Incident 196: Compromise of National Biometric ID Card System Leads to Reverification and Change of Status

Description: When the leader of the Afghan Taliban was found possessing a valid ID card in the Pakistani national biometric identification database system, Pakistan launch a national re-verification campaign that is linked to numerous changes in recognition status and loss of services.
Alleged: Pakistan National Database and Registration Authority developed and deployed an AI system, which harmed Pakistani citizens.

Suggested citation format

Perkins, Kate. (2013-09-01) Incident Number 196. in McGregor, S. (ed.) Artificial Intelligence Incident Database. Responsible AI Collaborative.

Incident Stats

Incident ID
Report Count
Incident Date
Sean McGregor


New ReportNew ReportDiscoverDiscover

Incidents Reports

Foreword: Pakistan is a leader in the application of identification systems and technology to a range of development issues. The National Database and Registration Authority (NADRA) of Pakistan has become a central player in a number of program areas and has been internationally recognized for its expertise, including winning many awards for excellence. Pakistan has pioneered applications of biometric technology, successfully administering smart card programs for disaster relief programs and financial inclusion schemes for the underserved. In addition to its novel applications, NADRA’s story is also one of effective programming to include traditionally underregistered communities, including tribal groups, transgender populations, and women. NADRA has experienced great success, but its example also shows some of the limitations to the effective deployment of technology when this confronts vested interests. Its experience offers many lessons for other developing countries. This is the story of NADRA and its objectives, business model, and programs, as recounted by Tariq Malik, its former chairman and the architect behind its international recognition, to me and Sneha Raghavan at the Center for Global Development. In November 2009 Tariq Malik was awarded the ID Outstanding Achievement Award at the Global Summit on Automatic Identification in Milan. He received one of the highest awards in IT, Sitara-e-Imtiaz (Star of Excellence), from the president of Pakistan in 2013 for innovative citizen-centric ICT application and services rendered for the state of Pakistan. Alan Gelb

Senior Fellow

Center for Global Development

My Inspiration for NADRA

In early 2008 I was working in the United States in the area of ICT when I was invited to accompany Benazir Bhutto on her fateful journey to Pakistan as an election observer. After she was killed in a murderous attack, I returned to the United States, reviewed her e‑mails, and thought deeply about my various talks with her. I noted 10 points on which she intended to take action if she came to power.

Benazir recognized that there was a trust deficit between citizens of Pakistan and the state and wanted to eliminate it. This involved, first, establishing a one-to-one relationship between citizen and the state. Second, she sought to empower women; third, to strengthen national security; and, fourth, to build a positive image of the country. Reflection on candid talks with her revealed a deep desire that she also wanted to strengthen democracy; improve government service delivery; reform governance; manage disasters effectively; combat corruption; and introduce transparency and combat crime.

It was said that Benazir might not have been killed had she not stood up in her car. This was a powerful metaphor for me – despite the risks, should she not have stood up for her objectives? Hers were objectives I believed in also. I realized that I wanted to find an avenue to implement them, so I started looking for jobs in Pakistan. One had become available in the National Database and Registration Authority (NADRA) as general manager of networks. In April 2008 I applied and was interviewed by the board of director generals and appointed, but when I joined they decided that I was overqualified. Within three months, they promoted me to the position of head of technology or deputy chairman. I presented these 10 objectives to the board and explained that NADRA could use its expertise in data and systems and could make a difference by setting our goals around these objectives. I argued that we should address each one by developing a technology product to address it. The board agreed, and I was given full authority to move forward.

The Origins of NADRA

To recap a little, Pakistan’s first registration office was established by Zulfikar Ali Bhutto, Benazir's father, in 1973 under Article 30 of the Second Amendment of the constitution of Pakistan to perform identification and maintain the statistical database of the citizens of Pakistan. It was stipulated that every person should have a state-issued ID. When I reviewed old documents, I was amazed – he did not want only an ID document to simply record evidence of identity (EOI), but a system that provided an automatic census for informed decision making and to improve strategic planning for the country. In 1973, in a parliamentary session, Bhutto stated in parliament to the people of Pakistan, “Due to the absence of a full statistical database of the people of this country, this country is operating in utter darkness.” The government started issuing national identity card (NIC) numbers to its citizens in 1973. His vision was, in fact, not too different from the 10 points expressed by his daughter.

In 1998 the Nawaz Sharif government took the initiative to conduct the door-to-door census of Pakistan with the help of the Pakistan army. In 1999 the Pakistan army started contemplating merging two institutions, the Directorate General of Registration Pakistan (DGR) and the National Database Organization (NDO), to computerize census data collected and use it to issue computerized cards. A NADRA ordinance was promulgated with ID forms, ID cards rules, and regulations. Preparation began in 2000, but from 2000 to 2002 the program experienced difficulties. The “NADRA ordinance” came into being on March 10, 2000, by finally merging the DGR and NDO, and was introduced with the aim of reducing government interference in the process of registering people. The ordinance is comprehensive; it lays out a blueprint for how to register individuals. The government also provided some seed money on loan to enable NADRA to become financially self-reliant.

Registration was voluntary at the time and covered only adults 18 or above. They recorded manual thumbprints, but this proved to be inadequate for accurate identification. It was a paper-based personal identity system. Pakistan had many Afghan refugees, who started to pour in from 1979, and Bengalis dating from the 1971 separation of Bangladesh; there were also many Iranians living in the country. Many Afghans were able to obtain ID cards by deceiving the paper-based system, and there was no easy way to distinguish these false cards from the real ones. The years between 2001 and 2005 saw serious problems in the area of identification. Many genuine Pakistani citizens were without ID cards, many nonnationals possessed national ID cards, and there were many fake identities.

In response to these difficulties, my predecessor, Brigadier Saleem Moeen, had started researching the deployment of digital biometric technology, and by 2007 he had built a good foundation for a stronger system. The database was run in batch mode; if a person applied for an ID card, the batch to process the application ran during the night, comparing the captured thumbprints of the new applicant against the legacy data. The program imposed some basic business rules, for example, that the age of a person should be less than the duration of the marriage of his or her parents.

By 2008, we improved the data architecture to include the full set of 10 fingerprints and a digital photograph. This technology was powerful enough to enable full de-duplication of the national database and greatly reduced the prevalence of dual identities and identity theft. We asked the people to register and framed this in terms of a strategic partnership with the state, which in turn would recognize them as citizens. Registration was still technically voluntary, but people could not open a bank account without an ID card or obtain a passport or enter into any transaction with the state. An ID card was also needed to obtain a gas or electricity connection and to pay utility bills. These requirements made it very difficult to function without enrolling and caused people to register. By virtue of his post, the chairman of NADRA is also the registrar general of Pakistan, whose signature is on every ID. When I took over, NADRA was still based on an ordinance, but I requested parliament to approve it, and it became an act of parliament, under the 18th Amendment of the constitution. NADRA was lucky and blessed by very competent chairmen. I would give credit to Major General (Retd.) Zahid Ehsan, Brigadier (Retd.) Saleem Moeen, and Ali Hakeem for laying out the solid foundations of NADRA.

NADRA’s Financial Model

NADRA desperately needed a viable financial model. When I joined NADRA, it was left with only two months of salary. We decided that we would not ask for regular budget or for loans from the government as this would open the door to political interference. Instead, we would request a service fee rendered to the government and use our earnings to provide a cross-subsidy to the poor people of Pakistan to enable them to register free of cost. First-time applicants for an ID card would get the card free, but expedited service, within two weeks, would cost 1,100 rupees. We now earn money for a wide range of domestic services. For example, we charge banks around 35 rupees a head for authenticating an individual at the time of opening a bank account. Each individual fee is small, but when you have 55 million transactions, the amount generates significant income and in a sustainable way.

The business model has been very successful. In 2012–13 revenues were three times their level in 2007–8. NADRA now has 18,000 employees, 537 static registration centers, 15 registration centers for overseas Pakistanis, 236 mobile vans, and 74 semimobile units. We operate two data centers 150 miles apart, and a third one is planned to ensure that our data is secure. We run about 1,000 servers connected to 9,000 computers. We have issued 120 million identities and 97 million ID cards; the difference is the children, an area where we are still lagging. Our digital gallery includes 118 million facial images and 503 million fingerprints, making it one of the world's largest multi-biometric databases.

A public company was formed, NADRA Technologies Limited, registered with the Securities and Exchange Commission of Pakistan. It is wholly owned by NADRA and able to bid for contracts outside the country and earn revenues which could be plowed back to support our operations. We compete successfully for international contracts – income from foreign projects alone came to 17.6 million US dollars in 2013, which allowed us to further upgrade our infrastructure. In the last six years, we have been able to reinvest in our technology without accepting a regular budget from the government of Pakistan.

We also focus on employee incentives. I wanted people to be committed to working for NADRA, so we offered salary increases by creating a sliding scale salary structure that was better than that of the government. I introduced a reward and recognition system in which NADRA officers had incentives not to be involved in identity theft. One form of corruption in NADRA was that staff would often accept money in exchange for a token to jump ahead in a service line. To counter this, we created the system for fast-track service at a cost of 1,100 rupees and linked it to payments to payments to staff, who received bonuses from money collected through this system, as an incentive. . Staff could also boost their pay by facilitating more transactions. Unlike government offices, they could keep NADRA offices open for as many hours as they wanted, subject to the manager’s discretion. Between 2008 and 2013 the average salary at NADRA had increased by 131 percent. It is a perfect example how a just reward and recognition system can improve service delivery.

How We Increased NADRA Registration Rates – Especially for Women

In 2008 our rates of registration were not high enough to provide comprehensive identification. As soon as our resources allowed, I therefore made it a point to increase our representation to ensure that there would be an office in every district. We also made sure we would have the technical infrastructure to allow our mobile vans to go into far-flung areas to collect and transmit enrollment data on a daily basis, by making use of satellite communications. The static centers and van and semimobile units were organized into eight regions, each of which was headed by its own data infrastructure office, transmitting data back to the centralized database, which was located in Islamabad. This big data infrastructure through which the information could be exchanged with headquarters was named the Digital Video Broadcast System (DVBRCS).

Using this organization and technology, we rapidly increased registration from 54 million in 2008 to 98 million in 2014, including around 55 million men and 43 million women. While overall registration increased by 80 percent, the increase for women was 104 percent, relative to 65 percent for men. I made it a point in the registration campaign to stress that registration for women empowers them. If they wanted to be counted – as Benazir Bhutto had wanted – they had to let us count them. Overall we were very successful, but we faced some serious obstacles. In some regions we had to deal with the Taliban. We had to deal with terrorists. We had areas where a long-standing culture of male chauvinism would prevent people from letting us register women. Men would hold the hands of the women during registration, or not let us register them at all. I was persistent and sought to remove these objections. For example, we established 15 women-only registration centers, where all the staff were women, from the data entry operator to the manager. I would even send vans with women drivers to go into hujiras (area reserved for women only in a house) or areas ruled informally by warlords.

We incentivized women to register by making clear that if they were poor, they would be eligible to receive a subsistence grant from the Benazir Income Support Program (BISP) – a financial inclusion program geared towards the poorest of poor women. For a woman to register for BISP, they would need to register with NADRA for an ID card. They would also need this ID card if they wanted to go on their hajj pilgrimage, as it was a prerequisite to signing up for a passport. Similarly, we educated them on the need for ID to be able to exercise the vote, which involved running a campaign to assure them that they could vote how they wanted, even against the will of their husbands, and even when their husbands were not present. Many were thrilled to learn of these possibilities. When the first center was opened in Mardan, a region affected by the war, 3,000 women turned up on the first day.

To achieve universal coverage, we had to penetrate remote areas, beyond the reach of even mobile vans. We introduced motorcycle registration units and, where motorcycles could not go, man-pack units – I hired hikers, mountaineers, and skiers who could go into mountain areas and talk to the people there. Many living in rural mountainous communities did not see the need for an ID card because they had not interacted with the state in years. They were not interested in buying property or even going down the mountain to visit the towns. Through the hikers and mountaineers we explained that they could be eligible for social transfers and that, if needed, we would even open a BISP office on the mountain. We also explained that in the event of a disaster such as the 2005 earthquake, the government could use the ID system to provide a subsistence allowance. Once they understood the link between identification and social protection, they came in droves to register. We had to hire more mountaineers and hikers. It is pertinent to mention here that from June 2008 to December 2013, approximately 30 million Pakistanis got their ID cards free of cost.

In the process, we found some places where the state was totally absent. Azuchi, for example, was discovered by air force pilots in the foothills of the Kirthar Mountains. NADRA was activated and registered them quickly. They were thrilled as it was the first time the state actually bonded with them through the ID card. NADRA is often the first and only government office in certain districts, as we are present in areas that lack post offices, police stations, or any other representative offices of the government. To cite another example, Baluchistan experienced a tragic earthquake last year. We were the first responders because we had an office there, as well as mobile vans.

In addition to issuing general national ID cards in Pakistan, we issued special cards to citizens with special needs on the advice of the president, Asif Ali Zardari. We also wanted to document our diaspora. We registered 6.3 million identities of overseas Pakistanis through our embassies. This NICOP (National Identity Card for Overseas Pakistanis) card entitles them to transact with Pakistan and has some additional benefits. Similarly, the “Pakistan Origin Card” is issued to Pakistanis living abroad in states such as Germany and France that do not allow dual citizenship. They can transact with the government and enter visa-free despite relinquishing their Pakistani citizenship.

NADRA empowered minorities, as well, by easing registration for them. For example, Hindu marriages were recorded with the help of Mukhis (community leaders). Christian baptismal certificates were also accepted for registration. Similarly, Pakistan became the first Muslim country that allowed eunuchs to declare their identity as such (Khawaja Sara). All minorities are on voter lists, as well, and have equal voting rights. NADRA has empowered 1.3 million Hindus, 1.2 million Christians, and about 50,000 Sikhs, Buddhists, Parsis, Jews, and other minorities in the last four years. While registering minorities, my message was be counted, exercise your vote, and have your say in your Pakistan!

Towards More Sophisticated Smart Cards—and Universal Financial Inclusion

The Pakistani diaspora cards needed to be compliant with ICAO standards (9303 Volume 2) and with sufficient security features to be difficult to forge and hard to use for illegal purposes, such as human trafficking, illegal immigration, or other criminal acts through identity theft. I wanted to introduce the best identity document in the world. I researched and found that the German ID card was the best and at the time the most advanced card available. It became my passion to make Pakistan’s ID card more secure than the German ID card, which has 28 security features; ours was designed to have 36 features. The contents on the face of the card are also in its chip; it has three ghost images, a machine-readable zone (MRZ), the flag includes microtext, and the colors change when the card is tilted.

The operating system on the chip was written by our own NADRA engineers indigenously. It is a passive chip, so it is not possible to track the owner. The chip includes a match-on applet that stores and matches the four best fingerprints and the digital photograph on the chip. The system is very secure; when you match information from the chip, it does not leave the chip itself. All you know is whether or not it matches, based on the applet. This gives confidence to card owners that their information is under their control, and they have the authority to allow access to this information for a service or product that they are entitled to. The card has other technical features and also provides a platform through which services can be delivered. These services can be enrolled through digital certificates, PIN, and token methodology. These include financial services – Pakistanis living abroad can send remittances by partnering with banks as an alternative to the “Hawala and Hundi System” (undocumented system), where all too often undocumented money is sent to terrorists. The card satisfies the Know-Your-Customer (KYC) requirements of the State Bank of Pakistan for secure money transfer.

With programs like social protection and health insurance in mind, we then moved to introduce a similar smart card for citizens inside Pakistan. I introduced the Smart National Identity Card (SNIC) in October 2012 to convert Benazir’s dream of secure identification for biometrics-based voting, pension disbursement, social and financial inclusion programs, and other services like health insurance and keeping track of whether children are vaccinated or are not. This can be an effective vehicle to strike a strategic partnership between the state and citizens in a meaningful way. It can revolutionize governance by cutting intermediary channels and bureaucracy. As an incentive, we partnered with a national insurance company, the State Life Corporation of Pakistan, to offer insurance against accidental death for a nominal charge built into the ID card fee. I ran a campaign, offering citizens the choice of a simple plastic Teslin-based card or the new card with the built-in insurance policy. It was a huge success. One million people applied for smart cards in just eight weeks. Another selling point was that this card would be accepted as an identity credential in more than 100 international airports because it adhered to the ICAO rules for machine-readable travel documents.

Another feature is the QR code on the back of the card. It serves as another security feature: taking a picture of it and running a program decodes it to provide the same information as on the front and back of the card. The QR code also enables the card to be used for other applications such as e-commerce; platforms for iPhone applications were being developed when I was chairman. It does not require a lot of infrastructure to use; all that is needed is a machine that is capable of reading the card or a mobile phone that has the same capacity. I hope the current chairman will take it to the next level.

I strongly believe that by experimenting with information and biometric technologies and using NADRA’s citizen database to Pakistan’s competitive advantage, Pakistan has a unique opportunity to become the world’s largest and most inclusive digital financial services market, with every household in the country connected to the integrated digital financial platform. It is my passion to see this happen, as it would convert Benazir’s dream into reality. Benazir Bhutto was a strong believer that financial inclusion matters for the welfare of the poorest of the poor and for national economic development, and I believe this, too. Without formal financial services, the poor have to rely on informal mechanisms that are expensive, unreliable, and often exploitative. NADRA’s smart card solution triggered a branchless banking revolution in Pakistan. Digital financial inclusion must be a priority for the current government, as leveraging NADRA’s database can enhance its ability to optimize service delivery. The State Bank of Pakistan will be better able to monitor suspicious financial transactions when they are under the net of digital transactions, rather than outside in the untraceable cash economy. With citizen registration soaring to 98 percent, Pakistan has the necessary ingredients to achieve universal financial inclusion.

Identification for Social Protection

We soon began to use our ID technology to implement a variety of social protection programs. Since most of our beneficiaries were now uniquely identifiable, it seemed like a logical transition to use it in this way. Pakistan had not had a census since 1998, and although international organizations often conducted surveys to roll out their programs, there was no mechanism to preserve the data that they collected.

Internally Displaced People

The first opportunity came in 2009 when the government launched the military operation “Rah-e-Rast” to take action against terrorists in Swat and Malakand. In the wake of this operation, many people had to leave their homes, to be lodged in camps set up by government and welfare agencies and with host families. Many of their houses were destroyed, and many lost their livelihood. The government estimated the displacement cost at around 25,000 rupees per household. The problem was how to deliver assistance without pilferage and to offer an assurance to international donors that the money would go to intended beneficiaries. As stressed by the president, the recipients also needed to be able to preserve their dignity. This would not be possible if they were expected to wait in long lines for handouts of rations or clothing.

In response, we used the NADRA database, sorting by the address field to identify potential beneficiaries in the Swat and Malakand regions. The government had come up with a 3R strategy: Relief (immediate), Rehabilitation, and Reconstruction. We calculated that the total number of beneficiaries should be 396,653 and arranged to disburse 25,000 rupees to each internally displaced family using debit cards in tandem with biometric verification. We recorded the thumbprints of all who came to receive the first phase of immediate relief in the form of in-kind aid. The database had an architecture whereby it was easy to see who constituted the direct members of each family, so that we were also able to prevent two members of the same family from receiving the entitlement. The system would reject a member of the family who intended to claim assistance if another member of this family had already done so.

Not surprisingly, many people wanted to change their addresses to Swat in order to qualify for aid, so that we received a lot of applications for change of address. To resolve these cases, we reregistered people living in camps created by the army to ensure that they qualified. When people were reregistered, their fingerprints were taken again to check if there was a match with the citizen database. I also established a rule to the effect that for those critical three months nobody moving into the Swat-Malakand region would qualify to change their address – we did not believe that people would have valid reasons for moving into a war-torn area.

Eligible families were given ATM cards loaded with 25,000 rupees that were activated upon registration. Their purchases encouraged the private sector and helped boost the economy of neighboring districts that were not directly hit by the war. The program was viewed favorably by the United Nations and the World Food Program because it encouraged local distribution and saved them the cost of bringing in food and other in-kind relief by air. They calculated their contribution per family and went through us to allocate an additional 5,000 rupees on each card.

NADRA’s system allows for two addresses – permanent and current. When people move they can change their permanent address once they decide to stay in their new location. This prevents people from claiming benefits from places they used to live in, even if they do not live there anymore. An unexpected and welcome consequence of the relief program was that people in war-torn areas who received money no longer wanted to move despite the difficulties. The social disbursements encouraged them to start new businesses in that area and stimulated the local economy rather than having the money leave to other provinces.

Another lesson that I learned was the necessity of having an effective grievance redress system. The process was not perfect, because there can be a variety of family structures. There can also be children without parents. In these cases we had to reorganize the children as separate family units and to mark the eldest as the main beneficiary. We created similar special mechanisms for widows. I went on national television and asked for people to come to us with their grievances. Some complaints did not reflect deficiencies of the system. Because of the database, we were able to tell if members of the family had claimed the money without telling other members or if people were attempting to cheat the system by trying to claim it twice. The biometric system helped to deny access to relief funds to imposters. In one case, a person attempted to access relief 132 times but was denied and given authorized relief only one time, as was his right. The internally displaced people project was covered in the Economist of August 2009 in a story entitled “A Plastic Prop.”

Flood Relief

A second humanitarian crisis came in 2010. Pakistan experienced flash floods with over 20 million people affected. The prime minister called a Council of Common Interests – CCI – meeting with provincial chief ministers to discuss the potential use for NADRA and create buy-in among all provinces. We presented the possibilities and the limitations of the database usage. We wanted to use this database in a way that the personal information of the citizens would not be compromised. I compared cash disbursements in disasters all across the world, such as after Hurricane Katrina and the 2010 Haiti earthquake. None of these systems was perfect, and the most persistent problem appeared to be difficulty reaching the correct beneficiaries. My advice was for the affected regions to declare themselves as calamity-hit areas in accordance with Pakistan’s 1951 Calamity Act. Once again, we used the address field to sort by all the provinces that declared a calamity and created a consolidated fund with contributions from all of these provinces. The fund grew when international donors also chipped in, realizing that the program was transparent. The federal government offered to match an additional amount of money to each province that contributed to the consolidated fund.

NADRA set up 131 registration sites near the disaster-hit areas. In the event that people had lost their ID cards, as is likely to happen in the case of a disaster, we used our mobile vans to retake their fingerprints. The system would confirm who they were, so that we could issue them a new card. About 700,000 ID cards were reissued free of cost; 300,000 of the new cards were issued to women. Ironically, the disaster served as an opportunity to increase our registration numbers, as we registered many people whom we had previously struggled to reach. Under the Citizens Damage Compensation Program (CDCP), a partnership between Pakistan, the World Bank, and others, a total of 77 billion rupees was distributed to 2.84 million families, an average of over 27,000 rupees per family. We later repeated the process after another flood (2011), with similarly successful results.

In order to ensure that this system was sustainable, we wanted to create a structure to address future disasters. NADRA helped the government to create a division in the cabinet to use the best practices from NADRA’s own experiences and developed manuals of disaster management for future use. My intent was to leave a legacy of sustainable processes which any government can invoke for emergency relief or disaster management. This structure was activated under the National Disaster Management Authority (NDMA), an existing organization. Under the NDMA, we have Provincial Disaster Management Authorities (PDMAs), creating a link from the federal government to the provincial governments to respond in such situations. These PDMAs have the authority to work independently and to invoke NADRA in the event of a disaster.

NADRA’s model of cash disbursement using a biometric system was a big hit. Last year, the United Nations approached me to see if the same cash disbursement system could be deployed in post-conflict situations in Africa.

The Benazir Income Support Program (BISP)

Per Benazir Bhutto’s vision, the elected government wanted to establish a financial inclusion program geared towards women living below the poverty line. But who were these women? Even though it could establish unique identity, NADRA did not have a mechanism to find out who was living below the poverty line. In the first phase of the program, all elected members of the National Assembly (regardless of their party affiliations) agreed to go back to their constituencies with a form to collect data on poor women on which we would run some business rules to try to gauge the legitimacy of the claims. The World Bank opposed this approach because it thought it would introduce political patronage into the program. The World Bank was right, but the government was right, as well. It did not want to wait until the World Bank completed its poverty scorecard survey but rather some immediate relief to a vulnerable part of the community that was suffering under immense inflation shock. Previous governments have rolled out similar programs/surveys (Tameer-eWatan Program and Peoples Social Program) but failed because the survey results were not digitized, and, as such, their findings were not applied by local offices after the government’s departure. I made it a point that this time the surveys must be digitized and preserved in the form of a relational database for use by future governments. NADRA stepped in. We agreed that the data could be collected by elected members of the National Assembly in their constituencies (as an interim measure), and then we would use advanced data analytics to determine how political, how true, and how accurate it was! The World Bank started its poverty survey in parallel but this time with a difference – NADRA partnered with World Bank to digitize the household survey.

Initially, some 4.3 million beneficiary forms were received from the politicians, but after running data analytics and screening, only 2.3 million were determined to be eligible for assistance. I came under heavy criticism from all politicians because we were able to identify some people who were not eligible due to NADRA’s advanced data analytics. But when I explained why some out of the 4.3 million excluded (e.g., government employees, overseas Pakistanis, etc.) were not qualified, they agreed. I could feel the power of “big data” and data analytics. Real data was so convincing that it was even able to convince these politicians! When the World Bank completed its household survey, Pakistan’s first ever poverty database was born, and hence BISP switched to the World Bank’s authentic, nonpolitical and real data. With increased survey data coverage, the number is increasing; new surveys using the World Bank’s poverty scorecard have placed the number of eligible women at about six million. I suspect that this year this number could hit seven million.

Developing Payments Mechanisms for Social Transfers

The first tranches of BISP transfers were delivered with checks through the Pakistani post office system. However, bribes were frequently reported, and we sought a more efficient system of delivery; smart cards that would be loaded with money when a beneficiary was eligible for a transfer. At the same time, because we had 120 million mobile phones, we tested mobile-to-mobile transfers from the BISP office to the beneficiary. When the beneficiaries receive a code, they can go to an eligible store, which will provide them with the money. About 130 billion rupees were disbursed via money order, 16.7 billion via the BISP debit card, 5.25 billion using smart cards, and an additional 1.8 billion using mobile services.

But these new approaches had their problems, too. PIN numbers were involved, and people often shared them with others who would steal their transfers by pretending that the government had not loaded the card, under the guise of helping them get their money. Some people sold their cards to others, not understanding how to use them. I went on television and tried to educate people not to share their PINs and, if they had done so, to explain how they could deactivate their accounts. Some then started selling the cards and fraudulently deactivating them at the same time – the potential for deactivation created its own problems.

These services proved to be a catalyst for branchless banking in Pakistan. After 60 years of banking development, there was only one rural bank branch for every 20,000 rural people. Under BISP, some six million families will have been taken towards financial inclusion by way of Visa or debit cards or mobiles backed by bank accounts; BISP also provided a financial stimulus to the poor of some 108 billion rupees. Banks became increasingly interested in the possibility of educating a largely illiterate population on how to use smart cards, debit cards, and credit cards and incentivize savings for future needs.

Increasing Birth Registration

While Pakistan’s national ID system has moved towards comprehensive coverage of up to 98 percent of the adult population above the age of 18, the rate of birth registration is much lower. I believed that a centralized ID system with very good technology like that of NADRA might help to accelerate these low birth registration rates.

The roadblock to increasing the birth registration rate was that this is the responsibility of the union council, the smallest jurisdiction under provincial government. I therefore developed a program with NGOs and the telecom sector in Pakistan to develop a mobile application to facilitate preregistration. Pakistan has a high density of mobile phone users, so that any midwife or parent can use that application to create a preregistration in our database. NADRA’s infrastructure, such as mobile vans, can then be alerted to go and register the child and take this record to the union council. This is one way that children can be registered through the collaboration of NADRA and these union councils.

Another approach is to set up a standard that all union councils can follow in accordance with NADRA’s methodology. We created a program and provided union councils with both computers and software to register children. While our software is used, the certificate-issuing authority would still be the union council. All certificates would have the same standards. In Punjab, about 98 or 99 percent of union councils are issuing these standardized certificates; in Sindh, about 86 percent; in Baluchistan, 74 percent. In Khyber Pakhtunkhwa, around 98 percent are using the approach.

A birth certificate is also a social document that registers the child as a member of a family. We have relaxed our rules on the necessity of prior marriage or divorce registration to register a birth, and we also accept certificates from religious authorities in lieu of civil registration. There still needs to be a public awareness campaign around this, but we hope it is successful in increasing rates of birth registration in Pakistan.

In the case of children who have been displaced, lost their parents, or other similar circumstances, we have relaxed the definition of the parent. In these cases and similar ones, such as children born out of wedlock, philanthropic organizations may take care of the children. The institution taking care of the child can assign an alias name of a father that appears on the front end of the database, so that the child will not face any stigma from being parentless. On the back end of the database, however, we would know that this child is registered under a certain orphanage. In June 2013 we also started registering the orphanages and the children therein. I pushed back to register children at whatever age they are, so that we can have a snapshot of the demographic situation at any given time. If there is a child that we registered at age 5, when he comes back at age 18 for an adult registration, we would know that we registered him 13 years ago and would be able to compare data. Since biometric data do not become stable until the age of 14 or 15, we start taking biometrics at this age.

The Approach to Privacy

Security versus privacy is a very hotly debated issue these days. In Pakistan’s perspective, extraordinary security circumstances demand extraordinary steps to revamp governance. The state is eroding very fast, and it has to restore its writ to avoid collapse at the hand of outlawed and terrorist groups. At the same time, it is my strong belief that privacy of citizen data must be guaranteed in a fragile state like Pakistan. It is sad that Pakistan does not yet have an official data privacy law. In this situation I had to be especially careful to implement systems within NADRA in a way that protected the privacy of the citizens. I started to craft Pakistan’s first citizen’s privacy law with select lawyers and parliamentarians, but then I had to leave. I made it a practice to obtain permission from NADRA’s independent board for any developments or applications that would require data sharing. There is also a mechanism in place to obtain official approval for sharing NADRA data with other organizations, including government departments, on a case-by-case basis.

To take one example, when we were asked to help the Election Commission of Pakistan, we had to seek approval from the NADRA board on what fields we were allowed to share and what fields had to be kept private. Similarly, when the tax study was done for the Federal Bank of Pakistan, they already had a lot of data on who was paying their taxes and how much they paid but wanted us to help them identify trends on how many people were missing from this list and how much potential revenue. We have not shared the identities of these individuals with them, but rather provided aggregated data that can help them with policy making. In addition, data sharing needs to adhere to strict parliamentary processes. For instance, if there is a national security issue, NADRA does not itself evaluate the need for personal data to be shared with the law enforcement agency. In this case, the crisis management cell within the Ministry of the Interior makes this determination and submits a formal request for certain data fields. Such standard operating procedures (SOPs) allow for data sharing to be considered on a case-by-case basis and create a paper trail for the data that is shared. NADRA’s administration is answerable to its board. I was also responsible to the Standing Committee on the Interior in the National Assembly, as well as the corresponding Standing Committee in the Senate. All of these bodies have to be kept abreast of any information-sharing needs.

As far as we know, there has not been heavy censure from civil society groups about NADRA’s processes on privacy. We get a lot of positive feedback on the way we use this data and the positive implications these services have for minorities and marginalized populations. However, the case-by-case approach is not enough. More needs to be done on the legislative side, and it needs to take place quickly. I was working extensively on a law that protects the identity of the people. Had I stayed a little longer, I might have managed to see it through. My goal was to get the legislation completed in June 2014.

Towards this end, I also implemented software whereby every transaction is noted in the database. I was working towards my goal that each citizen would be empowered to check the database to see who has accessed his or her data. This is possible if all such transactions are recorded with the audit trail. For example, as an individual citizen, I should be able to put in a request to NADRA to see if the tax authorities have checked my data. This can help citizens to assert that they have paid their taxes or are eligible for a certain benefit. In addition, it also gives the right to citizens to know who has checked their personal information and for what reason. This also empowers NADRA management to oversee its employees to see whether they are involved in identity theft or not. I had a zero tolerance policy on identity theft and was criticized for firing 254 employees involved in violation of NADRA’s policies and procedures.

Conclusion: Optimizing Governance with Biometrics and Information Technologies

NADRA’s success story reflects how organizations can be turned around. NADRA has become the backbone of governance in Pakistan. The systems it has enabled enforce transparency in almost every facet of life where the state touches the lives of its citizens. From birth certificates, to ID cards, to marriage, divorce, passports, and death certificates, and electoral rolls – it maintains the identity, recording its unique evidences and helping the state to enforce transparency. It empowers citizens, minorities, vulnerable communities, and, above all, the state to enforce rights and to roll out targeted subsidies to vulnerable communities. I was able to do what I have done because of the passion ignited by Benazir Bhutto’s dream. Her dream resonates with that of Dr. Martin Luther King Jr. Though I took inspiration from her, I never let political influence guide my job. I ensured that NADRA allowed the government of Pakistan to treat all citizens equally, no matter what their political or religious orientation was. Governments are about people, who have unique identities, and NADRA helps the state of Pakistan to build its capacity to improve service delivery. My overall vision for NADRA has been for it to provide an effective identity management strategy for Pakistan. With unique identity as the foundation, we can roll out reforms more easily and transparently, eliminating any political expediency.

NADRA’s identity management system with its associated identity card is the single version of the truth, recording all evidence of personal identity. Systems for passports, civil registration, and other documents should all originate from a single version of the truth. This was very far from the reality we faced at the start, a reality that often originated from people’s attempts to cheat the authorities. A single identity truth could provide a mechanism to save a lot of money on social support and cash transfers – the ability to engage citizens and dispense these services directly to individuals is a huge benefit. We estimated potential savings of more than 50 billion rupees from eliminating cheating and false positive and false negative identification errors from such programs.

My vision is for NADRA offices to become citizen service centers, rather than just registration centers. They would activate health insurance, register vaccinations, help facilitate scholarships, and offer access to a range of other government services, cutting the intermediaries like bureaucracy and hence striking a direct partnership between citizens and state. Only the innovative use of a mixture of biometrics and information technology can achieve this result. This is the true manifestation of e-government. I wanted to use the same offices for computer literacy programs in the evening, to impart training and engage the citizens to make use of the Internet to better their own lives and careers. I wanted to make it easy for citizens to transact with the state and easy for the state to enforce the law, rather than to be an absent state.

Standard identity cards have limits but cannot be replaced overnight. We have to incentivize smart cards to make them more popular and encourage their adoption. For instance, creating a smart card for children allows international recognition and may facilitate government services specifically geared towards children, including vaccination and healthcare records as well as education, immigration, and building experience with financial transactions. “Big data” in general is an area where Pakistan has an opportunity to strengthen governance. Teaming up with the World Bank allowed us to develop the poverty scorecard, which can identify not only people living below the poverty line but also the lower middle class. This allows us to roll out tiered subsidies for different income groups, based on their own specific needs. Having a biometric database alone is a start but is not enough to create an effective state.

Technology in the Service of Development: The NADRA Story

ISLAMABAD - The Islamabad High Court (IHC) Friday accepted a petition challenging the jurisdiction of the National Database and Registration Authority (NADRA) to impound CNIC of a girl by purportedly adjudicating upon the question of her paternity.

A single bench of IHC comprising Chief Justice of IHC Justice Athar Minallah issued the verdict in a petition filed by a girl Urooj Tabani against NADRA’s decision of impounding her CNIC and excluding her from the family tree.

In the verdict, the IHC bench allowed the petition and declared the impugned orders and proceedings of NADRA as illegal, void, without lawful authority and jurisdiction. Justice Athar declared, “The Authority is not vested with the power and jurisdiction to, directly or indirectly, adjudicate upon or interfere with intricate contested family disputes, including paternity. In such eventualities a change in the particulars incorporated in a card issued under the Ordinance of 2000 shall be subject to a declaration by a competent court.”

He added, “The CNIC of the petitioner shall be restored and the change of paternity shall be subject to a declaration by a competent civil court. By disputing the paternity of the petitioner, the latter must have been exposed to unimaginable pain, agony and emotional distress as well as psychological trauma.”

Urooj filed her petition challenging the jurisdiction of the NADRA to impound her CNIC by purportedly adjudicating upon the question of her paternity. After hearing the parties, the same IHC bench directed the Authority by way of an interim relief vide order dated 11-10-2019 to unblock the CNIC of the petitioner.

She stated that the NADRA, instead of complying with this court’s order, issued a fresh CNIC showing some other person as the petitioner’s father. However, realising that this court had not passed such an order, the Authority unblocked and restored her original CNIC.

The counsel for the petitioner argued that the Authority could not have changed the paternity of the petitioner unless there had been a declaration to this effect by a competent court and adjudication of contentious and disputed questions of fact is outside the scope of the power and jurisdiction vested in the Authority.

The IHC bench observed in the verdict that the petitioner had unequivocally stated that she was prepared to take a DNA test so that the question of paternity is decisively settled. Similarly, the counsel for Yaqoob was asked that whether the latter was willing to accept the suggestion made by the petitioner. The counsel had sought time and on the next date of hearing Yaqoob did not consent to undergo the DNA test.

IHC accepts petition challenging jurisdiction of NADRA to impound CNIC of girl

ISLAMABAD: The Islamabad High Court (IHC) on Saturday has imposed Rs0.5 million as fine on the National Database and Registration Authority (Nadra) for blocking the CNIC of a woman over a paternity dispute.

The court, in its order, states that the authority is not vested with the power and jurisdiction to, directly or indirectly, adjudicate upon or interfere with intricate contested family disputes, including paternity.

Nadra’s proceedings have been declared “illegal, void, without lawful authority and jurisdiction. The court also imposed a fine of Rs0.5 million fine on Muhammad Yaqoob too.

“The CNIC of the petitioner shall be restored and the change of paternity shall be subject to a declaration by a competent civil court. By disputing the paternity of the petitioner, the latter must have been exposed to unimaginable pain, agony, and emotional distress as well as psychological trauma,” the court said in its judgment.

Nadra and Muhammad Yaqoob have been directed to deposit the imposed costs with the deputy registrar (judicial) of this court within 30 days. The amount will then be paid to the petitioner, according to the order.

Petitioner Urooj Tabani had approached the court and said that her father, Muhammad Yaqoob Tabani, has refused to acknowledge her as his daughter and got her CNIC blocked. Her name was also removed from his family tree.

Tabani’s mother, Firdous Fatima, had filed a petition in a Karachi family court and claimed that she married Muhammad Yaqoob in 1989 and her daughter was “born out of wedlock”. She asked for a “decree for maintenance, delivery expenses, and dower amounting to Rs500,000.”

IHC imposes 0.5m fine on Nadra for blocking woman’s CNIC

ISLAMABAD: The Islamabad High Court (IHC) has declared that the National Database and Registration Authority (Nadra) is not vested with the power and jurisdiction to, directly or indirectly, adjudicate upon or interfere with intricate contested family disputes, including paternity.

A single bench of Chief Justice Athar Minallah on Friday passed verdict against the Nadra’s decision of impounding Urooj Tabani’s CNIC and excluding her from the family tree.

The same IHC bench on 11-10-2019 as an interim relief had directed the authority to unblock the CNIC of the petitioner.

The court declared the impugned orders and proceedings of the Nadra as illegal, void, without lawful authority, and jurisdiction.

Justice Minallah declared, “The Authority is not vested with the power and jurisdiction to, directly or indirectly, adjudicate upon or interfere with intricate contested family disputes, including paternity. In such eventualities a change in the particulars incorporated in a card issued under the Ordinance of 2000 shall be subject to a declaration by a competent court.”

He added, “The CNIC of the petitioner shall be restored and the change of paternity shall be subject to a declaration by a competent civil court. By disputing the paternity of the Petitioner, the latter must have been exposed to unimaginable pain, agony and emotional distress as well as psychological trauma.”

Urooj Tabani has filed a petition challenging the jurisdiction of the Nadra to impound her CNIC by purportedly adjudicating upon the question of her paternity.

She stated that the NADRA, instead of complying with this court’s order, issued a fresh CNIC showing some other person as the petitioner’s father.

However, realising that this court had not passed such an order, the authority unblocked and restored her original CNIC.

The counsel for the petitioner contended that the Authority has acted malafidely and without jurisdiction, the Authority is bereft of adjudicating the paternity of the petitioner, the Authority is not empowered to adjudicate disputed and contentious questions requiring recording of evidence, the Authority had no locus standi to seek review of the order passed by the learned Ombudsman, Respondent No 3 i.e. Muhammad Yaqoob Tabani (father of Urooj), had never challenged those public and other documents relating to the petitioner wherein his name was recorded as the latter’s father, the domicile certificate and the educational certificates issued by the competent forums are public documents and a presumption of truth is attached thereto.

He argued that the Authority could not have changed the paternity of the petitioner unless there had been a declaration to this effect by a competent court and adjudication of contentious and disputed questions of fact is outside the scope of the power and jurisdiction vested in the Authority.

The counsel for Yaqoob Tabani placed reliance on the decree in the suit for jactitation of marriage and reliance has also been placed on copies of documents to establish that the paternity recorded therein was different.

He argued that the Authority had rightly impounded the CNIC and changed the family tree by excluding the name of the petitioner.

The IHC bench observed in the verdict that the petitioner had unequivocally stated that she was prepared to take a DNA test so that the question of paternity is decisively settled. Similarly, the counsel for Yaqoob was asked that whether the latter was willing to accept the suggestion made by the petitioner.

The counsel had sought time and on the next date of hearing Yaqoob did not consent to undergo the DNA test.

Nadra has no power to adjudicate upon paternity: IHC

Balochistan, Pakistan. Akhtar Mansour, head of the Afghan Taliban, finished his lunch at a roadside cafe, and was en route to the provincial capital of Quetta when his white Toyota Corolla was reduced to a smoldering mass of twisted metal by two Hellfire missiles, fired by a U.S. military Reaper drone.

Mansour was killed in an instant, his death now a footnote to America’s 20-year misadventure in Afghanistan. But he was survived by a shiny piece of mint green plastic, retrieved from the car’s charred remains: an identity card issued by Pakistan’s National Database and Registration Authority (NADRA) that identified him as Muhammad Wali, a Pakistani citizen.

For Pakistan’s government, the discovery that the leader of the Afghan Taliban had acquired this supposedly secure and unforgeable form of identification was a source of great embarrassment. In response, a nationwide identity “reverification” campaign was launched to root out foreigners posing as citizens, forcing 180 million people to prove that they were, in fact, Pakistani.

That was the summer when, with the War on Terror as a dramatic backdrop, a woman named Gulzar Bibi received a letter from NADRA informing her that her ID card had been blocked. She didn’t know it then, but the news would turn her life upside down and leave her living in fear for years to come.

Fifty-three years old and a mother to nine children, with a voice prone to swelling indignantly when launching into a story, Gulzar has lived in an informal settlement in the Pakistani capital of Islamabad for the past 40 years. She spends the monsoon months muttering Quranic verses, praying that the water rising in the garbage-choked sewers nearby will not wash her home away. For the rest of the year, she fights off threats of eviction from Islamabad’s municipal authorities, staring down bulldozers dispatched to raze her house.

Life is usually difficult for Gulzar, but NADRA’s decision to suspend her computerized national ID card (CNIC) made it impossible to do things most people take for granted. Her cell phone stopped working and she was unable to access welfare programs that provided food rations, state-subsidized medicines and free schooling for her children. Quickly, her eldest daughter realized that her ID card had been suspended as well. In official NADRA parlance, it had been “digitally impounded.” Then, all three of Gulzar’s sons followed, along with a brother in Lahore. Like dominos, the whole family fell.

The letter instructed Gulzar to visit a government office three miles away. A widow who barely makes ends meet by cleaning rich people’s houses, she grapples with a number of long-term health conditions. Years ago, she had been bitten by a pair of dogs. The infection festered, curdling into sepsis and debilitating her for life. “I walk two steps and I’m out of breath,” she told me. Still, she had to go. She could not survive without state support.

Gulzar’s predicament wasn’t an aberration. In October 2016, NADRA revealed that it had been blocking an average of 225 CNICs every day since September 2013 — throwing, by that count, a grand total of nearly 660,000 lives into chaos. Many have been reinstated but, as of March 2020, more than 150,000 identities remained suspended. Over the past two decades, the CNIC has come to underpin all aspects of Pakistani life. Since it is also an official marker of citizenship, an impounded card renders its holder, to all intents and purposes, stateless.

Established in 2000, NADRA has been internationally celebrated for designing and maintaining a national database that holds the personal and biometric information of 98% of the Pakistani population. The World Bank has referred to the organization as “the single source of truth for identification data” in the country. The authority — which falls under the jurisdiction of the interior ministry, but operates as an independent corporate body — has since helped to implement identity-related projects in Bangladesh, Kenya, Nigeria, Sri Lanka and South Sudan.

But, as thousands of Pakistanis can attest, NADRA is also a perfect example of the dangers of unchecked digitization, of how centralized databases can be wielded against people who don’t fit the state’s idea of a model citizen — to the particular detriment of women, working-class people and ethnic, sexual and religious minorities — and how such systems can push someone like Gulzar even further into the margins. The information collected by NADRA, staggering in its volume and increasing by the minute, is also maintained in the absence of legal safeguards, meaning that there is no way of knowing how it has been, will be, or could be used in the future.

Despite multiple requests, NADRA did not respond to the questions raised by this report.

Biometrics — deriving from the Greek “bios” (life) and“metron” (measure) — have formed part of identification systems for thousands of years. Evidence exists from Assyrian payment receipts to inky footprints on Chinese divorce records. In South Asia, however, the gathering and collation of this form of personal information has long been associated with ideas of criminality and state control.

In 1858, near the Hooghly River in West Bengal, India, the English civil servant William James Herschel ordered a local contractor named Rajyadhar Konai to stamp his palmprint on a piece of paper, in order to make an agreement they had reached binding and indisputable.

“I was only wishing to frighten Konai out of all thought of repudiating his signature hereafter,” Herschel later recalled. Herschel was struck by the unique yet highly reproducible nature of the human handprint and his decision marked its first modern use for official purposes. For administrators of the British Empire such as him, the native population of South Asia tended to blur together. Individual identity was, as one scholar put it, unfixed in the colonial gaze. Herschel believed that the uniqueness of a person’s biometric information could help colonial authorities to keep track of people on an individual level.

The practice of fingerprinting, writes historian Chandak Sengoopta in his book “Imprint of the Raj: How Fingerprinting was Born in Colonial India,” is a bit like curry powder: “Developed in India but not indigenous, British but not evolved in Britain itself… incorporated into British tradition and then gradually retransmitted to the world at large, blurring the simplistic distinction we often make between home and empire.” By 1897, the entire Bengal police force had taken up fingerprinting. Four years later, the London Metropolitan Police would begin using it in criminal investigations, too.

Milestones in biometric history

More than a century later, Gulzar stood in line at NADRA’s newly inaugurated “Mega Center” in Islamabad’s central business district to reinstate her card. She wasn’t aware of the murky history of identification on the Indian subcontinent, but she still couldn’t help feeling like a criminal. Not only does a blocked CNIC have immense material ramifications, it also takes a psychological toll. You can’t help but wonder, what did you do to provoke the blocking of your ID and what will the future consequences be?

This anxiety is especially pronounced for Pashtun people like Gulzar. Pashtuns account for the majority of blocked CNICs — 63% in 2017 — despite accounting for just 15% of Pakistan’s population. The Pashtun community has historically lived in southern Afghanistan and northwestern Pakistan, an area long unsettled by war and displacement and once again making worldwide headlines for those same reasons. It is unclear what your legal rights are if NADRA blocks your card, but many Pakistani Pashtuns fear being categorized as Afghan refugees and forcibly sent across the border, to a country they have never lived in.

At the NADRA office, Gulzar learned why her card had been blocked. NADRA’s database is organized as a network of family trees, with a man as the designated head of each registered household. One of her brothers had lost his CNIC and, when a stranger tried to pass it off as his own, the system flagged it and blocked all other linked IDs.

The officials were far less helpful when it came to fixing the problem. Gulzar would have to provide some sort of evidence that her family had been resident in Pakistan before 1978, the year the country amended its citizenship laws to account for East Pakistan becoming the newly independent Bangladesh. And, no, copies of her long-deceased parents’ papers would not do. Perhaps some sort of land record? Maybe a tenancy agreement from 40-odd years ago?

Gulzar’s heart sank. Although she now lives in Islamabad, she grew up nearly 125 miles away, in Peshawar, the capital of Khyber Pakhtunkhwa province. “My parents, grandparents, uncles are all dead,” she told the person handling her case. “The only land to their name is the graves in which they’re buried.”

The NADRA official shrugged. If Gulzar wanted her CNIC to be reinstated, heading back to Peshawar was the only way.

Gulzar’s worries over her suspended identity were not just for herself. “I’m an old woman,” she shrugged. “I’ll die soon enough.” Her biggest concern was that if she didn’t scramble to sort the matter, her children would suffer. With that in mind, but no plan in place, she boarded a bus and set off in search of documentation that would prove she was from the country she had lived in all her life.

Rather than fading away, the colonial legacy of individual identification came to be seen as increasingly necessary in South Asia after Partition. The fall of the British Empire and the creation of an independent India and the new state of Pakistan in 1947 was a bloody and chaotic process. Nearly 10 million people scrambled across hastily drawn borders in what remains one of the largest migrations in human history. Who was Indian? Who was Pakistani? Who was a refugee, requiring state assistance? Governments on either side wanted to know.

In Pakistan, a citizenship law was enacted in 1951. People born there after that year, those who migrated there before 1952 and others with at least one Pakistani parent were deemed to be citizens.

Similar questions arose in 1971, when Bangladesh declared its independence. Under then Prime Minister Zulfikar Ali Bhutto, who came to power on the socialist promise of “roti, kapra aur makaan” — food, clothing and shelter for all — a national registry began collecting data in 1973. In the absence of a “full statistical database of the people,” Bhutto declared, “this country is operating in utter darkness.”

Looking back, you can see the emergence of a fundamental tension. Did Pakistan’s government want to know who people were in order to provide them with welfare entitlements or did it just want to know who they weren’t: an Indian, a Bangladeshi, or a member of some other ostensibly undeserving group? Welfare and surveillance aren’t necessarily mutually exclusive motivations, but in the ensuing decades — especially the 1980s, when millions of refugees fled the Soviet invasion of Afghanistan for the relative safety of Pakistan — the gap between them widened.

Gulzar Bibi was a little girl in 1973, when Pakistan first began issuing photographic IDs. (The very first was issued to Bhutto himself.) She lived in Peshawar, in a house bursting at the seams with grandparents, uncles and aunts, nieces and nephews. The men worked as butchers, while the women stayed home.

Her family had all moved out decades ago but, when Gulzar stepped off the bus from Islamabad, she went straight to her old house and knocked on the door. When the owners answered, she recited the names of her grandfather, father and uncles. Did they sound familiar and did the new residents remember buying the house from any of them?

They did. Relieved beyond measure, Gulzar returned to the NADRA office in Islamabad, brandishing a land deed from the mid-1970s. The officials wrinkled their noses. The document needed to be attested by a senior police official from a local station, they said — someone who could vouch for its veracity and for Gulzar herself. So, back to Peshawar she went, another day’s wages lost. No one at the station would sign off on the document for her. “We don’t know you,” they shrugged, “so how can we vouch for you?”

As she stood there, ready to give up, a memory floated into her head: the name of a police officer who used to visit the family shop when she was a little girl. He had retired long ago, staff at the station said, but they knew where he lived. So off she went, knocked on his door and reeled off, once again, the names of her grandfather, father and uncles. Did he remember them?

The old man squinted at her.

“Of course I do. You used to throw rocks at me when I’d come to the shop. How you’ve grown!”

If you know how to read it, your CNIC card can reveal a lot about you.

A sequence of 13 numbers serves a similar purpose to a social security number in the U.S. Many of those individual digits signify particular personal details. The first indicates your province of birth; the second, the division within that province, and so on, all the way down to your specific union council, the smallest administrative unit in Pakistan. The last indicates your assigned gender.

Your photograph appears on the right, in monochrome, and below it, your signature. On the left is an embedded microchip and, on the back, a QR code. Hold it up to the sun and a pair of ghost images will shimmer into view: a tiny silhouette of Pakistan and your own face. The card is printed in layers, each with its own security features — 36 in total, including microtext, holograms, guilloche patterns and rainbow printing. NADRA says that it is among the most secure digital identity cards in the world.

The data on the card is stored in its microchip, along with scans of your irises and all of your fingerprints. Along with the details of 180 million other citizens, this information is collected in a centralized database in Islamabad, which NADRA refers to as the Data Warehouse. According to a 2018 World Bank report, that database is linked to at least 336 public and private services. Three years on, the number is probably higher.

If you go to a store in Pakistan to purchase a new SIM card for your cell phone, this is likely what will happen: the salesperson will ask you to place your CNIC into a card reader; the reader then authenticates itself to the card, after which the card will verify itself to the device. Following this exchange — think of it as an introductory handshake — the reader will ask for your thumb scan and match it with the print stored on the card. If, for some reason, the system can’t match your credentials, you can’t buy a SIM. Or access your bank account. Or collect social security. Or vote.

Broadly speaking, digital biometric identification is composed of three processes: enrollment, which establishes information about a person; authentication, which confirms their identity; and authorization, which determines what services can be accessed after authentication. Think of it as a series of questions:

  1. Enrollment: What do we know about you?

  2. Authentication: How do we know it is you?

  3. Authorization: What are you entitled to?

Proponents of biometric identification often invoke fraud prevention as a reason for its use, harking back to Herschel’s argument a century-and-a-half ago. There is, however, very little evidence to indicate that such systems do, in fact, curtail fraud in any meaningful way. As regards Pakistan, there is insufficient research to make an argument either way. Still, the idea has a remarkably firm grip on the popular imagination.

“There’s a way of thinking — it probably has strong colonial origins and isn’t unique to Pakistan, necessarily — that permeates Pakistani society, beginning with the elite,” said Haris Gazdar, a researcher who has worked on government social protection programs. “And that thinking is that people are opportunistic, that they’re liars and thieves, unless you can control them.”

It’s a Hobbesian view with a South Asian twist: left to their own devices, not only do people tend towards brute self-interest — in this part of the world, they are wily and have a peculiar gift for finding workarounds for almost any system or situation. (The North Indian and Pakistani concept of “jugaad,” or makeshift innovation, presents a more positive spin on this perceived trait.) Therefore, “anything automated,” Gazdar said, “anything that reduces the discretion of a person to do mischief, is deemed better.”

Digitization may hold within it dreams of a more streamlined, secure and scrupulous world, but it rarely plays out that way. In fact, it can make everyday life significantly more fraught, especially when systems do not work as intended. Sometimes — or in the case of Pakistan, often — there is no electricity or internet, meaning that card readers cannot work. Chip-based cards like NADRA’s are relatively secure: your biometric data isn’t transferred over a network, so it can’t be intercepted in that manner. They are, however, vulnerable to what are known as “man-in-the-middle” attacks, in which a malign party inserts themselves between two points of a digital conversation, making the legitimate participants believe that they are talking privately and directly to one another, when the exchange is actually being controlled by the attacker.

In addition, a significant number of people lack easily discernible fingerprints — most notably bricklayers and other manual laborers, but also some hairstylists, chemotherapy patients and older people — which complicates authentication.

Facial recognition isn’t foolproof either. Sometimes automated systems can’t tell two people apart, especially if they happen to be Brown or Black. Currently, a Pakistani Bengali man in Karachi is embroiled in a bizarre seven-year standoff with NADRA, which deploys facial recognition technology at its service centers. When he applied for a CNIC in 2013, he was photographed. When the card was collected in 2014, another photograph was taken, and the system verified that both images were of the same person. Except, they were not. The case appeared before an ombudsman, but remains unresolved. Since the case is ongoing, the man chose to remain anonymous for this story.

“The ombudsman was just as perplexed — he asked if this was a joke,” said Hiba Thobani, the man’s lawyer. “To the naked eye, it is clear that these are two separate people, but NADRA officials refused to acknowledge that their technology could be flawed.” The man in question still doesn’t have a functioning ID card.

Across the world, proposals for identity schemes often meet with robust opposition. In 2006, the British parliament announced plans that proved so contentious that they were repealed within five years. In India, successive governments have expanded the use of a controversial identity system known as Aadhaar, which contains the biometrics and personal information of over a billion Indians, despite opposition from a broad cross-section of activists, lawyers, researchers and politicians. When NADRA was established in 2000, however, there was no concerted movement against it.

One reason was that Pakistanis were familiar with the concept of a national ID card. Many had possessed a rudimentary paper version since 1973. Another was that NADRA, as a concept, absorbed people’s hopes and desires for Pakistan, even when they seemed contradictory. Some thought it would make the state more responsive to citizens, while others argued that it would deter criminals and other troublemakers. Some liked the idea of a more powerful state, others thought it would safeguard against state overreach, and everyone brightened at the thought of a more streamlined bureaucracy.

“At most, people said, ‘It’s an extra hassle,’” recalled Haris Gazdar. “Then they said, ‘OK, but at least it’s a one-window hassle.’”

The 2016 assassination of Akhtar Mansour and the discovery that he held a CNIC prompted the first widespread public scrutiny of NADRA’s processes. Farhatullah Babar was a senior member of the Pakistani senate at the time. “We raised this question in parliament: ‘Who issued this ID card and how? Was the state involved in any way?’” he said.

Mansour, an inquiry later revealed, had posed as a Pakistani since 2005. He had purchased property and flitted in and out of the country with ease. Was NADRA complicit or just incompetent? Would there be any real accountability? “We were told that NADRA had dismissed some lower-level functionaries,” Babar said. “The real issue, of course, was who greenlit his credentials, but the matter was not allowed to be pursued.”

Before the Mansour debacle, an underground market for NADRA data — including forged documents and counterfeit family trees — had proliferated across the country, reportedly in collusion with junior clerks at banks that used NADRA-provided verification software to steal people’s identities. It was only after Mansour’s death, Babar said, that there was any reckoning with the possibility of fraud actually being facilitated by NADRA or the use of its systems.

“The suspicion arose that, if certain state institutions could manipulate the national identity card for their own ends, then private individuals could do it too,” he told me.

That summer, NADRA began sending text messages to the head of every registered household, asking them to confirm the individuals on their family tree and to report any so-called “intruders.” Failure to do so could result in their CNICs being blocked.

It was around that time that Gulzar’s ID card stopped working. If you ask her, though, she’ll invoke a name rather than a year. “It was the era of Chaudhry Nisar,” she said, referring to Pakistan’s interior minister between 2013 and 2017. Nisar is a controversial figure, with an uncanny resemblance to Mr. Bean and a similar proclivity for public gaffes. In those years, the securitization of Pakistan was in full swing, and the sight of barbed wire, body scanners and sandbags on streets had become the new normal.

But, as terrorist violence surged — the December 2014 Taliban massacre in Peshawar of 149 people, including 132 schoolchildren, marking a particularly gruesome apogee — Nisar’s statements became emblematic of the state’s hamfisted approach. At one point he proclaimed that buying too many rotis might indicate a person’s involvement with terrorist activity and that any instances of such behavior should be immediately reported to the police.

Nisar also stated that the reverification exercise that followed Mansour’s killing — the cost of which was passed on to the public, with NADRA charging each participating household 15 Pakistani rupees — would be completed within six months. After that, the national identity database would, once more, be secure. Instead, many Pakistanis found themselves locked out of NADRA’s systems and forced to prove that they belonged in their own country. One woman, whose CNIC was blocked early in the process, threw a party for her whole neighborhood when it was reinstated — a full five years later. Some are still waiting.

It took Gulzar Bibi just four months to get her family’s cards unblocked, but the process left her with a permanent sense of dread. A few months ago, she found her son and daughter, Reza Gul, whispering heatedly in a corner. Unknown to Gulzar, they had gone to the NADRA office to apply for a passport for Reza Gul. When an official quizzed Reza Gul about her mother’s place of birth, she gave the wrong answer, sparking terror that the family’s CNICs had been suspended all over again.

“When they told me this, I swear to Allah, my head whirled — the curses that flew from my mouth!” Gulzar recalled.“I grabbed all my documents, put them in a plastic bag and went straight to that NADRA officer. I threw all our cards at him. I said listen, if you don’t want to issue her a passport, don’t — but how dare you block our cards again? Open them up right now or I’ll break every chair in this office. He was so terrified by all my fuss that he began apologizing. ‘You’re like our mother,’ he pleaded.” Gulzar refused to let up until the man confirmed that their IDs were free and functioning.

Gulzar’s youngest child, Saba Gul, a lanky 10-year-old with tousled hair, is fond of clambering upon the rubble that forms the boundary of the family home and peeping, like a solemn little soldier, over a tattered orange tarpaulin, strung up to replace a wall demolished in the city’s latest eviction attempt. While her mother talked, Saba skipped idly around the yard, squatting among the hens, trailing her fingers along the clothes draped on the line. From time to time, Gulzar looked towards her with an expression peculiar to mothers everywhere: one of stern affection.

Gulzar cannot enroll Saba in a government school. NADRA officials, she said, were refusing to issue the necessary documents because she was born after the death of Gulzar’s husband. The database flagged this tragic and inescapable fact as a system error.

“I didn’t know how to tell them she was already in my stomach when my man died,” Gulzar told me, drawing her shawl tight around her. “This is a conversation strictly for women, you know.”

Like any software, NADRA’s identification system operates within the limitations built into it by human designers. Given the system’s patrilineal structure, when a woman marries, her record moves from her father’s tree to her husband’s. When she renews her CNIC, her spouse’s name appears on the card, replacing the father’s. (NADRA has recently announced a relaxation in this policy but, either way, the records of Pakistani men require no such migration.)

Another rule, explained in a 2014 essay by Tariq Malik — head of NADRA at the time and chief architect of the organization’s biometric services — requires that the age of a person be less than the duration of their parents’ marriage. That stipulation assumes no children are ever born out of wedlock in Pakistan. Such rules are often presented as fixed attributes of the system, but they stem from choices made by people, based on a sense of what they believe Pakistani society should look like, not what it can and does look like.

Across the world, researchers are discovering the complex and occasionally counterintuitive consequences of database design. The United Nations High Commissioner for Refugees’ digital identity system, for instance, is also family-based. Unlike NADRA, however, the first and oldest person to enroll is registered as the head of the household. Accordingly, they then have the right to receive benefits on behalf of the family.

Emrys Schoemaker, a researcher whose work focuses on digital identities, explained that when families from South Sudan sought refuge in Uganda, women and children fled first. Designated as the official heads of their families, many women felt newly empowered. “They could make different financial decisions. They could invest in education,” Schoemaker said. “But their partners weren’t so happy about not being able to control household resources. Apparently, this was the biggest driver of domestic violence in refugee contexts.”

In the case of NADRA, a number of lawsuits illustrate how the messiness of real life collides with the constraints of an automated database. In 2013, the authority blocked the CNIC of Urooj Tabani, a young woman born in 1993. At the time of her birth, according to publicly available court documents, Tabani’s mother and father had been married for four years. A year later, another man surfaced, claiming to be her mother’s husband, alleging that they had married before then and never divorced. In turn, Tabani’s father filed for an annulment. Years later, in 2011, when Tabani turned 18 and got a CNIC of her own, he complained to NADRA and asked for her to be removed from his family tree. The authority complied, even though Tabani’s parentage was never in dispute. Under the logic of the rules governing NADRA’s database, the marriage had been voided, so Tabani couldn’t even exist.

Tabani sued NADRA in Islamabad High Court in 2019 and won. The court ordered NADRA and her father to each pay her half a million rupees ($2,886) in damages. But, in 2018, 22-year-old Tatheer Fatima had less luck. Fatima was making an opposite demand, petitioning the Supreme Court to remove her father’s name from her identity documents. As he had abandoned her at birth, stopped paying child support when she was a toddler and refused to facilitate her application for a CNIC or a passport, why should her identity be linked to his? Instead, she wanted to be known as “bint-e-Pakistan” — a daughter of Pakistan.

Over the years, NADRA has made accommodations for certain groups that do not fit its traditional notion of family. In 2014, following a three-year court struggle, it relaxed the definition of “parent” for orphans, allowing the head of an orphanage to become a child’s legal guardian. In 2017, with its hand once again forced by legal proceedings, NADRA clarified its policy regarding the “khwaja sirah” (trans or third gender) community. It would allow community leaders, or “gurus,” to appear in place of a parent on a khwaja sirah’s CNIC. A few years earlier, the community had locked horns with the authority, winning the right for its members to self-identify as a third gender on their CNICs and shutting down the authority’s initial recommendation that applicants wishing to be identified in that manner undergo a medical examination to prove that they were biologically intersex.

In Fatima’s case, however, the court took a proscriptively conservative approach. After hearing arguments from NADRA, which protested that it could not skip the father’s section in the database without installing new software, the court dismissed her petition altogether. The removal of her father’s name, it decreed, would be against both shariah and the constitution of Pakistan. In doing so, though, it sidestepped a fundamental question: why is paternity integral to citizenship?

And what if there is no father at all? In the late 2000s, a British-Pakistani woman moved to Islamabad with her five-year-old daughter. The child had been conceived in the U.K. via a sperm donor: a process not currently legal in Pakistan. When the woman sought to apply for the child’s Form B, NADRA officials were flummoxed. The system could not compute — literally — the existence of an essentially fatherless child. The woman, whose daughter is now 12 years old, had no desire to wade into a legal minefield, so she chose to rely on visa extensions to keep her child in Pakistan. It helped that her relatively affluent status meant that she could rely on private alternatives to government services and that she and her daughter had British citizenship to fall back on. For them, the CNIC problem simply became a vague annoyance to work around, like a missing step in a staircase.

Citizenship is an elusive concept in Pakistan. “So slippery in fact that,” in Urdu, “there is no word that adequately describes it,” says Aysha Siddiqi, a development and postcolonial geographer at Cambridge University. Colloquially, the word “shehri” — closer in meaning to “city-dweller” — is used. In 2012, Siddiqi began studying the aftermath of unprecedented floods in Pakistan. In the previous two years, heavy monsoons had caused the Indus River to burst its banks, inundating a full fifth of the country’s landmass. Nearly 2,000 people died. Another 20 million lost their homes and livelihoods.

This time, Pakistan, long dependent on international aid, had to rely on its own social protection mechanisms. Though the U.N. referred to the floods as the greatest humanitarian crisis in recent history, the global response was muted.

NADRA stepped in to help by using its biometric database to identify and assist those affected. The authority set up registration sites across flooded areas and used vans fitted with equipment to retake the fingerprints of individuals whose ID cards had been washed away. Some 700,000 CNICs were reissued, and 77 billion Pakistani rupees ($452 million) was distributed to nearly three million families. Many had never opened a bank account, so NADRA issued them with ATM cards to withdraw cash from temporary ones, opened by the government on their behalf.

Siddiqi describes what she saw as an example of “disaster citizenship.” A strengthening of the social contract between state and citizens in the aftermath of a crisis. “This was really the first time in most people’s living memory that they got a particular entitlement from the state, simply for being citizens — not because they had access to patronage or anything like that,” she told me. “The state reached out to them on this very universal platform in a very bureaucratized manner. The people I was spending time with wanted to be seen by the state and, in some ways, the NADRA card was giving them that.”

Slowly, and at least partly through schemes implemented by NADRA, citizenship was solidifying from an abstract notion to one with material benefits. According to Siddiqui, “to discount the borderline revolutionary potential of that would be disingenuous.” When an earthquake killed at least 800 people in southern Balochistan in 2013, NADRA officials served as first responders, simply because they had offices and vehicles in the region. For some Pakistanis, living in districts that had no other trace of the government — not even a post office or police station — their first encounter with the state, other than perhaps the military, was through NADRA.

The shimmering promise of NADRA as a great equalizer can be seen in a recent promotional video. A woman walks into a service center and sits across from a NADRA official. She looks into a lens, then presses her fingertips, one by one, on a biometric device. Walking away, with registration form in hand, her chin lifts slightly and she smiles. “I’ll have my own national identity card,” sings a voice in the background, “I’ll take every step with pride.” Other women appear on the screen: mothers, wives and widows, trans women, women in wheelchairs. “This card will be my honor.” They visit banks and hospitals, enroll in college, line up to vote and flash inky thumbs, beaming all the while. “We must fulfill our national responsibility and acquire a national identity card.”

The researcher Haris Gazdar understands that people often do not want to live “off the grid,” separated from the safeguards and benefits of citizenship. “Most people actually want to be on the grid, they want stuff from the grid. They would like to vote, they would like their children to go to proper schools, they would like to have bank accounts and phones and all of those things. And everything required to be on the grid — such as, say, a national identity card — they want access to that, too.”

Gazdar thinks of NADRA as an inevitability, something that the market would have created, had the state not. His concern is how it can become more inclusive and closer to the benevolent technological utopia of the promotional video. “Now that we have an instrument that is so important and powerful, how are we actually using it? And why are so many people out of it?” he said. “How is it that a lot of people have disputed claims? Why are we making life so difficult for people when errors occur?”

Most horror stories about ordinary people locked out of NADRA are part of an older, global narrative of bureaucracy as a brick wall. The difference is that, in most other encounters with the state, individuals do, at least, have some recourse. “The police have the power to arrest me and I have the right to contest that,” Gazdar said. “The procedures are laid out, even if they are frequently violated. With NADRA, the problem is that it’s grown so quickly and expanded into so many areas, because of the way that we use technology, that this conversation was never had.”

When digital rights advocate Nighat Dad moves through Lahore, Pakistan’s second-largest city, she always notices the CCTV cameras mounted above the streets, recording its inhabitants. “It’s pretty jarring, actually,” she said. “They flash abruptly when capturing an image. If you’re driving, it can be a safety hazard. People often complain about that flash, but I’ve never heard anyone ask exactly what is being recorded, where the data is being processed, who has access to it, when it will be destroyed. Those questions come to my mind, but an ordinary person won’t think that way. They’ve always been told that this system has been put in place to make us safer.”

The CCTV cameras — numbering at least 10,000 in 2,000 locations across the city, although nearly half reportedly do not work — are part of the Lahore Safe City project, one of several such initiatives linked to NADRA’s database being rolled out in urban areas across Pakistan. Most have been installed in partnership with the Chinese technology firm Huawei. (According to research from the Washington D.C.-based Center for Strategic and International Studies, Pakistan has signed more agreements of this nature with Huawei than any other country in the world.) The Safe City project first attracted scrutiny in 2019 when intimate images of couples in cars, captured by CCTV cameras, were leaked on social media, with license plates clearly visible.

Earlier this year, a California-based firm subcontracted to develop technology for the Lahore project sued Huawei in U.S federal court, alleging that Huawei had pressured it to build in a “back door” that would grant access to sensitive Pakistani data, including national identity records. Huawei denies the allegations.

In 2012, a Turkish hacker claimed to have accessed NADRA’s servers by creating backdoors to them. In 2015, intelligence reports warned of data leaks resulting from the government’s reliance on third-party technology, much of which is sourced from companies based in countries including France, Germany, Sweden and China. Last winter, the names, addresses and CNIC numbers of over 100 million Pakistanis were available for sale online, but both the interior ministry and NADRA denied responsibility for the breach.

Explaining the implications of compromised data to ordinary people is frequently challenging for digital activists. The threats often seem abstract and improbable. But, in Pakistan, the dangers are real and concrete. In late 2020, a 15-year-old girl went with her mother to a government welfare office to collect relief payments. An employee used the phone number on her records to harass her, then turned up at her house and raped her. A few months later, in a separate incident, a NADRA employee was arrested for harassing a woman over the telephone. He had retrieved her number from the NADRA database.

In the summer of 2021, as the Taliban reclaimed power in neighboring Afghanistan, it seized control of everything left behind by departing American forces, including military biometric devices and U.S.-funded Afghan government databases. Anxieties in Pakistan took a new turn. What if Pakistani data falls into the wrong hands at some point in the future?

“If the U.S. didn’t think about how that technology could be weaponized against Afghan citizens, do you really think we’ve thought about that?” asked Dad. When new technologies are introduced, she argues, they are always presented in a positive light, as crucial to national security and economic development. “But we’re never permitted to discuss its possible side effects. If you don’t give space to that discourse, then you’ll never consider the possibility of misuse. And you won’t have prepared for it.”

Dad then paused. Over the years, so much ground has been lost to technological evangelists that the comparatively small number of people with contrary views have been forced to reconfigure their positions. “Earlier, we were very outspoken in our opposition to a biometric database,” she said. “But now, so deep into the digital era, you sort of surrender yourself to its inevitability. We’ve come to the point where we’re like, ‘OK, biometric data is fine — but where’s the protection mechanism?’”

She is quick to point out that Pakistan still has no data protection law. A bill is currently under review. In its first iteration, government bodies were exempt from the stipulations contained within it. “Without a law, there’s just no way of holding anyone accountable,” said Dad. “We currently have no legal recourse, no way of holding to account Safe City administrators, telecoms companies, internet service providers — any public or private body that is handling our data, really.”

In 2016, Pakistan passed the Prevention of Electronic Crimes Act, a controversial law that ostensibly aimed to counter online harassment and terrorist activity. Instead, it has severely curtailed free speech and privacy. Journalists and bloggers critical of the state are frequently charged under the law and state agencies are authorized to collect and record their data in real time, without a prior warrant.

“We’re the only country in the world to pass such a law and, yet, we have no data protection measures in place,” said former senator Farhatullah Babar. “The result is that state agencies can play havoc with your data with absolute impunity.”

One example is a court order from 2017. While hearing a case on the proliferation of allegedly blasphemous content on the internet, a High Court judge ordered NADRA to maintain a database of individuals belonging to the Ahmadi community, a persecuted minority sect constitutionally barred from identifying as Muslim in Pakistan. The explicit purpose of the database was to ensure that Ahmadis do not hold public office. The court also ordered NADRA to provide details of people who officially changed their faith from Islam to other religions — a potentially life-threatening taboo in Pakistan — even though the country has no formal laws against apostasy.

Earlier this year, an anti-terrorism court ordered the authority to block the CNICs of the Pashtun civil rights leaders Manzoor Pashteen and Mohsin Dawar, the latter a sitting member of parliament. Pashteen and Dawar, who were charged with inciting sedition while addressing a rally in Karachi, were declared to have absconded when they failed to appear before the court in February. In response, a judge ordered that their ID cards be blocked. The order alarmed Dad, Babar and other observers. Unlike other instances of suspended CNICs, which could be argued away as technical or administrative errors, here was a clear example of the system being used against specific individuals in a pointedly punitive manner.

For the past two years, Hafiz Hamdullah, a former senator from Balochistan, has been contesting his blocked CNIC in court. NADRA said it digitally impounded Hamdullah’s card because intelligence agencies claimed that he was of Afghan origin, despite a long trail of Pakistani documents marking the milestones of his life. The Islamabad High Court noted that there was no proof he had not been born in the country. That alone, it said, made him a citizen under Pakistani law. In a detailed 29-page verdict, the court ruled that NADRA does not have the authority to decide citizenship. Its function is solely to furnish identity documents to eligible individuals. The process deployed by NADRA over the past decade — blocking CNICs while reviewing cases, leaving people in limbo for years — the document unambiguously stated, was illegal.

And yet, Dawar’s CNIC remains blocked, and the Hamdullah ruling can still be overturned by the Supreme Court. For the time being, at least, the system remains fraught with baffling contradictions.

There is a distinct sense of deja vu about Pakistan in 2021. Tariq Malik, the man who vastly expanded NADRA’s powers and influence as its chairperson between 2012 and 2014, is back in charge of the authority after a stint as chief technical advisor to the United Nations Development Program. One of his first moves upon reinstatement was to fire 47 NADRA employees for facilitating fraudulent ID cards. As of August 2021, amid allegations of four million fraudulent IDs circulating in the country, a new identity reverification campaign is now underway, with NADRA urging people to check for intruders lurking in their family records.

As the Taliban takeover of Afghanistan prompts an exodus of refugees, an old debate is also being reignited: what are Pakistan’s responsibilities towards its Afghan-origin communities, many of whom have known no other home? There are plans for a new ID for foreign residents in Pakistan, which will allow them to open bank accounts and do business. At the same time, though, Pakistan has drastically impeded people’s ability to move back and forth across the Afghan border.

Gulzar Bibi’s son-in-law is Afghan, a child of refugees. He makes his living driving a taxi in Islamabad. One year into the marriage, he and his family visited relatives in Afghanistan. His wife, Reza Gul, went too. “I warned them not to take her,” Gulzar said. “I only agreed to the marriage on the condition that she stays here, near me.”

The Durand Line — the boundary between Pakistan and Afghanistan — has historically been extremely porous, allowing Pashtun families, traders and fighters to freely move across it. But during their visit, in June 2016, Pakistan introduced a new policy: all Afghans wanting to cross into Pakistan would now require a valid passport and visa. Reza Gul wasn’t Afghan, but she had no way of proving she was Pakistani either: a legal minor, she didn’t have any official documentation, let alone a Pakistan identity card.

“She got stuck. She’d cry there. I’d cry here. I cried so much,” Gulzar recalled.

Eventually, the couple made it back to Pakistan. The first thing Gulzar did was march her daughter over to a NADRA office to sort out her documents.

On an overcast Sunday afternoon, Gulzar Bibi jiggled her foot on her bed, still seething. It had taken her all afternoon to recount her multiple run-ins with NADRA’s labyrinthine bureaucracy. In the yard, a rooster crowed repeatedly, as if in indignant agreement with her.

“At one point, I was ready to set fire to myself in front of the NADRA office,” she said. “This card, Allah, they treat it like some sort of national treasure, like gold.” She leaned forward. “Tell me, will we need an ID card to enter heaven now?”

Pakistan’s biometric ID scheme is stripping citizenship from thousands of people