They hadn't believed it at first, that the government was going to build a dam across the Narmada River. Obviously those surveyors and engineers hadn't seen the river in the monsoon, huge and powerful, surging through the canyons and across the plains. But the people of Gadher knew the power of the river. They had lived beside it, with it, for centuries. And even if it were possible to control such a force of nature, would anyone really dare to impound the most holy river in India? Shiva himself had named it Narmada, the "ever-delightful." As the old people say, you wash away all your sins by bathing seven times in the Yamuna or three times in the Saraswati or once in the Ganges, but the mere sight of the Narmada has the same effect!

Well, in the last few years they had seen the dam rising, and now they believed. There would be a dam, and the village of Gadher-along with hundreds of others like it-would be submerged. They had to move, and soon. It wouldn't be easy to find good land, especially with so many others looking at the same time, so it was fortunate that the government was helping them. The Land Purchase Committee had already found some land it thought the people of Gadher would like. With some trepidation, a group of men from the village went with the committee to look it over.

What they saw was a pleasant surprise. The soil was rich, there was water nearby, and the previous owners had planted lots of banana trees. There was no forest, unfortunately-so it would be hard to find bamboo for building, fiber for rope, herbs for medicine, or fodder for the goats. And it was so flat! With no hills to hide behind, how would one find the privacy needed to perform the most basic functions? No, it certainly wasn't home. But then, home-Gadher-would soon be deep under water.

They were fortunate, they knew, to be offered a fertile plot. Nonethe- less, it was strange buying land-not something they'd ever done. They were adivasi, the original dwellers. They lived and farmed where their par- ents and grandparents before them had lived and farmed. They had little money-hadn't had much need for it up to now-but the government had promised to pay them for the land they lost. They would get half in cash, and the government would use the rest to pay for their new land. Or so they understood. Actually, the government would only make a down payment on the land; they would have to pay the balance-in twenty interestfree annual installments-from what they could earn by selling their crops.

The officials who had brought them here were anxious to make a deal. They explained that the government would build a school and health clinic here, and put in roads, electricity, and a water pump. And because the men from Gadher liked what they saw, they agreed to sign the papers-well, not sign, because they couldn't read or write-but affix their thumbprints. They did so, and the officials congratulated them and said they were landowners now, and asked them how soon they could move. The men went back to Gadher to tell their families, and a few weeks later they were ready to go. There were about seventy of them, more than a tenth of the village- sad to quit their lifelong home, but hopeful, too. They understood that by forsaking the beloved land of their fathers they were ensuring a better future for their children.

Moving day came. The people loaded their belongings onto the government trucks and climbed aboard. After the convoy had traveled for an hour or so, the trucks stopped and the men jumped down, eager to show their families their new home. The first things they noticed were the corrugated tin huts, all close together in rows. With the hot sun beating on them, they weren't inviting, but they were only a temporary measure, to be replaced with wooden houses before too long. It was the land that mattered, and, as they looked around, it was the land that stunned them. There were no banana trees, there was no water; it was a wasteland! They walked around, confused, trying to orient themselves. They turned to the officials: This isn't the land you showed us. Oh, yes, came the reply. This is your land. Here are the deeds you signed.

No, no, no. It wasn't right. The villagers shook their heads. This is not it.

The officials looked at one another. Then one spoke in soothing tones. You're upset because you expected to see roads and schools-everything already in place. But you must understand, all that will take some time.

No, said the men. We were shown rich land; this land is useless. It is not what you offered us.

The man in charge sighed. Look here, he said. just give it a chance. See if you can make it work. If you really aren't satisfied in a year or so, then we'll find you some new land.

What could they do? They stayed.

That was In 1990. The people of Gadher stayed at the Timbi resettlement site for two years, struggling to make the saline soil produce. By the time I got there, in April 1993, they had abandoned the place and moved back home. The dam was already half finished and in a few months, when the summer rains came and the river backed up behind it, Gadher would be flooded. Nevertheless, the people had returned to Gadher because they couldn't feed themselves at Timbi and they had no place else to go. It was hard to tell that there had ever been a settlement at Timbi. The people had taken everything with them, including the sheets of tin under which they had lived for two years. The only evidence of their stay was a few blackened firestones, scattered among the weeds. Because the government had never supplied any of the amenities it had promised, there were no roads, no electric poles, no school or clinic, no well or handpump to mark the site of what had once been-not a village, exactly; more a refugee camp.

The people in a neighboring hamlet had seen the Gadher refugees come and go. "Oh yes," one old man said, "those poor families. When they arrived, we told them the land was no good, but by then the sale was complete." Someone else added, "If we had known that the land was going to be sold, we would have warned the buyers. The big landlords who had owned all this land had been trying to sell it for a long time, but no one around here would have it. It's full of salt, no good at all." The owners, however, had finally gotten lucky: the Land Purchase Committee agreed to act as their broker, to sell the land for them. But even with thousands of villagers desperate for land, the committee had had to resort to fraud to unload it.

There is, in the heart of India, a sacred pool surrounded by temples and shrines. From this pool rises India's holiest river, the Narmada. Since ancient times, pilgrims have come here to be blessed by the river as it begins its 800-mile journey westward through the hills, forests, and plains of three states to the Arabian Sea. Some go on to make the parikrama, the ritual circumambulation of the river, from its source to its mouth and back. This trip, from temple to temple along the riverbank, traditionally takes three years, three months, and three days to complete. Nonbelievers too have been drawn here. Rudyard Kipling set his jungle Book in these forests.

India began thinking about damming the Narmada, its fifth-longest river, in 1946. The official Narmada Valley Development Plan now calls for 30 major, 135 medium, and 3,000 small dams to be built on the Narmada and its tributaries over the next 50 years. The centerpiece of the scheme is to be the Sardar Sarovar Dam, stretching 4,000 feet across the river and rising to the height of a 45-story building. When its associated canals, irrigation works, and power transmission lines are taken into account, Sardar Sarovar is the biggest water development project in India, and probably in the world. The multibillion-dollar venture is intended to irrigate nearly 4.8 million acres (2 million hectares) of farmland and bring drinking water to 30 million people. It will also take the land of at least 320,000 people, many of whom are the indigenous or tribal people known in India as adivasi.

My first guide to the Narmada Valley was M. P. Shah, public relations officer for the local dam resettlement agency. He informed me that I could not visit Gadher. It, along with most of the other places I wanted to see, was off limits to journalists because of the danger from outside provoca- teurs opposed to the dam. Only two days earlier some local social workers had been stoned and their cameras stolen, he told me. In any case, I would be wasting my time, he added, because Gadher "is already vacated." The story I had heard, about the people of Timbi having returned to Gadher, was "false information," put out by opponents of the dam. No one, Shah said, had abandoned their resettlement site for their old village. I expressed a willingness to brave the stones of the antidam forces, but Mr. Shah was adamant. Gadher was too dangerous and too far away. Poor Mr. Shah-I was clearly not the first journalist who had pestered him with what he regarded as impertinent questions. He had recently been seconded to this job from the complaints division of the tax collector's office and I had the feeling that he would be applying to be transferred back soon. His exasperation was contagious, and I decided that it would be no hardship to wait to see Gadher until I was on my own.

Gadher, home to five hundred families, is one of the largest of the several hundred villages to be submerged by the dam. For centuries, the people of Gadher have lived peacefully as subsistence farmers, cattle and goat herders, and gatherers from the river and forest. It was, at least until the shadow of the dam fell across it, a thriving village, though poor by Western standards. Its thirteen hamlets are scattered across low hills, surrounded by forest, fields, and grazing land, and bordered by the river. There are no paved roads; oxcarts, bicycles, and a few motor scooters are the only transport. The houses, mostly wooden with thatched or tiled roofs, are spread out so that each family has room for a vegetable plot and a few animals. Our car kicked up a dust storm on the dry dirt road as we turned into the village. A young woman balancing a brass water pot on her head hurried out of our way. I was, as always in Indian villages, dazzled to see such heavy work being done so gracefully by such a slender creature. Her gold and red sari was probably synthetic, not silk, and it may have been the only one she owned, but in elegance of dress and bearing, she and the other women of this dusty village cast the women of New York or Paris into the shade.

As soon as we stopped, a crowd gathered, curious to discover our business-and to investigate our car, some sort of jeep. Outsiders-city people-are rare here, and the villagers are understandably wary of strang- ers. Fortunately, I was with Balraj Maheshwari, a lawyer who is well known locally. Mr. Maheshwari works with the Rajpipla Social Services Society, a local organization founded in 1975 by a Catholic priest and dedicated to helping the poor of the region. Once he vouched for my good intentions, people were eager to tell their stories. Within a few minutes, two young men appeared carrying a beautifully carved daybed draped with a woven rug. They placed it under the shade of a large fig tree and motioned me to sit down. Mr. Maheshwari, an old man of the village named Ramsin Pijiyu, and a few of the more daring children joined me on the couch. The other men arranged themselves around us, while the women and children peered from behind the fence of a neighboring house, turning away to giggle each time I smiled at them.

Several villagers expressed pleasure at having a chance to speak to an outsider. "When others come to get information, they go only to Kevadia [the dam headquarters]," said Ramsin Pijiyu. "Then government officers come and take one or two persons from here to tell the visitors that everything is fine. Sometimes we have to pretend that we are from a different village, or that we are resettled someplace. You are the first from abroad to come here. It is good because we can tell you how we are." A cynic from the back of the crowd called out, "But if they find out someone from abroad has been here, they'll be patrolling in project trucks tomorrow."

By April of 1993, when I visited Gadher, most of its residents had been moved out of the village-split up among thirty-one different resettlement sites. But roughly a third of them had returned to Gadher because of intolerable conditions in the resettlement colonies, ranging from barren land to polluted drinking water and outbreaks of cholera. The returnees-and the fifty or so families that had never left the village-were under pressure to move before the summer monsoon began. But they were wary: Twice lately, officials had taken groups from Gadher to see potential resettlement sites and each time they had discovered that the land they were shown was not the land they were to move to. "We've been cheated once, so we look very carefully now," said a middle-aged man named Kanti Bhoga who had recently returned from Timbi. "If good land is given and we could cultivate the land and our life is happy, then we should go. But without good landif they want to cheat us only-then how can we go?"

It's not only the quality of the land they have been offered that con- cerns the people of Gadher. What they have tasted of the refugee existence has deeply discouraged them. They have discovered that in leaving their land, they are leaving behind a way of living. "Life in the other place was very congested-all the houses side by side," Mr. Pijiyu explained. "If we have a choice between the two, we will choose this. All the traditions, all the social life that we know and enjoy are here. If we move a hundred kilometers away, the traditions will be different. So, will ours survive? That worries us. At the other place when someone died, we were not allowed to burn the body in the open in the natural way. We had to bury it on our land. That was a very bad experience. Here we can get everything free: Water is free; wood and other things from the forest are free. It means that our livelihood is very easy and we are not in trouble. But if we go elsewhere, we will have to buy everything-water, forest produce, everything. So it's not good to leave our homeland. Whether this dam is good for someone else or not, we don't know, but for us it's very bad."

Work on Sardar Sarovar began in 1961. On April 5 of that year India's first prime minister, Jawaharlal Nehru, flew to the remote dam site to lay the foundation stone. The helipad on which he landed was built-as were the dam's offices and the guest quarters for visiting dignitaries-on land ob- tained by the forcible eviction of at least eight hundred families. As it turned out, the villagers had been forced from their homes somewhat pre- maturely: Construction was stopped for twenty years by an argument over how the costs and benefits of the project would be divided among the three states through which the river flows-Gujarat, Madhya Pradesh, and Maharastra.

In 1969 the government appointed a tribunal to settle the quarrel. After ten years of testimony and debate, the tribunal awarded most of the project's irrigation and drinking water to Gujarat, the driest of the three states. The tribunal also ruled, however, that Sardar Sarovar is not viable on its own. It requires three more reservoirs upstream so that the massive amounts of water that flow through the river during the monsoon season can be stored and released later in the year. Known collectively as the Narmada Sagar Projects, these three upstream reservoirs will displace another 200,000 people and cost another $1.6 billion. But because planners say this additional storage capacity is essential if Sardar Sarovar is ever to recoup its cost, the tribunal stipulated that the Narmada Sagar dams should "be completed at the same time as, or prior to the [Sardar Sarovarl dam." All four dams, said the tribunal, "should be viewed as a technically and economically interdependent project.

The World Bank had long been interested in helping to finance Sardar Sarovar, but it could do little while the tribunal was still debating the matter. Once the tribunal ruled, however, the Bank was swift to act. Bank staffers, working closely with Indian officials, spent several years reworking the project, trying to maximize its financial and technical viability and minimize its negative side effects. Once the project plans were finalized, four delegations of Bank staffers and consultants-missions, in Bank par- lance-visited India to appraise the technical and economic aspects of the project. They did not, however, consider social or environmental issues, an omission that worried the Bank's tiny environmental office. There was cause to worry. India's resettlement record is disturbing, to say the least. A conservative estimate of the number of Indians forced from their homes by large dams since Independence is 11 million, another 4 million having been displaced by mines, industrial developments, and wildlife sanctuaries. Some authorities put the figure at 20 million or more. Three-quarters of these people were not "rehabilitated"-bureaucratese for returned to their previous standard of living. As a result, millions of poor but self-sufficient peasants have ended up as beggars in the slums of the nearest big city.

The Bank too has had many bitter experiences with resettlement. Ac- cording to its own experts, Bank-funded development projects across the world have displaced millions of people, pushing many into destitution. In 1980, belatedly recognizing the harm it had done, the Bank announced that all new projects must "ensure that, after a reasonable transition period, the displaced people regain at least their previous standard of living." Two years later, the Bank looked specifically at the impact of displacement on the world's tribal peoples. It found that "tribal people are more likely to be harmed than helped by development projects" and stated that "the Bank will not assist development projects that knowingly encroach on traditional territories being used or occupied by tribal people, unless adequate safeguards are provided." The Bank also said it "would not be prepared to assist with a project if it appears that the project sponsors had forcibly 'cleared' the area of tribal people beforehand." Thus, by the time the Narmada loan was under consideration, the Bank had already adopted policies and standards designed to prevent forcible relocation and to ensure that displaced persons, especially tribal people, would be protected from the negative side effects of development.

India has no national laws governing resettlement, but the tribunal had imposed resettlement standards on the Narmada project. Ousted landowners were to receive not merely financial compensation but "land for land"-at least 4.8 acres (2 hectares) of irrigated land per farming family- and each son eighteen years old or older was to be given his own land. In addition, though the tribunal's ruling ignored the plight of the many thousands of tribal farmers who have no legal title to their traditional lands, India is a signatory to a 1957 United Nations convention that recognizes the rights of tribal people to their traditional land and requires governments to provide displaced tribal people with "lands of quality at least equal to that of the land previously occupied." But these enlightened pledges were worthless, because India-with the Bank's knowledge-violated them from the beginning.

After 1961, when the first six villages were razed to make way for the helipad and dam offices, construction stopped for twenty years. Then in 1981, five more villages were destroyed to make room for the holding tanks for the main irrigation canal. Jai Prakash Associates, the contractor building the tanks, was also given the task of clearing the land of people. Construction workers and company officials pressured the villagers to affix their thumbprints to deeds of sale, or to blank sheets of paper to be filled in later. Some relinquished their land this way; others refused. The homes of those who resisted were demolished. When some villagers appealed to the Gujarat High Court for a stay against forcible removal, the contractors stepped up the pace of demolition so that the stay, when granted, was useless. A few of the four thousand oustees were paid for their property according to a formula that a World Bank-sponsored study says "invariably undervalues land." Three-quarters, however, received no compensation at all. Years later, the Bank reported that many of them "could be seen camping in extreme poverty at the edge of what remained of their lands."

Aware of this sorry history, the Bank's environmental office sent Thayer Scudder, a Cal Tech socioeconomist well known for his studies of the impact of large engineering projects on poor people, to investigate the current resettlement plan for the Narmada dam. Scudder found that there was no resettlement plan and that the government had not even begun to collect the information needed to make a plan. No one even knew how many people the dam would displace. After Scudder reported that no reset- tlement plan existed, the Bank dropped its insistence that India submit a resettlement plan as a condition of receiving the loan.

All Bank loans must be approved by its board of directors, which bases its decisions on the formal Staff Appraisal Report (SAR) prepared for each proposed loan. The report for Sardar Sarovar acknowledged that the proj- ect had already harmed thousands of people, but it offered assurances that India would correct those wrongs and treat all future oustees fairly and in accordance with Bank policies on resettlement. It did not mention that India's Ministry of Environment had refused to approve the dam on the grounds that more information was needed about its environmental and social impact. On March 7, 1985, the board of directors of the World Bank unanimously agreed to lend India $450 million for the Sardar Sarovar dam. The loan agreement gave India until December to come up with plans for resettlement and for dealing with the dam's environmental problems.

In 1985, a thirty-year-old social worker from Bombay named Medha Patkar arrived in the Narmada Valley to conduct a study of some of the villages that were due to be submerged. Soon, however, she put her professional work aside and made it her mission to help people being displaced by the dam get the land and compensation to which they were entitled. At that moment the future of Sardar Sarovar was rewritten. With her master's degree in social work and her urban upbringing, Medha Patkar is an unlikely candidate for the revered leader of tens of thousands of tribal people and peasant farmers, though that is what she has become. She is, however, from a family of activists: Her father, a trade unionist, fought against the British for Indian Independence; her mother works with a women's organization. Medha is a small woman-she stands just five feet tall and weighs only ninety pounds. Blessed with an almost photographic memory for facts and figures, she is a formidable debater and strategist. As a public speaker she is passionate without being melodramatic.

Wearing canvas sneakers and a faded sari, Medha Patkar traveled thousands of miles up and down the Narmada Valley. She learned the languages of the tribal people, ate their food, and lived their spartan life. In village after village, she exhorted people to organize and demand just treat- ment. She wasn't the only-or even the first-person to try and help those the dam would harm. But it was Medha whom the crowds trusted; she whom they came to hear.

After several years of agitating for better treatment of oustees, Medha Patkar concluded that the government had neither the land nor the will to resettle people properly. Others were coming to the same conclusion. In August of 1988, a number of local groups that had been working to secure better resettlement decided to oppose the dam outright-on economic and technical as well as social grounds. Thousands of villagers rallied to the call for nonviolent mass protests. A year of strikes and rallies followed. In response, the Indian government imposed the Official Secrets Act on the region around the dam-a move that effectively banned gatherings of more than five people. Several hundred people were arrested for violating that ban.

In 1989, local opponents of the dam Joined with national groups to form a national organization, the Narmada Bachao Andolan (Save the Narmada Movement). The Andolan staged huge protests all around India, using the Gandhian tools of mass rallies and nonviolent civil disobedience. In September 1989, 60,000 people representing 250 organizations came to a remote part of Madhya Pradesh to rally against "destructive develop- ment." The following January, 5,000 men and women marched on the Narmada Valley Development Authority offices, forcing them to close. In March, 10,000 protesters blocked the highway from Bombay to Agra for two days, and in May 2,000 people staged a sit-in outside the prime minister's house in Delhi.

The most significant, though not the biggest, demonstration against Sardar Sarovar began on Christmas Day, 1990. In what came to be known as the Long March, three thousand oustees and supporters set out to walk down the valley to the dam site. After a week of walking, the protesters reached the Gujarat border but were barred by police from going farther. Medha Patkar and six others then embarked on a hunger strike, demanding that the government suspend work on the dam and hold an independent review of the entire project. Two of the hunger strikers were hospitalized, but the government refused to budge. After twenty-two days, the remaining hunger strikers broke their fast. The Long March did not stop the dam, but it made Narmada a national issue and Medha Patkar a national heroine. It is a role she disowns, pointing to the many others who have devoted their lives for the past several years to the antidam movement.

By the time of the Long March, Murlidhar Devidas Amte, affectionately known as Baba Amte, had joined the fight against the dam. A seventy- six-year-old social worker whose work with leprosy patients has earned him the admiration of the country, Baba Amte decided to devote his last years to saving the holy river. He moved to the small village of Chhoti Kasrawad on the bank of the Narmada and vowed to stay with the village and drown if the dam were built. Gradually the dam was accumulating a roster of distinguished opponents, men and women the government could not dismiss as manipulated peasants, opponents of progress, or dupes of subversive forces.

The year 1990 was an especially bad one for the project. Shortly after Baba Amte threw his considerable moral weight into the fight against the dam, B. D. Sharma, the government's commissioner of scheduled castes and tribes, appealed publicly to the president of India to suspend the proi- ect. "The execution of Sardar Sarovar, like many other projects, is not in keeping with the provisions of the law and the law itself is violative of the Constitutional safefuards provided for the tribal people, as also of human rights," he said. "The people are determined not to move out from their homes in the Valley and prefer a watery grave than disorganization and destitution in an unknown world which has been the story of all displaced so far."

Also in 1990, India's Department of Environment and Forests declared that the conditional approval it had given Sardar Sarovar in 1987 "must be deemed to have lapsed," since the "studies, surveys and action plans" due in 1989 had not been produced. And last but not least, the World Bank reported in 1990 that 40 percent of the 105 legal covenants attached to its loan agreement with India had been fulfilled only partially or not at all. This finding had little impact, however, because it was kept hidden from the public and the Bank's directors.

Since Independence, India-with considerable help from the World Bank-has built more than 1,500 large dams, structures that Nehru extolled as the temples of modern India. India is now one of the most dammed nations on earth-and many more dams and irrigation projects are planned or underway. So far most have been failures. Seventy percent of the large irrigation projects started since Independence are still incomplete. Referring to this state of affairs, Prime Minister Rajiv Gandhi said in 1986, "We can safely say that almost no benefit has come to the people from these projects. For sixteen years we have poured money out. The people have got nothing back, no irrigation, no water, no increase in production, no help in their daily life."

I saw an example of this not far from the abandoned Timbi resettlement site. Next to the road was an empty canal, intended to carry water for the Karjan Dam (another Bank project). That dam was completed by 1989, but the canal-indeed the entire irrigation system-has never functioned. The cement lining was badly cracked-in places it was nonexistent-and weeds sprouted from the fissures. The underlayer was made not of the requisite concrete, but of rough sand and rocks. As we stood looking down at this shoddy piece of workmanship, a passerby on a motorbike dismounted and joined us. He was a local man and seemed amused by our interest in the canal. "They're relining it-for the second or third time," he remarked. "But the material is weak; it's not what they promised." Balraj Maheshwari, the lawyer from the Rajpipla Social Services Society, who had been shaking his head silently, said, "This project is meant only for the rich people to collect money from abroad and to put it in their pockets. The bigger the project they don't execute, the more money goes into their pockets. That is the politics." As Father Joseph, the Catholic priest from south India who founded Rajpipla Social Services Society, expressed it to me, "The profits of a dam are running long before the water runs."

A 1987 study by the Gujarat State Legislature found wide discrepancies between the amount of land that planners said would be watered by irrigation projects and the amount that actually was. Panam Dam, in northern Gujarat State, for example, was expected to deliver water to 118,490 acres, but in fact irrigated less than half that area. Damanganga Dam, also in Gujarat, was to irrigate 135,900 acres of land; in fact only 1,336 acres were ever actually irrigated. Both these dams were funded by the World Bank. The reasons for the shortfalls are many. Planners overestimate the amount of water required by or available to the system; contractors carry out substandard work; and, in the Bank's own words, "most irrigation schemes have been poorly operated and maintained." Frequently the first farmers along an irrigation network commandeer most of the water, leaving only a trickle for those downstream. A 1991 World Bank review of its Indian irrigation projects found that "economic viability [has] been poor all along for most projects." These problems led the Bank to conclude that India should spend the next ten years improving the performance of existing irrigation systems rather than building new ones.

But new schemes, like Sardar Sarovar, continue to be started-if not finished. On the one hand it seems that such schemes proceed against all logic and reason. On the other, their progress is entirely rational in view of the opportunities they present for their backers to enrich themselves at the public's expense. With a large project like Sardar Sarovar, engineers and bureaucrats can look forward to kickbacks from those whose land they overvalue for compensation purposes or purchase at inflated prices to resell to oustees, as well as from contractors who increase their profit margins by using shoddy materials or simply not doing the work at all.

Big projects enable politicians to acquire power by exercising patronage and money by taking bribes. Thus it often happens that the ruling party supports such activities, while the opposition rails against them. The Times of India cites one such example: the fundamentalist Hindu Bharatiya Janata Party in Madhya Pradesh. "When in power, they have supported [Sardar Sarovar]," noted the paper. "But when out of it, they have opposed it."

Some argue that the Narmada dam must be built precisely because India's record of irrigation projects is so poor. "Most existing irrigation schemes are not operational," Amar Gargesh, the project's press officer, told me. "That's why we need this one." Gargesh also maintains that Sardar Sarovar should not be subject to a conventional cost-benefit analysis. "It's not fair to evaluate it like a commercial project," he said. "With our droughts we need this project." Both the Indian government and the World Bank say that Sardar Sarovar is a good investment, but their calculations are based on many assumptions-such as those about sedimentation rates, the availability of water, and the impact on the environment-that their own experts say are wildly overoptimistic. By the government's own admission, half the country's large dams haven't come close to their projected economic rates of return. "All their claims for Sardar Sarovar are based on one assertion only," says Girish Patel, a Harvard-trained Gujarati lawyer who has studied the project closely: " 'Forget the past; in the future it will happen differently.' But we say that you cannot forget the past. You can improve on it slightly, but you cannot go from 40 percent efficiency to 100 percent in one jump."

Perhaps the strongest economic argument against Sardar Sarovar is that the three Narmada Sagar dams, which the tribunal said were essential to its success, are not being built. R. C. Singh Deo, Madhya Pradesh's former Irrigation Minister, says that "it is impossible for Madhya Pradesh to complete Narmada Sagar. . . . It has had to beg from the Central Gov- ernment to pay its share in the Sardar Sarovar project." The World Bank has admitted that there is no guarantee that the upstream projects will be built. Without them, however, Sardar Sarovar will irrigate 30 percent less land and provide 25 percent less power than projected. That alone makes Sardar Sarovar a money-losing investment.





Billboards extolling, in Hindi and English, the good things Sardar Sarovar will do for India line the road leading to the dam site. But though the signs assert that the people of India will benefit greatly from the project, they don't say precisely how. Is Sardar Sarovar primarily an irrigation project, a drinking-water supply project, or a power project?

Until recently, promoters pointed to irrigation as the chief rationale for the project. The project brochures, which are many and glossy, are enthusiastic about the benefits to farmers. One, called "You Be the judge," declares that "Sardar Sarovar will mean richer farmers . . . Sardar Sarovar will mean increase in agricultural produce . . . Sardar Sarovar will mean eternal smiles on millions of faces." Actually, however, most of the land to be irrigated by the project already has access to water, and most of the region's poor, dry, and tribal areas will receive no irrigation water. Even if the planned irrigation system is completed, it would deliver water to only 28 percent of Gujarat's drought-prone districts and 11 percent of its tribal lands.

There is disagreement in the Bank as to what the real purpose of the dam is. Officially it is an irrigation project with a power component, but some say it only makes sense as a power project, with irrigation added on to make it politically palatable. One staffer who has analyzed the project for the Bank says, "The hydro benefits from that project are of tremendous benefit to India and to millions of Indians, but it turns out that the irrigation part doesn't really pay. The Bank thinks of it as an irrigation project, but it's a hydro project." Another Bank employee who has worked on the project confirmed this view: "Providing hydro power for industry wouldn't be popular, so irrigation was added. But the irrigation is money-losing-in fact, they discovered they could boost the rate of return by reducing the amount of land to be irrigated."

At the heart of these issues is the matter of the dam's height. The tribunal considered limiting the dam to a height of 436 feet, which would have been enough to provide the desired amount of irrigation water. But a higher dam would produce more hydro power-and hydro is the only benefit Madhya Pradesh, the state with the most people to be displaced by the project, will get from it. The tribunal, therefore, opted for a 455-foot- high dam. The extra 19 feet will produce 10 percent more electricity. But it also increases by 50 percent the amount of farmland to be submerged and by 70 percent the number of people to be displaced in Madhya Pradesh by the reservoir.

Originally only urban centers were to get drinking water, but, as criti- cism of the project mounted, officials expanded the drinking-water pro- gram to include villages. Planners now say that by the year 2025, some 8,000 villages will have water piped directly to them, and "You Be the Judge" promises that "Sardar Sarovar will mean no drinking water problem in all villages." In this drought-plagued region, an adequate supply of drinking water would be an enormous boon. Amar Gargesh emphasized this aspect of the project when I spoke to him in his office in Gujarat's capital, Gandhinagar. The government office blocks that dominate the town are arranged around a massive statue of Gujarat's most famous son, Gandhi, sitting cross-legged at his spinning wheel. From my position by the window, I gazed out on the Mahatma as Mr. Gargesh spoke of Gujarat's desperate need for water. "You have to see the problems these people have with water to appreciate them," he said. "In some places the groundwater is so depleted that the wells are empty. People spend all day collecting a single bucketful. With Sardar Sarovar, not a single village will be left without drinking water. That's official."

If true, this would be a powerful argument in favor of the dam. But a brief investigation into these promises reveals that there is little behind them. The list of villages to be serviced was assembled so hastily that it includes several hundred that India's census says exist only as names on out-of-date maps. A Bank report points out that, although the number of people the authorities say will receive drinking water from the project has increased by more than 25 percent since 1991, there has been "no corresponding increase" in the amount of water allocated to domestic purposes in the project plan. Although project officials count the supply of drinking water to thousands of villages among the benefits of the project, they do not count the hundreds of millions of dollars such an undertaking would require among its costs. The project has no funds budgeted for village water supply. Nonetheless the promoters of the dam continue to insist that it will bring water to millions. "It's cruel," says Girish Patel, the lawyer who has represented in court the villagers forced out by the dam. "The people in those areas are just waiting for water, when the project officials know it's not going there."

Neither the Bank nor the Indian government ever claimed that the Sardar Sarovar project was entirely benign. In the Bank's words, "The argument in favor of the Sardar Sarovar Project is that the benefits are so large that they substantially outweigh the costs of the immediate human and environmental disruption." There is nothing sinister about setting costs against benefits. We do it every day when choosing what to eat, how late to stay up, and whether or not to take the risk of, say, smoking or driving a car. But there is a dilemma when one person or group gets the benefits of a given activity, while another pays the costs. This is the case with Sardar Sarovar. As S. C. Varma, the chairman of the Narmada Valley Development Authority, put it, "The family getting displaced . . . undergoes hardship and distress and faces an uncertain future so that others may live in happiness and be economically better off."

In some cases, taking from one group and giving to another makes sense. A society may decide, for example, to assist its neediest citizens at the expense of its wealthiest ones, in order to ensure a minimum standard of living. Many countries do this to a certain extent through taxation or social welfare schemes. But large irrigation projects tend to do the reverse: to take from the poor and give to the rich. The Ukai irrigation project, just south of Sardar Sarovar, is a striking illustration of how this happens.

Ukai is the largest functioning irrigation project in Gujarat. Built with World Bank funds, its dam and irrigation works displaced 70,000 people, mostly tribals. The farmers of the area had previously grown their own food, mostly millet, barley, and corn. With irrigation it became possible to grow more "demanding" but saleable crops, such as sugar cane and wheat.

Most small farmers could not afford the fertilizers and pesticides that such intensive, irrigated agriculture requires, but the largest and wealthiest landowners were able to take advantage of the possibilities of irrigation and in doing so increased their earnings considerably. As a result, the value of all the irrigable land in the area rose-as did taxes and the price of even basic farm supplies-and the small farmers found that they could no longer afford to keep farming in the old way. Most small farmers eventually sold out to wealthier men who could capitalize on the new conditions. Vast sugar cane plantations now dominate the area. Balraj Maheshwari, the lawyer who took me to Gadher, knows the Ukai area well. "I could show you prosperous landowners who were turned into laborers, step by step," he told me. "Their children went to Baroda University. Now they're living in the slums of Baroda." Some have Joined the 100,000 wretched souls who work on the big cane estates from May to October every year, living in miserable conditions and sleeping in the street. Girish Patel, who has represented the cane workers in their legal appeals for better working conditions, calls them "migrant slaves."

"The saying is 'Unless we have a cake we cannot share it. So let us produce a cake,' " says Father Joseph. "But the sharing never comes. As the cake is produced, about ten to fifteen percent of the population gobbles it up." Where once the Ukal region was more or less uniformly poor, now it is divided into a dominant class of rich sugar cane magnates and a subclass of near-destitute sugar cane workers. "We are not enamored of the old way and it is not our objective to see people living in the same circumstances as in the past," says Girish Patel. "But these circumstances are worse."

By 1988, it was already difficult for surveyors, researchers, or anyone connected with the project to work in the Narmada Valley. Local hostility to the dam was so intense that large areas of the submergence zone were effectively off limits to project authorities and Bank officials. On several occasions, Bank officials canceled or cut short their trips to the valley on the advice of the Indian government, which warned that they were in danger of being physically attacked if they entered the submergence zone. Though many villages were militantly uncooperative, one consultant asserts that India exaggerated the dangers in order to keep researchers out of the area.

Construction proceeded, however, and as the dam grew higher and the time for flooding drew nearer, pressure mounted for villagers to leave the submergence zone. Many were willing to leave, if only they had someplace to go. Nearly everyone had heard some discouraging tale from friends or relations who had already moved. In 1989 about a hundred people left Gadher for a place called Amroli. They were told they would get houses, a school, a dispensary, and free electricity. Because the land was not irri- gated, the government also promised to sink a well on each five-acre plot. After three years, however, the people still had no houses-and no land on which they could build their own houses. They had no electricity, no school, no dispensary, and no source of water fit for humans or animals.

When a World Bank team visited Amroli in 1992, they found fewer than forty people still living there; the rest had returned to Gadher. A number of men complained that the land they received was not the land they had been shown-just as in the case of Timbi. Moreover, many people had fallen ill since coming to Amroli. One women told the visitors, "This place is like a crematorium. Our children suffer from vomiting, diarrhoea, and dehydration." Indian officials discount such reports. "There has been talk of malaria. There has been talk of dysentery. But it has not become a problem," Amar Gargesh assured me. "No one is resettled where there is no good drinking water. But it may become contaminated later-that we cannot help."

News of such disasters-along with the Andolan's insistence that by refusing to move, the people could defeat the dam-convinced tens of thousands of people in the reservoir area to defy orders to leave their homes. Many, including Patkar, took a vow to drown as the waters rose about them rather than leave the threatened villages. Clashes between vil- lagers and the police sent to evacuate them became increasingly common. On several occasions government agents used bulldozers to destroy homes in villages about to be flooded. Villagers accused the police of harassment and brutality, and police charged that Andolan members were inciting violence.

In 1992, the human rights group Asia Watch sent a team to the Narmada Valley. The team confirmed reports of government mistreatment and concluded that "these abuses appear to be part of an increasingly repressive campaign to prevent the groups from organizing support for protests in villages affected by the dam and disseminating information about the social and environmental consequences of the dam."

As resistance was spreading across the valley, the Andolan's supporters in Europe, Japan, Australia, the United States, and Canada were pressuring the Bank directly and through their parliamentary and congressional repre- sentatives. The international campaign, coordinated by Lori Udall of the Washington based Environmental Defense Fund, turned Sardar Sarovar into a worldwide symbol of all that is wrong with the development industry, big dams, and the World Bank.

In 1989, the campaign brought Medha Patkar and other key figures in the Narmada struggle to Washington, D.C., to testify before a special House hearing on the World Bank's role in the project. In 1990, the campaign organized a major symposium in Tokyo attended by members of the Diet as well as academics and journalists, which led the Japanese government to take the unprecedented step of canceling its $200 million financial and technical support of the project (having first, according to one Bank staffer, taken the precaution of designing the dam to take only Japanese turbines). The campaign also sent several Andolan activists on a speaking tour of Europe, bringing the issue to the attention of European Parliamen- tarians and the European members of the Bank's board of directors. And it lobbied hard for an independent review of the project.

The furor over Sardar Sarovar was doing nothing to improve the Bank's reputation as a champion of the poor, or its chances for future financial support from Japan, the United States, or its European members. In 1989, the Bank, under pressure from outside, reluctantly asked Thayer Scudder to look once again at the resettlement situation. His findings, as presented in a report by the Bank's India country department, indicated that resettlement was finally progressing well, especially in Gujarat, where all major problems had been resolved and resettlement targets would soon be met. In reality, Scudder had reported nothing of the sort. He had found the situation highly unsatisfactory and recommended that the Bank stop the loan until the major resettlement problems were resolved. When Scudder's own report was leaked, it became apparent that the Bank's management was so wedded to Narmada that it would ignore or distort the advice of its own experts if that advice was unwelcome.

Within the Bank, the Narmada project was being pushed by the India Country Department, whose director is known in Bank jargon as "head of India." The India Department finds it difficult to say no to India. A staffer who works closely with the department explains why: "The India Department has one job-to lend money to India. Of course every country department pushes its own country, but for years, during the Cold War, India held a special position as the largest nonaligned democracy. The donor community treated it with kid gloves for years-and the India Department still thinks that way."

India has long been the Bank's most important client. For nearly forty years, says one long time Bank official, "India was the jewel in the Crown of the World Bank. . . . The reputation of the Bank tended to be mea- sured in terms of what it could do for India." Over the years, the Bank has lent India $40 billion, nearly twice as much as it has lent any other country. As greater and greater sums of money passed between them, the Bank succeeded in putting its imprint on India, but India's imprint on the Bank is just as deep. The fact that India owes the Bank so much money only serves to increase its power in the institution. As Bank officials are fond of observing-on condition of anonymity-if you owe a bank $100,000, you worry; if you owe it $100 million, the bank worries.

The India Department apart, concern about the handling of the Narmada project was growing within the Bank. It was nothing new for the Environment Department to complain about a particular project, but sud- denly it had allies in the legal and agricultural divisions. The debate was a gentlemanly one, no open quarrels or threats of resignation, but by 1991 it had spread so far throughout the Bank that it was impossible to keep it private. Internal memos were promptly leaked, and journalists and activists had no trouble getting damaging quotes from anonymous Bank officials. Moreover, the executive directors-the twenty-four-member board whose job it is to oversee the Bank's management-were being fiercely lobbied by environmental groups on the practical and moral shortcomings of the project. Some directors were beginning to question the official version of events. "It was really becoming an embarrassment to the Bank, and there were tremendous tensions between the executive directors and the senior management," remembers one close observer.

Even Barber Conable, the Bank's president and thus its top manager, wasn't sure where the truth lay. There was simply too much conflicting evidence. Opponents and supporters of the project agreed on only one thing-that Sardar Sarovar was far from the worst of the Bank's projects. "India has two or three hundred other big projects-some financed by the Bank-with worse resettlement problems. Sardar Sarovar is probably the best, certainly one of the best," remarked one staffer, echoing the words of many people I spoke to on all sides of the issue. The spotlight was on Sardar Sarovar not because it was the most egregious example of destruc- tive development but because of the remarkable alliance of determined villagers, local activists, and international groups that fought it.

In 1991, Conable agreed to commission an independent review of the project's social and resettlement problems. Conable, a Republican con- gressman for twenty years before moving to the Bank, asked his former colleague, Bradford Morse, to lead the review team. Morse had spent six terms as a Republican congressman before becoming head of the United Nations Development Program. His deputy was to be Thomas Berger, a distinguished Canadian jurist, a man admired for his even-handed reviews of several controversial projects. It was an unprecedented step for the Bank-and a much more radical one than Conable realized at the time.

The Bank had never asked-nor indeed allowed-outsiders to critique an entire project. It often hires consultants. but thev have limited remits, typically spending only a few days at the project site, a few weeks on an assignment. Their reports are edited by Bank staff and circulated privately. This was all the Bank wanted from Morse. As one participant in the negotiations said, "The Bank had in mind a couple of weeks in India and a typewritten document, stapled at the corner."

But Morse, after consultations with Berger, who was experienced in conducting independent reviews, and with critics of the dam who warned him about the pitfalls of doing business with the Bank, laid down tough conditions for accepting the commission. He demanded access to all rele- vant documents, total freedom from the Bank, and complete control of a million-dollar budget. He also insisted that the team retain copyright in its report and publish it independently. The Bank was taken aback, but Morse prevailed on all points. He and Berger selected two other full-time team members: Hugh Brody, a British anthropologist, and Donald Gamble, a Canadian environmental engineer.

In September 1991, the team embarked on nine months of research in India and Washington, sorting through old reports and speaking with poli- ticians, bureaucrats, scientists, environmentalists, and the oustees them- selves. Unlike normal Bank missions, the Morse team-whose appointment had been welcomed by the dam's opponents-was able to enter the "refusenik" villages as well as the resettlement sites. Its members also spoke privately to many people in the Bank and examined old project papers. "People in the India Department would talk about how their hands were tied by politics, by India's special status," reported one team member, who added, "I was amazed at what I saw of how the Bank works. I thought I had understood the Bank and I was a great admirer. I felt that they were setting a precedent in many of their environmental issues, and some of the social issues. On paper it's terrific. In practice, it's another story."

The review team found that there had been no environmental impact assessment of the project and that neither India nor the Bank had consid- ered the effect such a huge project would have on people living down- stream. This despite the fact that the Bank's own experts had warned that the dam would result in increased salinity in the lower reaches of the river, destroying one of India's most lucrative fishing grounds, that it was likely to silt twice as fast as was being assumed, and that the project would be able to deliver only half as much irrigation water as had been stated. The team also discovered that only a few months earlier a Bank consultant had described parts of the project as "death traps," as "ideal breeding sites" for malarial mosquitoes, and had said it was "taking malaria to the doorsteps of the villagers." "People have died," concluded the team, "yet the Bank's status reports simply say that the preventive measures required by the formal [loan] agreements seven years ago are 'not yet due.' "

With regard to the key question-resettlement-the team found that India still had neither a comprehensive resettlement plan nor the data on which to base one. Each time India had missed a deadline for submitting a resettlement plan, the Bank had simply set a new one. A succession of deadlines, in December 1985, December 1988, March 1989, and June 1989, had all slipped quietly away, unmet. In June 1990, the Bank changed its tack and began to request only partial plans. Those too failed to materi- alize. The review concluded that "there developed an eagerness on the part of the Bank and India to get on with the job. Both, it seems, were prepared to ease, or even disregard, Bank policy and India's regulations and procedures dealing with resettlement and environmental protection in the hope of achieving the much-needed benefits."

Most disturbingly, the review team found that virtually all the flaws and problems it discovered in the project had already been reported to the Bank by staff technicians and consultants, and simply ignored by those higher up the chain of command. "All this information was in the Bank's files. Morse told [Lewis] Preston [Conable's successor as president] they could have saved a lot of money by just looking in their own files," said a member of the review. "This will happen again, only next time it'll be worse-and if the Bank doesn't realize that, they're dreaming in techni- color. The Bank thinks its problem is in India. It isn't in India, it's in Washington, D.C."

In June 1992 the team published its findings in a 385-page paperback book. It fell on the Bank like a hammer blow. "There wasn't a soul in the [India] department that had any idea what an impact the Morse Report was going to have," one insider told me. "It was like watching a ship hit an iceberg in slow motion. The Bank thought they could ride roughshod right over it. They were unconcerned up until the day it arrived." The report's detailed description of the project and of the Bank's part in it was devastating. The review team had been asked to suggest ways of improving the project. Instead it said that the Bank should "step back from the Project and consider [it] afresh."

Ten weeks after receiving Morse's report, the Bank made its response. In a document called "Narmada: Next Steps," the Bank assured its directors that since the completion of the Morse Report, the Indian Government had adopted "a comprehensive set of actions in line with the recommendations of the Bank mission" and had demonstrated its "seriousness of purpose . . . by several important discussions." There was therefore no reason to "step back" from Sardar Sarovar as Morse had recommended. The document angered many who had expected the Bank to use the Morse Report as an excuse to withdraw or suspend funding for the project. One long-serving staffer said at the time, "If Narmada is not suspended, it's the death knell for the environment in the Bank. If they can't suspend this one on environmental grounds, they'll never suspend anything."

Shortly after "Narmada: Next Steps" was published, more than 250 environmental, human rights, and grass-roots development groups signed a full-page open letter to Lewis Preston in the Financial Times under the headline "The World Bank Must Withdraw from Sardar Sarovar Immediately." Failure to do so, the letter said, "will confirm that the Bank is beyond reform" and will lead to a campaign to cut off funding to the Bank.

When the members of the review team saw "Narmada: Next Steps," they too were furious. Though Morse was in the hospital suffering from a bout with emphysema, he and Berger held a telephone conference call with Joseph Wood, the vice president responsible for the India Department, and several executive directors. Morse told the directors that the document ignored or misrepresented his report's main findings. Berger and Morse also put their charges in writing in a letter to Lewis Preston. Citing numerous examples of misrepresentation, they concluded: "The Bank may reject our finding that its incremental strategy has failed. . . . The Bank may decide that overriding political and economic considerations are so compelling that its Operational Directives are irrelevant when decisions have to be made about the Sardar Sarovar Project. But it should not seek to reshape our report to support such decisions." The letter was also sent to each executive director, despite Preston's last-minute plea to the bedridden Morse not to do so.

On October 23, 1992, ten days after Morse and Berger sent the letter, the board met to consider the future of Sardar Sarovar. Several directors from industrial countries-notably those from the United States, Canada, Japan, Germany, Australia, and the Scandinavian countries-called for a suspension of disbursements. But the rest sided with the developing countries, and the board voted to give India five more months, until April 1, 1993, to comply with the terms of the loan. Patrick Coady, the U.S. executive director, sternly warned his fellow directors that voting to continue the loan was a "signal that no matter how egregious the situation, no matter how flawed the project, no matter how many policies have been violated, and no matter how clear the remedies prescribed, the Bank will go forward on its own terms."

On March 31, 1993, the day before the Bank's deadline was to expire, the government of India asked the Bank to cancel the Sardar Sarovar loan. This was a face-saving maneuver, agreed ahead of time with the Bank, which otherwise would have had to cancel the loan the following day on the grounds of noncompliance. Medha Patkar hailed the cancellation as a giant step toward victory: "The exit of the World Bank from Sardar Sarovar has given a great boost to the Andolan and our supporters and has shown that the dam is not a fait accompli." Supporters of the project also claimed to be pleased, describing the move as evidence of India's unwillingness to be dictated to by foreigners. "The government felt the Bank was encroaching on the sovereignty of India. We needed to put our foot down," said Amar Gargesh.

The immediate financial impact was only moderate, because India had already received all of the Bank's $250 million low-interest loan and $30 million of the $200 million the Bank had lent on normal terms. India's move also killed another $440 million worth of loans for the project that had been in negotiation, but soon after the Sardar Sarovar loan was can- celed, the Bank announced eight new loans, totaling $2.3 billion, for India.

The Bank would like to put Narmada behind it, but that may be im- possible. "Just because India has said 'We don't want any more World Bank money' doesn't mean the Bank's responsibility ends," argues Smitu Kothari, one of the founders of the Andolan. "They legitimized this dam for years, financially and politically. They made it possible." The Bank's Legal Department agrees. In an internal memo, the Bank's General Counsel expressed concern that the directors had been given the impression that the Bank is no longer involved with the Sardar Sarovar project. "This is not the case," it said. The government of India is still "legally obligated towards the Bank to carry out its obligations under the loan agreement."

Though the Narmada Valley Development Authority is determined to complete the project, the loss of Bank financing may have doomed it. The authority is nearly out of money; it faces huge debt service payments on its outstanding loans; and it has not been able to raise any more funds, public or private, from abroad. In any case, construction of the project has been halted since January 1995 by a ruling of the Indian Supreme Court. Each monsoon season, however, the half-built dam floods nearby villages. Despite police efforts to cordon off the submergence zone, every year scores of people who have taken the "drowning pledge" manage to enter those villages and barricade themselves in the low-lying houses. So far, the dam's interrupted construction -and the monsoon rains have combined to flood most areas only to chest height. A higher dam or heavier rains will change that.

A year after the loan's cancellation, Lewis Preston reflected on the lessons of Narmada. It was, he acknowledged, a public relations disaster. "The Bank has paid the highest price for not recognizing the importance of the environment," he said. "That mistake, I think, has, in terms of the criticisms of the Bank, eroded some of the support that the Bank is entitled to." Otherwise, however, he found little to be concerned about. "The Narmada project is not one to be ashamed of, in terms of what it will eventually do for the people of India. . . . It will help two million people and that is significant in anybody's language."