Oregon

The registration table carries both columns side by side: Non-Affiliated at 1,142,360 and Independent at 153,045. The first means no party; the second is the Independent Party of Oregon, a registered party with members, ballot access and its own nominations. They differ by a factor of seven and mean opposite things. Every schema that maps party strings to a canonical vocabulary, every minor-party rollup, and every assumption that independent implies unaffiliated will collapse them.

Non-Affiliated is 37.1% of the state, ahead of Democrats at 31.9% and Republicans at 23.9%. Oregon registers voters automatically through driver licensing and records them with no party unless they choose one, so the plurality is produced by a default rather than a preference. A model that treats party registration as expressed intent, or that estimates support from registration shares, is reading a system artefact as a political fact for more than a third of the electorate.

Beyond the majors the table carries Constitution, Independent, Libertarian, No Labels, Green, Progressive, People, Families and Other as separate columns — nine of them, with People holding 446 registrations statewide and Families 8,362. Collapsing these into a single other bucket destroys the distinction between a party with 446 members and the residual category, and Other is itself already one of the columns, so the rollup would be nested inside a column of the same name.

The registration report is published by county and again by congressional district, and the district tables list counties as their components — the 1st is Clatsop, Columbia, part of Multnomah, and Tillamook. Multnomah shows 124,234 registrations inside the 1st against 583,460 countywide, so most of it lies in other districts. Any join that treats a county as belonging to one congressional district, or that sums district registration by matching county names, double-counts the split counties and misplaces the rest.

The non-affiliated share runs from Malheur at 49.1% and Umatilla at 46.3% down to Wheeler at 27.0% and Wallowa at 29.6% — a 22-point spread, and the extremes are rural counties at both ends rather than an urban-rural gradient. Applying the statewide 37.1% per county to size a partisan universe is wrong by up to twelve points, and the direction of the error is not predictable from population.

The Census counts 1,528 local governments in Oregon: 1,029 special districts, 223 school districts and only 240 municipalities. Districts outnumber cities more than four to one, and special and school districts together are 82% of all local government. Modelling the local layer as cities captures under a sixth of the state's governments, and the district layer carries its own elections that no municipal harvest reaches.

The Secretary of State publishes a registration report every month, and each one carries this month, last month, the same month a year earlier, and both percentage changes on the same page. So registration is a time series rather than a snapshot, and a figure quoted without its month is not interpretable — the August 2026 total of 3,076,893 is up 0.22% on July and 0.77% on August 2025. A pipeline that stores one number and refreshes it in place destroys the only view that shows movement.

Oregon administers 36 counties and the Census counts 36 county governments. Across this corpus the two disagree only where consolidated city-counties exist — Colorado, Montana, Nevada and California — and agree everywhere else. Oregon is a clean case, so a reconciliation against Census unit counts is safe here, but the same reconciliation is short by one to two in four neighbours and the difference is knowable in advance rather than discovered.