68494 Scaling up Infrastructure: Building on Strengths, Learning from Mistakes Infrastructure Network January 30, 2006 Scaling Up Infrastructure: Building on Strengths, Learning from Mistakes 2 INTRODUCTION INFRASTRUCTURE IS A RISKY BUSINESS ... WITH POTENTIALLY GREAT REWARDS. Infrastructure is complex, controversial, Why the decrease? The Bank was In our shift, we forgot about the synergies We therefore embrace the complexity and often risky. There is potential for listening to critics, including civil society between poverty alleviation, growth, and and risk of infrastructure investments, white elephants, environmental damage, opposition to big infrastructure projects. infrastructure: in order to make a difference for these loss of livelihood, and corruption. We expected the private sector to take people, including the 1.6 billion who lack the government’s place as the primary n Infrastructure is the backbone of a access to electricity, and the 1.1 billion We know that because we’ve seen it, �nancier of infrastructure, and shifted functioning economy, facilitating who lack access to clean water. and we’ve been on the wrong side of the our assistance to sectors, such as health economic growth, without which equation at times, making some mistakes and education, that on the surface seemed sustainable poverty reduction is The past criticism of the Bank’s projects along the way. to be more directly linked to the Bank’s impossible. Without decent roads, and our own evaluations of program poverty reduction objectives. And, in farmers cannot deliver their products effectiveness helped us to learn from our From 1998 to 2002, the Bank vastly cut some sectors, like water supply, our track to consumers. Without reliable mistakes and improve the quality of what its infrastructure lending from $9 billion record of results was poor. electricity, manufacturers cannot we do. Today, we are again scaling up to about $5 billion, which—at less than compete in today’s markets. our engagement in infrastructure to make 30 percent of Bank lending—was an all- Since then, however, private-sector greater impact where it matters most. time low. investment has not �lled the gap. In n The 2.7 billion people who live developing countries, total infrastructure on less than $2 a day need access investment would need to be equivalent to clean water, sanitation, energy, to 5.5 percent of GDP, while actual transport, information, and expenditures are only about 3.5 percent. communication. Children cannot And private-sector investment in get to school because they spend infrastructure briefly peaked at about their time fetching water. Rural $110 billion a year in the late 1990s clinics cannot store vaccines and use before declining to about $60 billion a modern diagnostic tools because they year later. lack electricity. 1 Scaling Up Infrastructure: Building on Strengths, Learning from Mistakes 3 BOX 1.1 BALANCING GROWTH WITH ACCESS: THE CASE OF VIETNAM’S INFRASTRUCTURE INFRASTRUCTURE LESSON #1 Background. is a good example where Results. Vietnam’s total investments in the Bank’s infrastructure strategy is balancing infrastructure have been close to 10 percent BALANCE GROWTH WITH ACCESS. growth and access. Several recent Bank-�nanced of GDP in recent years, which is a very high projects there have promoted economic growth level by international standards. As a result, We need to balance infrastructure investments by eliminating infrastructure bottlenecks. These Vietnam’s road network has more than doubled that promote economic growth with those that include two projects that have enhanced the in length since 1990, and its quality has improved target enhanced access for the poor. national road and electricity networks, and substantially. Access to clean water grew from one that is improving transportation and flood 26 percent of the population to 49 percent protection in the Mekong Delta. Other Bank between 1993 and 2002, and during the same projects are promoting access directly—such time, access to hygienic latrines grew from 10 as rural electricity, rural transport, urban water percent to 25 percent of the population. More supply and sanitation, and urban upgrading—by generally, per capita growth has averaged 5.9 using participatory approaches, connection percent, and poverty (measured at the $1 a day subsidies, and innovative low-cost solutions. threshold) has fallen from 51 percent of the In the past. In the early decades of long to show results. More recently, population to just 8 percent. Bank activity, investment mostly the lack of participatory mechanisms focused on large-scale infrastructure and unsustainable, poorly targeted managed by central government subsidies undermined the success agencies, and aimed at facilitating of the “basic needs� approach. BOX 1.2 economic growth. Poverty reduction was We learned. It is a false choice to PROVIDING ACCESS TO WATER: MOVING FROM CENTRALLY DRIVEN TO COMMUNITY secondary, pursued via “trickle-down� pursue an infrastructure strategy DRIVEN APPROACHES IN INDIA bene�ts. During the 1980s, the Bank that favors either growth or access, adopted the “basic needs� approach Background. Until the mid-1990s, Bank- shift, with government agencies changing from and we need to balance both. �nanced rural water supply projects in India providers to facilitators and rural communities as a counterpoint to previous “trickle down.� It aimed at bringing a standard Our approach has changed. The relied on large government agencies to gaining control over �nancial resources and design and implement water supply schemes. driving project implementation. package of infrastructure services to Bank’s infrastructure strategy is The Maharashtra Rural Water Supply and the poor, but was still centrally driven combining support for infrastructure Environmental Sanitation Project (approved in Results. The Uttar Pradesh Rural Water Supply and lacked community involvement. investments that promote economic 1991) was emblematic of this centrally driven, and Sanitation Project (approved in 1996) is a good growth with a more careful delivery of engineering-dominated approach. In the interest example of the new community-driven approach. The result was. Each phase had its The project empowers village communities to services to the poor, tailored to their of serving the “basic needs� of the poor, the weaknesses: during the early years, project provided perverse incentives to higher- make design choices and procure materials, needs and ensuring affordability. services, and civil works.They are supported by the “trickle down� approach took too level government agencies to build costly water supply systems without much consideration NGOs that assist with community mobilization given to community preferences and affordability. and private �rms that provide technical design, Village- and district-level authorities were inspection, and monitoring services. Investments unwilling to take over the operation and in water supply and sanitation are complemented maintenance of the schemes, arguing that their by programs promoting health awareness, technical skills and �nancial resources were women’s development, and non-formal education. not up to the task. Because of the disconnect The project has achieved full cost recovery for between service providers and bene�ciaries, operation and maintenance, and partial cost many of the schemes fell into disrepair after a recovery for capital costs—major improvements few years. In 1995, the Bank began working with over past practice in the Indian water sector. the Indian authorities to develop a sustainable Recent evaluations of sustainability have shown approach to the delivery of improved water that 92 percent of the water supply schemes and supply and sanitation to the rural population. 100 percent of the latrines �nanced by the project The new approach represented a paradigm are fully functional and in use by the bene�ciaries. 2 Scaling Up Infrastructure: Building on Strengths, Learning from Mistakes 4 BOX 2.1 WATER PRIVATIZATION IN BOLIVIA: GETTING THE POLITICAL ECONOMY WRONG INFRASTRUCTURE LESSON #2 Background. The case of Bolivia’s water to new users in the poor area of El Alto. But sector in the 1990s is a good example of in 2000, the national regulator refused the ENGAGE ALONG THE ENTIRE SPECTRUM flawed partnerships between the public and increase, and instead increased the price for OF PUBLIC-PRIVATE SOLUTIONS. private sectors. Bolivia’s Major Cities Water new connections. By 2004, the price for a and Sewerage Rehabilitation project covered new connection stood at $450 for water and The choice between public and private provision 60 percent of the country’s urban population, sewerage, six months’ wages for poor Bolivians. should be driven by local conditions, not ideology. located in the three main cities. In Santa In early 2005, faced with public protests over Cruz, a participatory consumer cooperative the high connection prices, the government Private sector participation is not the goal, but helped results surpass expectations. The Bank cancelled the concession. it can be a useful tool if the circumstances are required the national government to privatize right. What matters most is results. the water and sewage utilities in La Paz and Results in Cochabamba. In Cochabamba, Cochabamba before receiving the project loan. the municipality challenged the national This privatization provoked social unrest and a government’s transformation of the utility political impasse that resulted in the cancellation into a state corporation, and the privatization transaction was declared void. Municipal efforts In the past. For several decades, the We learned. Private participation is of each concession. to privatize the utility produced a tender Bank supported public investments in not a panacea—governments remain Results in La Paz. Privatization focused for which no bids were received. Finally, in infrastructure, rarely turning its attention central to the delivery of infrastructure on low-income neighborhoods and achieved late 1999, an unsolicited bidder, Aguas del to private provision. Political pressures services, either as providers or enablers. positive results initially. Average tariffs were Tunari, was awarded the concession: tariffs often undermined service quality No other actor can compensate for increased by 35 percent prior to the transaction. were increased by 35 percent, but without and inflated costs, and many public government weakness in strategy, Post-transaction tariffs included a cross- subsequent improvements in service quality. utilities became �nancially distressed, regulation, or risk management. In subsidy from existing consumers to pay for The tariff increase triggered riots and the rapid new connections, which helped extend service cancellation of the concession. lacking resources for maintenance and the proper environment, private sector service expansion. By the mid-1990s, innovation and ef�ciency can produce the Bank became reluctant to lend to great results. However, depoliticizing BOX 2.2 governments for public infrastructure prices is dif�cult, particularly at the ENGAGING ALONG THE ENTIRE PUBLIC-PRIVATE SPECTRUM: PRAGMATISM IN MOROCCO projects; some believed the private retail interface with electricity and water Background. In the Middle East and North provided two loans to Morocco to rehabilitate sector could eventually replace the consumers, and building up regulatory Africa (MNA) region, Morocco has pioneered and extend the services provided by public public sector as the main �nancier. capacity takes considerable time and public-private partnerships in the provision utilities in several cities and towns. effort. Reaching out to those who will of infrastructure services. This effort began in The result was. After a quick expansion, be affected by the changes, especially the early 1990s with the Maghreb Gas Pipeline Results in Morocco. Morocco’s approach to private infrastructure investment peaked Project, when the Bank provided technical public-private partnerships has been pragmatic, the poor, and promoting transparency, aiming at complementing public investments in at $114 million in 1997 and dipped to are essential to building public support. support to structure the transaction and new facilities, as well as increasing the ef�ciency $56 billion in 2003. It was strongly Risks should be carefully allocated �nancial assistance in the form of a Partial Risk and quality of existing infrastructure services. concentrated in a few countries, with Guarantee. Since then, the government has between the public and private sectors. extended private participation to a full range The methods differ across sectors, and include the top �ve recipients receiving half licensing of new private operators, concessions of infrastructure services, including electricity of all private infrastructure investment Our approach has changed. Solutions generation and distribution, water supply, solid and management contracts for municipal services, from 1984 to 2004. Moreover, it was are being tailored to the speci�cs of each waste management, telecommunication, urban and the privatization of state-owned companies. also concentrated by sector, with country and sector. Fostering partnership transport, and port operations. Whereas the The program has proceeded in harmony with telecommunications receiving almost between public and private actors is a Bank’s �nancial support has only been required the government’s ability to create a proper regulatory environment and to monitor results. half of all private infrastructure long-term process, and countries should for a small share of transactions (mostly in the energy and telecommunication sectors), the Overall, Morocco has attracted more than $13 investment, and water and sanitation move at their own pace. The Bank is billion in private investments in infrastructure, helping them to address capacity and government has frequently called on the Bank’s just 5 percent. In addition, a number of by far the highest amount among developing expertise, particularly to learn from other high-pro�le privatization transactions lost institutional constraints, facilitate the countries’ experience. Meanwhile, the Bank did countries in the MNA region. At the same public support and became distressed, sharing of experience, and assist with not neglect those companies that remained time, Morocco also managed to improve the especially in the more dif�cult sectors the management and mitigation of risks. publicly owned and managed. In the water supply performance of public infrastructure providers. such as water and electricity. The Bank is also supporting purely and sanitation sector, for example, the Bank public provision, when this represents the right infrastructure solution. 3 Scaling Up Infrastructure: Building on Strengths, Learning from Mistakes 5 BOX 3.1 FAILING TO SAFEGUARD PEOPLE AND NATURE: NARMADA DAM IN INDIA INFRASTRUCTURE LESSON #3 Background. Perhaps no project has damaged Results. Neither the Bank nor its partners the Bank’s reputation like the Sardar Sarovar suf�ciently consulted with the affected people: FROM THE OUTSET, DESIGN PROJECTS Project (Narmada) in India, where neither the what were their concerns and priorities? The THAT SAFEGUARD PEOPLE AND NATURE. Environmental Assessment nor the Resettlement Indian state governments resettled some of Plan was prepared prior to appraisal—as the displaced, but the process was flawed: Assess social and environmental impacts carefully, required by Bank policies. The Bank’s �rst land was poor and compensation rates were and integrate such assessments into project design. resettlement specialist went to the �eld after the inadequate. Unsurprisingly, several Indian NGOs project was appraised. He found that the three led protests. In 1992, the Bank commissioned an With the knowledge that most infrastructure participating state governments did not know independent review (Morse Commission), which projects create winners and losers, ensure that the number of people that would be affected, concluded that the Bank did not comply with its project bene�ts are channeled to potential losers. how many were tribal peoples, or where they Operational Manual Statements on Environment, wanted to relocate. There were no feasible plans Resettlement, and Tribal Peoples. Moreover, or institutional capacity for the resettlement of it concluded, appropriate resettlement and 100,000 people from the submergence area and rehabilitation was not possible under prevailing In the past. Projects were designed Our approach has changed. Since the 140,000 people affected by the canal system. circumstances. The Commission recommended �rst, and the mitigation of their mid-1990s, the Bank has been conducting However, in 1985, despite this specialist’s severe that the Bank step back from the project. environmental and social impacts environmental and social analyses and misgivings, the Bank agreed to �nance the dam’s In 1993, India requested cancellation of the construction (approving an IBRD loan of $200 $165 million undisbursed loan balance, and was handled as an afterthought. consulting with affected people early in million and an IDA credit of SDR99.7 million). subsequently completed the dams with its own the project cycle. Strengthened internal funds in collaboration with the three relevant The result was. At appraisal, there were oversight through a central quality- state governments. often no feasible plans for environmental assurance and compliance unit has rehabilitation and the resettlement and improved consistency and effectiveness compensation of displaced communities. in the application of safeguard policies. Implementation was delayed, projects The Bank also established transparent were restructured, and in some cases BOX 3.2 accountability mechanisms, including BRINGING BENEFITS TO PEOPLE AND NATURE: THE BOLIVIA – BRAZIL GAS PIPELINE cancelled. The Bank’s reputation was recourse to an independent Inspection damaged in the eyes of its clients Panel for affected parties. We have also Background. The Bolivia-Brazil Gas Pipeline habitat. Moreover, the project supported the as well as other stakeholders. become more adept at engaging civil Project (approved in 1997 and closed in conservation of biodiversity. A trust fund of 2001) incorporated strong measures to $1 million was established to protect Bolivia’s We learned. Environmental and social society, most importantly at the local address potential negative environmental and Kaa-Iya National Park, which is co-managed dimensions should be integrated into level, to take their views into account as social impacts in the design, preparation, and by an indigenous NGO in collaboration with project preparation, appraisal, and we help governments design sustainable implementation stages. The project worked Bolivia’s National Protected Areas Agency. supervision. This is best done through projects. That said, we have also become closely with community-based organizations Other examples of innovative solutions included early engagement on safeguard issues more decisive in supporting dif�cult but and consulted the public on draft regulations the use of semi-domesticated birds for pest high-reward projects, taking risks once and the project’s environmental assessment. control, since the application of chemical in the identi�cation and design stage. Resettlement was treated not as a problem insecticides was forbidden in the early stages of we conclude that the design is sound and but as an opportunity. Affected Indigenous construction. that proper arrangements are in place People negotiated a share of project bene�ts in to make sure that results are achieved. exchange for their acceptance of the pipeline Results. The project received the passing through their lands. Their priorities Environmental Award of the International were more secure land rights (land titling) Association of Impact Assessment in 2001. In and funds they could manage for their own only four years, the project constructed the development. The involvement of Indigenous largest gas pipeline in Latin America, through People in vegetation recovery in Bolivia was 3,150 kilometers of diverse terrain—and initial also important. They actively participated in results suggest that both people and nature sowing and harvesting native plants for their were respected in the process. subsequent reintroduction into their natural 4 Scaling Up Infrastructure: Building on Strengths, Learning from Mistakes 6 BOX 4.1 LOSING CONTROL DURING IMPLEMENTATION: THE CASE OF SULAWESI IN INDONESIA INFRASTRUCTURE LESSON #4 Background. The Indonesia Second Sulawesi Results. The Bank, the central government, Urban Development Project illustrates the and the four provincial governments lost CONFRONT CORRUPTION DECISIVELY. ill-effects of corruption when project design and control of the project. Local project of�cials supervision arrangements fail to include adequate took advantage of this, and diverted funds Infrastructure is particularly vulnerable to �duciary controls in a country with major for their personal bene�t. A �duciary review corruption, which degrades quality, increases governance weaknesses. During negotiations in and a subsequent forensic audit conducted costs, keeps honest investors away, and 1996, the Bank agreed to a major expansion of in 2002 found widespread de�ciencies in project scope (from 20 to 41 participating local contract documentation and clear patterns undermines public trust. We should �ght it at all governments, with more than 150 implementing of fraud and corruption. The Bank declared levels—in the projects we �nance, in the sectors agencies), in spite of the already very complex— many contract packages as misprocurement, we engage, and in the countries we support. multi-sectoral and multi-provincial—project initiated debarment sanctions against more design and weak provincial oversight capacities. than 100 �rms and individuals, and cancelled the This expansion made supervision an almost remaining loan amount. impossible challenge, even before the Asian crisis In the past. For most of the Bank’s However, we also learned that more and the political implosion of the country in 1998. history, corruption was not addressed needs to be done both at the project A very ambitious decentralization legislation explicitly. The Bank was not proactive, and sector levels, and that we still don’t (passed in 1999) unfortunately lacked clear implementing regulations, and added to an already and the Bank’s procurement and know enough about the relationship chaotic situation. �nancial management policies did between corruption and institutional not have explicit anti-corruption arrangements in infrastructure. features. No analytic work was carried BOX 4.2 Our approach has changed. We are out on the incidence and impact employing new tools to �ght corruption DESIGNING PROJECTS TO GUARD AGAINST CORRUPTION: BANGLADESH RURAL of corruption in infrastructure. in Bank-�nanced projects, including ELECTRIFICATION The result was. We had little knowledge public debarment, investigations by the Background. In Bangladesh, a Bank-�nanced years. A good performance record as a meter about the scope and nature of corruption, Department of Institutional Integrity, project (approved in FY02) is supporting reader can lead to a linesman or other job, either at the level of individual projects, and in a pilot stage, voluntary disclosure. the work of the Rural Electri�cation Board and this job expectation is a strong incentive or more generally in the infrastructure We are also cooperating with a range (REB) and its rural electric cooperatives to to maintain a good track record. sectors. We knew that corruption was of public and private partners, for expand electricity services in rural areas. The n The cooperatives use independent hurting project results, but we were example, in the context of the Extractive cooperatives have protected themselves from the corrupt practices commonly seen in other consultants to survey rural areas to identify unable to mount a proactive response. Industries Transparency Initiative. Work potential consumers, design the electricity is underway in about 20 countries aimed power sector utilities through a number of In the second half of the 1990s with the distribution network, and estimate the innovative arrangements: launch of the anti-corruption strategy, at improving government accountability revenue to be generated by each proposed the Bank’s approach radically changed. for the use of resource revenues. At n The Board of each cooperative is elected line. The list of lines to be constructed each the same time, we are integrating by consumers. This Board and REB year is disclosed to the public on the notice We learned. Shareholders and clients improved revenue management and management approve the salary structure for board, reducing the risk of corruption and welcomed the Bank’s determination to governance into the design of all Bank- the cooperative, which is usually market- favoritism in investment decisions. �ght corruption head on. Strengthened based. Every year the management of each �nanced extractive-industry projects. cooperative negotiates a set of performance Results. By expanding electricity service �nancial management and procurement And we are launching a systematic to 0.6 million new consumers per year, REB’s targets with REB. If the targets are met, systems helped to prevent some of effort to deepen our knowledge of the cooperatives reached 45,000 villages with 6.5 management receives a bonus. Not meeting the abuses at the project level, and the corruption-infrastructure interface, the PTA targets results in penalties. million consumers at the end of 2005. The promotion of improved governance and with the goal of developing and cooperatives maintained low system losses (13% the rule of law raised the awareness of n Since meter reading is a common source on average) and high collection performance implementing an effective anti-corruption of corruption, meter readers are hired (97%). Rural electri�cation increased agricultural decisionmakers and the general public. program in the infrastructure sector. on contracts of only one year. With good productivity, raised education levels, and performance record the contract may be improved the quality of health services. extended, but never can it exceed three 5 Scaling Up Infrastructure: Building on Strengths, Learning from Mistakes 7 BOX 5.1 NEGLECTING THE BASICS: ALGERIA URBAN WATER SUPPLY AND SEWERAGE INFRASTRUCTURE LESSON #5 Background. The Algeria Urban Water Results. The government failed to increase Supply and Sewerage Rehabilitation Project water tariffs, resulting in the deterioration of DON’T FORGET THE BASICS. (approved in 1994 and closed in 2003) illustrates the utilities’ �nancial situation. The project the consequences of inadequate appraisal of failed to achieve any of its objectives: it did not While integrating the Bank’s new agenda, we need technical design, �nancial sustainability, and complete the intended rehabilitation of water to ensure that the basics of project preparation implementation capacity. The implementing networks in the cities of Oran and Algiers, the and appraisal—technical design, economic agencies were largely weak and poorly utilities did not become self-�nancing entities, coordinated. Inef�cient procurement and nor were there improvements in leakage control and �nancial analyses, and implementation contract management led to long delays. and the reduction of unaccounted-for-water. arrangements—don’t get crowded out. Engineering studies signi�cantly underestimated Furthermore, the physical rehabilitation of the the cost of network rehabilitation. Moreover, wastewater treatment plants had to be dropped low and �nancially unsustainable water tariffs because no institutional arrangements for their were a long-standing issue in Algeria at the time management were put in place. of project appraisal, and the Bank’s assessment In the past. Historically, Bank We learned. It is vital to ensure that of the likelihood of future tariff increases was infrastructure project teams comprised technical design is appropriate, demand overly optimistic. an engineer, an economist, and a estimates and �nancial projections �nancial analyst. As the Bank’s are reliable, and implementation agenda expanded, more and more staff arrangements are prepared suitably. with specialized skills were added This requires a well-balanced team and to the core team, including �nancial constant diligence from management. management, environment, poverty, Our approach has changed. We gender, participation, and communication strengthened accountability through specialists, all of which were key to BOX 5.2 real-time assessment of project quality improving quality, sustainability, and at entry, quality of supervision, and GETTING IT RIGHT: THE CASE OF ROMANIA’S RAILWAYS impact. However, as project preparation country and sector portfolio performance. and appraisal became more complex, Background. The Romania Railway Results. Financial targets developed during We are reskilling our teams through Rehabilitation Project (approved in 1996 and appraisal and agreed during negotiations proved some of the basics were crowded out. recruitment and training, always aware closed in 2003) aimed to increase the capacity to be realistic; the Bank helped apply a costing The result was. Many infrastructure that further scale up will depend on and reliability of the Romanian railways system. model that improved ef�ciency. Moreover, Bank projects with unsatisfactory outcomes having the right skills to get the job done. Project design adapted lessons from successful supervision during implementation also helped failed because of shortcomings in railway rehabilitation and restructuring resolve differences between contractors and programs in other countries. The Bank helped the railway company. By the time the project technical design, economic and establish an Integrated Railway Information was completed, the Romanian railways achieved �nancial analyses, and implementation System, providing technical assistance in compatibility with European Union standards. arrangements. And, with the move implementation, and organizing training for Ultimately, the Railway achieved signi�cant away from infrastructure, we lost many railway management and staff. The Bank helped productivity and ef�ciency gains, which reduced of the very technical sector experts to identify necessary legal and institutional the Railway’s net de�cit from $214 to $44 that we needed to get this right. improvements, especially at the interface million between 1995 and 2002. between daily operations and management Scaling Up Infrastructure: Building on Strengths, Learning from Mistakes 8 CONCLUSION TOGETHER, WE CAN MAKE PROGRESS. The World Bank’s infrastructure team to balance growth and access; to balance lending could reach $10 billion annually We recognize that there is still a lot to is integrating what we’ve learned over public and private sector engagement; to within a couple of years, up from $7.4 learn, and self-examination will remain the years. Continuing to invite scrutiny, balance newer pro-environment and anti- billion last year and about $5 billion just an important part of what we do. we are opening our doors, con�dent corruption systems with more traditional a few years ago. that criticism will continue to improve project requirements; and to balance the Moving forward, we believe these lessons us. We especially value input from the complexity of our agenda with the need By linking our lending to country will help us negotiate the challenges people who live closest to the projects to keep projects simple and focused. performance and learning from past ahead and produce tangible results. We themselves—without their insight and mistakes, we have already improved the do not fear the risk, and remain hopeful participation we cannot succeed. With this sense of balance, borne of quality of our projects—for example, of bringing great bene�ts to the poor in a more intense collaboration with while half of all water projects failed in developing countries. The above lessons are all mutually critics, partners, and bene�ciaries, we the 1990s, more than three out of four reinforcing, and balance is the common are scaling up our activities to have have succeeded lately. theme that unites them. We are learning more impact. The Bank’s infrastructure