The register underneath the biometrics

There is a familiar sequence in national identity programmes. A government secures financing, procures an enrolment platform, runs a deduplication campaign, and issues credentials. Somewhere in parallel, and usually with a fraction of the budget and attention, a civil registration modernisation project moves at its own pace. The two are described as complementary. They are rarely designed as one system.

Vital Strategies has published two case studies, on Rwanda and Thailand, that put a name to the cost of that separation. Their argument is that birth and death registration should be treated as foundational digital public infrastructure rather than as a standalone administrative function, and that digitising civil registration on its own is no longer sufficient. What creates value is the link: connecting civil registration to the national identity system produces a population register that other government systems can actually trust.

The failure mode is undramatic, which is why it is so often tolerated. When the civil register and the identity database are not properly bound, they diverge. A birth recorded in one is absent from the other. A death registered locally leaves an active identity record in the national system. Names, dates and parentage differ between two official sources, each of which believes itself authoritative. None of this announces itself on the day it happens. It surfaces years later, at the counter, when a citizen is refused a service because two arms of the same state disagree about who they are.

The observation that should travel furthest is that no amount of biometric sophistication compensates for this. Accuracy in matching answers a narrow question: is this the same person as the one enrolled. It says nothing about whether the enrolment was correct, whether the person is still alive, or whether the record reflects a registered legal event. A perfectly performing algorithm sitting on a weak register will confirm bad data very reliably.

Rwanda’s case is instructive because the sequencing was deliberate. Digitised birth registration was integrated with the national identity system, which enabled real time verification and data sharing across sectors that had previously operated in isolation. The technology is not the interesting part. The institutional decision to treat one register as the source of truth, and to wire the others to it, is.

For a government team currently designing or reviewing a programme, this converts into a small number of uncomfortable questions. Which institution owns the population register, and is that mandate written into law rather than assumed. What is the documented reconciliation procedure when civil registration and the identity system disagree about an individual. Is civil registration a funded, scheduled component of the identity programme, with its own deliverables, or is it the phase that begins once the cards have been issued.

Answering those questions costs very little at design stage. Answering them after ten million credentials have been issued costs a great deal.

We made a version of this argument in June, tracing the distance between the ghost children reporting of seven years ago and Benin’s C’est moi card, where the principle that mattered was not any particular platform but the unique identifier assigned from birth. The platform debate absorbs most of the oxygen in this field. The register question decides whether the platform was worth buying.

Source: Birth registration emerges as foundation for digital public infrastructure, Biometric Update, 13 August 2026.

Want to comment, ask a question, or join the discussion? Continue the conversation on LinkedIn.