Last month saw the publication of a report on the World Bank’s ill-conceived approach to digital ID, described as “essential reading for all concerned about human rights and development” by former UN Special Rapporteur on Extreme Poverty and Human Rights Philip Alston. As the press release summarizes:
“Governments around the world have been investing heavily in digital identification systems, often with biometric components (digital ID). The rapid proliferation of such systems is driven by a new development consensus, packaged and promoted by key global actors like the World Bank, but also by governments, foundations, vendors, and consulting firms. This new ‘manufactured consensus’ holds that digital ID can contribute to inclusive and sustainable development—and is even a prerequisite for the realization of human rights.”
The report argues that India’s digital identification system has been central to the formation and promotion of this consensus. This has also been increasingly clear to me in my experience as an economist and identity management consultant who has provided advisory services to the World Bank. For the World Bank—and particularly its Identification for Development (ID4D) cross-sectoral practice—the Indian system, named Aadhaar, has become the singular answer to development and a key source of inspiration. This continues irrespective of the body of evidence which shows how poorly a “fit” the Aadhaar system is for identity management in India, and even more so elsewhere. Aadhaar represents a mirage: it is not evidencing the universality, inclusiveness, unprecedented enrollment speed, meaningful legal identity, nor accuracy that it is claimed to represent.
The World Bank’s own data on the completeness of ID systems displays the “20/80-rule”: the overwhelming odds are that digital ID systems not building on a functional civil registration system (in which births, deaths, marriages and so forth are recorded) will exclude 20% or more of (mostly vulnerable) people, or they will take at least 80 years to cover all. Many developing countries often abandon underperforming ID-systems obtained at great cost, only to launch new and even more sophisticated systems. Instead of using existing service infrastructure for civil registration, new digital ID systems are rolled out through a quick fix “mobile campaign,” held once or twice, with mobile enrollment kits and temporary enrollment staff. But this invariably leaves a coverage and service void behind.
But what about Aadhaar, then? Hasn’t Aadhaar enrolled almost all of the Indian population (1.29 billion by March 2021, out of 1.39 billion), in just a decade (from September 2010), at minimal cost (USD $1.60/enrollment)? If one believes the data from the Unique Identification Authority of India (UIDAI), then yes. But independent data are unavailable; UIDAI controls the message—even the Comptroller and Auditor General of India (CAG) had to use UIDAI data for its first ever audit of Aadhaar. Still, CAG found that UIDAI’s operational and financial management have been utterly deficient. Claims about Aadhaar’s impressive coverage and universality might, then, be questionable. Neither is the database accurate: the Aadhaar system has no way of weeding out dead enrollees (about 80 million in 10 years) or people leaving India (including Indian citizens). CAG also found UIDAI’s digital archiving and its collection and storage of the physical documents that back up enrollments to be inadequate.
Furthermore, claims about the uniqueness guaranteed by biometric technologies within Aadhaar are also illusory. There is no uniqueness for the approximately 25 million children under five years old enrolled in the database. Multiple Aadhaars were issued to the same persons, while different Aadhaar numbers associated with the same biometric data were issued to multiple people. Fingerprint authentication success for 2020-21 was only (an unverifiable) 74-76%. This may well be the canary in the coalmine, indicating exaggerated coverage claims for Aadhaar. Indeed, a Privacy International study explains the very statistical impossibility of a unique biometric profile in a population of 1.39 billion people. Rather, each Indian person has an average of 17,500 indistinguishable biometric “doubles.”
These claims about the benefits of biometrics have far-reaching implications as Aadhaar is linked to other areas of governance. A new law provides for the use of Aadhaar to verify the electoral roll. Weeding out “ghost entries” when the uniqueness and de-duplicated nature of the Aadhaar database is disproved is a doomed exercise, and represents another potential threat to India’s democracy.
Aadhaar’s “big numbers” are a mirage too. Proponents claim that over a billion were newly enrolled at record speed at low cost. But this is not as unprecedented as is suggested. For elections in India, 900 million voters are registered or verified every five years—which tops Aadhaar’s enrollment accomplishment. And India’s bureaucracy has long provided multiple forms of documentation; for proof of identity, date of birth, and address, enrollees can choose from a menu of no less than 106 valid documents. Less than 3 in 10,000 enrollees lacked valid ID prior to Aadhaar enrollment by 2016. The Aadhaar system is a duplication which simply adds on biometrics—which, as we saw, are not the holy grail they are claimed to be. To suggest that other countries, which do not have this multitude of breeder documents and existing enrollment capacities, can copy the Aadhaar approach and obtain widespread coverage, is an illusion.
In respect of claims that Aadhaar brings down costs and increases efficiencies: these costs are applicable only in India. I have found that digital ID systems in many African countries cost 5 to 10 times more per capita than India’s ID system. The high failure rates of ID-systems in many developing countries add to the unbearable costs for poorer countries and their more vulnerable people.
This cries out for a better identity management model—one that is centered around citizenship, with civil registration as the foundation, which seeks to guarantee rights. A model closer to northern European identity management systems comes to mind, or one that is already in use in South Africa. Such systems stand in contrast with Aadhaar, which seeks to side-step the “pesky political issue” of citizenship. This is perhaps the most serious and dangerous element of the mirage: Aadhaar only provides an “economic identity” (with rights limited to government hand-outs, and “voluntary” use for private services), which aims to facilitate economic transactions and private sector service delivery. The UIDAI, then, insists that Aadhaar has “nothing to do with the citizenship issue.”
But Aadhaar’s “citizenship-blindness” is make-believe. Enrollment into Aadhaar was selective in Assam state, for example, where the issuance of digital ID was linked to citizenship determinations. Suddenly, Aadhaar proved to be an exclusionary “citizenship ID” after all. Aadhaar has dangerously played into worrying trends, such as the Citizenship Amendment Act and widespread lack of proof of citizenship—all while proponents claim that it is a model of how to achieve “legal identity for all.”
Aadhaar proves to be a mirage that we see while traveling on “the road to hell,” which is paved with imaginary intentions and is leading to a deadly development destination. Its presentation as a “model” digital ID system should be urgently reconsidered.
