Reforming the Investment Climate: Lessons for Practitioners

Most people agree that a good investment climate is essential for growth and poverty reduction. Less clear is how to achieve it. Many reforms are complex, involving more than technical design and content. They are both political, facing opposition from organized and powerful groups - and institutionally demanding, cutting across different departments and levels of government. Reform thus requires paying as much attention to understanding the politics and institutional dimensions as to policy substance, which is the goal of this paper. Drawing from more than 25 case studies, it shows that there is no single recipe or manual for reform, given diverse contexts and serendipity in any reform effort. But three broad lessons emerge. The first is to recognize and seize opportunities for reform. Crisis and new governments are important catalysts, but so is the competition generated by trade integration and new benchmarking information. The second is to invest early in the politics of reform. Central to this process is using education and persuasion strategies to gain wider acceptance and neutralize opponents. Pilot programs can be valuable for demonstrating the benefits and feasibility of change. And the third is to pay greater attention to implementation and monitoring. This does not require full scale public management reforms. Reformers can draw on private sector change management techniques to revitalize public institutions responsible for implementation. Given the cross-cutting nature of reform, new oversight mechanisms may be needed to monitor and sustain reform. The paper concludes with an emerging checklist for reformers and identifies areas for future work.


Foreword
Economic policy-making remains more an art than a science, in common with many other areas of public policy. We know a few things about the technicalities of reform, for instance how to streamline customs procedures or simplify the process of business registration. But it is less clear how one gets these and other more complex policies agreed upon and implemented. Some believe that we can study the matter and develop typologies and recipes, almost a manual. I very much doubt this makes sense. The management of societies may not lend itself to engineering-type approaches, where we can hope to discover principles of how things work and then design manuals. Fundamentally, the process of reform and implementation is about humans agreeing, compromising and acting together. In the process, innovation is possible. At the same time credibility and some level of predictability matters. The equivalent of immutable physical laws remains elusive. For reform we cannot write a blueprint. Like for change management in firms, the best we may be able to do is to generate interesting case studies that help sharpen judgment and inform policy-makers about the process and impact of reforms.
What interest do policy-makers have to sharpen their judgment and learn from other experiences? Here, competition matters. Not competition between firms, but between jurisdictions. When some countries pull ahead, when they see incomes rise -that makes policy-makers in other places wake up. Statistics that allow us to compare the performance of different jurisdictions make success and failure visible. GDP statistics show us who earns more, who grows. The benchmarking of "competitiveness" gets policy-makers' juices flowing, often under pressure from the media who highlight deficiencies revealed by statistics. Just consider the impact of the OECD PISA study on comparative performance of education systems or the World Bank/IFC reports on "Doing Business".
It is in that perspective that we have embarked on a review of reform episodes. Our hope is that the experiences and lessons of the cases analyzed in this paper will indeed help sharpen judgment. Additional work is being done on reform episodes monitored by our "Doing Business" project. Between our work and that of others there will be close to one hundred cases written up. The challenge is to find useful ways of sharing experiences among relevant policy-makers and to develop the best possible ways of learning -maybe in the spirit of business school case study teaching and, just maybe, in the more distant future by using emerging multi-player games to explore governance and policy problems.

Michael Klein
Vice-President, Private Sector Development, World Bank-IFC Chief Economist, IFC

I. UNDERSTANDING THE CHALLENGES TO REFORM BACKGROUND
Recent analytical work has led to a broad understanding among policymakers and development practitioners that microeconomic reforms aimed at strengthening property rights, unleashing competition, and reducing the cost of doing business are critical in creating a sound investment climate and achieving growth and poverty reduction (World Bank 2004b, 2005aLewis 2004). Also recognized is that these changes need to be credible and sustained for private firms to respond by increasing investment and production.
Much less understood is how to manage such reforms. Investment climate reforms are challenging for three main reasons. First, the list of possible reforms is long, and figuring out the right priorities and sequencing can be a daunting task. Second, such reforms are often politically contentious because of their distributional consequences: groups that lose in the short term are often well organized and in a position to resist, while those that benefit usually do so only in the long run and are too dispersed to matter politically. And third, investment climate reforms can be institutionally challenging: their implementation typically requires coordination among diverse departments and levels of government as well as fundamental institutional changes in organization, incentives, and behavior that can be difficult to achieve.
So, in tackling the investment climate would-be reformers face many technical, political, and institutional challenges. Key among these challenges: How to identify the most important constraints, and how to sequence reforms? How to overcome opposition from interest groups and get reform onto the agenda? How to package reforms so that they are both credible and feasible? How to mobilize support for reform? How to create the incentives and capacity to implement reform? And how to create institutional mechanisms to drive, monitor and sustain reform?
This paper addresses these questions through an analysis of the literature and the findings of more than 25 case studies of reform commissioned by the World Development Report 2005 team and by the Foreign Investment Advisory Service (FIAS) of the World Bank Group (see annex A). 1 The case studies, drawn mostly from developing countries, cover the main broad areas of the investment climate: product markets, 2 land markets, labor markets, and capital markets. They focus on countries that adopted and implemented a 1 The World Development Report cases can be found on the Web at http://econ.worldbank.org/wdr/. 2 Product market reforms include reform of the policies and institutions governing the way in which products and services are produced and sold. They tend to be industry specific. The main such regulations are those governing: entry (general business registration, industry-specific licensing, and restrictions on foreign direct investment); the import and export of goods (quotas, tariffs, and customs); and production (price regulations, product standards and government certification processes, intellectual property rights, investment incentives, subsidies, and competition policy or agency). particular investment climate reform with positive results, though many also include negative episodes from which lessons can be learned. 3 Though reform is a non-linear, dynamic process, for analytical purposes we distinguish between two main stages of reform, each with its own set of challenges and processes: Initiating and designing reform-which involves identifying priorities for reform, getting reform onto the agenda, developing reform proposals, and building acceptance for reform; and Implementing and sustaining reform-which involves strengthening the incentives and capacity to implement reform and creating institutional mechanisms to monitor and sustain reform.
We also distinguish between two sets of factors that influence reform processes: contextual factors and those that are within the control of policymakers. A country's political system and culture affect what reformers can achieve. 4 Together they explain a large part of the variation in the outcomes of reform and in its impact on economic performance. But while reformers must take political institutions and the structure of domestic interest groups as given, within this framework--and short of fundamental political reform--they have much scope to design and manage the reform process in ways that get the process going and achieve favorable outcomes. We focus on those factors within the direct control of policymakers, with the goal of highlighting the menu of strategic choices at each stage of the reform process.

WHY INVESTMENT CLIMATE REFORMS ARE CHALLENGING
Macroeconomic reforms are often carried out in times of crisis by a stroke of the penachieved by administrative decree and a few key actors. The benefits are usually immediate, visible, and spread across the population, with losers or potential losers often too dispersed or too small in number to be of political importance. 5 Investment climate reforms are a different matter-and face very different challenges: Getting the priorities and sequencing right can be technically challenging. Identifying the right package and sequence of reforms can be a difficult task when reformers face a long list of constraints and possible reforms. The list of possible reforms is indeed long: it is now known that developing countries have to improve the way they regulate capital, labor, land, and product markets; reform 3 Detailed discussion of the reforms and their impacts can be found in the case study reports, cited throughout the paper. Summaries of selected cases are presented in Annexes C through H. 4 The role of political systems in development is an important and much broader topic that is not covered in any great detail in the paper. Fukuyama (2006) discusses in greater detail how the different types of political institutions and political culture shape the ability to generate reform decisions and enforce outcomes. 5 See Krueger (2000) and Navia and Velasco (2002) for more detailed analysis and comparisons of policy reform in macro vs. microeconomic areas. This section draws from their work. their tax and judiciary systems; protect investor and property rights; invest in infrastructure and education; and tackle administrative barriers and deeply rooted governance issues. The challenge is to identify the reforms that will stimulate a quick, strong supply response while laying the groundwork for the next round of essential reforms.
The reforms are often politically contentious. They involve significant transfers of income and opportunity. Moreover, those who lose in the short term tend to be powerful interest groups or politicians capable of blocking reform, while those who gain tend to do so only in the long term and are often too dispersed to matter. Thus while removing barriers to entry, for example, favors new entrants, it may be resisted by incumbents, feeling threatened by the prospect of competition and changes in traditional ways of doing business. The involvement of many players and political institutions also mean much slower and less predictable decisionmaking processes.
The reforms can be institutionally challenging. Their implementation can be demanding, requiring detailed technical and administrative work and the development of institutional mechanisms for sustaining reforms that tend to be ongoing rather than one-shot events.
So, in tackling the investment climate would-be reformers face many technical, political, and institutional challenges. Key among these challenges: How to identify and sequence the most important reforms?
How to overcome opposition from interest groups and get reform onto the policy agenda?
How to achieve reforms that are both credible and feasible?
How to mobilize support for reform and mitigate opposition?
How to create the incentives and capacity for implementation?
How to create institutional mechanisms to monitor and sustain reform?

DIFFERENT REFORMS AND COUNTRY CONDITIONS, DIFFERENT CHALLENGES
Identifying priorities for reform is a first challenge and a process that is highly country specific (as we discuss in greater detail in the following section). But once priorities are determined, the reform process tends to confront common political and institutional challenges. Not all reforms of the investment climate face these challenges to an equal extent, however. Indeed, there is much variation across different reforms (figure 1) as well as across different country conditions.

Political challenges
The political science literature uses a number of analytical approaches in studying political impediments to reform. Chief among them are: collective action by organized interest groups; the institutional setting in which policy decisions are made; and political competition and incentives for politicians to reform.

Interest groups
Interest groups include political parties and representatives and institutions of the state bureaucracy at the central, state, and local levels. Depending on the area of reform, they also include a range of external actors-private firms, business associations, labor unions, consumer groups, judges, notaries.
Policy preferences and the importance attached to reform vary significantly among these groups, depending on the distribution of the costs and benefits of the reform. Their relative power-the extent to which they are organized and have access to the political process-also varies. Groups likely to lose in the short term, even if small in number, tend to resist reform. They also tend to have greater influence in the process, through "insider" connections or formal ties to decision-makers and political parties, than do the beneficiaries-new entrants and other larger and more diffused groups for whom the benefits tend to materialize only in the medium to long run. Even if there are no clear

Administrative complexity
High Low losers, uncertainty about the consequences of reform can generate suspicion and hostility or weaken the incentives for potential beneficiaries to organize.
Some investment climate reforms face greater challenges from interest groups than others. Reforms that are more technical or administrative, such as business registration or inspections reform, tend to affect few clearly identifiable groups capable of mobilizing against reform and so may be politically less contentious. The main resistance comes from departments and bureaucrats at different levels of government seeking to preserve their roles. Reforms with wider distributional consequences are more likely to provoke a response from interest groups. These groups can have a positive impact when they mobilize support and help get reform onto the agenda. But more often they act to block reform, seeking to preserve existing benefits. That makes it difficult to obtain support for reform from politicians who benefit from ties to these groups and who are often preoccupied with short-term costs and electoral horizons rather than long-term benefits. Two examples illustrate: Labor market reforms often meet resistance from "insiders" such as unions and formal sector workers, who benefit from existing laws and tend to be well organized, compared to new entrants or informal workers who benefit from reform and lack voice. Moreover, labor market reforms may be opaque to ordinary people: voters may not understand or care about their impact and are difficult to mobilize in support of reform. These factors help explain why in Latin America during the 1990s unions and their parliamentary allies repeatedly stalled legislation to extend the use of fixed-term contracts or to reduce severance payments. While 24 countries in the region liberalized trade and the financial sector, only 6 substantially amended labor laws (Gill, Montenegro, and Domeland 2002, p. 5).
Land reform is similarly contentious. Most large-scale redistributions of land have taken place only under exceptional circumstances: either under pressure of revolt and revolution or in the aftermath of war and foreign occupation. Even where de facto property rights are recognized, formalizing land titles and creating land registries can run into resistance from interest groups. Notaries and other registry personnel may oppose modernization for fear of losing their jobs or opportunities for bribery. In Peru local public officials resisted land titling reform because they saw it as encroaching on their jurisdiction over urban planning (Endo 2004, p. 19).

Institutional setting for decision-making
A second political challenge is the institutional setting in which policy decisions are made. Reforms which entail fundamental institutional change often require legislation to signal government commitment and limit the scope for deviations or reversals in implementation. And securing legislative approval often requires the involvement of a wide range of political institutions and "veto players"-typically the executive (president or prime minister), political parties, the legislature, and the judiciary. (Interest groups may have special ties to veto players enabling them to influence these players, but they do not normally exercise veto power directly.) The political science literature suggests that the more veto players there are in a political system, the more credible and stable are its policies-but also the more difficult and time-consuming it is to initiate and pass reform.
One common finding: initiating reform appears to be easier in presidential than in parliamentary systems, but securing approval can be time-consuming. The reason is that presidential systems typically grant the president formal agenda-setting powers, including the right to initiate reform by executive decree. Presidential authority seems to have been important in initiating regulatory reforms in such countries as the Republic of Korea and Mexico. 6 But presidential systems may also be less effective at generating decisions because they have another veto point: unlike in parliamentary systems, the executive and the legislature are independently elected and may be of different political parties. Tensions between the executive and the legislature help explain why, for example, many Latin American countries with presidential systems found it difficult to agree on labor market reforms during the 1990s. In Colombia, labor reform that first appeared on the agenda in 1999 took nearly four years, several working drafts, and two failed attempts to achieve legislative approval in 2002 (we discuss this example in greater detail in Section II below).

Political competition and incentives
A third strand of analysis focuses on the dynamics of political competition and the extent to which it provides incentives for politicians to reform. 7 In democratic countries elections allow voters to express their political opinions and hold politicians accountable for results. But their role may be undermined by political market imperfections. One such imperfection is the lack of information, which makes it difficult for voters to judge the performance of individual politicians. Even in developed countries politicians can evade electoral punishment through obfuscation, particularly when they control media outlets, or by spending public money on highly visible but socially worthless projects, or by targeting narrow voter groups at the expense of the majority. These problems are compounded in developing countries where literacy and press freedoms are generally lower. They are also more pronounced in countries with greater social polarizationwhere voters may value politicians' ethnic identity or other attributes over their performance and where politicians make promises to these groups at the expense of wider benefits-and in countries where patronage and clientelistic politics predominate. The lack of competition between political parties and the short time horizon of politicians can be further disincentives to deal with reforms with long-term benefits and payoffs.
These shortcomings help explain why many investment climate reforms are not passed even when they might benefit a majority of citizens and voters. They also help explain the persistence of corruption and other forms of narrow self-interest among elected officials. If politicians know that there is little chance of being held accountable, they have little incentive not to be corrupt. But politicians may be less likely to take the risk if voters are kept informed about the process through which political decisions are madethrough decentralization, independent validation and communication of the results of reform, or other means to encourage participation (Keefer and Khemani 2005, p. 22).

Institutional challenges
Investment climate reforms, often ongoing rather than one-shot events, demand technical and institutional capacity to plan, implement, and sustain change. Moreover, their implementation often involves many different departments and levels of government. Inefficient and often corrupt bureaucracies are normally responsible for implementation. And monitoring mechanisms that ensure accountability are often lacking. Overcoming these challenges requires making fundamental and difficult organizational change or creating new institutions with appropriate incentives and capabilities to implement and sustain reform.
These challenges are especially prominent in reforms that are ongoing or cross-cuttingsuch as regulatory reform aimed at assessing the stock and flow of business regulations or inspections reform cutting across a wide range of inspectorates. Other reforms, such as labor market reform, face a different type of problem: while less institutionally demanding, they often suffer from delays in implementation or lack of policy enforcement. * * * * * Technical, political, and institutional challenges can pose important obstacles to investment climate reforms. Yet reformers have great scope to design and manage the reform process in ways that overcome these challenges and make reform possible. As we show in the subsequent sections, deliberate choices and strategies arise at each stage of the process. A better understanding of these choices can help would-be reformers and development practitioners deal with the challenges and improve the prospects for change.

II. INITIATING AND DESIGNING REFORM
Identifying priorities for reform and countering resistance from interest groups, politicians, and political parties are often the biggest challenges in the early stages of the reform process. Our case studies suggest that reformers can deal with these challenges by: relying more on new diagnostic tools and practices to identify priorities for reform; seizing windows of opportunity to get reform onto the policy agenda; sequencing and packaging reforms so that they are both credible and feasible; and using various measures to build coalitions for supporting reform.

IDENTIFYING PRIORITIES FOR REFORM
Investment climate reforms cover a wide agenda, and no country can tackle everything at once. Establishing priorities is especially important for less developed countries, which tend to face all the possible problems but have the least capacity to tackle them. In the past, choices were based on a standard list of reforms known as the "Washington Consensus" (macroeconomic stabilization, privatization, trade liberalization, protection of property rights). But this approach has not always triggered an adequate supply response. Reformers often tried to do too much at once, and often failed, or started with reforms that were not of high priority; and, sometimes, reform in one area created unexpected distortions in another, often leading to weak or even adverse effects (Hausmann, Rodrik, and Velasco 2006, p. 12).
Over the past few years, however, the process of identifying priorities for reform has become much more context-specific, and increasingly more targeted. Reformers can make use of a growing body of new analytical tools and information to identify priorities with greater precision, and in a way that helps deliver results and build momentum for more reform. These tools include: Surveys of investors that help identify some of the most important issues; Detailed benchmarking data, such as those provided by the World Bank Group's Doing Business indicators, which benchmark and rank the cost and quality of business regulations for key cross-cutting investment climate issues; Market information on prices and return on investments which provide important information. For example, high real interest rates indicate that access to finance is a key constraint, while low returns to education suggest that aggregate skills are not, at least for now (Hausmann, Rodrik, and Velasco 2005); and In-depth industry analyses which can be used to identify industry-specific policy issues, such as entry barriers, undue protection and subsidies, inadequate price and product regulations, and issues relating to government certification and protection of intellectual property rights (Palmade 2005).
Complementarity between reforms and the role of policy learning also need to be considered. Reforms in one area, such as trade and product markets, may amplify the benefits of reforms in other areas, such as the labor market. Moreover, labor market reforms may allow for the compensation of losers from trade and product market reforms. 8 Synergies in economic impact (for example, reforming land markets and contract enforcement also improves access to financing) and in political economy (combining trade reforms with social protection can diffuse opposition) also come into play. And reform priorities often emerge through a process of policy learning and experimentation. China is a striking example of a country that designed and sequenced reforms by learning from its own policy experiments and, to some extent, from the experiences of other economies as well (see page 18 below).

SEIZING OPPORTUNITIES TO GET REFORM ONTO THE POLICY AGENDA
Once the priorities for reform are determined, what makes reform possible? One popular view is that reform is possible only in times of crisis or under a new government, when reformers benefit from strong political and popular support that makes it easier to counter opposition from interest groups. Indeed, our cases suggest that policymakers more often than not use these events to sharpen the focus on a problem and bring it onto the policy agenda. But crisis and political change are not the only factors that make reform possible.
Reformers often take advantage of periods of growth to make hard choices, because public support is high and resources are available to compensate losers. Two other factors also appear to play an important role (figure 2): Increasing competition, created by global integration and technological changes more broadly, and by spillovers from other policy initiatives such as trade liberalization or international agreements; and New information and diagnostics that expose a policy problem and create awareness of the need for reform, as well as new policy ideas generated from past experiences and an ongoing process of learning and diffusion.
Our analysis also suggests that identifying a single cause of reform can be difficult: only a few of our cases highlight just one factor as being important (see annex B). More often it is a combination of factors-and players-that helped create a window of opportunity to get reform onto the agenda. Indeed, in all our cases, political leaders played a critical role in recognizing a problem and seizing the brief windows of opportunity to get reforms onto the agenda. Moving quickly was seen to be important, given that windows for reform remain open only for brief periods before policymakers feel they have addressed the problem, before events or personnel change, or before interest groups coalesce against reform (see Kingdon 1995 for a detailed discussion of policy windows in the agenda setting phase of reform).

Using political change and crisis as catalysts
Political change and crisis are often considered important catalysts for reform. New governments or new political leaders benefit from a "honeymoon" period in which opposition parties and interest groups are less likely to oppose reform and reformers have enough time before new elections to win back the support of groups that might be alienated by reforms. And crisis is seen to help focus attention on policies that are not working, induce legislators and citizens to grant executives greater discretionary authority, and weaken the influence of interest groups opposed to reform (Haggard 2000;Tommasi 2002).
Indeed, political change played a big part in catalyzing reform in half of our cases. Newly elected governments, especially those that gained political control, used their first year in office to launch or pass difficult and unpopular reforms. In Colombia labor market reforms came onto the agenda in 1999, but it took a new government to get the reforms passed in 2002 amidst a severe unemployment crisis. In office for less than a year, the new government had strong popular support and faced little threat of political defeat. The appointment of a new minister of labor who championed the reforms proved critical. Rapid approval of the new labor code was also aided by the fact that many of the policy proposals had been prepared and widely discussed during two previous attempts in 1999 and 2001 (Echeverry and Santa Maria 2004, p. 30). New governments in Poland and the Slovak Republic also passed difficult labor market reforms well within the first year of taking office, building on preparatory work that had already been done.
In non-democratic settings reform came about as a result of new political leadership rather than electoral change. In Vietnam pressure from the domestic private sector for market institutions and clear regulatory rules had been growing since private activity was

Percentage of cases
Major Minor legalized in 1990. But the turning point came with changes in the country's one-party leadership in 1997. The appointment of a new prime minister, president, and party secretary general led to early public announcement of an action plan to improve the business environment and the subsequent passage of a new Enterprise Law in 1999 (Mallon 2004, p. 15). A similar change in political leadership in China in 1978 initiated the country's vast and sustained program of pro-market microeconomic reforms, including the decision to pilot major land market reforms in Shenzhen in the mid-1980s.
Often coinciding with political change, crisis played a big part in 40 percent of our cases. Fiscal and financial crises were most often the catalyst, leading policymakers to bundle investment climate reforms with a larger package of macroeconomic and structural reforms aimed at addressing the crisis. In the Slovak Republic the 1998 fiscal crisis created the political support the new government needed to pass a comprehensive reform package that included a range of controversial labor, social welfare, and tax reforms (Jurajda and Mathernova 2004, p. 8). The government also benefited from a relatively homogeneous coalition of four center-right parties and from the relative underdevelopment of trade unions and other groups opposed to reforms. In Korea the 1997 Asian financial crisis enabled the government to move away from a piecemeal approach and adopt a radical deregulation initiative as part of a broader economic recovery program; the reforms called for eliminating 50 percent of business regulations and deepening regulatory reforms that had so far yielded little result (FIAS 2005f, p. 6).
But it was not always a fiscal or financial crisis that provided the impetus. Other types of crisis also helped propel reform onto the agenda: unprecedented levels of unemployment in the case of labor market reforms in Colombia and Poland; big corruption scandals in the case of regulatory reforms in Korea and Italy; and two major safety accidents in the case of inspections reform in the Netherlands.

Taking advantage of spillovers from trade and product market reforms
Growing competition from global integration and rapid technological change is a broad force driving change. But investment climate reforms are also pushed onto the agenda as a result of spillovers from more specific policy initiatives. Trade and product market reforms proved to be a major driver of other reforms in virtually all our case studies. By increasing competition, such reforms helped shift the incentives of incumbents once opposed to reform while creating new constituencies for change. In Mexico trade liberalization through the North American Free Trade Agreement (NAFTA) induced business associations to lobby the government for reductions in the regulatory burden to help them compete (Salas 2004, p. 6). And in Colombia greater openness and competition led employers to become vocal supporters of reforms aimed at increasing labor market flexibility (Echeverry and Santa Maria 2004, p. 13). That an economy's openness is significantly associated with institutional change is among the main findings of the IMF's World Economic Outlook, September 2005 (box 1). Conversely, the regions that have done the least on the investment climate front (such as South Asia and the Middle East and North Africa) also tend to be those with the highest barriers to trade and foreign direct investment.
Reformers also took advantage of international agreements to build popular support for a broader goal and create the climate for change. Such agreements were a major catalyst for reform in more than a third of our cases. Integration with the European Union (EU), which provides tangible political and economic benefits for member countries, was the main impetus in all the cases in Central and Eastern Europe. 9 In Latvia political consensus in favor of joining the EU combined with a push from a strong, well-organized private sector made it easier for the government to adopt regulatory and inspections reform as part of the broader EU harmonization program (Coolidge, Grava, and Putnina 2004, p. 9). To a lesser extent such countries as China and Vietnam leveraged the prospect of joining the World Trade Organization to deepen market reforms aimed at leveling the playing field between the public and private sectors (Mallon 2004, p. 20).

Box 1. What drives institutional change?
The International Monetary Fund's World Economic Outlook, September 2005(IMF 2005 investigates the main factors driving such institutional transitions as strengthening property rights, lowering corruption, and improving accountability. An econometric analysis based on transitions in some 65 countries uses a probit model linking the probability of transition to several possible explanatory factors. The findings suggest that while a country's institutions are shaped by a combination of history, economic structure, political system, and culture, they are not immutable. A range of factors can help bring about institutional change: Trade openness is significantly associated with a greater likelihood of institutional transitions and with greater institutional quality. A move from complete autarky to full liberalization increases the probability of transition by about 15 percentage points. Greater openness allows a greater role for export sectors that are rent-proof and require innovation, creating momentum for positive institutional changes. It also reduces the ability of domestic producers to sustain monopolistic rents, which impede improvements.
Transitions are also more likely in countries with high levels of press freedom, a broad indicator of political accountability. With the political leadership answerable to a broad cross-section of the population, policies are more likely to be beneficial to the broader economy. Greater accountability is also associated with higher institutional quality. Improvements are more likely in countries whose neighbors have higher institutional quality. This is consistent with the view that a strong regional effect exists: transitions are more likely to happen in clusters of countries within a region around the same time, reflecting competition and the demonstration effects of regional success stories.
By contrast aid appears to have a negative impact on the probability of transition. This may reflect the fact that countries that receive more aid are also those that suffer from disadvantageous initial conditions that impede change. The empirical evidence on the net effect of aid on institutions is mixed.

Generating new information to create demand for change
New information about a specific policy problem provides another impetus in creating demand for policy change. In Colombia the 2002 labor market reforms originated in a 1999 study led by a small group of technocrats in the Ministry of Labor supported by private think tanks and academia. By benchmarking regulatory performance against that of neighbors and OECD countries for the first time, the study helped spark a national debate and laid the groundwork for the passage of the new labor law in 2002 (Echeverry and Santa Maria 2004, p. 12). In Vietnam detailed diagnostic studies by the Central Institute for Economic Management, a government think tank, exposed the costs of doing business (including corruption) under the old Company Law. These analyses, providing the first measures of the potential growth and employment gains from reform, were used to counter opposition to the new Enterprise Law from vested interest groups (Mallon 2004, p. 30). In all the cases of land market reform new information about the costs and benefits of improving access to land among other issues helped focus attention on the problem and build broad consensus for change among a wide group of stakeholders.
International benchmarking indicators and cross-country comparisons are important in identifying priority areas for reform, but they are also valuable tools for fostering competition and promoting reform. The country rankings recently published by the World Bank Group's Doing Business project (World Bank 2005a) triggered calls for change in more than 20 countries. Sub-national benchmarking can help foster reform by creating competition between jurisdictions, as it is doing in India and Mexico. In these and other cases diffusion of ideas between national and sub-national levels has also helped jump-start the reform process. In Mexico federal regulatory reforms helped promote fast-track business start-up programs in several states and more than 20 cities (Salas and Kikeri 2005), while in China policy experimentation with land reforms in Shenzhen municipality led to changes at the national level (discussed in greater detail on page 18).
Reform often comes about through the accumulation of "intellectual innovations"-or new policy ideas-and through a dynamic process of policy learning and diffusion of best practices among specialists or by international and other organizations (Orenstein 2000, Stern et al 2005). Regulatory reforms at the federal level in Mexico were prompted in part by the example of the United States and by advice on best practices from the Organization for Economic Cooperation and Development (OECD). In other cases the direct exchange of experience helped convince policymakers of the need for reform.
Pakistani reformers learned about capital market reforms through visits to Australia and Malaysia, and mainland Chinese officials garnered ideas about land reform on frequent visits to Hong Kong (China).
The spread of new information has been aided by the information revolution of the past decade. The rapid proliferation of the Internet, mobile communications, and other technologies allows reformers and citizens in developing countries to learn what has been tried elsewhere, what has worked, and what has not. Especially in countries with a free press, these new sources of information have helped put pressure on governments to reform; indeed, some analysts have argued that the explosion of information was largely responsible for the spread of market-oriented reforms in Eastern Europe (Kedzie 1997).

Leading the way through political leadership
Getting reforms onto the agenda appears to be a largely top-down process: the policy reform literature suggests that political leaders (heads of government and their political appointees) play a key role in seizing opportunities and promoting reform, drawing on their institutional and organizational powers and their ability to command public attention (Kingdon 1995). But technocrats and policy entrepreneurs, both in the government and outside (think tanks, NGOs, academia), also perform a crucial role in this process, helping to shape and continuously push for reform.
The case studies underscored the importance of political leadership in seizing reform opportunities and getting the process started. In Mexico a powerful president who controlled the party in power was instrumental in launching regulatory reforms in the late 1980s that had been designed and promoted by a small group of technocrats in the Ministry of Economy and Trade. And in Peru presidential support was crucial to getting the second round of land market reforms onto the agenda in the early 1990s, aided by the persistent diagnostic and advocacy efforts of the Institute of Democracy and Liberty. Just as persuasive are the many converse examples from those countries where, despite no shortage of high-quality analysis and proposals, reforms failed to get onto the agenda because of a lack of political leadership.

Donor role
Donor pressure to initiate reforms was not an important factor in our cases, though it is often perceived as one in developing countries. Instead, reform was largely a domestic process driven by local champions. Part of the reason may be that many of the reforms we reviewed took place in the 1990s, when donors focused more on macroeconomic and structural reforms than on microeconomic ones. Investment climate reforms are still relatively new to the agenda, and donors, like policymakers, are on the learning curve.
Where donors were involved, in such countries as Hungary, Latvia, Pakistan, and the Slovak Republic, they focused largely on providing diagnostic support rather than (as in earlier reforms relating to macroeconomic or privatization policy) on imposing conditionality, given the different political and institutional challenges and timeframes of such reforms. Even so, the presence of donors in these cases was seen as strengthening the hand of reformers against their opponents.

STARTING WITH REFORMS THAT ARE BOTH CREDIBLE AND FEASIBLE
The accumulation and diffusion of knowledge and perspectives among policy specialists will generate a range of policy options in any area of reform. Which of these options should be considered? Should reformers start with radical options that may have bigger payoffs, or incremental steps that may provide smaller payoffs but are easier to implement? Can pilot and sector-specific programs help iron out technical difficulties and build support through demonstration effects? Different circumstances may call for different strategies and tactics. Reformers with strong political support (as a result of crisis, a new government, new leadership) may find it easier to "strike while the iron is hot," embracing faster and bolder reform. But reformers contending with divided politics, strong interest group resistance, and weak capacity may face constraints on the extent of change. In these cases policy learning through pilots and other pragmatic measures that target specific constraints, build on previous experiences, encourage the development of constituencies, and are easy to implement can help jump-start the reform process. 10 Our cases suggest there are simply no hard and fast rules. What matters more is to achieve policy credibility by recognizing and targeting the main policy constraint, adopting a clear strategic direction, and undertaking a steady and sustained process of change. 11

Setting the pace of reform-radical or incremental?
Agenda change can often be sudden and radical, aimed at signalling a shift in direction and securing political support (Kingdon 1995). But the process of reform tends to be more incremental-though crisis or new governments can create opportunities for more radical shifts. 12 Financial crises in Hungary and Korea for instance provided reformers strong political support to adopt far-reaching regulatory reforms, while the 1998 fiscal crisis did the same in the Slovak Republic. Similarly, in Mozambique the resettlement crisis made possible the complete revision of the Land Law in 1995, designed to eliminate the most important distortions and demonstrate credible government commitment to land reform (Tanner 2002, p. 15).
But radical change, while it may be desirable, may not always be feasible, especially in systems with many checks and balances, fragile coalition governments, strong interest group opposition, or weak implementation capacity. In circumstances such as these, the process of reform tends to be more incremental. Reform may start by adopting the easier measures first-to demonstrate feasibility, achieve initial successes, and develop the constituencies and capacity for deeper changes over time. Vietnam's enterprise reforms in 2000 for instance began by simplifying business start-up procedures. That helped expand the private sector and create a constituency for the more difficult licensing reforms, which then enabled reformers to offset resistance from line ministries and provincial authorities (Mallon 2004, p. 26). Peru's land market reforms began with the politically easier titling measures. These led to quick, tangible gains, helping create support and capacity for the more difficult and longer-term program to formalize property (Endo 2004, p. 10). And Mexico's deregulation unit started its ambitious program to reduce red tape by reviewing the regulations of simpler, more cooperative line ministries. In this way it slowly built experience and credibility to tackle the more complex ministries (Salas 2004, p. 16).
Incremental or partial reforms can be risky, however: if they produce no effects or even adverse effects, they can undermine the credibility of the entire reform effort. Before adopting a bold land reform program, Mozambique first took an incremental approach; Korea and the Slovak Republic did the same in regulatory reform. In these cases the incremental reforms were unsuccessful-and costly to taxpayers. Similarly, piecemeal inspections reforms in the Philippines and the Russian Federation were no more than short-term palliatives, and they quickly became subject to backtracking and reversals (Coolidge forthcoming).
Moreover, short-term winners from partial reform may act to preserve their gains by blocking further reform. A study of partial reform in the early stages of post-communist transitions shows that financial-industrial conglomerates that gained from privatization used their power to block new market entry, while new "entrepreneurs cum mafiosi" undermined efforts to establish a viable legal system to support the market economy (Hellman 1998). The study suggests that the success of partial reform depends on both creating winners and "constraining" them. Winners can be constrained by increasing political competition and expanding political participation to include short-term losers and a wider range of constituencies in policymaking. In this case post-communist states with greater political participation and competition were able to adopt and sustain more comprehensive reforms than those largely insulated from mass politics and electoral pressures.
Identifying the most important constraints at the beginning of the process is one way to minimize these risks and achieve credibility. Another is through the use of targeted, pragmatic interventions that seize opportunities for reform but are also politically and technically feasible. 13 Pilots and sector-specific approaches are among such interventions.

Using pilots and sector-specific reforms as learning and demonstration tools
When reform faces technical and political challenges, pilot programs can help by providing learning and demonstration effects. In China, where pilots have been at the center of the reform strategy, states and provinces have often functioned as "laboratories of reform," with policy experiments in one location feeding into the agenda-setting process for others (box 2). In Jordan customs reform was first developed and tested in the Aqaba special economic zone before being rolled out to the rest of the country. In Peru land reforms began with a series of pilot projects in the early 1990s that provided useful learning about constraints and effectiveness for the nationwide program (Endo 2004, p. 14). And in South Africa the property valuation approach using market assessments was applied to the entire country in 2004 (with Parliament's adoption of the National Property Tax Act) only after being piloted by Cape Town in 2000 (FIAS 2005h, p. 6).
Identifying and supporting sectors with high growth potential is another way to get reform started, as suggested by recent work outside of our case studies. In the Dominican Republic, for example, tourism and the maquila sector (the assembly of imported parts for re-export) were the main drivers of growth. Like the rest of the economy, they faced problems relating to infrastructure, security, and trade protectionism. But rather than tackle these difficult problems across the economy, the government insulated these two sectors and provided targeted solutions, improving security and infrastructure around the main tourist areas and giving the maquila sector special trade policy treatment (Hausmann, Rodrik, and Velasco 2006, p. 15).
Pilots or sector-specific experiments can be important catalysts for nationwide reform rather than enclaves, provided they are properly designed. Reformers need to design these initiatives from the outset to be replicable elsewhere. They also need to tackle institutional and other constraints to the expansion of reforms beyond the pilot areas. And they need to keep their attention focused on the pilot throughout the process-from design through implementation and monitoring-to learn the lessons offered for broader reforms. Most importantly, while pilot and sector-specific approaches help elicit information about what works, they also require willingness and clear criteria or mechanisms for cutting-off failures before they become too costly.

Box 2. Special economic zones as the laboratories for China's land reforms
In China special economic zones served as the laboratories for developing nationwide land reforms in the 1980s. The experiments with land started with the establishment of the Shenzhen special economic zone. When foreign firms started to invest in Shenzhen under China's "open door" policy, demand for land grew rapidly. Because the existing legal framework did not allow the transfer of land, the local government tested transfers of land use rights by allowing the allocation of state-owned land to serve as local firms' contributions to foreign joint ventures. Prompted by continued growth in demand and by learning from Hong Kong (China), the provincial government took the experiment further: in 1981 it issued regulations allowing the transfer of land use rights to investors and establishing fee standards for land use. By separating land use from land ownership rights, the new policy allowed foreign investors for the first time to lease land for a given period, with up-front fees for use rights. It also increased revenues for the provincial government and preserved its land ownership.
The experiment continued. In 1987 the provincial government declared the end of the old free land use policy and the Shenzhen municipality conducted the first sale of land use rights to a local public company. And in 1988 provincial ordinances officially recognized that land use rights could be transferred, assigned, or mortgaged. The successful experiment in Shenzhen led to a national constitutional amendment allowing transfers of land use rights (leases) "in accordance with the law"-the critical step that permitted land market reforms to spread to the rest of the country.
Source: FIAS case study of reform in Shenzhen, China, forthcoming.

BUILDING COALITIONS TO SUPPORT REFORM
Reform is most likely to succeed in a supportive political climate. Creating such a climate can be a central challenge of the reform process. It requires understanding the attitudes and influence of different stakeholder groups. It also requires the painstaking process of building coalitions for reform. And that requires strategies to leverage and mobilize supporters, both powerful and not so powerful, while diffusing resistance, especially from influential interest groups (figure 3).

Leveraging and empowering supporters
Supporters or potential supporters of reform (A and B in figure 3) are often overlooked or taken for granted. But mobilizing them can help increase the strength or density of support for reform, which in turn helps weaken the opposition and influence of interest groups.

Creating champions of reform
Strong and influential proponents of reform can be easily identified and mobilized to become champions for reform. In Mexico the small, empowered group of technocrats leading the first deregulation unit (UDE) became a major lobbying force for later transforming it into a commission backed by law (Cofemer). UDE and Cofemer in turn leveraged the support and influence of major industrial players and other private organizations by giving them a seat on the board of the Deregulation Council.

Oppose
Other reformers created small, independent teams as the focal point for challenging the status quo and promoting reform-usually an inter-ministerial commission, a special task force, or a working group reporting to top political leaders and comprising a mix of reform-minded technocrats and external experts. In Korea a group of presidential advisers chosen from outside the traditional bureaucracy championed the 1998 regulatory reforms (FIAS 2005f, p. 16). In Hungary a small, ad hoc task force in the Ministry of Finance closely linked to the prime minister's office led the 1994 regulatory reforms (FIAS 2005e, p. 24). And in the Slovak Republic a small team of advisers to the deputy prime minister for economic affairs-mostly trained overseas nationals lured back by the prospect of EU membership-worked to champion reform (Jurajda and Mathernova 2004, p. 19).

Informing and educating the public
Mobilizing less influential groups and the public at large poses a bigger challenge. Many investment climate reforms are opaque or even an object of suspicion to voters. Negative public opinion can impede or even prevent reform. But policymakers fear that educating the public takes too much time and that publicizing bad economic situations can risk making the problems worse. Reformers also are reluctant to undertake public education programs because they lack information about the programs' benefits.
Yet communications and education through the media and other means can help allay public fears, persuade citizens that change is in the national interest, and build broad public support for reform-with individual politicians and interest groups no longer able to resist public pressure for change. Public education is especially important in new or controversial reforms that are not widely understood. To reduce opposition to land reform from municipalities, Peru's government mobilized support from informal urban settlers, the main beneficiaries, business associations and political parties by sharing information about the reform and its objectives and holding public hearings and debates in communities (Endo 2004, p. 20). Other reformers built support by linking new reforms to broader goals that already enjoyed wide support. Korea, for example, framed regulatory reforms as an anticorruption campaign. Vietnam linked enterprise reforms directly to employment growth. And in Cape Town, South Africa, reformers tied property tax reforms to "common equity objectives," which secured support for a potentially divisive reform on the eve of local elections.

Encouraging public participation
Going further, reformers can empower potential supporters by seeking their active participation in the reform process. In Hungary the commissioner for public administration, responsible for deregulation, organized conferences and technical workshops throughout the country and invited citizens and public employees to submit deregulation proposals, promising a monetary prize for the best one (FIAS 2005e, p. 30). In Korea reformers solicited inputs on a wide range of regulatory policy initiatives from NGOs and broad coalitions such as the Citizens' Coalition for Better Government (FIAS 2005f, p. 12). And in Vietnam the government distributed drafts of the Enterprise Law to mass organizations and the public at large for comments and feedback (Mallon 2004, p. 36). These efforts helped foster national dialogue and generate public support--and provided a valuable citizen feedback loop in the reform process.

Diffusing stakeholder opposition
Diffusing resistance from opponents of reform (C and D in figure 3), particularly influential individuals or interest groups (D), is part of the process. Leveraging and empowering supporters and potential supporters are critical in this regard, but dialogue and compensation also have a part to play.

Establishing a process for dialogue and consultation
In the early stages of reform, public education and persuasion can generate "buy-in" for reform. Some reforms may also require narrower and more structured dialogue with the private sector and other stakeholder groups on concrete policy alternatives. Consultation is particularly important for policies which will have wide distributional consequences and are likely to impact multiple stakeholder groups. In such reforms, failing to consult can lead to missed opportunities for "discovery" or the vital exchange of information that can inform policy-making. 14 It can also create suspicion and defer resistance to a later stage, as it did in Korea and Mexico where the failure to build a sufficiently broad base of stakeholder support at an early stage hampered implementation of regulatory reforms (for more on this, see the section on creating oversight mechanisms). Moreover, dialogue and consultation can help make the outcome more credible, even at the expense of delays and compromise. And it can provide an incentive for interest groups to participate in positive ways while weakening their ability to directly influence government.
Consultation proved to be important in both democratic and non-democratic settings, though the extent of it differed from case to case depending on the type of reform and initial circumstances. Several rounds of negotiations and compromise among multiple players were needed to smooth the way for politically-laden labor market reforms in Colombia (box 3) and Poland. Consultation also played an important part in countries commonly perceived as having top-down government with limited opportunities for political participation. In Vietnam reformers secured approval of the Enterprise Law through a formal process of public-private dialogue on various drafts with private sector groups and associations. The process also helped create a framework for discussions with members of the National Assembly, the legal establishment, and grassroots organizations (Mallon 2004). Similarly, in China the government of Hangzhou municipality set up a "hearing system" to elicit feedback from external experts, business groups, and the public on proposals to change the investment approval system. By doing so, the government refined its proposals and built broad support that helped weaken the resistance of line agencies (Yufei, Lei, and Yu 2004).

Box 3. How consultation helped pave the way for labor market reforms in Colombia
Colombia's far-reaching labor reforms, passed in 2002, were, from the outset, deeply political and generated intense opposition from strong and powerful interest groups. Formal sector workers with the most to lose were vocal and active in their opposition through strikes and other means. Fierce opposition also emerged from other organized players who derived substantial benefits from the status quo: unions of the large public and private agencies that administered "parafiscal" taxes and services (a key non-wage labor cost imposed on employers to finance social welfare programs for the general population, amounting to 8 percent of payroll), which together managed about $750 million per year and employed nearly 25,000 people throughout Colombia; unionized blue collar workers who directly benefited from the social services; and member of congress, mayors, state governors and other career politicians who used the agencies for patronage and other political purposes. These groups also mobilized other more powerful groups not directly affected by the reforms but with considerable disrupting power, including the teachers' union, university students, and pensioners. Meanwhile the main beneficiaries of reform, the unemployed and informal workers, were too dispersed and too weak to lobby for reform.
Managing the opposition required efforts to slowly build consensus through stakeholder discussions. The government established a "discussion table" in 2000 with representatives from the most important groups-government, the main labor unions, key business federations, the main political parties, and academia. Their wide-ranging views made initial discussions of the reform contentious and confrontational. Agreeing on a diagnosis and solution proved challenging. To make the case for reform, the reform team presented the underlying economic analysis and empirical evidence. But selling the story proved difficult: interest groups rejected the evidence, questioned the underlying model, and successfully stigmatized the reforms as leading to unemployment and a reduction in income. Despite the difficulties, the continuity and conviction of a core group of technocrats proved crucial in keeping the discussions going.
The consultation process not only helped build awareness. It also exposed key players to reform ideas, and helped reach compromise on a set of proposals that were politically feasible and allowed for quick congressional approval. While many of the original proposals remained in tact, bargains were reached on two of the most controversial elements of the reform package. First, the payment of parafiscal taxes was made more flexible rather than reduced, as originally proposed and defeated in an earlier attempt in 2001. With some restrictions, the law exempted payment of such taxes for new blue collar workers for a maximum of four years, and for students under 25 years of age working part-time. And second, while the 2001 bill contained a provision for deepening the "salario integral", an instrument that allows employers to hire workers under a contract that does not oblige them to pay certain non-wage costs, the 2002 package of reforms was passed without this provision. While the reform package did not include any direct provisions to compensate losers, it included a host of broader social protection measures, including the contemplation of an unemployment insurance scheme, the strengthening of public pension and health programs, and the creation of a micro-credit program for small enterprises. The main lesson from Colombia's experience: it can take years of slowly "selling" a reform until an initiative actually gains momentum and is passed.
Source: Echeverry and Santa Maria 2004. The process of negotiating and bargaining with multiple groups can be difficult and confrontational, requiring a delicate balance between achieving political compromise and meaningful policies that do not jeopardize the benefits of reform. Our cases showed that the process requires support and commitment from the top political leadership. But it also requires the support of technocrats, and sometimes external moderators, to focus on the policy issues, maintain continuity and political neutrality, and ensure accountability and legitimacy of the process. In Vietnam leaders from the Central Institute for Economic Management and the Viet Nam Chamber of Commerce and Industry-the central institutional players in the process-played key roles in engaging with policymakers, business, and the media. In Latvia technocratic staff at the Latvian Development Agency and Bureau of Public Administration Reform supported and sustained the process throughout. And in Peru the involvement of the Institute of Liberty and Democracy was perceived as having made the process objective, fair, and transparent.
At the same time consultation may have its limits in difficult reforms facing intense opposition from narrow interest groups. In such cases policymakers have often confronted opponents, fearing that consultation can be counterproductive because it gives too much influence to specific groups, takes too much time and delays reform, or leads to minimal reform proposals. For these reasons port reform in Colombia and customs reform in Mexico were both largely top-down affairs: rampant corruption and inefficiency, broad public support for change, and a brief window of opportunity led reformers to proceed with little input from stakeholders.

Compensating losers
A few investment climate reforms, such as infrastructure reform, can involve clear groups of losers-for example, state enterprise workers with high job security and generous pay and benefits. In these cases mitigating opposition may require directly compensating such groups. The port reform in Colombia for instance included generous severance packages to minimize opposition from workers, while telecommunications reform in Uganda was accompanied by a package of guarantees and new pension benefits for workers transferring to the privatized entity. Direct compensation programs may raise concern that payments are being captured by politically powerful groups, but such concerns can be minimized by demonstrating the net benefits from the reform for society as a whole. The programs also can entail high up-front costs and can backfire if not carefully managed. In Colombia, the severance program helped mitigate opposition but also became a problematic part of the reform program, creating favorable conditions for corruption and fraud in its administration, which eventually led to the imprisonment of corrupt officials (Navarrete 2004).
Most other investment climate reforms have little or no immediate and direct impact on a clearly identifiable group of losers who feel sufficiently threatened or are sufficiently organized to object. In some product market reforms (such as opening sectors to competition, reforming business registration, and reducing red tape) those who stand to lose are firms exposed to the normal risks of doing business, individuals or bureaucratic groups engaged in corruption, or local politicians wanting to retain local authority functions. While these groups may resist reform, they pose a smaller political challenge and so usually are not directly compensated. Moreover, uncertainties about who will lose, who will gain, and by how much together make direct compensation difficult.
Many investment climate reforms therefore involve indirect forms of compensation, often through political or economic concessions in the design of the reform or through policies to offset its effect. Unions in Poland became more inclined to support the 2002 labor reforms when they could make a "deal" protecting their position against the rise of new unions: measures aimed at promoting labor market flexibility were accompanied by restrictions on dismissals for members of the three main trade unions. Also important: the costs of reform were not concentrated in any one group, and policymakers were able to convince the public that short-term benefits for employers would translate into better jobs for all. Similarly, in Colombia, combining the 2002 labor market reforms with a broad package of social protection measures (including pension and health programs) made it possible to reduce hiring and firing restrictions while meeting union concerns about the lack of social safety nets. * * * * * Once policymakers have identified the priorities for reform, gotten reform onto the policy agenda, developed credible and feasible proposals, and built coalitions to support the reform, they need to move on to the next stage: implementing the reform and sustaining it over time. That raises an entirely new set of challenges.

III. IMPLEMENTING AND SUSTAINING REFORM
Implementing and sustaining investment climate reforms is a critical part of the reform cycle. Yet this stage is often neglected. And it is at this stage that reform most commonly fails, for three main reasons. First, reforms tend to become overly personalized. Policy initiatives that depend heavily on a few key individuals often collapse once these individuals move on or are replaced. Second, investment climate reforms typically cut across different agencies and levels of government and, as a result, lack oversight and coordination. Third, the lower-level organizations and officials responsible for implementing reforms tend to have interests and objectives that differ from those of the policymakers who designed the reforms. Local governments and civil servants often revert to old practices once the political pressure for change has subsided. And they often have only weak capacity to carry out reform.
How can these institutional challenges be addressed? Key among the measures: Strengthening the incentives and capacity of the national and local officials and agencies responsible for implementing new regulations.
Creating institutional mechanisms to provide oversight and sustain reform.
Paying close attention to measuring and monitoring results, among the most important but weakest aspect of the reform process in all of our cases.

STRENGTHENING INCENTIVES AND CAPACITY
Reforms typically confront technical and administrative constraints at all levels of government, but especially at state and local levels, which carry the burden of implementation. This is also the stage where bureaucratic power typically exerts itself. These constraints are symptomatic of a need for broader public reform, even in relatively well-developed countries. Indeed, many of our cases highlight the interdependence of investment climate and public management reforms. Yet they also point to ways to overcome incentive and capacity constraints short of waiting for fundamental, long-term public sector reform.

Bringing in new leadership and reform teams
Implementing reform may require new leadership, as existing managers often find it difficult to reinvent themselves, are defensive about proposed changes, or suffer from a lack of credibility with stakeholders. Replacing obstructive officials in positions of influence with more reform-minded individuals helped smooth the way for regulatory reform in Hungary and the Slovak Republic. Bringing in new leadership and expertise from the private sector also proved to be a critical first step in reforming business registration and capital markets in Pakistan, and contributed to the success of customs reform in South Africa.
Developing small reform teams with staff recruited on a meritocratic basis is another way to bring in the expertise needed to implement reform. Including the core group of technocrats who helped design the reforms in the first place also helps ensure a sense of ownership and continuity, and can enable the process to persist despite changes in the political landscape. In Colombia the core team involved in formulating the 2002 labor market reforms later joined the ministry charged with implementing them. In Mozambique and Peru the small technical teams that designed land reforms continued to facilitate dialogue and implementation. And in Latvia the involvement of the core group of technocrats at the Latvian Development Agency and Bureau of Public Administration Reform allowed implementation to outlast several changes of government.

Revitalizing institutions
Creating the right incentives and capacity in several cases required transforming the implementing agencies into new, more service-oriented ones. Pakistan converted the government body responsible for business registration into the Securities and Exchange Commission of Pakistan, a higher-powered, semiautonomous agency backed by a legislative act (box 4). South Africa combined the Customs and Excise and Inland Revenue Departments to form the South Africa Revenue Service. As a semiautonomous agency, the new entity has the authority to set its own human resource policies, establish competitive pay packages, acquire and dispose of property, and impose fees or charges for its services (FIAS 2005j, p. 9). In both cases fresh resources from outside helped bolster the commitment and quality of the existing staff.
In a few cases governments created new implementing agencies to bypass existing agencies that were considered difficult to change. In Peru, where corruption and resistance from bureaucrats and notaries prevented reform of the national land registry in the early 1990s, the government created a new, parallel agency (Cofopri) in 1996 to register informal property and develop the national formalization program. The goal was to shift responsibility for property registration from the municipalities, which had been resisting reform, to a central jurisdiction with decision-making power (Endo 2004, pp. 11-13). Cofopri became an executive branch semi-autonomous agency chaired by a minister of state, reporting directly to the president.

Box 4. Transforming organizational culture to support reform in Pakistan
Created by law in 1998 to implement the Companies Ordinance, the Securities and Exchange Commission of Pakistan (SECP) has been markedly successful-largely thanks to a radical overhaul of organizational culture. In a difficult and unprecedented step, it dismissed (with paid salaries) 80 of the 380 staff it inherited from the old Corporate Law Authority (CLA). That provoked fierce resistance and legal challenges, but most of the 80 have since found other civil service posts. The 300 remaining employees were given a choice: continue as civil servants with their job security assured regardless of performance, or become employees of the SECP.
Becoming an employee would approximately double their pay and benefits, but it would also expose them to the risk of being made redundant if not deemed up to the job-and force them to compete for promotions purely on merit. All but one accepted the new terms of employment. The SECP supplemented these staff with 40 international and domestic experts, a group that brought along outside standards of service and integrity and became the driving force behind the cultural change. Unlike the CLA, where cases tended to be referred up a bureaucratic structure, the SECP encouraged employees to take responsibility for making decisions. It emphasized staff training, both in Pakistan and at regulatory bodies in other countries. And it set up a vigilance cell in the chairman's office to deal with complaints, within a year practically stamping out the staff corruption that had been entrenched in the CLA. Another critical step was the creation of a supervisory policy board to serve as a conduit between the government and the SECP. This proved crucial in overcoming the government's resistance to relinquishing the powers it had exercised through the CLA.

Creating performance targets and incentives
Aligning managerial incentives with reform objectives is important. One way to achieve this is by translating these objectives into specific performance targets or indicators which can then be independently monitored (monitoring aspects are discussed in greater detail in the section below). Examples include output measures (e.g. the number of land titles secured), input efficiency ratios (e.g. the number of companies registered per staff in the business registry), softer process measures (e.g. the incidence of employee and investor complaints), and outcome measures (e.g. a reduction in the incidence of workplace fires). In Korea, for example, President Kim Dae Jung's call for a 50 percent reduction in government red tape proved a useful goal for implementing regulatory reforms. 15 Similar targets were set in Australia, Hungary and the Netherlands, while EU accession requirements provided a clear set of targets for prospective member states in the Central and Eastern European cases.
Performance incentives help attract and retain skills, and foster compliance. Peru's new land management agency introduced productivity-based promotions to provide incentives for staff, while South Africa's new Revenue Service established competitive pay scales and performance bonuses that helped attract qualified and motivated customs officers (FIAS 2005j, p. 9). But these examples are among the few exceptions. More often, established civil service procedures made it difficult to implement performance incentives. In Mexico sanctions for poor performers were developed but never used. In the Philippines civil service constraints made it impossible for the customs reform to address organizational and human resource issues, and as a result reforms were consistently undermined by customs personnel and other special interest groups (FIAS 2005a, p. 1). When modifying civil service rules is difficult, other mechanisms can be considered. Hangzhou municipality in China used public recognition of managers as a motivating tool. It also tried to attract younger and better-educated employees and create a service culture that placed a higher value on experts and technical staff (Yufei, Lei, and Yu 2004).
Training can help improve performance incentives, by raising skills and increasing job satisfaction. In Latvia training for inspectorates was an integral part of the reform, with trained inspectors given a business advisory role in addition to their traditional roles. In South Africa, Cape Town's property tax reform program involved a major push to train data collectors and mass appraisal modelers, which helped create a conducive work environment and team commitment to the project's success (FIAS 2005h, p. 5). Yet training is often not provided or not pursued seriously enough. In Hungary and Korea insufficient training at provincial and local levels led to slower implementation than planned (FIAS 2005f, p. 25).

Contracting out reforms to the private sector
When it comes to implementing certain reforms, the private sector often has greater expertise and capability than the public sector. The Jordanian and Mexican governments both took advantage of private sector strengths-the Jordanian government by contracting out port and airport management and the Mexican government by contracting out environmental auditing as part of its inspections reform. In both cases the initial results are promising. The challenge in these cases was to align the short-term incentives of private contractors with the longer-term objectives of the government institution responsible for implementation, and to build up the government's capacity to oversee and support private contractors.

Harnessing information technology
Information technology is transforming the dynamics of the implementation process and can be a potent tool for sustaining change. Pakistan automated and electronically linked all its major business registries. Vietnam introduced business registration by Internet. And Mexico introduced electronic systems to help simplify environmental and customs inspections. All these information technology solutions helped simplify procedures, provide access to information (such as on laws, regulations, and procedures), and inform the public about its rights and about how to participate in the process. They also helped remove discretionary powers, improve transparency and accountability, and facilitate monitoring. Moreover, information technology solutions can be easy to replicate and scale up; in Jordan the customs system initially developed for Aqaba is now being used throughout the country. The effectiveness of these solutions depended in large part on simplifying the underlying processes and integrating them into broader change management efforts.

CREATING OVERSIGHT MECHANISMS TO SUSTAIN REFORM
Many investment climate reforms are ongoing and cross-cutting, involving many different departments and levels of government. Sustaining reforms can therefore demand special efforts to make the reforms permanent, insulate the process from political and bureaucratic interference, and ensure transparency and accountability. That can entail a new set of functions, including providing continuous oversight and advocacy, fostering policy coordination and compliance, supplying technical support to local levels, and monitoring results.
The need for such new functions led to the creation of oversight mechanisms in 60 percent of our cases, largely in cross-cutting regulatory and inspections reforms and in infrastructure reforms (to regulate sectoral policies and tariffs). Among the oversight mechanisms put into place to guide and sustain reform, the most common were independent commissions, followed by inter-ministerial coordinating committees and by new units created in an existing government department (table 1). Oversight mechanisms vary by type of reform in their detailed structure and mandate, but they tend to have common features aimed at ensuring independence from traditional insiders (such as political groups or business lobbies) and maintaining accountability. Key among these features: Credible mandates and objectives, often made explicit by legislation to signal government commitment, limit the scope for deviations or reversals by line agencies, and define the broad principles and guidelines surrounding the policy change. Reforming countries passed new "enabling" or umbrella laws that were difficult to argue against and not specific enough to provoke opposition from interest groups. In Korea the passage of a new Basic Administrative Law in 1996 made it difficult for line ministries to resist reform. Most OECD countries have adopted similar laws to support specialized, independent units involved in reforming business regulations.
Participation of a wide range of stakeholders, including high-level representatives from the government and, importantly, from the private sector and other groups. Stakeholders are brought in to contribute knowledge, experience, and ideas and to ensure transparency, minimize the risk of capture, and pressure reluctant line agencies to reform. The United Kingdom's Better Regulation Task Force drew its members from a range of backgrounds-including business, labor unions, and consumer groups-all appointed by the prime minister and charged with advising the government on the consistency of and compliance with new legislation (FIAS 2005d). Similarly, Mexico's Deregulation Council, created in 1995, brought together representatives from the government, the private sector, labor, academia, and other interest groups, such as agricultural organizations.
Development of transparency and accountability mechanisms, including full public disclosure of the regulatory process through the Internet. In most cases reformers publicly disseminated draft proposals and regulatory impact assessments to allow feedback from consumers and expose reluctant agenciesoften a sharp break from past practices of showing drafts only to selected interest groups. Such practices were designed to ensure that reform did not benefit favored groups over others. The oversight institutions themselves were held accountable by the participation of a wide range of stakeholders in their governing councils.
A need for greater integration with finance and planning authorities. Several cases showed the need for greater involvement of finance and planning authorities to help integrate reform into line ministries and provide incentives for compliance. In Korea a reform team was set up in the office of the prime minister to support the Regulatory Reform Commission, but without budgetary threats the group had little leverage over reluctant line ministries (FIAS 2005f, p. 21). Similarly, in Mexico Cofemer's location in the secretariat of economy rather than finance made it harder for the commission to exert its powers because of the secretariat's narrow scope and its inability to provide consistent oversight and to assess the budgetary impact of proposed measures (Salas and Kikeri 2005, p. 4).
Obtaining political backing and building broader ownership proved to be just as critical in implementing reform as in initiating it. Oversight agencies need political backing to do battle with the many vested interest groups opposing change, exercise leverage over reluctant line ministries and local governments, and ensure compliance. But sustaining political interest proved to be a challenge, especially under changing political circumstances. In Mexico presidential support was critical to the ability of the first deregulation agency (UDE) to remove price controls, repeal entry barriers, and simplify cumbersome commercial court procedures. But elections in 2000 fragmented Congress and weakened the president, making it harder for the new commission (Cofemer) to use its powers with line ministries, despite its new legal backing. In 2003, for example, Cofemer waived its right to issue an opinion on regulatory proposals for the underperforming telecommunications sector (Salas and Kikeri 2005, p. 4). A large part of the problem also stemmed from the fact that policymakers had worked only with the most senior public administration officials, leaving most of the bureaucracy feeling alienated from the reform effort. Korea's regulatory reforms similarly suffered as political support deteriorated, allowing different government departments to pull in different directions and resulting in piecemeal change (FIAS 2005f).

IMPROVING MONITORING
Measuring and monitoring results, though an essential part of the reform process, is often overlooked or not systematically followed up. This function involves several tasks: evaluating the potential costs and benefits of the reform, translating reform objectives into specific performance targets or indicators (as discussed above) that can then be monitored, and reviewing compliance and outcomes once implementation begins.
In only a few cases was monitoring effective. Australia was the most positive example. The Australian program included monitoring agreements from the start, setting out quantitative targets and requiring state governments to submit annual reports to the National Competition Commission detailing progress made and difficulties encountered. Latvia introduced "report card"-type surveys to monitor inspections reform (box 5). Measuring initial conditions made it possible to later benchmark the progress of reform against target indicators.

Box 5. Monitoring and evaluation helping to build trust in Latvia
In 1999, early in the process of EU accession, the new government of Latvia identified removing administrative barriers to investment as a priority for reform. It made inspections reform a critical part of this, prompted by complaints from business about burdensome, arbitrary, and harassing behavior by government inspectors. The government issued a new "instruction" aimed at improving transparency and accountability in all inspectorates, created an Inspections Coordination Council, and initiated training in a new "client orientation" for inspectors.
In 2001 the government was eager to evaluate its reform program. Earlier, progress had been assessed through focus groups and anecdotal evidence. Now the government wanted a stronger statistical basis to monitor progress and to determine which reforms were working as intended and which were off-track and in need of a new strategy. Supported by the Foreign Investment Advisory Service (FIAS), it carried out a self-assessment of progress. The evaluation used templates to gather official information about administrative procedures, a business survey to collect information about experiences with such procedures, and public-private dialogue to discuss the implications of the data and to use the data to guide revisions to the program.
The 2001 survey confirmed that inspections were no longer a serious problem for businesses. It also made it possible to establish a baseline, using hard data about the frequency and duration of inspections, the incidence of bribery, and the perceived quality of each inspectorate. To take the gains further, the government used the data from the templates to fine-tune the reform strategy.
A second survey, in 2003, documented a solid reduction in the inspections burden. The frequency and duration of inspections had declined, and the perceived quality of most inspections had improved. Interestingly, the survey results showed a significant difference between national-level inspectorates (fire, worker safety, sanitary), which had participated in the reform program with clear signs of improvement, and municipal-level inspectorates (construction, municipal police), which had not participated and showed no improvement.
In 2005 the government requested FIAS support for a third survey, including assistance to make the effort fully local. Local survey firms had been used since the beginning, but there was concern that a government-sponsored survey would not be credible with the business community. A task force was organized to oversee the survey, with representatives from business, academia, civil society, and the government. FIAS provided basic training in sample design, quality control, and safeguards to protect the anonymity of respondents. While the data from the 2005 survey have not yet been fully analyzed, the participants, including key stakeholders from both the public and the private sector, are pleased to have a mechanism that both sides can trust to monitor the impact of ongoing reforms.
Elsewhere systematic monitoring was absent, was poorly enforced, or ultimately bore little relation to the objectives of the reform. Even the United Kingdom had little systematic evaluation of outcomes. Instead, monitoring was largely left to outsiders such as the OECD and to academics (FIAS 2005d, p. 22). In Hungary the scope and breadth of reforms combined with a lack of administrative and analytical capacity undermined evaluation. And strong opposition from ministries made the process even more difficult (FIAS 2005e, p. 32). In Korea the problem was more a lack of coherence between targets and desired outcomes. When the regulatory agency was required to cut the number of regulations by 50 percent, it responded largely by focusing on less important and controversial ones.

IV. SUMMARY AND CONCLUSION
Our analysis of the literature and case studies on reforming the investment climate leads us to conclude with a summary of key lessons and issues that deserve greater analytical and practical attention.

MAIN LESSONS
There is no standard process for reform. There is and can be no "how to" manual. Different reforms involve different stakeholders and different mixes of technical, political, and institutional issues. Reform is also shaped by a country's politics and capacity. The best we can do is to highlight common insights and lessons which emerge:

1.
A growing set of diagnostic tools and proven good practices is making it easier to identify priorities for reform. Benchmarking indicators, country rankings, business surveys, and industry-specific analyses are now available to help reformers more easily identify and target the key constraints in their country's investment climate and therefore the priorities for reform. And as the case of China shows, reformers can also determine priorities through policy experimentation and learning from their own experiences and the successes of other countries.

2.
Exposing the economy to international competition through trade and product market reforms is a good place to start. Exports and competition not only are among the most important drivers of growth; they also help create domestic pressures for tackling the other important investment climate reforms-in land, labor, and capital markets. By increasing competition, trade liberalization in Mexico compelled business associations to lobby the government for reforms aimed at reducing regulatory red tape. In Colombia the greater competitiveness resulting from global integration led employers to become vocal supporters of labor market reforms. In Central and Eastern Europe reformers deliberately used the process of accession to the European Union (EU) to push through regulatory reform, building support for the reform by linking it to the broader goal of joining the EU. That an economy's openness is significantly associated with institutional change is among the main findings of the International Monetary Fund's World Economic Outlook, September 2005(IMF 2005. Conversely, the regions that have done the least on the investment climate front (such as South Asia and the Middle East and North Africa) also tend to be those with the highest barriers to trade and foreign direct investment.

3.
New information plays a powerful role in exposing a policy problem, fostering competition between jurisdictions, and creating demand for change. Generating new and specific information about a policy problem can be an especially important catalyst for contentious reforms involving multiple stakeholders. In land market reforms in China (Shenzhen), Mozambique, Peru, the Russian Federation (Veliky Novgorod), and South Africa (Cape Town), reformers used detailed diagnostic studies to identify problems of access to land and, for the first time, to measure the costs and benefits of tackling it through reform. These studies helped foster debate and overcome interest group opposition. In Vietnam detailed studies exposing the costs of corruption helped develop support to confront the many vested interests opposing reform of the old Company Law. Benchmarking and rankings provide another powerful impetus. The World Bank Group's recent publication of the Doing Business country rankings (World Bank 2005a), for instance, prompted calls for reform in more than 20 countries. In such countries as India and Mexico, sub-national benchmarking is playing a vital role in fostering reform due to increased competition between states.

4.
Crisis and political change provide opportunities to push through bold reforms. A deep fiscal crisis together with a newly elected government generated the support reformers in the Slovak Republic needed to adopt and carry out tax, labor, and other regulatory reforms far more ambitious than the piecemeal efforts of the past. Financial crisis helped thrust more ambitious regulatory reforms onto the agenda in Hungary in the mid-1990s. Other types of crises also helped propel reform: in Mozambique the resettlement crisis made radical overhaul of the land code possible; in Italy and the Republic of Korea major corruption scandals helped launch regulatory reforms; and in the Netherlands two major safety accidents pushed inspections reform onto the agenda. Strong political support for dealing with these crises made it easier to overcome resistance and gain wide acceptance for reform.

5.
Pilots and other pragmatic steps can provide important learning and demonstration effects which help get the reform process going, especially when a reform faces great uncertainty or strong opposition. Reform often comes about from a process of policy learning and experimentation. Pilots, in particular, can help create momentum and provide a testing ground for deeper change down the line. Peru and South Africa used pilots to test the technology involved in property registration and valuation, assess the difficulties larger-scale programs would face, and convince politicians that such programs were administratively and politically feasible. China has put pilots at the center of its reform strategy, using special economic zones to test market-oriented policies such as land use rights before extending them nationwide.

6.
Leveraging and empowering supporters through education and dialogue can help mitigate interest group opposition. Supporters and potential supporters are often taken for granted, but educating and mobilizing them early on helps create coalitions for change and minimize the incentives and ability of narrow interest groups or politicians to obstruct the reform process. Greater discussion may be needed for reforms that will have wide distributional consequences and are likely to involve multiple stakeholder groups. These strategies appeared to be important in both democratic and non-democratic settings: the shift to market-oriented policies in China and Vietnam involved greater discussion and attention to public opinion than is commonly perceived. Yet consultation may have its limits even in democratic settings, especially when reform is resisted by narrow interest groups. Under pressure to change, policymakers pursuing port reform in Colombia and customs reform in Mexico opted for rapid, top-down change, made possible by rampant corruption and inefficiency, broad public support, and a brief political window of opportunity for reform.

7.
Incentives and capacity for implementation can be created short of fundamental public sector reform. Technical and administrative constraints-particularly at state and local levels which carry the burden of implementation-highlight the interdependence of investment climate and public management reforms. Yet they can be overcome without undertaking long-term public sector reform, by leveraging change management techniques from the private sector. To support business registration and capital market reforms, Pakistan transformed an existing government body into a new, high-powered commission, driven largely by leadership and skills brought in from the outside. The new team helped introduce flexible financial and administrative procedures, foster a service culture, and train staff in all the regional offices. Others contracted out implementation directly to the private sector-Jordan the management of ports and airports and Mexico its environmental auditing. Information technology solutions also were instrumental. In customs, business, inspections, and land registration reform, such solutions helped in simplifying procedures and alleviating capacity constraints, but also in removing undue discretionary powers, minimizing corruption, and forcing transparency.

8.
Investment climate reforms, more cross-cutting and continuous than one-off events, call for special efforts to make the reforms permanent, insulate the process from political and bureaucratic interference, and ensure transparency and accountability. That can mean a need for new functions-providing oversight and advocacy, fostering policy coordination and compliance, and supplying technical support to local levels. This need led to the creation of new oversight mechanisms-often in the form of dedicated and empowered teams or institutions-in more than 60 percent of our cases. Regulatory reforms in Hungary, Korea, and Mexico for instance adapted the oversight mechanisms established in most OECD countries, supported by new umbrella laws specifying the standards and processes for reviewing the stock and flow of regulations.

9.
Monitoring should be an integral part of the reform process, not an afterthought. Monitoring was effective in only a few cases, even among more developed countries. Among the few positive examples were Australia and Latvia, which built in clear performance targets and monitoring agreements at the start. Elsewhere monitoring was poorly enforced, or bore little relation to the underlying objectives of the reform. Monitoring fell short even in higher-capacity countries such as Korea and Hungary, where lack of analytical capabilities and opposition from ministries made the process difficult. In low-capacity countries monitoring proved to be even more difficult, and harder than initiating reform. Yet monitoring is critical-not just to evaluate impacts and outcomes, but also to ensure transparency, provide a feedback loop to adjust course as needed, allow citizens to hold reformers accountable for results, and build support for sustaining reform.

10.
Above all, getting the reform process right is just as important as ensuring the sound technical content of the reform. Paying attention to the way in which reforms are initiated and carried out builds the legitimacy, support and ownership needed to achieve policies and outcomes that are both desirable and sustainable. As the economist Joseph Stiglitz has noted, reform process and in particular "implementation and political much is known, with the partial exception of labor and land market reforms, about the politics and institutional aspects of specific reform areas such as business registration, licensing, investor protection, contract enforcement, and bankruptcy. More case studies of how reforms play out under different political and institutional country conditions could also be useful in answering questions such as: How do reform processes differ between democratic and authoritarian regimes-and between presidential and parliamentary systems? What is the tradeoff between speed of decision-making and the credibility and sustainability of reform? And how do countries with weak governance/low capacity go about reform compared to countries with more accountable governments and better institutional capacity?
Second, deeper analysis of complementarities across reform areas could be helpful in thinking about how to package and sequence reforms, particularly in countries with least capacity or in post-conflict. Which reforms can be bundled and which should not beand which reforms are unlikely to produce results unless supported by action in other areas? More important, given limited capacity, what measures are needed to provide clear signals of policy credibility or policy certainty?
Third, much remains to be learned about building constituencies for change. More indepth analyses of how persuasion and bargaining strategies work in practice to influence and change stakeholders' preferences over time could be useful, for instance through more analysis of: public information campaigns; side payments or outright buyouts; grandfather clauses and phase-ins and phase-outs of policy that spread the pain over time; and the involvement of private business associations, policy think tanks, academia, and watch dog and other civil society organizations in launching and sustaining reform. The role of the media and of a free press in making the public an active participant and pressure group for reform is another important area for more analysis. Future work along these lines would help in moving from a static to a more dynamic understanding of building and supporting agents of change.
A fourth issue pertains to institutional arrangements for reform. Investment climate reforms often require an organizational overhaul of the existing bureaucracy or the creation of new oversight or advocacy mechanisms to lead and sustain reform. But more detailed analyses, especially of experiences in low-income, low-capacity countries, are needed to ensure that proper arrangements are in place. In particular, more work is needed to shed light on such questions as these: How can incentives and capacity for implementation be strengthened? Are new institutional arrangements needed? If so, what types and under what circumstances? How do such institutions build ownership of reforms at the local level? What tools and approaches can and should they use to ensure transparency and accountability-and what are the main factors in ensuring their effectiveness?
Fifth, monitoring, often the most neglected part of the reform process, is an area that most invites in-depth work. Much of the problem stems from the lack of clearly defined performance indicators and monitoring and evaluation systems. Thus more work is needed to develop reform-specific performance indicators and assess what it takes to put into place effective monitoring and evaluation systems and capabilities. All this is central to evaluating impacts and outcomes, ensuring transparency and accountability, and providing a feedback loop to adjust course as needed.
The role of development partners, not explicitly addressed in this paper, is a final issue that deserves more emphasis. In particular, what role can donors and international financial institutions play? Can reformers involve the private sector and civil society organizations (grassroots organizations, advocacy groups, service providers) in the delivery of investment climate services, building on the experiences and models from public service sectors such as health? Would regional or sectoral networks of practitioners help in information sharing and capacity building, and what lessons can be drawn from other reforms such as private participation in infrastructure where such networks have played a supportive role?