’m not sure, but let’s explore this more and perhaps we can generate some ideas for interventions that “fit” your reality’, or ‘Let’s problem-solve together’. It is certainly difficult to tell a politician or a manager requesting assistance that the Bank is exploring and experimenting with responses to public sector reform. And it goes without saying that many in the organization are also keen on rigor, uniform standards, and accountability, and may not be particularly interested in complex logics about historical, political, bureaucratic, or even cultural feasibility. So, while the document effectively captures important perspectives, it raises a number of challenges related to its implementation, about getting from it to the real world.
Problems in Anyland
Given these reflections on the document, what is going on within development research circles these days, and the challenges these factors pose for activities in the real world – for getting from here to there – a simple thought experiment might be instructive. How might a practitioner apply the Public Sector Management Approach in a particular situation?
We might begin by imagining this situation. For example, we can imagine a government of a country – let’s call it Anyland – asking for World Bank support to improve the management of its two most important social service delivery ministries, health and education. We can assume that this country has few resources and high demand for them. Possibly, the system of government in Anyland is based on an extensive and deeply embedded patronage system, in which people get jobs in government based on their personal, political, regional, and ethnic identities and relationships. We might also assume two ministries with very serious problems of inefficiency, corruption, and poor performance; with decayed and out-of-date infrastructure; with no effective internal and external communication and cooperation linkages; with multiple and overlapping programs and units; with large numbers of employees only marginally motivated to do any work; and with a multitude of reasonable explanations for their dysfunctional performance. Anyland might be found in many regions of the world.
From a practitioner’s perspective, then, the question is what might be done to improve this situation in Anyland. Following a traditional approach, considering what interventions are needed might begin with a catalogue of all the things that need to be fixed, focusing on the multitude of things that aren’t working. This would include lack of capacity, poor technology, faulty organizational structures, a dearth of job descriptions and salary scales, and many other issues that constrain performance. In such an approach, a practitioner might then envision reforms in several areas so the ministries would be able to carry out their missions and functions more effectively.
But it might also be considered that none of these problems could be fixed without attacking a more fundamental problem of the organizations – the ingrained patronage system filling them with inappropriate and corrupt people and ensuring incompetence. In this situation, the Bank might want to recommend replacing the patronage system of hiring, promoting, and firing employees with a modern, merit-based civil service system, perhaps incorporating some of the ideas of the new public management.
Yet there is good reason to question if this is a feasible approach to the problems of health and education in Anyland. Indeed, the history of many now-developed countries indicates that the creation of merit-based civil service systems took decades to introduce and implement. A clear example comes from the United States. Demand for such a reform started to emerge in the 1830s; some initial and failed initiatives were introduced in the 1850s and were reintroduced in the 1870s; finally, in 1883, after the assassination of a president by a disappointed office seeker, the Pendleton Act was introduced. It was not until the 1920s, however, that as much as 75 percent of the federal public service was incorporated into the civil service, a proportion that fell to 60 percent in the 1930s and 1940s as government leaders recruited a cadre of public servants to help face the challenges of the depression and war.
This was not an isolated case – some European countries needed the entire nineteenth century to consolidate their civil service systems. Some developing countries have actually deconstructed such systems over time, something that was also attempted over several decades in the United States. Moreover, a historical perspective would underscore the reality that civil services create their own pathologies that eventually can constrain development.
Returning to Anyland, but now armed with the knowledge that civil service reform would be incredibly difficult to put in place in a project cycle of three or five years, are there other options if we continue to suspect that an important first step has to be to deal with how people go about their jobs and the incentives they face for performing poorly or well?
Following the suggestions in the new Public Sector Management Approach, it would certainly be appropriate to invest in a participatory diagnostic process in which discussions focus on the functional problem rather than the solution. Ministry staff might be encouraged to investigate and discuss internal incentive systems and why officials behave the way they do. This might lead to a list of many problems, but in the course of the discussion, it might become more apparent why and how people are selected to work in the ministries. This might lead to an understanding of the patronage system as an institution that produces important benefits for those who control and manage it. It would become clearer why this system is tolerated even though it produces many dysfunctional outcomes for the delivery of health and education services. In effect, it would become clear why its demise might be strongly resisted by those who benefit from it. This collaborative work might even lead to a question of whether the existing patronage system could be turned to the advantage of better performance. This is really thinking outside the normal box of public sector reform.
Indeed, although patronage systems are usually considered to be synonymous with incompetence and corruption, in fact how patronage-based organizations perform is strongly affected by the motivations of those who control patronage appointments. 5 That is, their purposes can be varied and flexible because those who make appointments have great discretion in terms of what they seek to achieve through those appointments. Certainly patronage appointments can be used to build family dynasties, enrich certain ethnic groups, and deliver on election promises to individuals, parties, and regions. They can be totally divorced from any productive results.
But they can also be used to bring new technical expertise to bear on a problem, to build the professional profile of an organization, to support and implement a reformist agenda. This is part of the explanation why many countries in Latin America were able to move fairly quickly and decisively to introduce market-oriented reforms in the 1980s and 1990s – reformist presidents and ministers used their very wide appointment power to bring in cadres of young, well-trained technocrats and to place them in positions of influence within government. Similarly, in the history of the United States, those who worked hard to improve professionalism and performance in many public sector bureaus in the late nineteenth and early twentieth centuries were in fact patronage appointees. High-performing organizations in Brazil have had a similar but more recent history. Over time, appointments such as these have done a great deal to improve the level of professionalism in many public sector organizations.
Now, in the case of Anyland, it is important to understand how the patronage system within the ministries is organized and the purposes it is set to achieve – how it reflects the important power relationships and tradeoffs that are made in the service of political stability, for example. But that knowledge is not hard to come by because of the approach that encourages working with local collaborators who carry this knowledge with them – they may even be beneficiaries of the patronage system. They will also understand that if the goal is improving the performance of the ministries through the ‘capture’ of patronage, advocates of change will need to count on a reform-oriented president – or one under extensive pressure to improve the delivery of health and education services because of concerns about elections and political stability. He or she will need to have the capacity to recruit ministers and a few other high-level officials who share his concerns.
Given these conditions, the stage would then be set for these officials to recruit like-minded and well-prepared people into positions below them and to encourage these people to look for like-minded and well-prepared people to work for them, and on down, always allowing, of course, for the use of additional criteria such as party affiliation, ethnicity, or regional identity. The experiment would be that, through these reformers, the other problems of the organizations could be addressed more easily and that at least some of the challenges and bottlenecks to change would be eased.
Of course, the use of the patronage system for reform purposes would not produce a magic bullet for the organization – organizations almost never function as well as we would like – but it might create the internal dynamic that would make change in other areas – capacity building, reorganization, corruption control, professionalization – more feasible. It would also be clear that any changes introduced by the reformers could only be institutionalized over time, so that in the near term it would be important to think of how incentives and values could be stressed that reward professional and responsive behavior – even without the structure of a merit-based civil service – in ways that embed these behaviors and norms as much as possible.
This may all be very far-fetched, the result of far too much imagination and optimistic thinking. Nevertheless, this is an example of moving from the new Public Sector Management Approach to the real world. It requires that practitioners think out of the box about appropriate responses to particular problems, it uses knowledge of a local situation as a basis for problem-solving, it engages a network of co-analysts, and it is no doubt also experimental and subject to much change as it unfolds. The lesson of this thought experiment is not that the use of patronage is a good solution to public management issues, but simply that this is an example of how solutions to local problems have to be worked out in terms of the feasibility of local solutions.
The World Bank and the Public Sector Management Approach
In addition, if this kind of hypothetical case could be imagined, it suggests an interesting problem. If this is what a situationally specific response to a development challenge might look like, if it responds to the idea of flexibility, situational diagnostics, creativity, learning as we go, and taking politics seriously – would it fly in the Bank? Would a Bank official be able to propose this kind of approach for Anyland without being accused of going along with corrupt practices or of not having a firm idea of the problems or the solutions needed? Could the official be accused of betting on a short-term result without thinking about sustainability?
In short, does this kind of a response correspond to the job that Bank practitioners are expected to do? One thing noticeable about the description of Anyland is the almost total absence of management-speak and discussions of metrics and timelines and in the presence of a somewhat ‘loosey-goosey’ conception of results. Is there, perhaps, a fundamental contradiction within the Bank between the pressure for results and the expectations about flexibility, creativity, and thinking out of the box that are incorporated into the new Public Sector Management Approach? Do management specialists in the Bank have any incentives to build their careers by promoting this kind of approach? That is, if we pick up on some of the ideas that are circulating in academic circles – that situationally determined responses to specific problems are needed – has the Bank altered how it encourages and rewards its staff so that this approach is rewarded? These are the questions that can link some very good ideas and guidance in the document with the very real world of its implementation.