Rebooting India: Realizing a Billion Aspirations (6 page)

The situation was ripe for culture clashes between those who had spent years in the hierarchical, highly formal bureaucratic structure of the government, and those from the much more informal private sector background, where shorts were considered office wear and it was not just acceptable but expected that you called the CEO by his first name. One world changes at glacial speeds, while the other metamorphoses overnight. On the first day he joined UIDAI, Viral attended a meeting at the Department of Posts. ‘Here I was, having just returned to India after a decade in the US, sitting in the conference room surrounded by senior bureaucrats from the Indian postal service. Then the chairperson of India Post walked in, and to my absolute amazement, everyone in the room stood up!’ he recalls. He had to learn the rules as he went along; you couldn’t just barge into an office in Delhi, but had to enter according to order of seniority. In a meeting, you couldn’t just speak up whenever you had a question; your rank determined when it was appropriate for you to talk.

Viral himself is fairly outspoken, bluntly direct where bureaucrats tend to be oblique and hierarchical. Only half in jest, Ram Sewak Sharma bestowed upon him the title of ‘free radical’, both a description as well as a clever chemistry pun. This nickname stuck and became widely adopted, and today Viral laughs about it. ‘I was allowed to do a lot of things because people said, “Oh, this guy’s a free radical, he’s going to come up with something different!”’

The bureaucrats were equally perplexed by these brash private sector fellows, who didn’t seem to understand the concept of expense records, public audits or paper files. Each group had their own impenetrable jargon—talk of government financial rules, committees, questions from members of Parliament and due process, was countered with the language of cloud storage, servers, bandwidth and data encryption. Those in the private sector dashed off quick emails to communicate with team members, largely unconcerned with building institutional memory, whereas the holy grail for communication and record-keeping in the government continued to be the quaintly anachronistic file containing green sheets and ties on either side, wending its ponderous way through the bureaucratic hierarchy. One
set of people wanted to do things first and worry about them later, whereas the other set was always mindful that they would be held accountable for every step in the decision-making process.

Despite the potential for a comedy of misunderstandings, the entire organization overcame some initial hiccups and settled down to function smoothly.

Both sides realized that they had to learn each other’s vocabulary, and more importantly, earn each other’s trust. Given the magnitude of the challenge at hand, they could not afford to function in silos, and adopted a collaborative model of work where shared expertise helped to drive innovative decision-making. Viral recollects Nandan telling him at one point, ‘When you set yourself a goal that large, everything appears small in comparison.’

The organization continued to grow, and eight DDGs were appointed, reporting to the DG. Reporting to each DDG were two to four assistant director generals (ADGs). Eight regional offices were created around the country to work closely with the state governments for Aadhaar enrolment and customizing the rollout to the local requirements. These offices were set up in Mumbai, Bangalore, Hyderabad, Ranchi, Guwahati, Chandigarh, Lucknow and Delhi. At its peak, the entire organization consisted of about 200 government officials, and another 100 or so consultants, volunteers, those on sabbaticals, and interns.

In the meanwhile, the technology team in Bengaluru was working furiously in the Ring Road apartment, trying to meet the goal of issuing the first Aadhaar number by 2010. While the core of the technology team was in place and design decisions were being finalized, the programmers who would build the entire system had yet to be brought on board. The sense of urgency to achieve the one-year target was signalled by the fact that the DG, Ram Sewak Sharma, didn’t waste time waiting for programmers to be hired to work on the enrolment software; he rolled up his sleeves and wrote the entire thing himself in a matter of days, working after office hours and on weekends.

In 2013, roughly four years after the birth of the UIDAI, Ram Sewak Sharma stepped down from his post as DG. This marked the
end of the start-up phase and the next DG, Dr. Vijay Madan, worked to scale up Aadhaar across the country. Enrolments were increased in states that were lagging behind, and the UIDAI worked with a number of ministries and state governments to build Aadhaar applications.

Amelia, in her early twenties, had one simple desire—she wanted to buy a two-wheeler. Money was not the problem, since she had scrimped and saved for the past three years. The problem was identity. Sold into prostitution before being rescued, she lived in a protection home in Chennai without identity documents of any kind. Fellow resident Priya wanted to work at a hotel in the Middle East, but couldn’t get a passport since she too had no identity documents.
²

Both of them were among the nearly 500 women and children, part of a special camp organized by the UIDAI, who received the first-ever identity document they had possessed—an Aadhaar number. In Jharkhand, women who had received Aadhaar cards felt more empowered; they could now collect government benefits on their own, without relying on their menfolk. Their status in the community went up, and some women used this boost to tackle other social issues in their villages, like getting their husbands to stop drinking
.
³

How was it that all these people qualified for an Aadhaar number, despite being unable to obtain any other form of identification? The answer is that Aadhaar was deliberately designed to be as inclusive as possible, especially targeting those who had fallen through the cracks of all other government identity schemes. Long before a single piece of data was collected from any Indian resident, the UIDAI put in hours at the drawing board, planning exactly what type of data it would need to collect to succeed in its mission of inclusiveness.

The first question was as simple as asking whether the UIDAI needed to enrol any residents at all. Was there an existing database they could use for the job? The answer was, not quite. While some official databases did exist—voter rolls, the list of people who fall below the official poverty line and qualify for benefits—they had several major
drawbacks. One, the data they contained was not sufficient to ensure a truly unique identity for every Indian resident, and was often beset with errors. Two, they did not cover the entire population of the country. And three, each of these data sets had their own structure and format; integrating them all would be a logistical nightmare. All in all, it was clear that the UIDAI would be better off building its own de novo database, with all data being collected in one shot.

The next decision to be made was as to the kind of data to be collected. It was clear that there would be two types: demographic and biometric. To handle the demographic side of things, the Demographic Data Standards and Verification Procedure Committee was set up, chaired by N. Vittal, former central vigilance commissioner, which subsequently came up with an exhaustive set of guidelines for demographic data collection.

Given that every field collected would have to be repeated 1.2 billion times, the effort was to keep things as simple as possible. But nature abhors a vacuum, and various government departments jumped in to try and fill the perceived gaps in the system. The health ministry suggested that the UIDAI collect the blood group data of each individual; other ministries wanted information on caste, disability status and other attributes. Ram Sewak Sharma reminisces, ‘A common mistake that government applications make is to collect data which they are not likely to use at all, creating long and complicated application forms—the pension form for government servants is like a booklet, with some absolutely irrelevant information like religion and ethnicity. This has happened because the designers of these forms have no idea why they want the extra information, but feel that since it might be required later for some purpose, it’s better to collect it now. When we were designing the input form for UIDAI, there were many suggestions from other stakeholders, largely government departments, to collect extra information. The department of minority affairs wanted religion to be collected, and the rural development department wanted income-related information to be collected. The UIDAI adopted a policy that it would collect only minimal information which was necessary and sufficient to establish the identity of a person—
nothing more and nothing less. We just wanted to establish that X is indeed X. We also needed his address to communicate his Aadhaar number to him.’

Eventually, the only fields made mandatory were a person’s name, address, date of birth and gender. These mandatory fields are sufficient for the purposes of verifying one’s identity and address. The committee report also specified exactly how all this information was to be collected. As a result, data verification and capture have been converted into a standardized process, easily implemented by any one of the many partners in the Aadhaar ecosystem.

Each of these fields was thoroughly discussed and debated before inclusion in the final list. The goal was to make the demographic data as inclusive as possible so that nobody was left out of the database because they didn’t fit into a standard category. For example, the Aadhaar number was one of the first government-issued IDs that officially recognized ‘transgender’ as an option when specifying the gender of the person being enrolled, at a time when most government programmes still forced transgendered individuals to identify themselves as male or female. This option was made available years before the Supreme Court of India officially recognized ‘transgender’ as a third gender category.

Similarly, it was not mandatory for women to supply their husband’s or father’s name while applying for an Aadhaar number. Although the field was included in the enrolment process, a woman could choose to leave it blank; she could be issued an Aadhaar number without requiring the data of a male relative to be provided.

Pramod Varma was present at the Demographics Committee meetings where these decisions were taken. He tells us, ‘Almost everybody in the committee was initially in favour of including the husband’s or father’s name while enrolling women. I raised my hand, and the chairman, Mr Vittal, said the young man has something to say, let him speak. I asked, “If this information is made mandatory for all, what about special cases like orphans or homeless people? In practice, anyone who doesn’t fit neatly into the mandated categories is often overlooked and eventually excluded from the system.” After I finished
speaking, I remember Mr Vittal giving a beautiful five-minute speech on why governments become so bureaucratic towards the end user. After some debate, the committee agreed that the best way forward would be to make such details optional rather than mandatory.’

Although the demographic data was restricted to just four fields, capturing even this minimal information turned out to be a Herculean task. The diversity of India manifests itself in the names of its people; there is no standardized naming convention that applies nationwide. People in the south often do not have surnames, but may incorporate the name of their ancestral village and family deity into their names instead. Some people include their father’s name as their middle name, others don’t. As a result, a person’s name may have anywhere from one to five components. Addresses in India aren’t standardized either. They often incorporate the name of local landmarks—houses of worship, schools, cinemas, shopping complexes—as points of reference. In India’s villages, where everybody knows everybody else, a postal address that says ‘behind the banyan tree’ or ‘deliver to X, son of Y’ is often sufficient for a letter to reach its intended recipient.

While the deliberations around demographic data collection were ongoing, the same process had begun for biometric data as well, undertaken by a Biometric Committee set up in September 2009.
⁴
The committee consulted several biometric experts and academics to decide the best type of biometric data to collect and the associated costs. They also closely studied other programmes around the world that used biometric data for various purposes, like the US VISIT programme that collects fingerprints for border security reasons. In parallel, the UIDAI carried out pilot studies to test the validity of fingerprint data, using over 250,000 fingerprint images collected from 25,000 people living across Delhi, Uttar Pradesh, Bihar and Orissa, many of them in rural regions.

Putting together all this information, the best strategy that emerged was to capture a combination of a person’s photograph as well as all
ten fingerprints and iris scans. The pilot study showed that the agrarian population often has bald fingerprint ridges. If we only used fingerprint data, we would have either excluded the exact section of people most in need of Aadhaar, or our database would have a huge number of errors, since even small errors get magnified with 1.2 billion people. As Ram Sewak Sharma puts it, ‘Including iris data in the biometric set was in my opinion one of the most crucial decisions that allowed us to reach 99.99 per cent accuracy when it came to de-duplication. In iris scans, we now have a very powerful authentication factor which especially benefits poor people and in some cases is even better than fingerprint authentication.’

Once the data collection standards were in place, the next order of business was to decide how exactly a resident could go about enrolling for Aadhaar.

The goal was to grant residents the freedom of choice to enrol in the manner most convenient to them, and the Aadhaar enrolment ecosystem that eventually arose was designed to meet this goal. The entire ecosystem can be thought of as a pyramid, with the UIDAI at the top. Next come a class of organizations known as ‘registrars’, whom the UIDAI appoints. These organizations include all state governments as well as government agencies like the Registrar General of India, India Post, the National Securities Depository Limited, public sector banks and oil companies. The registrars all have one thing in common—they engage directly with residents. Below the registrars in the pyramid are the enrolment agencies, appointed by the registrars after a competitive bidding process in the open market. It is the job of the enrolment agencies to provide the hardware and the operators who will actually carry out the enrolments in the field. Over 100,000 operators have undergone this certification, and at its peak, there were 30,000 simultaneous enrolment stations deployed.