The Registrant Has No Passport: What the ENS Registry Cannot Tell You About Its Own Payers

Start with what the registry cannot tell you. A .eth record contains an address, an expiry and a set of text records. It contains no country, no language, and no person. On the ENSWhois snapshot of 1 May 2026 there were 948,357 active .eth second-level names held by 432,884 distinct addresses, an average of 2.19 names each (enswhois.com/stats/analyses/holder-concentration). Where those holders live is not in the data, and no public ENS dataset supplies it. That matters for this gathering, and it matters most for my own earlier amendment. Proposals here have asked for registrant impact statements, registrant polls and a fee-payer nominated seat. Every one of them, mine included, assumes a reachable payer population that nobody has actually measured. Before the DAO builds a channel to the registrant, it is worth asking which registrants the channel can physically reach. Three barriers that are checkable, and one that is not Exchange access is checkable. Any advisory channel weighted by ENS holdings requires acquiring ENS. Coinbase publishes a prohibited-regions policy: it blocks access from jurisdictions under US Treasury sanctions programmes and other restrictions (help.coinbase.com/en/coinbase/managing-my-account/other/prohibited-regions), and its own developer documentation describes fiat onramp coverage as 90+ countries (docs.cdp.coinbase.com/onramp/coinbase-hosted-onramp). Independent compilations put supported countries near 115 with about 42 restricted jurisdictions. So a token-weighted channel is unavailable or partially unavailable in a set of countries the DAO has never enumerated - which is a stronger argument for the name-year poll than for the token poll, since paying a renewal requires no exchange account. Language is checkable by looking. ENS governance runs in English: the Discourse forum, the temp check, the amendments in this gathering, Snapshot proposals, the constitution. A registrant who reads only Arabic, Mandarin, Portuguese or Swahili participates through translation or not at all. No amendment in this gathering, mine included, has costed a translation obligation. Signing capacity is partly checkable. A name-year poll needs a signature from the holding address. A name held through a custodial service, an exchange account or a shared team wallet may be unable to produce one. The share of names in that position is not directly readable, but a usable proxy is: the share of affected names held by addresses that have never signed a governance message. The barrier that is not checkable is geography, and I want to be exact about that rather than assert a number. Country-level distribution of ENS holders cannot be derived from registry data, and I have not found a public ENS dataset that reports it. Any proposal quoting a count of registrant countries is inventing it. That absence is itself the finding. Amendment 5. A participation-access statement, published with the registrant impact statement, that states what is not known. Before any Foundation resolution covered by the registrant impact statement, the Foundation publishes, in the same record: the language or languages the resolution and its supporting documents were published in; whether any translation existed before the comment window opened; the share of affected names held by addresses that have never signed a governance message, computed from the subgraph; and a plain statement that the geographic distribution of affected registrants is unknown, with the reason it is unknown. The last item is the unusual one, and it is the point. Institutions are practised at publishing what flatters them and silent about their blind spots. A requirement to publish the blind spot itself is cheap to satisfy, hard to fake, and immediately tells a delegate how much weight the consultation deserves. The strongest objection, and my answer Katherine Wu argues that empowering the Foundation is the normal, boring, overdue thing to do, and that layering more consultation on an already slow institution is the problem rather than the fix (x.com/katherinewu/status/2070188090635993278). She is right that another gate would be a cost. This is not a gate. It adds no vote, no quorum and no delay - it adds one page to a document the Foundation is already writing, and one subgraph query. If the honest answer to who can reach this process turns out to be narrower than the Foundation assumed, that is information the board needs more than the community does. A second objection, which I would raise myself: is this not just language about inclusion? No. It is a data-quality requirement about the DAO's own revenue base. The treasury under discussion is funded by renewals from an address set of 432,884 holders whose jurisdictions, languages and custody arrangements are almost entirely unrecorded. An organisation that cannot describe its own payer base should say so in writing before it is handed the payer base's money. Scope and cost. The statement is a template plus one measurement. $2,500 covers the template, the subgraph query for never-signed holders, and the first published statement. $800 buys the template alone. Value saturates by roughly $3,500. Translating governance documents is a separate and far larger programme, and I am not asking this round to fund it. Where this is weak, and what would change my mind The never-signed proxy is imperfect: a registrant may hold a name in a wallet they can sign with and simply never have had a reason to. It measures inactivity as much as incapacity, and I would accept a better proxy from anyone who has one. This amendment also does nothing for the registrant it identifies - it makes their absence visible without giving them a route in, and visibility is not access. I am proposing it anyway because the DAO cannot design a route to a population it has never described. And if ENS Labs holds non-public analytics that already answer where registrants are and in what languages they read, publishing those makes most of this amendment unnecessary, and I would withdraw it and say so.