UID: from inclusion to exclusion


back to issue

AMONG the many stated goals of the Unique Identification (UID) project, which aims to provide a unique number to all residents identified by their biometric features, was to facilitate and promote inclusion into government programmes. It hoped to facilitate inclusion by providing a recognized ID to all residents and to promote inclusion through ‘de-duplication’ of beneficiary databases.

One of the most appealing claims of the UIDAI project was to enable inclusion of the millions of Indians into various government programmes from which they were often wrongly excluded. Any project that helps overcome the problem of exclusion from social welfare programmes is welcome, especially since these programmes can play an important role in narrowing social inequalities. A closer scrutiny shows that the UID project’s initial claims were based on a basic misunderstanding of the causes for exclusion. Worse, the experience of the past five years of the project suggests that the project itself has become a source of exclusion from government programmes.

As far as the inclusion agenda of the UIDAI project is concerned, there were two important objectives: one, to facilitate inclusion by providing identification documents to all residents. In UIDAI’s documents, the lack of an ID was seen as an important source of exclusion. Thus, the provision of ID documents was closely tied to improving access to government benefits. Two, the project aimed to promote inclusion by reducing corruption (especially duplication in databases). There can hardly be any argument about the desirability of either of these two objectives.


A commonly held perception is that the lack of ID documents is very widespread. However, the UIDAI does not provide any reliable data on how many Indian residents are without ID documents. As far as I am aware, there is no reliable estimate of this even today. To get a sense of this issue, in a small survey of 2200 rural households in ten states in 2013, we asked about possession of different identity cards (such as a voter ID, ration card, NREGA job card, etc.). Between 85-95% of respondent households had one of these IDs. Just over 80% had either a bank or post office passbook. (At that time, only around 15% had aadhaar numbers.)

Even if the proportion of population without any ID is very large, there is the question of what might be the best way to remedy the situation. One option could be to launch a new ID-giving project (such as the UID project). Another option is to focus on existing ID-giving procedures: e.g., improving the birth registration system,1 enhancing the accuracy of other ID databases, such as voter ID or ration cards, and then mandating that these be accepted for various purposes, such as opening bank accounts. Intriguingly, these options were never debated.

The UID project was designed to give aadhaar numbers to both categories of people: those who had no prior ID documents and those who could produce proof of ID and of residence (to get their third, an aadhaar number). To reach out to those individuals who did not have any ID documents, the UIDAI put in place an ‘introducer system’. The UIDAI’s own records on how many aadhaar numbers were issued through the introducer system are damning. Out of nearly 850 million aadhaar numbers issued so far, only 0.219 million (i.e., 0.03%) have been issued to people through the introducer system.2 The remaining aadhaar numbers were issued to people who already had two pre-existing ID documents.

The small number of those who got aadhaar without any prior ID document could mean one of two things: that the number of people without any ID is quite small (less likely) or that the proportion of such people is large but that the UIDAI has (so far, at least) failed to reach out to them. The circularity of the ID-giving exercise and contradiction in claims versus the actual experience has received little attention so far.


Apart from the provision of ID documents, another concern for the UID project was exclusion from government social welfare programmes such as the PDS, NREGA, social security pensions, among others. UIDAI documents suggest that they perceived the lack of identity documents to be the root cause of exclusion from government programmes.

As pointed out by Khera,3 this understanding of the UIDAI is flawed. Eligibility for these programmes requires a person to satisfy the conditions of these schemes: e.g., for social security pensions such as old age or widow pensions, the elderly are required to produce proof of age and a widow will be required to produce a death certificate of her husband. A widow with, say, an aadhaar number, but without her husband’s death certificate will continue to be excluded. Possession of any ID does not guarantee inclusion; applicants need to possess the scheme-specific document.

Further, even among those who did have these documents, a major source of exclusion – perhaps more than the lack of ID documents – is that the scale of these programmes is capped by the government. Even if a person meets the eligibility criteria and is able to produce all the necessary ID documents, she/he may not be included simply because the government has already exhausted its target coverage for that particular scheme. The cure, then, is to expand these schemes.

The only way in which the UID project could have contributed to this was by de-duplication. Here again, there are no reliable estimates of the extent of duplication in government records. In the PEEP Survey of 2013, a door-to-door verification of 3789 pensioners in ten states found only one case of duplication.4


We have seen how the UIDAI project had two important goals – provision of ID documents to everyone, and better inclusion into government programmes. While both goals are desirable, the experience so far tells a different story. On the first (providing ID documents to the ID-less), the project has achieved next to nothing so far (only 0.22 million out of nearly 830 million aadhaar numbers were issued to the ID-less). On the second, the UID project cannot do much. The UIDAI’s understanding of the source of exclusion was inaccurate. The agenda of greater inclusion is better served by expanding the scope and reach of government programmes by removing the artificial caps on their coverage, which are driven by financial concerns. To some extent, this has happened over the past decade.


Having analyzed the intent of the UID project and outlined the limited extent to which it could contribute to greater inclusion, we now look at its actual record of implementation. Perhaps to the horror of its promoters (and in ways unimaginable to them), there is now a lot of anecdotal evidence showing that the UID has become a source of exclusion from these programmes – the very opposite of what it set out to achieve.

Under UPA-II, following the announcement of ‘direct benefit transfers’ in late 2012, there was massive pressure to make aadhaar de facto compulsory in various government programmes involving transfer of cash – including scholarships, pensions, NREGA wages and, to an extent, even in the PDS. The idea was that cash would be transferred into bank accounts (often opened with the help of aadhaar numbers) and beneficiaries would authenticate themselves at the last mile using aadhaar-biometrics, using banking correspondents wherever required. Letters were issued by concerned ministries to states and districts asking them to enter the aadhaar number against the name of each beneficiary of these schemes (referred to as ‘aadhaar-seeding’). Often the pressure took the form of imposing impossible deadlines to complete ‘aadhaar-seeding’ of beneficiaries.

In September 2013, the Supreme Court intervened and directed the government to ensure that benefits were not denied for lack of an aadhaar number. However, the court’s message, either intentionally or unintentionally, did not adequately percolate down to the field. After the Supreme Court’s interim orders (in September 2013, March 2014 and 2015), government letters have carefully avoided the use of the term ‘compulsory’ or ‘mandatory’, but between the lines the message is clear (e.g., instead of entirely removing aadhaar as a requirement, the circulars now state that those without aadhaar should be helped to enrol). Some state officials claim that the pressure is now conveyed orally at review meetings (e.g., video conferences) that take place between the Centre and states.


Bureaucrats in favour of the aadhaar linkage maintain that its use is necessary for de-duplication of government records. They point to a reduction in the number of beneficiaries since aadhaar-seeding began as proof of duplicates who have been weeded out. On the other hand, there are field reports which suggest that at least some of the reduction in numbers is the result of a pressure to show that 100% aadhaar-seeding has been achieved. When beneficiaries of various schemes do not present their aadhaar numbers to add to the database, the response of field staff has been to simply strike off their names.

In Jharkhand, due to the pressure for aadhaar-seeding the NREGA database, job cards were cancelled for workers who did not have aadhaar numbers.5 The cancellation or deletion of names from databases is the first way in which aadhaar has become a source of exclusion from NREGA. This was applicable to the PDS in Delhi where aadhaar was compulsory for new ration card holders under the National Food Security Act.6 This has also happened in the case of scholarships and pensions in several states.7


In Andhra Pradesh, it has become a source of exclusion due to the errors in the aadhaar-seeding process.8 The third reason why aadhaar is a source of exclusion is when biometric authentication is made compulsory but there are issues with matching biometrics. For example, if the electronic point of the sale machine at the ration shop erroneously rejects a person’s fingerprint, it may become necessary to re-enrol their biometrics.

The insistence on showing the aadhaar number to register demand for work under NREGA is the fourth source of exclusion. In Bihar (Katihar and Araria) and Jharkhand (Latehar), reports from the field show that demand for work under NREGA is not registered unless aadhaar numbers are provided.

In many cases, while people have not been excluded, the process has caused some hardship – getting an aadhaar number (often paying money to get enrolled), having it seeded with the relevant administrative authority, making repeated trips when biometric authentication fails (either due to lack of connectivity or biometric mismatch) and so on. People are routinely illegally charged for enrolment into the UID database.9 In a nutshell, the UID project has ended up excluding some people from existing entitlements and has complicated paperwork for accessing government programmes rather than simplifying it.

The role of social welfare programmes in reducing social inequalities is well recognized. In India, legal entitlements under the National Rural Employment Guarantee Act (NREGA) and the National Food Security Act (NFSA), as well as welfare schemes such as social security pensions and scholarships, have played an important role. The UID project was expected to be an enabling technology which would contribute to this goal by enhancing inclusion in welfare programmes, simplifying access and improving implementation. However, the experience of the past five years with UID has shown that it has not done well as far as enhancing inclusion is concerned. On the contrary, perhaps unintentionally, it has ended up being another source of exclusion from these crucial programmes.

The unintended consequences of the UID project point to a larger concern. It is important to remind ourselves that the UID project continues to operate without any binding legal framework. (The bill prepared by the UPA-II government was rejected by the concerned standing committee.) This opens up the possibility of the project being used in unanticipated and unintended ways.10 If the project is to continue, the very least that must be done immediately is to put an adequate legal framework in place in order to prevent its abuse in the future.



1. The civil registration system has shown significant improvements over the years. According to Census estimates, in 2012, 84% of all births were registered (see http://www.censusindia.gov.in/2011-Documents/Reg_formula/Level_Registration.pdf).

2. This data is based on a reply to a right to information application filed by Ujjainee Sharma.

3. Reetika Khera, ‘The UID Project and Social Welfare’, Economic and Political Weekly 46(9), 2011.

4. Jean Drèze and Reetika Khera, ‘Water for the Leeward India’, Outlook, 24 March 2014. Accessed online on 25 May 2015. http://www.outlookindia.com/article/Water-For-The-Leeward-India/289801

5. Ankita Aggarwal, ‘The Slow Destruction of MGNREGA: Evidence from Jharkhand’, India Together, 16 March 2015. Accessed online on 20 May 2015. http://india together.org/destruction-of-nrega-with-evidence-from-jharkhand-government

6. Aarefa Johari and Mayank Jain, ‘Why Poor People in Delhi are Desperate to Get Their Babies Uniquely Identified.’ Accessed online on 20 May 2015. http://scroll.in/article/709399/Why-poor-people-in-Delhi-are-desperate-to-get-their-babies-uniquely-identified

7. The Pioneer, ‘Aadhaar Now Must for Post-Matric Scholarship.’ Accessed online on 20 May 2015. http://www.dailypioneer.com/state-editions/bhubaneswar/aadhaar-now-must-for-post-matric-scholarship.html

8. The Hindu, ‘Many go Without Ration Card in Agency Due to Mistake in Aadhaar Seeding’, 19 May 2015. Accessed online on 25 May 2015. http://www.thehindu.com/todays-paper/tp-national/tp-andhrapradesh/many-go-without-ration-in-agency-due-to-mistake-in-aadhaar-seeding/article7 222362.ece

9. Ankita Anand and Nachiket Udupa, ‘How to Get Married Without Aadhaar’, The Caravan, May 2015.

10. Jean Drèze, ‘Unique Identity Dilemma’, The Indian Express, 19 March 2015. Accessed online on 25 May 2015. http://indianexpress. com/article/opinion/columns/unique-identity-dilemma/; G. Sampath, ‘Missing the Big Picture on Big Data’, The Hindu, 20 May 2015. Accessed online on 20 May 2015. http://m.thehindu.com/news/national/the-big-data-conundrum/article7224734.ece/