Great Score. Proud Government. Angry Citizens. What Exactly Is OECD Measuring?

This does not prove the ranking is wrong. Convergence between a capability score and a human outcome can exist, and in some countries it may. But convergence has to be demonstrated, not inferred from the label. When a ranking calls itself "User-Driven", a reader will reasonably assume it says something about users. On the evidence, it says something about governments: what they have built, what policies they have adopted, what mechanisms exist on paper to consult citizens. Those are capabilities. They are useful to know. They are not the experience.

It is worth being precise about what kind of problem this is, because it is tempting to reach for the wrong one. Nobody here is alleging that Spain, or Portugal, or anyone else, gamed the OECD. There is no suggestion of manipulation or bad faith, and none is needed. The more accurate and, frankly, more uncomfortable point is architectural: the benchmark can be satisfied by capability and process evidence without independently proving the complete customer outcome that its label invites readers to infer. A government can answer every question honestly, score brilliantly, and still preside over a journey that leaves its citizens exhausted. The instrument permits it.

The customer does not live in the process steps

Consider how a public service is usually assessed. A digital identity exists: tick. An online form exists: tick. The form is mobile-friendly, accessible, available in two languages: tick, tick, tick. Each step, examined on its own, is a success. Then a real person tries to register a business, or renew a residence permit, and discovers that step three requires a document only issued after step five, that the appointment system has no appointments, and that the helpline refers them back to the website. Every component passed. The journey failed. The customer does not live in the process steps. The customer lives in the whole thing, from the moment they need something to the moment they have it — including the parts that fall between agencies and therefore between everyone's metrics.

Some will point to the European Commission's 2025 eGovernment Benchmark as the answer. It is a genuinely different and more demanding layer of measurement: mystery shoppers work through roughly one hundred key public services across nine life events and record what they find. That is a real improvement over self-reported capability, and where its results diverge from the OECD's, the divergence is instructive. But it, too, tests defined services against defined criteria. It shows whether a service can be completed by a trained tester. It does not, and does not claim to, establish the full end-to-end human outcome for an ordinary person with an unusual case, a missing document, and a deadline. Use it as contrast. Do not mistake it for the proof that is missing.

Why this is more than an academic quibble

Authoritative rankings do work in the world. A top User-Driven score becomes a line in a ministerial speech, a slide in an investment-promotion deck, a data point in a consultancy's country risk assessment. It shapes how a government sees itself and how businesses judge a country's operating environment. That is where the label starts to cost money. A company choosing where to open an office may read "User-Driven: 0.94" as a proxy for low bureaucracy and smooth cross-agency journeys. It is nothing of the sort. Businesses should treat a high User-Driven score as evidence of digital capability and nothing more; independent proof that a country is easy to deal with must come from somewhere else — ideally from people who have actually tried.

There is a public-money question, too, and it is not a small one. International comparisons of this kind are funded, directly or indirectly, by member states — that is, by the public. They carry political weight, reputational weight and, as argued above, potentially economic weight. If public resources produce a ranking with those effects, the least one can ask is that the evidence supports what the ranking appears to mean. At present, for the User-Driven dimension, it does not.

What the OECD said

DYRÍGO put its material questions to the OECD and gave it the opportunity to respond before publication. No substantive evidence was received that changes the conclusion of this analysis. That should be stated carefully: silence is not proof that DYRÍGO is right. It is proof only that the problem remains unanswered, which for a ranking that governments quote with such confidence ought to be uncomfortable enough.

Conclusion

The OECD Digital Government Index contains genuinely useful capability data, and nothing here argues for throwing it away. It argues for honesty about what the data can bear. Capability should be reported as capability. Customer outcomes should be measured independently, end to end, by people who are not the government being measured. And the current standalone, performance-like User-Driven country ranking, which invites readers to infer an outcome it does not establish, should be discontinued in its present form. Three sentences will do.