The I.D. Man

Nandan Nilekani is counting every Indian for the world’s largest biometric database.Illustration by Mark Ulriksen

One afternoon last December, inside a long tent pitched beside a busy road in a southern district of New Delhi, a punchy young drunk was shouting, “Why are you not including me? Take my photograph!” The tent was a winter homeless shelter used by migrant workers. That day, it was doubling as a government office. A table, cluttered with laptops and cables, stood in a pool of electric light, with two chairs on one side of it and a white screen behind the chairs. Twenty or so people stood around; most of them were slight, youngish men wearing wool hats and bright scarves looped loosely around their necks. The tent’s interior was dark and dusty, but a few balloons hung from the walls.

The men were standing in line for I.D. numbers, in a country with a shortage of I.D.s. They were being drafted into a public scheme, launched by Nandan Nilekani, a genial software billionaire, which intends to create a national biometric database ten times larger than the world’s largest existing biometric database. The aim is to help reduce the extraordinary economic distances between those who have benefitted from India’s boom of the past two decades and those who have not. The effort has been called “the biggest social project on the planet,” and enthusiasts have described it as a project of national transformation; skeptics see a threat of intrusions by the state, or detect patriotic vanity—a craving for the world to see India as a society capable of the achievement. The scheme is known as Aadhaar, which means “foundation” in Hindi. Aadhaar’s logo is a red fingerprint, in the guise of a rising sun.

I spoke to Farooq Ali, a day laborer in the produce market across the road. He was from Uttar Pradesh, to the east, where his wife and four children still lived. He was thirty-five or thirty-six—he wasn’t sure. Like the others in the tent, Ali was keeping an appointment made for him by SPYM—the Society for Promotion of Youth and Masses, an N.G.O. that works with the homeless. He had never had a photo I.D. When his turn came, he took off a maroon zippered jacket and sat in front of a computer screen, looking amused. A technician, sitting at a right angle to him, and staring at his own screen, typed Ali’s name and date of birth: “1975 (Approximate).” He took Ali’s photograph and scanned his irises with a rectangular black device the size of a thick hardback. He then pressed Ali’s fingertips against the glass square of a fingerprint scanner. When the software rejected the prints, he wiped Ali’s hands on a grubby orange cloth, then pressed the fingers again. Manual labor can make fingerprints hard to read.

A few moments later, as four lanes of traffic raced by, Ali showed me the receipt from his enrollment. An I.D. number would follow in the mail, but for now he had a piece of paper with his name printed on it—an object of significance in a country that has no equivalent of Social Security numbering, and where just thirty-three million Indians, out of 1.2 billion, pay income tax, and only sixty million have passports. Hundreds of millions of Indians are barely visible to the state: they have either no I.D. at all or a weak form of it, issued by local authorities. Consequently, they can’t easily open bank accounts or buy cell-phone SIM cards, and they can’t secure state services owed to them. Their official opacity hampers economic growth and emboldens bureaucrats who steal from welfare funds or harass citizens for bribes. I asked Ali about the likely effect of an I.D. number on his life, and he said, in an easy, patient way, “Maybe some benefits.” Beside him, Gulzar Khan, another market worker, added, “I never had I.D., so I never thought about what it would mean, but now that I have it I think that, at least, people will know who I am if I’m killed in an accident.”

In the summer of 2009, when Nilekani left Infosys—the Indian outsourcing company that he had co-founded, nearly thirty years earlier—a streaming video of his going-away party was broadcast, live, on the company’s internal network, which is available to more than a hundred thousand employees. At the event, Nilekani gave a speech that described the invitation to launch Aadhaar, which was made by Prime Minister Manmohan Singh, as an offer that he could not refuse. He went on, “My only identity is Infosys. I will be going to lead a program to give identity to every Indian. But today I am losing my identity.” He departed with a gift of a Toshiba laptop, as well as his holding of 1.45 per cent of the company—at today’s valuation, about half a billion dollars.

I saw Nilekani this past winter, eighteen months after he left Infosys, as the Aadhaar project—an experiment in the limits of technocratic governance—was rolling out across the country. I found him in two contrasting workplaces. The headquarters of the agency built around him, the Unique Identity Authority of India (U.I.D.A.I.), is on a low floor of a tower in central Delhi. Nilekani is known here, with unvarying formality, as the Chairman. In the corridors, men in dark suits carry brown files tied with string. A door is marked “Officers’ Toilet.” During meetings, junior staff members are not encouraged to speak.

Each evening, a driver collects Nilekani and brings him home to a white, colonial-era government villa, in a neighborhood of other such villas. Here, peacocks move coldly across his huge lawn, between two ficus trees whose trunks are each ten feet in diameter. On the other side of the road is the house where Indira Gandhi lived, and where she was killed by her bodyguards; the garden path on which she took her last steps has been overlaid with glass.

Toward the end of each week, Nilekani takes a two-and-a-half-hour flight south, to Bangalore, where the winter weather is Californian. “Bangalore is more laid-back—not in the sense of afternoon siestas, but less aggressive,” according to Nilekani, who likes the music of Neil Young. Bangalore, India’s technology capital, is where the U.I.D.A.I. has its tech headquarters—run by Indian engineers with backgrounds at Google and Intel—and a data center whose servers may eventually hold more than ten billion Indian fingerprints. Nilekani was born in Bangalore in 1955, when the city was greener and sleepier, and his main home is still here. He lives with his wife and his mother; his daughter and son join them when they are on break, from Harvard and Yale, respectively. At a time when Mukesh Ambani, the chairman of Reliance Industries, the largest Indian conglomerate, has just moved into a cantilevered, twenty-seven-story, four-hundred-thousand-square-foot house in Mumbai—an egopolis, with nine elevators—Nilekani’s low, sleek home, on a quiet side street, makes a statement not much louder than a West Village brownstone. Nilekani told me that he was too lazy to run a building as complex as Ambani’s. Besides, he added, “I don’t want to stand out too much from the crowd. If the crowd all had planes, then maybe I’d have a plane.”

Nilekani, who has often called Aadhaar “humongous” in scale, sometimes likes to meet with colleagues on the sofas in his Bangalore home. On a day I was there, buoyant young men and women, wearing shorts or Puma T-shirts, called him by his first name as they discussed encryption and tap-to-pay technologies; Nilekani’s spaniel napped in a patch of sun. “What you’re seeing is worlds colliding,” Nilekani said to me after his guests had left, as a member of his domestic staff took away a whiteboard and dismantled a projector screen. (Nilekani’s feet were bare, and he held identical phones in each hand.) The collision was between the worlds of bureaucracy and of information-technology entrepreneurship, the old Indian élite and the new. As Nilekani’s friends point out, if he succeeds in establishing an I.D. protocol—thereby, perhaps, helping to create an inclusive, low-corruption identity for India itself—it will be because he drew Delhi and Bangalore into an awkward embrace.

Nilekani has a round face, as if reflected in the back of a spoon, and the wry, detached manner of someone so taken by visions of the future that the present seems too mundane to merit his full attention. Although he is often persuasive, he is not always eloquent; sentences often fall away to phrases like “the branding thingy.” He is known as a great salesman, but he sells softly, blithely, by making it seem almost an imposition to have to explain the benefits of his product.

The intention of Aadhaar, as described by Nilekani, is to improve the efficiency of government services, “and all that stuff,” while giving people the means “to ask for what they deserve,” no matter where they are in India, even if they are far from where they were born. “The notion that an Indian lives and dies in his village is passé,” Nilekani said. His task is to assign everyone in India a random twelve-digit number that is unique to him or her—nobody gets more than one number—and to link the person to the number with a photograph, fingerprints, and iris scans. Though many countries have long traditions of people-listing—in records connected to baptism, taxation, or conscription—India does not. Nilekani’s digital effort is what he calls “leapfrogging stuff.” If the project is successful, India would abruptly find itself at the forefront of citizen-identification technology, outperforming Social Security and other non-biometric, and not fully randomized, systems. Nilekani informed me, disapprovingly, that seven of the eleven digits in identity numbers issued in Finland refer to a person’s date of birth.

Nilekani, whose closest equivalent in America is probably Michael Bloomberg, said, “Scale, speed, cost-effectiveness, quality, innovation capability—these were our goals. But the other challenge was: how do I bring the best people from government and from the private sector into one place?” Nilekani occasionally had to suppress his startup instincts. Not long after he began working in Delhi, he was urged to break the habit of dropping into the offices of subordinates: he risked disrupting treasured hierarchies. “The job has not been without its challenges,” he said, in a tone underscoring the scale of his discretion.

Infosys, which is worth nearly thirty billion dollars, makes its money by hiring young, technologically minded people, providing them with a comfortable workspace on one of the company’s eleven Indian campuses—vivid lawns, subsidized meals, gyms, and pools—and then selling their skills abroad. In the company’s core business, technicians write and support custom software that performs unglamorous tasks, such as inventory and payroll, for large overseas corporations. Clients have included Amazon, Apple, Ford, and Bank of America. A friend of Nilekani’s recently said, jokingly, of the back-room dullness of the coding, “It’s shitty work—I could never do it.” On reflection, he added, “But they do build beautiful processes.”

The company was founded in 1981, when Nilekani was a laconic twenty-six-year-old. He comes from an educated, middle-class, English-speaking family. (Today, when Nilekani has to make a speech in Hindi or in Kannada, his regional language, he needs time to prepare.) His father, the manager of a Bangalore textile mill, was a leftist—“a socially conscious citizen,” according to Nilekani. His mother, a college graduate, stayed home. When Nilekani was twelve, and his older brother was at college, his father lost his job, and his parents moved away from Bangalore in search of work. For the sake of stability, Nilekani was sent to live with his uncle, in Dharwad, a town north of Bangalore. From there, in 1973, he won a place to study electrical engineering at the prestigious Indian Institute of Technology, or I.I.T., in Mumbai. He was an average student but a social leader, a wide-ranging reader, and a star of the college’s quiz team. “I had a network,” he said.

Ramachandra Guha, a writer and historian known best for the book “India After Gandhi” (2007), first met Nilekani when their quiz teams faced each other. Guha, who attended St. Stephen’s College, in Delhi, told me, “I.I.T. is a place for nerds. Nandan was unusual because he was extremely laid-back. He wasn’t ferociously competitive, talking about how to get to the States, a scholarship to M.I.T., that kind of thing.” Nilekani’s wife, Rohini, a writer whom he met when he was in college, recalled, “Compared to my rather staid, middle-class family, he was a little on the wild side.”

Upon graduating, Nilekani became a trainee at a Mumbai software-engineering firm. His interviewer was Narayana Murthy. Two and a half years later, Murthy left the firm to found Infosys, and invited Nilekani and five others to come with him. According to Murthy, they started with two hundred and fifty dollars. In the early days of the company, Rohini Nilekani invested several thousand dollars; this has matured into a half-billion-dollar holding similar to her husband’s. Like him, she now earns millions a year in dividends and is philanthropically active.

Infosys was unusually committed to principles of fairness and transparency. Vijay Nilekani, Nandan’s older brother, who works for the Nuclear Energy Institute, in Washington, D.C., told me that Nandan’s young company was put at risk by its refusal to bribe the Bangalore telephone department. “They were not going to feed the system,” he said.

“And just where did that extra vowel come from?”

For much of the nineteen-eighties, Nilekani lived in the U.S., embedded with Infosys clients in Union City, New Jersey, and Skokie, Illinois, among other places. In 1993, he and his co-founders became rich when Infosys went public in India. By then, the government, which had long leaned toward socialism, had begun to enact pro-business reforms. (For two decades, Infosys and its competitors have paid no tax on overseas earnings.) In 1999, Infosys became the first Indian company to be listed on a North American stock exchange. Three years later, Murthy stepped aside, and Nilekani became C.E.O. The company’s rate of annual growth accelerated, reaching fifty per cent in 2005. “He pressed on the pedal,” Murthy said of Nilekani’s tenure, when we spoke on the phone. Murthy also noted, with perhaps only an accidental hint of myth containment, that “the Nandan period was really the golden period, as far as the market was concerned. And he did a good job. He’s very good at networking, he’s very good at schmoozing.” In these years, Nilekani became a board member of the foundation that runs the Davos summits, and he was lauded in Thomas Friedman’s “The World Is Flat,” in which he is credited with inspiring the ruling metaphor by saying something rather different (“Tom, the playing field is being leveled”). Nilekani told me, with a shrug, “I became a member of the global superclass, as they call it.”

Nilekani handed over his C.E.O. position to a successor in 2007, but he stayed at the company as co-chairman, and began writing a book, “Imagining India,” a smart, social-democratic reading of the country, in which his anxiety about national shortcomings—education, inequality, urban infrastructure—was offset by confidence about the reforming potential of information technology, open markets, and a bulge in the segment of the population about to enter the workforce. (China has no such bulge.) Nilekani, who had become a modern Indian archetype—the outsourcing tycoon—was now positioning himself to represent a new story, India’s overdue adoption of good governance.

While he was writing, he met regularly with Ramachandra Guha at a corner table at Koshy’s café, Bangalore’s Les Deux Magots. “He was bored, racking up the numbers for Infosys,” Guha told me one afternoon, in the same café. “And he began to realize that the challenges he sought were not business challenges.” Guha went on, “He’s the most brilliant Indian I know,” and he repeated the phrase. “He has a finer and more intuitive understanding of the dynamics of caste, of religion, of governance, the functioning of political parties, than the most highly skilled sociologists and historians.”

“Imagining India” was published at the end of 2008. Pratap Bhanu Mehta, the head of the Center for Policy Research, a Delhi think tank, recently described it as “not just a C.E.O.’s random reflections.” To Mehta, the book represented a new kind of Indian optimism: “It’s only in the last ten years or so that you have a generation in India that genuinely seems to believe that the future will be much better than the past.” But he also detected something old-fashioned in Nilekani—“an odd romance about state-building that we haven’t seen since the nineteen-fifties.” Contemporary talk of reform tended to be “about getting the government out of where it doesn’t belong. But here is a guy who’s saying, ‘Look, I’m going to build a state.’ ” Rohini Nilekani described herself as more left-wing than her husband, but added, “He was always a social liberal. He was never a market fundamentalist.”

In the book, Nilekani made a case for unique I.D. numbers. As he noted, a right-of-center Indian government, which was in power from 1999 to 2004, began to introduce a national I.D. scheme, largely driven by anxiety about border security. Nilekani rethought the idea as a more liberal, and more economically oriented, one. Such a program would remind citizens of their “rights, entitlements, and duties,” and would oblige the state to improve services. It would boost the national economy, by allowing hundreds of millions of Indians without bank accounts to open one, and it would cut government losses from corruption.

Nilekani’s book became a best-seller. Hawkers sold pirated copies at city intersections. In the spring of 2009, Nilekani travelled to the U.S. to promote it. He appeared on “The Daily Show,” and Jon Stewart called him “the Bill Gates of India,” adding, “You’re a very kind and lovely man. I welcome you as my new overlord.” Nilekani returned home, he told me, “with the intention of continuing my role at Infosys, when all of a sudden this whole thing developed.”

In May, 2009, India reëlected the government that had been in power for five years: the United Progressive Alliance, a coalition dominated by the Indian National Congress, with Manmohan Singh again as Prime Minister. “Sometime in May, I got the first inkling that these guys would like me to play a role in the government,” Nilekani told me. If there was surprise in his voice, he was being coy: his book, after all, was a manifesto; Murthy and others had been urging the government to make use of him; and he had some public-policy experience, including time with an organization called the Bangalore Agenda Task Force, which had held dreams, largely unfulfilled, of turning that city into a new Singapore.

Nilekani was offered a place on the Planning Commission, a body that guides national social policy. This did not appeal to him. According to several associates, he put himself forward as a possible education minister, but the idea was rejected. So the conversation turned to I.D.

Though the immigration-driven plans of earlier years had foundered, a new policy, more in Nilekani’s image, had begun to form, and in 2008 the Singh government had approved the creation of an I.D. agency. Many people suppose that, without Nilekani, this body would not have prospered. But he argues that the innovation was inevitable. Economic growth had increased tax revenues, allowing the government to expand India’s welfare state. Most spectacularly, and at an annual cost of about nine billion dollars, it had introduced the National Rural Employment Guarantee Act, or NREGA, which guaranteed every household in rural India a hundred days a year of unskilled manual labor in public works. But NREGA funds were getting lost to corruption; local officials often claimed reimbursement for wages that had not actually been paid. Indians in the eastern state of Orissa, who were surveyed by two economists from Harvard in 2008, had worked no more than a third of the days claimed on their behalf. “It’s only when you start spending big money on these programs that you start thinking about their efficiency,” Nilekani said. And, just as Social Security had been enabled by I.B.M.’s computing advances, today’s technologies offered new ways to enforce efficiency: biometrics, mobile connectivity, cloud computing. “In some sense, the demand met the supply,” Nilekani said.

He agreed to take the job for five years, on the condition that he would be given a cabinet-level appointment—to protect him from departmental turf wars. This also entitled him to a villa in Delhi, with parquet floors and a guard who snaps to attention when his car pulls in. (“My other life,” he announced, with faint embarrassment, when he took me there.)

“It wasn’t fuzzy,” Nilekani said of the I.D. project. “Five years from the day I took it, you would be able to say I succeeded, if people got numbers, and you’d know I screwed up, if people didn’t get numbers. So it was a zero or a one.”

One afternoon last winter, in the offices of the U.I.D.A.I.’s technology center, on Bangalore’s southern ring road, a young computer operator was watching a monitor on which photographs of Indians were passing before him at the rate of one every two or three seconds. He examined each image and its accompanying text: name, gender, date of birth. His job was to vet: Was this anxious-looking person, in fact, a man? A seventy-year-old man? He clicked “correct” or “incorrect,” and scrolled to the next person. That day, he had already inspected more than five thousand photographs, and he had clicked “incorrect” three hundred times: men listed as women, children as adults, photographs with two heads in them. Nine other operators were doing the same.

They were working just outside the glass-walled corner office of Srikanth Nadhamuni, the U.I.D.A.I.’s head of technology. Nadhamuni, who is in his late forties, worked in Silicon Valley for fifteen years before returning to India, the country of his birth, in 2002, to launch, with Nilekani, a nonprofit group that helps strengthen the public databases of city governments—records of births and deaths, for example. “I decided that e-governance would be very good in India,” Nadhamuni told me. “We’re good at I.T., and we’re bad at governance, and we can use one to improve the other.”

In the summer of 2009, Nadhamuni took a call from Nilekani, who is a friend. They began weekend discussions in Nilekani’s Bangalore home, joined by Sriram Raghavan, an entrepreneur whose company had been working to bring information technology to rural India. Nilekani recently described their challenge: “It’s not that there are no other databases with a billion things in them. Facebook may have five hundred million users, and Google may process a few billion e-mails, all the time. Internet systems have shown scale. The difference was, our system had to have the robustness and reliability of an enterprise system. In an Internet application, if an e-mail doesn’t go, it’s not the end of the world. But this was like a credit-card system. You can’t afford to lose transactions.”

After Nilekani left Infosys, in July, 2009, he began commuting between Bangalore and Delhi. The process of building a government department from scratch was “a huge dislocation from my comfort zone,” he said. His first hire in Delhi was Ram Sewak Sharma, a veteran government administrator, who became the U.I.D.A.I.’s director general, and was described to me as “Nandan’s Rahm Emanuel.” Nilekani moved fast. “The only way I can reduce risk is by compressing time,” he told me one day. Political will can slip away. And a program to reduce corruption risks being sabotaged by those who are corrupt. The U.I.D.A.I. had to look solid almost from the start.

Within weeks, he and a small team of civil servants and tech experts had established much of the project’s architecture. Aadhaar would subcontract enrollment. It would pay state governments, or other approved partners, about a dollar per citizen; these partners would, in turn, hire private enrollment companies. Aadhaar would exist only to receive data and supply numbers. (“I just want to be the plumber,” Nilekani has said, many times, on Indian television.) Enrollment would be voluntary, and available to all residents. And there would be no I.D. cards—just I.D. numbers.

A card, carrying a photograph and other biometric information, can confirm identity offline; it’s a database of one. But cards can suggest authoritarianism, and they create a market, for they can be bought and sold. Moreover, as Sriram Raghavan put it recently, “Everything should be in the cloud, right?” Indian cell-phone connectivity was already good enough for Aadhaar’s planners to imagine almost universal online access to a national registry. A cell phone connected to a cheap fingerprint reader could authenticate identity anywhere, and could even become the basis for a simple A.T.M., allowing banks to expand into the countryside. Nilekani called Aadhaar’s decision not to release cards an “epiphanic moment.”

“Nandan is a socializer of ideas,” Raghavan said recently. Nilekani flew from state to state, where he met with local politicians and created allies. In the view of Narayana Murthy, Nilekani’s Infosys mentor, the political obstacles were always larger than the computing ones. “Technologically, it’s a very simple project,” he told me. (Nilekani disagrees.) “The challenge is in making sure that literally hundreds of thousands of officers fall in line, they rally to his call and march to his tune.” Murthy admires Aadhaar and hopes for Nilekani’s complete success, but he finds it easy to imagine a half-success—something that is neither a zero nor a one. “I would not hold him responsible if this project did not take off, if it did not scale up as well as he wants.” He added, “Most things in India muddle along.”

While Nilekani schmoozed, Nadhamuni and his Bangalore colleagues began designing what was expected to be the world’s largest biometric database. (The current largest, built by the Department of Homeland Security’s U.S.-VISIT program, holds a hundred and twenty-six million biometric identities; the F.B.I.’s database holds about seventy million sets of fingerprints.) As Nadhamuni recalled, “We talked to Yahoo, Microsoft, I.B.M.—a lot of them are around here. My chief product manager is from Google, and one of the things we heard from them is that, when you design systems on a huge scale, you’re not going to get it right.” Google’s advice was to build a database whose code could be “refactored.” To use Nadhamuni’s metaphor, he had to make a car whose engine could be replaced, while running, without any loss of speed.

The database’s everyday function would be to authenticate: to give instant yes-or-no answers to the question “Am I not me?” But the greater technical challenge was insuring that nobody was counted twice during enrollment. “Authentication, that’s manageable, although we’re expecting one hundred million hits a day,” Nadhamuni said. “Authentication is a one-to-one match: one I.D. number matched to one fingerprint. But enrollment—I’ve got to take your fingerprints and match them against the whole country. And say, Yep, he’s not in here yet. So that is a one to n, with n being the size of the database, and in India n is 1.2 billion, assuming everybody’s already in. That will take four years, eight years. And we’re talking seven days a week, twenty-four hours, computers, thousands and thousands of them, just crunching.”

Fingerprints don’t fall into a sequence. “I cannot say, ‘My fingerprints are greater than, or less than, yours,’ ” Nadhamuni explained. “That’s what Google can do with any words that you give them. And that gives them very fast searchability.” An electronic fingerprint system first identifies a print’s minutiae—places where something interesting happens in the pattern—and then plots their relative positions. A landscape is reduced to a map of its landmarks. (Something similar happens with iris scans.) But the computerized comparison of these maps is still hard, because two prints taken from the same pressed finger can’t be expected to have the same level of detail, or the same orientation, or the same dimensions. Before a computer can rule out a match between two prints, it may have to align one against another in millions of different ways. Then the computer will move on to the next possible match in the database, and the next: millions of alignments each time. Private companies have developed proprietary algorithms that allow this to happen at a reasonable speed. Aadhaar has signed up with three such consortia—led by Accenture, L1, and Morpho—with the option of switching among the three at Aadhaar’s convenience. Aadhaar agreed to pay only for the fingerprints and iris scans that were processed, and held on to the original print images. Nilekani called this “an entrepreneurial approach to problem-solving. Everyone is paying the next guy only on a successful result.”

“You know that thing where you stand like a statue, then move real fast, then stand like a statue again? You totally stole that from me.”

In September, 2010, Rajana Sonawane, a forty-year-old woman in a remote village in the western state of Maharashtra, became the first holder of an Aadhaar number. She was presented with it in front of the Indian media and Sonia Gandhi, the president of the Congress party. “Wherever I go, it will come in handy,” Sonawane said of her I.D. number, although she later added, to a newspaper reporter, “Damn the honor, we don’t have anything to eat.”

When I met with Nadhamuni in Bangalore, three months later, he was contending with the wave of enrollees now starting to reach his servers; the intake had risen to forty thousand a day, and a graph on his laptop predicted a flow of a million a day by 2014. “This is the hockey stick I’m petrified by,” he said. “It’s taking off like this!” He lifted up his arm. The national total was then four hundred thousand. (By late September of this year, thirty-three million numbers had been issued, but there was a backlog of roughly thirty-eight million packets of I.D. data waiting to be processed.)

Last winter, Nadhamuni also had the mundane problem of typos. He gestured to the young employees outside his door clicking “correct” and “incorrect.” They had just been appointed, after an audit discovered that three per cent of the data sets arriving from field enrollment stations contained errors. “We’re eyeballing a hundred per cent of the transactions,” he said. The data flow would soon increase to a point where such vigilance would become unsustainable. But, with a guilty laugh, Nadhamuni recalled something that had been explained to him in childhood about elephant-taming. If the leg of an infant elephant was, for a while, tied to a stake by a chain, then in later life the elephant would behave as if restrained whenever it felt a chain around its leg—even if the chain wasn’t attached to anything. “What a powerful idea!” he said. “Not a real positive one, but the point is we have to send a message that we will not put up with bad data.”

The television stand in Nilekani’s Delhi office was still wrapped in plastic on the day I visited, in early January. A muted news channel was showing developments in a vast corruption scandal involving a former minister of communications. Outside the office door, young men stood ready for errands. Beyond, the corridors were filled with the sounds of construction, as rooms were readied for new hires. At his desk, Nilekani signed papers. I glanced at a copy of Forbes India on his coffee table; an article detected rising levels of opposition to Aadhaar, and was illustrated with a fingerprint made of barbed wire.

Nilekani has faced quiet resentment and resistance from inside the government. But, in the country’s broader establishment of N.G.O.s, think tanks, and universities, opposition has been more full-throated, and derives, in part, from the faction of India’s left wing which regrets that the government has shown waning interest in public ownership. Although Aadhaar is operated by a state agency, it is run by a corporate celebrity, and its work may bring about a change in the relationship between the state and the poor which, to some, resembles privatization. Aadhaar has been called “the Whitest of White Indian Elephants,” and mocked as an institution run by “techie fools.” Critics have distributed video demonstrations of fingerprints allegedly being faked using melted wax.

The previous day, I had met Jean Drèze, a development economist, at the Delhi School of Economics. Drèze was born in Belgium, became a naturalized Indian, and is now a significant figure on the Indian left. He helped design NREGA, the work-guarantee program, and was a member of the National Advisory Council, which is chaired by Sonia Gandhi. In his frugally underheated office, he had described ways in which corrupt officials could easily outmaneuver Aadhaar, and talked of the risks of disrupting NREGA while that program was still nascent. He had also expressed worry that Aadhaar posed a threat to government schemes such as the Public Distribution System (P.D.S.), which supplies subsidized food and other goods to the registered poor. Aadhaar opens the way for cash transfers to replace the P.D.S.: eligible citizens would receive money, paid into bank accounts, instead of access to cheap grains. They would then buy goods at the market rate. The state would get out of the rice-trucking business, and citizens would no longer be subject to whims and scams at their local P.D.S. outlets. The government is drawn to this idea, and so is Nilekani. But Drèze argued that Indian banks were unprepared for it. And, with traditional food handouts, he said, “you’re giving people what you know they need.”

In Nilekani’s office, I quoted Drèze’s description of Aadhaar as an “invasion” of NREGA.

“What am I, a virus?” Nilekani said, amiably.

When I asked about cash transfers, he said, “Rice or cash—that’s a public-policy decision I don’t want to get involved in now. I’m building the pipes.” Such responses frustrate Usha Ramanathan, a legal researcher and an anti-U.I.D.A.I. activist. “He keeps saying, ‘Don’t look at those consequences, just look at what I’m asking you to look at,’ ” she told me. “That’s an advertiser’s mode.” In turn, Nilekani detects frustration and technophobia in his opponents. “They don’t believe that technology can solve problems,” he said at one point. “They say, ‘We’ve been looking at these things for decades, and we haven’t solved them, and who are you to tell me you’ll solve it in three years?’ There’s some of that.”

Alongside these arguments about social policy, there is also some disquiet, among Indians, about Aadhaar’s threat to privacy. The fears are of two kinds. One is that the government will pry, illegally, into Aadhaar’s database. The other is that India will lose the natural privacy protections of chaos. In a country where databases are messy and unconnected, identity theft is tough work. It’s hard enough to prove who you are, and even harder to prove that you’re someone else. But a trusted I.D. number, like a Social Security number, inevitably links one database to another, and it can streamline fraud.

Nilekani has compared Aadhaar’s risks and rewards to those of nuclear energy. In his office, he told me that he supported proposed data-protection legislation, but added, “When you look at the challenge—at the few hundred million people out there who are outsiders . . . the upside is humongous.” He said that there were two possible responses to technology: “One is to go back and live in a cave. The other is to say this stuff is useful, but we must put in place checks and balances.” He accused Ramanathan, the activist, of not owning a cell phone. “That’s what I heard,” he said, smiling.

Later, I flew with Nilekani to Bangalore. At the Delhi airport, a friend or a stranger came up to shake his hand every few minutes. We talked about his wealth—at one point, he said, “I’m ridiculously rich”—and about how he frustrated his wife by having no taste for reminiscence. It was dark when his chauffeur picked us up in Bangalore and drove us into the city, past Yahoo billboards and mutton stands. His tone relaxed, and he sounded like a man who had already finished counting a sixth of the world’s population. “We understand how the railroads changed the U.K. and the U.S.,” he said. “Or how the Internet changed everything. Nobody disagrees with that. So a more ubiquitous, lower-cost, more efficient way of reaching people and reaching money to people—why would it not work? Of course, there’s a lot of execution risk; it has to scale; there’s the whole issue of adoption, training. But, fundamentally, I can’t see a reason why it can’t happen.”

The next day, he gave a public lecture on the campus of a technical institute in the northwest part of the city. Outside the hall, a dozen protesters distributed flyers. Inside, in front of several hundred people and a row of potted plants, Nilekani spoke for half an hour—unique numbers, enrollment, authentication—using the tone of an office manager bringing his people up to speed on new parking rules. He then found a more impassioned voice. “Whatever public service you want, your name has to be on a list,” he said. “If you don’t have any form of identity, if you don’t have any acknowledgment of your fundamental existence, then you’re essentially shut out of the system. You become a nonperson.”

Protesters in the hall began holding up signs, one of which read, “Beware. Big Brother Is Watching You.” When Nilekani asked members of the audience for questions, the skeptics spoke up: Will Aadhaar end up giving rights to illegal immigrants? What about a man who “meets with an accident and loses both his hands?” Didn’t I.B.M. do data processing for the Nazis?

After fifteen minutes of this, Nilekani seemed weary. “Come up with a faster, better, cheaper solution and we’ll implement that, yes?” he said. “What is your solution?”

A few days after I watched homeless workers being fingerprinted in the roadside tent in Delhi, I met Mohammad Intezar Khan, who had enrolled in Aadhaar a month earlier, as part of the same outreach program. He was living in a different shelter in Delhi: a concrete building off an alley in the largely Muslim neighborhood of Nizamuddin Basti. Khan was an engaging man in his early forties. Born in Bihar, five hundred miles to the east—where he had a wife and children—Khan had been in Delhi, on and off, for seventeen years. He worked on construction sites. He told me that he rarely had much more than a dollar in his pocket.

I was introduced to him early one evening in the shelter’s office. In an adjoining dormitory, an action film played noisily as men prepared to sleep, shoulder to shoulder, on blankets on the floor. The office was crowded with enrollees who had recently received their I.D. numbers by mail. (The letters read, in Hindi, “Aadhaar: Rights for the Common Man.”) The men had come to take the next step: to collect information about their new bank accounts. No more than a third of Indians have accounts—in part because of national regulations that require customers to show solid I.D. Aadhaar automatically opens a basic account for enrollees who don’t have one.

For some of the people picking up their bank passbooks, the appointment was dutiful. Of fifty accounts that had been opened, only four had funds in them, in amounts ranging from two dollars to a hundred dollars. I learned something about Indian expectations of privacy when someone at the shelter showed me a list of the fifty names, with account numbers and balances.

That evening, I spoke with Samsul Haque, a Learish man under many layers of clothing, who was wearing a hat embroidered with the word “Hero.” His Aadhaar notification card said that he had been born in 1967, but he told me he was actually sixty-two; he had been processed before the “correct” and “incorrect” technicians had been hired in Bangalore. Formerly a laborer, he was now a beggar at a nearby Sufi shrine. He made about fifty cents a day. “Sometimes I’m starving—how can I save?” he asked, and then added, with more satisfaction than regret, “I am just by myself. If I starve, I’ll starve. I’m not responsible for anyone.”

Khan, the construction worker, was in a different mood, in a way that suggested the potential of even a muddled or incomplete I.D. He was certain of the value of his new account. When we sat down to talk—a conversation that ended in the dark, after the power cut off—he described his present finances. The money in his life was of two kinds: cash that he sent home to his family and cash that he immediately spent. He earned about a hundred and ten dollars a month. He sent a little more than half of that home, through the bank account of a relative. He spent about two dollars a day on himself. The shelter was free, but he spent about seventy cents on an evening meal; he bought two or three cups of sugary tea, each day, at about ten cents a cup.

Now that he had a bank account, he planned to save a little—not by sending less to his family but by cutting back on tea, and by trying to economize on food. Thirty cents saved each day would accrue to ten dollars a month. He was fired up with a business idea: when he had saved about two hundred dollars, he intended to buy trousers and T-shirts in bulk and take them by train to Bihar, where clothing was expensive; he was confident that he could sell items at double what he’d paid for them. He was full of optimism about the opportunities ahead. He might even get a passport and find work in the Persian Gulf. One of his sons was already working as a craftsman, in Delhi, and was doing fairly well. “My other son will be in sales or marketing,” he said.

Khan had no possessions at the shelter, and carried nothing but a wallet, in the form of a small plastic bag. He showed me its contents: an address book, a Ramadan prayer timetable, and a decade’s worth of rail tickets. The tickets were on smooth, thin, gray paper, like newsprint. They recorded every journey that he had taken between his home village and Delhi. This had been his version of I.D. If he was challenged by the police, the tickets would help establish that he was not a thief but a kind of commuter. He held up the bundle of tickets: “I can say, ‘Look, this is who I am.’ ” ♦