MANAGING SUCCESSFUL GOVERNANCE REFORMS:

LESSONS OF DESIGN AND IMPLEMENTATION

SYNTHESIS AND DESK REVIEW

Mark Robinson

I. INTRODUCTION

This paper presents the findings of three country case studies of successful

governance reforms in Brazil, India and Uganda and a desk review of Bank-funded public sector governance projects from the mid-1990s as a basis for distilling fresh analytical insights and deriving operational lessons for the World Bank. In focusing on the political and institutional factors that contribute to successful outcomes the paper highlights the feasibility of various approaches to improving public sector governance in different country and regime contexts.

The synthesis of the findings of the case study research is guided by the hypotheses and analysis in the conceptual framework paper (Goetz 2005). It examines four areas of governance reforms addressed by the country case studies: public financial management (fiscal management and tax administration), anti-corruption, civil service reform, and innovations in service delivery. The synthesis builds on country specific observations about the conditions for successful reform as a basis for framing generalisations on the factors that influence reform outcomes, rooted in comparisons of political dynamics and institutional factors. The desk review was informed by a range of reports and documents produced by the Bank from ten countries in Africa, Asia and Latin America with a view to generating lessons and insights from diagnostic work and lending activities.[1]

The third part of the paper considers the implications of the research for analytical work and Bank lending operations. It seeks to complement economic and sector work in the Bank by highlighting the political and institutional factors that shape successful reform outcomes in different reform contexts. This can enhance understanding of the risks and feasibility of different types of governance reform and how their design and implementation might be improved.

II. CASE STUDIES OF SUCCESSFUL GOVERNANCE REFORMS

This section summarises the central findings of the case studies of governance reforms in Brazil, India and Uganda with reference to the propositions advanced in the conceptual framework paper (Goetz 2005). It begins by setting out a working definition of governance reform and then outlines the main elements of the conceptual framework. Next it summarises the nature of and the Bank’s support for governance reforms in the three countries, and then attempts to draw some broad conclusions in line with the approach set out in the conceptual framework.

The conceptual framework

The definition of governance adopted for this study builds on the formulation proposed by Campos and Pradhan as the manner in which the state acquires and exercises the authority to manage public goods and services (2004: 1). The governance reforms examined in this paper are reforms to the institutions that influence the behaviour of state actors, providing them with incentives to act in ways that improve the management of public goods and services (Goetz 2005: 4). They encompass reforms that are intended to promote structural changes in state institutions, though improvements in public expenditure management, the delivery of services, and promotion of accountability.[2]

The conceptual framework developed for this study is founded on the premise that successful design and implementation is a function of the formal and informal institutions that shape the incentives for decision makers to initiate governance reforms. Politicians’ decision to embark on a governance reform strategy entails a careful assessment of the potential risks and benefits. The principal risks are in the form of a loss of patronage resources (public sector employment and rents) and a possible erosion of political support (in response to unpopular measures). These are balanced by the potential benefits in the form of enhanced electoral dividends resulting from improved economic performance and management of goods and services. In this framework, the extent to which politicians are prepared to abandon their patronage methods and resources is a critical determinant of reform outcomes. By extension, the incentives that discourage bureaucrats from impeding reform initiatives are integral to successful implementation.

The incentive structures governing reform implementation are fashioned by three sets of political factors that form the point of reference for this analysis:

a)The formal and informal politicalinstitutions that shape the choices made by facing political and bureaucratic actors, distinguished by the extent to which they form part of the state apparatus and the degree to which rules and practices are codified or formalised. Informal institutions founded on kinship, ethnicity and religion operating outside and within the state apparatus can discourage the formation of horizontal alliances in support of reforms and undermine formal incentive and accountability systems.

b) The nature of the connections between state and society through which the compliance if not active support of non-state actors to reform is garnered with particular emphasis on the role played by political partiesand civil society organisations (such as trade unions and business associations).

c)The political agency required to package reforms, moderate their scope and pace, identify levels and arenas at which to begin, so that resistance is undermined and support cultivated. This highlights the importance of political leadership and the political skills required to initiate and shepherd risky reforms.

Drawing largely on the experience of large federal or successful developmental states, the conceptual framework focuses attention on the key institutional and policy design features that enable reformers to downplay political threats, broker the formation of pro-reform coalitions, cushion the shock of reforms, and make reforms more palatable by delivering tangible benefits. The structural features of politics and society influence the capacity of decision makers to embark on reforms, either in undercutting the privileges of elites accustomed to seeking rents through the state, or in enabling reformers to generate support from groups likely to benefit from reform. Variations in reform design can reduce the risk burden of decision makers, dissipate or pick off opposition, trigger an interest in monitoring reform in civil society and political opposition, and generate new constituencies of support.

Structural features of politics and society:

a)Institutional depth. The longevity, flexibility, adaptability and legitimacy of formal and informal institutions through which agreements are reached between contending social groups, or through which losers are compensated, generates stronger support for reform and lowers the cost of innovation.

b)Composition of governing elites. The extent to which traditional (especially rural landholding) elites can hold back pro-poor efforts. Governing elites relying mainly on clientelism can inhibit the emergence of pro-poor coalitions and the ascendancy of new social groups committed to reform.

c)Composition of civil society. A diverse civil society with institutions capable of developing horizontal solidarities can provide incentives to reformers by responding positively to reforms and offering new sources of political support to offset the loss of established constituencies that are opposed to reform.

Design and implementation variables:

d)Sequencing, timing and pace of reform. Reforms can be designed to generate early ‘winners’ who can support follow-on reforms or dissipate resistance through a gradualist approach that builds public support. Rapid and ambitious reforms can attract public support but also provoke political and bureaucratic resistance and thus entail a higher level of risk.

e)Technical capacity.Public sector capacity constraints can blunt the implementation and impact of reforms that are championed by political leaders and command public support. The creation of autonomous state institutions with high technical capacity can foster effective implementation but also weaken the capacity and commitment of existing government bureaucracies.

f)Extent of decentralisation. The devolution of responsibility for some reforms to lower levels of government (e.g. sub-national states or provinces in federal systems) can deflect some of the opposition to reform, but also encourage experimentation and competition between different levels of government to maximise the gains from reform.

g)Monitorability of reform. The commitment of governments to reform measures is contingent on their openness to public scrutiny and legislative oversight. Public accountability measures may assist in sustaining reform momentum by containing opposition from vested interests.

In sum, the framework rests on a simple proposition: the design and implementation of governance reforms depends on the way that decision makers respond to incentives shaped by political and institutional variables that are specific to context and reform type. Strategies range from the active pursuit of reform to diminish patronage and leverage political benefits to a more risk-averse approach that favours incremental change and continued, selective investment in patronage systems to minimise potential dissent and the derailing of reform.

Governance reforms in Brazil, India, and Uganda

Brazil, India and Uganda were purposively selected to illustrate successful reform implementation in three different political and institutional settings, with differing levels of engagement on the part of the Bank and other aid donors. A range of reforms was examined in each country to reflect variations in the type of reform and the political context. Two sets of reforms centre on public expenditure management, in the form of changes in tax policy and administration in Brazil and Uganda, and the Fiscal Responsibility Law in Brazil. Civil service reform and anti-corruption initiatives are the focus of two of the Ugandan case studies. The case studies from two Indian states principally focus on innovations in service delivery, which also have implications for tackling corruption and improving accountability.The policy goals, reform areas and focus of the case studies are shown in the accompanying table.

Public sector governance programs: policy goals, reform types and case studies

Goals / Reform type / Case studies
Fiscal stability / Public expenditure management
Tax reform / Brazil – Fiscal Responsibility Law
Uganda Revenue Authority
Brazil – tax policy and administration
Managerial efficiency and improved service delivery / Executive agencies
Performance contracts / Uganda Revenue Authority
Andhra Pradesh – Metro Water, Hyderabad
Uganda – Civil Service Reform Program
Capacity building / Information technology
Decompression and pay reform / Karnataka – Bhoomi scheme
Uganda – Civil Service Reform Programme
Public accountability / Anti-corruption agencies, public complaints commissions, ombudsmen
Service delivery surveys
Oversight institutions / Uganda – Inspectorate of Government
Karnataka – BATF and PAC
Andhra Pradesh – DWCRA
Uganda – Auditor General, Public Accounts Committee

The case studies focus on examples of successful reform, characterised by improved outcomes (such as rates of tax collection and the quality of service delivery) or institutional innovations that have the potential to contribute to improved outcomes (for example anti-corruption agencies, tax authorities, and fiscal policy initiatives). In some cases reforms primarily intended to bring about improvements in one area (such as service delivery) also produced other benefits (such as improved accountability and transparency). In several cases the observed successes were qualified in certain respects: increased tax capacity but worsened efficiency and equity (tax reform in Brazil); significant institutional innovation but modest improvement in outcomes (the Bhoomi scheme in Karnataka); and lack of sustainability and/or reversals (civil service reform in Uganda).

The initiatives varied in their breadth, in terms of the number of agencies or sectors they were intended to affect, and the degree of change they sought to induce, ranging from small-scale incremental reform to large-scale structural change.[3] They also varied in the speed and intensity of implementation, from incremental reforms that built up over time to ambitious reform initiatives introduced relatively quickly. These dimensions of reform have a bearing on trajectories of implementation in different political and institutional settings.

In Brazil the success of the Fiscal Responsibility Law lies in an institutional innovation which promotes improved fiscal management. In contrast the Brazilian tax reforms led to significantly enhanced revenues but left the underlying institutional conditions unchanged. The reforms in the two Indian states were primarily intended to improve service delivery but also had a positive impact in other areas. The proposals of the Bangalore Agenda Task Force (BATF) led to some improvements in municipal taxation and public accountability. The reform of Metro Water in Hyderabadimproved water supply and the management of public finances. The Bhoomi scheme in Karnataka greatly enhanced transparency and the efficiency of the administration of land records. The DWCRA scheme in Karnataka provided rural women with improved access to credit through the formation of self-help groups, and led to improved accountability of local government officials. Civil service reforms in Uganda reduced the number of civil servants and ministries and raised salaries, although the efficiency gains are hard to gauge. The creation of the Uganda Revenue Authority contributed to a significant increase in revenues, but did not contain the problem of corruption. New institutions designed to tackle corruption in Uganda were successfully established but produced modest results.

The cases of successful governance reforms in Brazil, India and Uganda can be situated in the broader context of reform efforts in each country. They are intended to be representative of the major thrust of governance reform efforts from the late 1990s and embody the more successful initiatives pursued by the respective governments.[4] The World Bank has been closely involved in governance reforms in all three countries, and several of the initiatives examined for this study received World Bank support, especially in Uganda. In Brazil and India the role of the Bank and other aid donors was marginal though the reforms are consistent with the broader thrust of Bank policy and support in each country.

A key governance challenge in Brazil emanates from the considerable powers granted to states for taxation, expenditure and administration under the 1988 Constitution. Devolution of fiscal responsibilities to state governments resulted in steady increase in the public service payroll and mounting debt as they struggled to perform functions previously assigned to the federal government. Effective monitoring of state fiscal policies by federal authorities was undermined by weaknesses in the political party system which encouraged self-interested behaviour by state politicians. A related governance challenge lies in the system of tax administration and the difficulties faced by the federal government in mobilising revenues in the face of high levels of state fiscal autonomy and in redressing the highly regressive nature of the tax system. This provides the context for the two case studies of the Fiscal Responsibility Law and reform of tax administration and underlines their centrality in governance reform efforts in the country.

The Brazil Country Assistance Strategy (CAS) emphasises the importance of financial sector reforms and financial management in the public sector as the mainstay of the Bank’s approach to governance reform. Loans for fiscal reform are intended to provide the policy framework for Bank assistance, with a focus on macro-fiscal management and structural reforms combined with technical assistance for public expenditure management and budget reforms (World Bank 2004b).[5] The CAS also acknowledges the importance of tax reform to reduce the inefficiencies created by the existing tax system, despite significant increases in tax revenues in recent years. While the Bank is not a major actor in Brazil, the emphasis in the CAS on fiscal adjustment and management and tax administration is consistent with the focus in the Brazil case study on the Fiscal Responsibility Law and taxation reforms (Schneider 2004).

Governance problems in Indiaassume many different forms. Weaknesses in public sector management and service delivery are exhibited in fragmented and unresponsive bureaucratic structures, lack of flexibility in staff transfers, low rates of productivity, poor targeting, and systemic corruption. The Bank’s strategy in India acknowledges the fundamental importance of problems of governance and public sector management as constraining the realisation of economic growth, effective service delivery and poverty reduction. In the Bank’s view the central challenge is the need for improvements in governance to tackle problems of weak institutional capacity, over-staffing, corruption, and poor budgetary management. From the late 1990s the Bank’s assistance strategy developed a sharper focus on poverty reduction, a more selective approach to support for reforming states, and greater attention to governance and institutions (Zanini 2001).[6] Two reforming Indian states that have sought to introduce governance initiatives since the late 1990s are Andhra Pradesh and Karnataka in south India, some of which are examined in depth through the case studies (Manor 2004).