The Quiet Violence of Digital ID: How Vulnerable Populations Are Locked Out

We like to talk about digital identity systems as if they’re neutral—clean, modern, inevitable. Governments and corporations sell them as tools of inclusion: a way to give the undocumented a legal existence, to make services more efficient, to cut down on fraud. But the architecture of these systems, and the quiet assumptions baked into them, can just as easily become a mechanism of exclusion. For people already living on the margins, a digital ID requirement isn’t a door to opportunity. It’s a wall.

This isn’t a hypothetical worry. It’s a pattern that repeats across continents, across technologies, across populations. When we design systems that demand a smartphone, a stable address, a birth certificate, or even a consistent facial appearance, we’re making choices about who counts as a person. And too often, those choices quietly erase the very people these systems claim to serve.

The Myth of Universal Access

Digital ID systems are often framed as a fix for the “identification gap”—the estimated one billion people worldwide who lack official proof of identity. The logic sounds reasonable: if you don’t have a paper ID, a digital one can be easier to get and harder to lose. But this framing skips over a more basic question: why do people lack ID in the first place?

For many, the absence of identification isn’t a simple administrative oversight. It’s a symptom of deeper structural exclusion. Indigenous communities may never have had their births registered because hospitals are far away or because the state doesn’t really recognize their existence. Refugees may have fled with nothing, their documents destroyed or confiscated. People experiencing homelessness may have lost everything in a fire, a theft, or a bureaucratic loop that requires an address to get an ID and an ID to get an address. Women in patriarchal societies may never have been registered because their fathers or husbands didn’t see the need. For these groups, the barriers to getting a foundational ID aren’t technical—they’re social, economic, and political.

A digital ID system that simply digitizes existing registration processes will reproduce those same barriers, often with higher stakes. If you couldn’t get a paper ID because you lacked a birth certificate, you won’t magically be able to get a digital one. And if that digital ID becomes mandatory for accessing food aid, healthcare, or shelter, your exclusion becomes more absolute.

When Biometrics Become a Barrier

Biometric authentication—fingerprints, iris scans, facial recognition—gets sold as the most secure and inclusive form of identification. After all, everyone has a face, everyone has fingers. But that claim falls apart under scrutiny. Not everyone’s body fits the template.

Manual laborers may have worn-down fingerprints that scanners can’t read. Elderly people often have faint prints because of thinning skin. People with certain disabilities may not be able to position their hands or eyes correctly for a scan. Facial recognition algorithms have well-documented biases across race and gender, misidentifying or failing to recognize people with darker skin, women, and transgender individuals at significantly higher rates. A 2019 U.S. government study found that many commercial algorithms were up to 100 times more likely to misidentify Asian and African American faces compared to white faces. When your access to food or medicine depends on a machine recognizing your face, an algorithmic error isn’t a technical glitch—it’s a denial of your basic needs.

Then there’s the question of consent. Biometric data is intimate and permanent. You can change a password; you can’t change your irises. For vulnerable populations—refugees, detainees, people in crisis—the power imbalance is stark. When a humanitarian agency offers aid conditional on biometric registration, is that consent freely given? Or is it extracted under duress, with no meaningful alternative? The World Food Programme has piloted iris-scanning systems in refugee camps, framing it as a way to prevent fraud and ensure fair distribution. But critics point out that refugees have no real choice: refuse the scan, and your family goes hungry. This isn’t inclusion. It’s coercion by design.

The Digital Divide Is Not Closing

Many digital ID systems assume a baseline of connectivity and device ownership that simply doesn’t exist for the most marginalized. In India, the Aadhaar system—the world’s largest biometric ID program—has been linked to everything from bank accounts to ration cards. But to authenticate yourself, you often need a smartphone, a stable internet connection, and a working biometric scanner at the point of service. In rural areas, where connectivity is patchy and scanners frequently fail, people are turned away from ration shops empty-handed. A 2017 study in Jharkhand found that 30% of ration cardholders faced authentication failures, with the poorest and most remote communities hit hardest.

This isn’t a temporary glitch. It’s a structural feature of systems that digitize access without digitizing infrastructure. When you move a service online, you shift the burden of access onto the individual. The state no longer needs to maintain a physical office in every village; instead, every villager needs a smartphone and a data plan. For those who can’t afford either, the service effectively disappears.

Digital ID systems also often require a fixed address, a phone number, or an email account—things that homeless individuals, transient workers, and people fleeing domestic violence may not have or may need to change frequently. A system designed for stability punishes instability. And the people whose lives are most unstable are precisely those who most need the protections that ID is supposed to provide.

Administrative Violence and the Error Loop

There’s a term that scholars of bureaucracy use: administrative violence. It describes the harm done not by malice, but by systems that are indifferent to individual circumstances. A digital ID system that flags a mismatch and locks an account without a clear appeals process is committing administrative violence. A system that requires you to travel to a distant center to resolve an error, but you can’t afford the bus fare, is committing administrative violence. A system that suspends your benefits because your name in the database doesn’t exactly match the name on your lease—a common issue for people who use patronymics, have multiple last names, or whose names were transliterated inconsistently—is committing administrative violence.

These errors aren’t rare. They’re the normal operation of rigid systems encountering messy human realities. And the consequences cascade. A denied ration leads to hunger. A suspended health ID leads to delayed treatment. A frozen bank account leads to financial ruin. For people already living on the edge, a single administrative error can be catastrophic.

The appeals process, where it exists, is often inaccessible to the very people who need it most. It may require online submission, legal documentation, or travel to a government office—all barriers for someone who is poor, disabled, or geographically isolated. The system is designed for the average user, but the average user doesn’t exist. We’re all, at some point in our lives, outliers: when we’re sick, when we’re displaced, when we’re grieving and can’t think straight. A humane system would account for this. Most digital ID systems do not.

Surveillance and the Chilling Effect

Even when digital ID systems function technically, they can exclude through fear. When every transaction—buying food, seeing a doctor, crossing a border—is logged and linked to a unique identifier, the state gains an unprecedented ability to track individuals. For undocumented migrants, this creates a terrifying dilemma: register and risk deportation, or stay invisible and forgo essential services. For political dissidents, registering means creating a permanent record of their existence that can be used to target them. For women fleeing abusive partners, a digital trail can lead violence straight to their door.

This chilling effect isn’t hypothetical. In Kenya, the government’s attempt to create a centralized digital ID system, Huduma Namba, faced fierce opposition from civil society groups who warned it would be used to target minorities and dissidents. The system was eventually blocked by the courts on privacy grounds, but the fear it generated among marginalized communities was real and lasting. People avoided hospitals, schools, and aid distribution points because they didn’t want to be registered. The ID system, intended to include, became a reason to hide.

Surveillance also changes the relationship between citizen and state. When you know you’re being watched, you self-censor. You avoid certain places, certain services, certain behaviors. This isn’t freedom. A digital ID system that tracks every interaction creates a panopticon, and the people most vulnerable to state power—minorities, activists, the poor—are the ones who will feel its weight most heavily.

What Would a Truly Inclusive System Look Like?

If we’re serious about inclusion, we need to design ID systems from the margins, not from the center. That means starting with the people who are hardest to reach and building outward, rather than designing for the median user and then patching exceptions. It means accepting that no single system will work for everyone, and that flexibility isn’t a flaw but a requirement.

First, any digital ID system must have a dependable, accessible, and free offline fallback. If the biometric scanner fails, there must be a human being authorized to verify identity through other means—a photograph, a witness, a community-based attestation. This fallback can’t be a degraded, second-class option; it must carry the same legal weight and grant the same access. Otherwise, we create a two-tier system where the digitally authenticated get full rights and the rest get scraps.

Second, enrollment must be proactive and mobile. Waiting for people to come to a government office is a recipe for excluding those who can’t travel, can’t take time off work, or can’t navigate bureaucratic spaces. Enrollment teams must go to remote villages, homeless shelters, refugee camps, and hospitals. They must accept alternative forms of identity proof: sworn affidavits, community leader verification, NGO records. The burden of proof must be lowered, not raised.

Third, systems must be designed to handle exceptions gracefully. A name mismatch should trigger a human review, not an automatic denial. A lost phone shouldn’t mean a lost identity. There should be multiple recovery pathways, including in-person options that don’t require original documents. The system should default to trust, not suspicion.

Fourth, privacy must be built into the architecture. Data minimization is essential: collect only what’s necessary, store it only as long as needed, and share it only with explicit consent. Individuals must have the right to know what data is held about them, to correct it, and to delete it. Surveillance features—location tracking, transaction logging, biometric databases—should be opt-in, not mandatory, and never a condition of accessing basic services.

Finally, we need independent oversight. No government or corporation should be the sole arbiter of how an ID system operates. Civil society, privacy experts, and representatives of marginalized communities must have a seat at the table—not as advisors, but as decision-makers with real power to veto harmful features. Transparency reports, algorithmic audits, and accessible complaint mechanisms should be mandatory and public.

The Cost of Getting It Wrong

The stakes aren’t abstract. In 2021, a report from Human Rights Watch documented how Kenya’s digital ID system had left thousands of people in the arid northern regions unable to access food aid during a severe drought. Their biometric data had been enrolled incorrectly, or their cards had been lost, and there was no functioning appeals process. People died. This isn’t a system failure; it’s a system functioning exactly as designed—efficiently excluding those it wasn’t built to serve.

In India, Aadhaar-linked ration systems have been linked to starvation deaths. In 2017, a 65-year-old woman in Jharkhand died after being repeatedly denied rations because the biometric scanner couldn’t read her fingerprints. Her family said she had been surviving on wild leaves. In Rajasthan, an 11-year-old girl died of hunger after her family’s ration card was canceled due to an Aadhaar authentication error. These aren’t isolated incidents. They’re the logical endpoint of a system that prioritizes technological integrity over human life.

We have to ask ourselves: what problem are we trying to solve? If the goal is to prevent fraud, we need to weigh the cost of that fraud against the cost of exclusion. In India, the government claimed Aadhaar would save billions by eliminating duplicate and ghost beneficiaries. But independent researchers found that the savings were vastly overstated, while the human costs—people denied food, wages, and healthcare—were systematically ignored. A system that saves money by starving people isn’t efficient. It’s cruel.

Frequently Asked Questions

Why are digital ID systems considered more inclusive than paper-based systems?

Proponents argue that digital IDs can reach people who lack paper documentation because they can use biometric data—like fingerprints or iris scans—to establish uniqueness without needing a birth certificate or other foundational documents. They also claim that digital systems are harder to lose, destroy, or forge, and that they can enable remote authentication, reducing the need to travel to government offices. However, these benefits depend on infrastructure, accessibility, and design choices that are often absent for the most marginalized populations.

What specific groups are most at risk of exclusion from digital ID systems?

The groups most at risk include homeless individuals, undocumented migrants, refugees, people with disabilities, elderly people, manual laborers, indigenous communities, women in patriarchal societies, and people living in remote areas with poor connectivity. Each of these groups faces distinct barriers: lack of a fixed address, inability to provide standard identity documents, physical characteristics that biometric scanners cannot read, or fear that registration will expose them to state surveillance or violence.

Can digital ID systems be designed to avoid these exclusion risks?

Yes, but it requires a fundamental shift in design philosophy. Systems must be built with the most vulnerable users in mind, not the average user. This means providing offline fallbacks, accepting alternative forms of identity verification, ensuring human oversight and accessible appeals processes, and embedding strong privacy protections. It also means involving marginalized communities in the design and governance of these systems from the outset, rather than treating them as an afterthought.

What can individuals do to advocate for more inclusive ID systems?

Individuals can support civil society organizations that monitor and challenge exclusionary ID practices, such as the Electronic Frontier Foundation or local digital rights groups. They can also push their governments to adopt privacy-protective and inclusive design standards, and to conduct independent human rights impact assessments before deploying new systems. At a community level, people can document and share stories of exclusion to make the human costs visible—because data points don’t generate empathy, but stories do.

Digital ID isn’t going away. The push toward digitization is backed by powerful interests: governments seeking control, corporations seeking profit, international agencies seeking efficiency. But we’re not powerless. We can demand that these systems be built on a foundation of dignity, not surveillance. We can insist that no one be forced to choose between their privacy and their survival. And we can remember, always, that the measure of a system isn’t how well it serves the majority, but how it treats the people it’s most likely to leave behind.