A full accounting of how voices are collected, verified, geocoded, and aggregated — and what we do and do not claim about the data.
Voices are collected through the Landocracy intake form at /join. The form asks five questions: why the land matters to the respondent (multi-select), location (city, state, ZIP), primary conservation concern (single select from 8 categories), an optional place description (free text), and contact details (first name and email address).
The form takes approximately 90 seconds to complete. It is available in English only at launch.
Upon submission, the respondent receives a confirmation email from Landocracy. The voice is not counted in the census until the respondent clicks the confirmation link. This email confirmation step prevents form spam and ensures that each registered voice corresponds to a real email address.
We do not verify that an email address is associated with a real person beyond the confirmation click. We do not require government ID or any additional verification.
ZIP codes are mapped to state legislative districts (state house and state senate) and U.S. congressional districts using publicly available TIGER/Line shapefiles from the U.S. Census Bureau, updated at each redistricting cycle.
We use ZIP code as the primary geocoding input because it is the location data respondents most reliably know. Where a ZIP code spans multiple districts, we assign to the district containing the majority of the ZIP code's population.
Geocoding accuracy is bounded by ZIP code resolution. We do not claim precinct-level or census-tract-level precision.
Public census data is aggregated: we display total voice counts and state-level breakdowns. District-level data is available only in structured briefings to verified legislators and, in anonymized form, to approved conservation organizations.
Minimum cohort size: we do not report any geographic or demographic breakdown with fewer than 50 voices, to prevent re-identification of respondents in low-density areas.
We claim: Landocracy represents a verifiable, geocoded record of expressed conservation concern, filtered to confirmed email addresses.
We do not claim: statistical representativeness of any geography's general population, voter registration status of any respondent, or that registered voices constitute a majority opinion.
Landocracy is a record of people who chose to register — not a random sample or a poll. It should be interpreted accordingly.
Landocracy is designed as a living census, not a one-time snapshot. A voice entered in 2026 should still reflect a real, current person in 2030. To support this, each registered voice carries a last_confirmed_at timestamp that records when the respondent most recently affirmed their registration.
On the anniversary of each registration, we send an annual confirmation email with a signed, time-limited link. The email asks the respondent to confirm their voice is still current, update their response, or remove their entry from the census. Voices that go five years without any confirmation are flagged as unverified in our internal data and excluded from briefings delivered to legislators and organizations — though they remain in the public count with a transparency disclosure.
Confirmation rates and the methodology used to handle lapsed registrations are reported in our annual transparency report. We believe a smaller, verified census is more valuable than a larger, stale one.
Methodology v1.0 — March 2026. This document will be updated if collection methods, verification procedures, or geocoding sources change.
Questions about methodology? Contact us. See also: Data Pledge.