World Bank President Paul Wolfowitz has called corruption the single most important obstacle to development and has ratcheted up the fight against graft in Bank projects. While this effort is welcome, it is being undermined by the bank’s simultaneous increase in infrastructure lending. The experience of Pakistan’s water sector shows that the bank’s self–interests facilitate rather than discourage corruption in infrastructure development. Unless the World Bank addresses upstream the corruption incentives that drive infrastructure decisions, the poor will continue to be deprived of access to essential services.
After closely following the script of his predecessor, in early 2006 the new President of the World Bank, Paul Wolfowitz, finally revealed his own vision for the embattled development institution. Identifying corruption as the single largest obstacle to development, he held up loans to India, Bangladesh, Kenya and Chad because of corruption concerns and increased the budget of the Bank’s anti–corruption unit. “This is about making sure that the bank’s resources go to the poor and don’t end up in the wrong pockets,” Wolfowitz told US News & World Report. “It is about fighting poverty.”
Critics have long accused the Bank and other donors of turning a blind eye to the leakage of development funds, leaving corrupt contractors and officials flush with cash, governments saddled with “white elephant” projects and odious debt, poor people devoid of essential services and the environment unprotected. The World Bank began to address the “cancer of corruption” under President Wolfensohn, and Paul Wolfowitz’s pledge to “move from talking about corruption to dealing with corruption” is welcome. Yet the world’s largest development institution still attempts to treat the symptoms and not the cause of the disease. In fact, the Bank’s current lending strategies might even be fueling the corruption epidemic.
Just as the bank vows to get tough on corruption, it has simultaneously announced a big increase in its support for infrastructure, the sector perceived to be the most corrupt globally according to Transparency International. In fact, approximately half of the World Bank anti–corruption unit’s investigations that have led to specific corrective actions were linked to infrastructure projects.1
Massive, centrally planned and financed water, energy, transport and other public works projects are particularly prone to corruption, thanks to their complexity, capital intensity and high price tags. They offer larger spoils than small–scale projects and programs to increase the efficiency of existing infrastructure. Unless corruption is checked in the earliest stages of the planning process, corrupt politicians, government officials and construction companies will always favor large–scale projects to address a country’s infrastructure needs.
Development efforts can only be effective if they reflect a country’s own priorities. The World Bank has acknowledged the importance of “country ownership” in recent years. Yet it has tended to equate country ownership with government ownership and government ownership with ownership by finance and infrastructure ministries. The bank has limited the opportunities for civil society input in the development of infrastructure strategies, and cut down the preparation time for infrastructure projects.
Combined, the bank’s push into infrastructure, the emphasis on government ownership and the limited accountability to civil society are creating large opportunities for corruption in a sector in which graft is already endemic. If the World Bank does not address the incentives for corruption upstream, fighting graft in individual contracts will be a losing battle. If its fight against corruption continues to be focused reactively on specific projects, infrastructure development will remain distorted, the poor will be deprived of essential services in many countries, and the environment will continue to be neglected.
Infrastructure, corruption, and development failures
Building infrastructure projects in the developing world is a $200 billion business that provides a plethora of opportunities for corruption. Bribes are paid to secure concessions and kickbacks are provided in exchange for contracts. Bid rigging occurs, shell companies are established, and procurement documents are falsified. Sub–standard materials are used in construction, regulators are paid off, and prices for infrastructure services are inflated. Compensation for forcibly displaced communities ends up in the pockets of bribe–seeking local officials. The World Bank acknowledges these corruption risks, but it has not figured out what to do about them. A recent bank report about infrastructure acknowledges that “anti–corruption is the area where the largest gaps remain in our understanding of what works and what does not.”2
Given the enormous potential pay–offs, it is not surprising that there are often powerful vested interests behind big, new public works projects. Peter Eigen, the founder of Transparency International, argues that corruption in the construction sector not only plunders economies; it shapes them: “Corrupt government officials steer social and economic development towards large capital–intensive infrastructure projects that provide fertile ground for corruption.”3 Paul Collier and Anneke Hoeffler explain: “If budget decision–makers themselves are corrupt, they may decide to skew the budget towards infrastructure spending so as to increase the opportunities for corruption. If roads are more capital–intensive than primary education, the budget may be skewed towards roads… and if there is more opportunity for corruption in road construction than in road maintenance, then roads may be built, allowed to fall apart, and then rebuilt.”4
Similarly, large dams, massive irrigation systems and river diversion schemes, some constructed with World Bank support, are often touted as the solution to the water and energy needs of the poor. Rarely, however, are alternative options assessed. The World Commission on Dams (WCD) pointed out that “the pressure on development aid agencies to move large amounts of capital … argued for large–scale solutions such as large dams.”5 Even as large dams have provided fewer than anticipated benefits, forced tens of millions of people from their lands, and destroyed rivers and river–based livelihoods, they have tended to prevail over alternative options. The WCD noted: “Decision–makers may be inclined to favor large infrastructure as they provide opportunities for personal enrichment not afforded by smaller or more diffuse alternatives. The consequences frequently directly affect the poor and the environment.”6
The political economy of infrastructure development “doesn’t just line the pockets of political and business elites; it leaves ordinary people without essential services,” according to Peter Eigen.7 The push for big projects diverts resources from decentralized, community–based options and from the maintenance of existing infrastructure. Ultimately, local people are stuck with the economic, social, and environmental costs of infrastructure projects that may not be the best option for providing water or energy services – or may not even be providing them at all. These two problems, namely corruption and unmet needs for infrastructure services, are closely linked.
The Pakistan case
Pakistan’s Indus Basin Irrigation System, the world’s largest water diversion scheme with more than 1.6 million kilometers of watercourses, is a prominent example of how corruption pervades economic development and distorts the priorities of infrastructure investment. It also shows how the World Bank’s business model and development paradigm encourage rather than counteract the pervasive dynamics of corruption.
Pakistan’s irrigation system has been shaped by the World Bank’s approach to water infrastructure for five decades. In the 1950s, the bank brokered a water treaty between India and Pakistan which created the foundation for irrigating the Indus Basin. It helped devise the policies and institutions of Pakistan’s water sector in a series of master plans and reports, and has loaned almost $20 billion (in 2005 prices) for projects in the sector.
The Indus Basin Irrigation System is a central planner’s dream turned concrete. Its corner stone, the Tarbela Dam, was the largest manmade structure on earth at the time of its construction. Tarbela is just one of 19 dams that block and divert the basin’s mighty rivers. Large canals, drainage highways and more than 100,000 distributaries crisscross the Indus basin.
Today, the Indus Basin Irrigation System serves an area the size of Bangladesh, and generates more than one fourth of Pakistan’s electric power. Yet the system is in deep crisis. The irrigation network operates extremely inefficiently, and sedimentation is rapidly reducing the capacity of its reservoirs. More than 60% of irrigation water is lost from the canal head to the root zone, and a lot of water is wasted on thirsty crops such as sugar cane that are not suited to the arid Indus Basin. Average crop yields are much lower than in neighboring India.
The construction of reservoirs and canals caused the forcible displacement of more than 200,000 people in Pakistan. Decades after they were moved, thousands of families are still living in misery. A report prepared for the World Bank argues that the lack of replacement land and corruption in the system are “creating extreme hardship for people.”8
Pakistan’s irrigation network has always served the privileged elite at the expense of the poor. World Bank and government programs have consistently favored feudal landowners. When the irrigation system was established, the government failed to recognize the land rights of the original inhabitants and allotted irrigated plots to rich landowners and military personnel. While large and very large farmers control 66% of all agricultural land in Pakistan, almost half of all rural households own no land. A World Bank evaluation noted in 1996 that the bank’s projects “provided large and unnecessary transfers of public resources to some of the rural elite.”9
The top–down engineering approach to Pakistan’s water sector has also caused massive collateral damage downstream. The Indus Basin Irrigation System starves areas of Sindh province – and particularly the Indus Delta – of water and sediment. And because the sediment trapped in the reservoirs does not replenish the delta, close to 5,000 square kilometers of farm land have already been lost to the sea. Meanwhile salt water is intruding 100 kilometers upstream in the Indus. The lack of water and sediment is destroying flood plain forests that are home to hundreds of thousands of people and mangrove forests that help protect the coast against storms.
While the downstream areas suffer from a water shortage, wasteful water use is wreaking environmental and economic havoc in the command area. Over–irrigation and inadequate drainage have caused the water table to rise across a large area. As a result, about 60% of all farm plots in Sindh are plagued by water logging and salinity.
Corruption in Pakistan’s water sector
Pakistan’s water sector, like many of those around the world, is fraught with large and small–scale corruption. According to a 2003 survey by Transparency International, Pakistan’s Water and Power Development Agency is perceived to be the second most corrupt institution in the country.10 Close to half of the more than 31,000 complaints received by Pakistan’s anti–corruption ombudsman in 2002 were related to this one institution.11 As the World Bank’s 2005 Pakistan water strategy admits, top positions in the country’s water bureaucracy are sold at a high price.
Corruption works in a variety of ways in Pakistan’s water sector. After paying high sums to secure senior government positions, officials need to recoup their costs in the form of kickbacks. They can do so primarily through projects that serve construction companies and large landowners, not through improved maintenance programs and low–cost projects that serve the poor. This is why the water bureaucracy, as the World Bank puts it, suffers from a “build–neglect–rebuild” syndrome, and “has yet to make the vital mental transition from that of a builder to that of a manager.”12
Even resettlement programs are a source of patronage, which rewards rather than penalizes large–scale displacement projects such as dams and canals. “Pakistan has well established corrupt practices in the revenue departments that hurt the interests of those who are resettled,” notes Pervaiz Amir, a consultant to the World Bank on large dams. “The manner in which resettlement and rehabilitation is handled becomes susceptible to patronage and corruption and it becomes difficult to ensure that every affected person is treated fairly and receives his or her due share.”13
Many officials in Pakistan’s water sector also allocate irrigation water to the highest briber and not necessarily to the most needy or productive farmers. “Payments to irrigation officials to ensure the delivery of sanctioned water supplies were reported as routine and endemic” the World Bank found in 2002, and “water availability clearly depends on efforts to bribe irrigation officials.”14
Corruption is allowed to flourish because Pakistan’s water sector lacks transparency and accountability. Water allocations at all levels of the irrigation system are not disclosed to the public, for example. The World Bank concludes: “In the shadows of discretion and lack of accountability, of course, lurk all sorts of interests – of powerful people who manipulate the system for their ends, and of those in the bureaucracy who serve them and are rewarded for this service.”15
Brick–and–mortar investments in centrally managed dams and canals are not the only way to address Pakistan’s water and energy needs. Because the existing infrastructure is not being properly maintained and so much water is being wasted, the efficiency of the irrigation system could be greatly increased. Plugging the leaks of the existing system is environmentally more benign than building new dams and canals.
It is also more economical. A World Bank evaluation found in 1996 that water conservation measures saved more water than the largest new dam in Pakistan’s investment program could have stored, and at one–fifth the cost.16 The Asian Development Bank estimates that an additional 4.7 million acre–feet of water could be provided either by conservation measures at a cost of $1.7 billion, or by a new dam with a price tag of $4.5 billion.17
Decentralized and nonstructural solutions to Pakistan’s water crisis also exist. The Indus Valley has huge groundwater reservoirs, which could store many times as much water as all future dams. Recharging these reservoirs would require more sustainable flood management practices which allow the Indus to overflow its banks temporarily rather than confine it within massive embankments.
Farmers still irrigate thousands of square kilometers of land through traditional techniques outside the modern canal system, and without support from the government or World Bank. Rainwater harvesting and simple, affordable treadle pumps provide a steady supply of water to farmers, without the added costs of bribes for water officials or diesel pumps. Drip irrigation kits apply water directly to the roots rather than the furrows, and use only half as much irrigation water in the process. An innovative way of planting rice without standing water (called the System of Rice Intensification) allows rice – a particularly thirsty crop – to be grown using only half the amount of water while boosting harvests. Such soft approaches have been adopted with good success around the world, and are being introduced in Pakistan. Shifting control over water resources from bureaucrats and absentee landlords to poor farmers would ensure a more economic use of water, reduce poverty, fight corruption, and protect the environment at the same time.
In 2003, the World Bank argued that a “genuine paradigm shift” emphasizing the proper management of water resources rather than new infrastructure was needed in Pakistan. Yet the bank’s new water strategy for Pakistan does not reflect this paradigm shift. It asserts that “Pakistan has to invest, and invest soon, in costly and contentious new dams.”18 The 2005 strategy recognizes the potential for efficiency gains, but does not address the maintenance gap in the water sector, and the serious social and environmental impacts of the current approach. In January 2006, General Musharraf announced that his government would soon start construction of the Bhasha and Kalabagh Dams. The two dams will cost more than $20 billion, will displace an estimated 160,000 people, and will further reduce downstream flows.
The World Bank prepared its water sector strategy for Pakistan without any input from civil society. It argued that “while all voices must be heard, much greater weight must be given to the voices of those who have responsibility and face the voters, and less to those who are self–appointed or who represent small special interests.”19 This is a remarkable statement about a country that is marred by corruption, in which top government positions are for sale, and which is run by a self–appointed military ruler.
Pakistan is a prominent example for the pervasive impacts of corruption on development planning. Yet as Eigen, Collier, Hoeffler and others have pointed out, the mechanisms that distort the development of Pakistan’s water sector are widespread. White elephant projects that made no economic sense and failed to deliver any developments benefits — like the Bataan nuclear power plant in the Philippines, India’s Dabhol power plant and the Turkwell dam in Kenya — can be found around the world.
Why are governments and the World Bank so obviously flouting the lessons of the past? The bank has always been good at evaluating its own performance, but is notorious for ignoring evaluation findings in subsequent operations. And although bank managers frequently speak out against corruption, the institution’s self–interests align with and reinforce the interests of corrupt borrowers and contractors in various ways.
The bank covers its administrative costs from the profits it makes by lending to middle–income countries. It has to continue lending to these countries in order to sustain its own business model. Since middle–income countries can raise capital on the private market, the World Bank must keep its lending costs low so as to not be out–competed by private banks. It is easier and cheaper for the bank to invest in large brick–and–mortar projects than to process loans for small, decentralized irrigation schemes, or for cheap but institutionally complex programs to improve the maintenance of existing infrastructure.
The interests of the World Bank’s member governments have helped define those of the institution’s bureaucracy. Northern governments favor loans that pay for the contracts of international consultants and construction companies. Borrowing governments prefer bulky projects that yield ribbon–cutting opportunitiesand political prestige, support centralized bureaucracies, and offer spoils for patronage. The bank’s institutional self–interests translate into an incentive structure that rewards staff for pushing money out the door quickly, and not for achieving lasting developing impacts. For example, Paul Wolfowitz recently promoted the author of the Pakistan water sector strategy to become the bank’s country director for Brazil.
The World Bank’s preference for brick–and–mortar projects has undermined efforts to improve the performance of Pakistan’s water sector before. In the 1980s, the bank approved four projects to rehabilitate the existing canal system and stem water losses. When the water bureaucracy resisted change and misused the loans for building new canals, the World Bank looked the other way. As an internal evaluation found, “the Bank did not insist on the implementation of the agreed strategy against the pressures of special interests,” and “[its] concern to keep disbursement flowing reinforced this focus. … The Bank helped to further this distortion of objectives by making it plain that construction progress was the highest priority.”20 In Pakistan and elsewhere, the bank’s self–interests conflict with its own development objectives and will continue to thwart its efforts to fight corruption.
Peter Bosshard is Policy Director of International Rivers. Shannon Lawrence is International Policy Analyst at Environmental Defense. A shorter version of this article appeared in the May 2006 issue of the Far Eastern Economic Review.
1. World Bank, Infrastructure: Lessons from the Last Two Decades of World Bank Engagement, January 30, 2006, p. 40
2. World Bank, Infrastructure: Lessons from the Last Two Decades of World Bank Engagement, January 30, 2006, pp 38f
3. Transparency International, Global Corruption Report 2005, pp. 1f
4. Collier and Hoeffler in TI GCR 05, p.13.
5. World Commission on Dams, Dams and Development, 2000, p. 178
6. World Commission on Dams, Dams and Development, 2000, p. 187
7. TI GCR 05, pp. 1f
8. Pervaiz Amir, The Role of Large Dams in the Indus Basin System, March 2005, p. 27
9. World Bank, Irrigation Investment in Pakistan, OED Precis, 9/1/96
10. Transparency International Pakistan, Nature and Extent of Corruption in the Public Sector
11. Transparency International, National Integrity Systems, Pakistan 2003, p. 78
12. World Bank, Pakistan’s Water Economy: Running Dry, Draft of June 23, 2005, pp. 11, 79
13. Pervaiz Amir, The Role of Large Dams in the Indus Basin System, March 2005, p. 28
14. The World Bank, Pakistan Poverty Assessment, October 28, 2002, pp. 88, 90
15. World Bank, Pakistan’s Water Economy: Running Dry, Draft of June 23, 2005, p. 59
16. World Bank, Irrigation Investment in Pakistan, OED Precis, 9/1/96
17. See Pervaiz Amir, The Role of Large Dams in the Indus Basin System, March 2005, pp. 41f.
18. World Bank, Pakistan’s Water Economy: Running Dry, Draft of June 23, 2005, p. 10
19. World Bank, Pakistan’s Water Economy: Running Dry, Draft of June 23, 2005, p. 112
20. World Bank, Operations Evaluation Department, Pakistan – On–Farm and Command Water Management and Irrigation Systems Rehabilitation Projects, Impact Evaluation Reportnumber 15863, 06/28/1996, paragraph 36