Empowered Vote · Volunteer Briefing · October 13, 2026
Every number on this page comes from the live database, verified git history, or our product analytics, not from estimates. We build tools for shared facts; our progress report should be one. Each briefing is written fresh and kept at its own dated address, so you can read what we said before and check it against what we say now.
This is a three-day cycle, and almost all of it is Treasury. 8,925 governments now have budget data, up 778 — 61 New York cities outside New York City, and 717 New York school districts covering fiscal years 1995 to 2025. The school districts alone carry 82,564 budgets and 2.4 million line items, which is why the line-item total went from 2.7 million to 5.3 million over a weekend. A school district is not a city, a county or a state; it is a single-purpose government that taxes and spends, and until this weekend this platform held none.
Nothing announced them. A catch-all bucket did. The entity breakdown further down this page names seven classes and sweeps everything else into an “other” row that is computed, never typed. That row went from 17 to 734, and reading it is the only reason anyone knew a new class of government had arrived. This is the second time that bucket has earned its keep: it was added in September after a four-class list printed 3,502 of 7,507 entities as though it were the whole roster. A breakdown that cannot surprise you is not measuring anything.
The other half of the report is that the stance line stopped dead. Not one row has been written to the stance table since Friday the 9th — the headline above is the same 32,000 it was then, to the position. Last edition closed a review queue and called it paid off once rather than solved; three days later nothing has refilled it. A flat number is a finding, and this one is the whole state of the research programme this cycle.
Each update is a dated snapshot. We do not edit one after publishing it.
This page is the current briefing, and it always lives at empowered.vote/briefing. Every edition also keeps a permanent address of its own, so a number you read here can be checked against the date it was published on. There is a picker at the top of this page that jumps to any of them. Each update is written from scratch: work that runs for a month stays the headline, but the sentences around it are rewritten every time. We never edit a published edition. When an earlier one got something wrong, the next one says what it corrects.
What the edition of October 10 got wrong, and this one corrects. It said of Civic Trivia that “the slot-C fix has not landed” and that it “needs a week of someone’s attention and has not had it”. The fix had landed ten days earlier. It was merged to the trivia repository’s published branch on September 30 and is wired into every path that writes a question — the replacement cron, both election generators, the seed and the generation prompts. The claim was written from the cycle’s commit list, which did not contain it, instead of from the branch, which did. The October 10 page stays exactly as published; the correction lives here. What the fix actually did is in the Inform table below, and it is not a clean win either.
It also set a test, and the test has finally moved. The trivia row had returned the same twelve people on the same nineteen game starts for three pulls running. It now reads 13 people on 20 starts. That is a movement of one person on each side of the ratio, which at this volume means the stall has ended and nothing more. We are not calling it a recovery.
Open to everyone, no login. Evidence-based, engaging, anti-partisan.
| Feature | Status | Journey | Where it stands | What's next |
|---|---|---|---|---|
| Civic Trivia Championships | Live | alpha |
Where it standsThe answer-slot bias is gone from the questions a player can actually be served. Across the 3,283 live questions in 43 active collections the correct answer sits in slot A 25.1% of the time, B 26.0%, C 24.8%, D 24.1% — even, where one collection once answered A every single time. The skew survives where the questions do not get served: 34.1% slot C among archived questions and 50.5% among expired ones. That shape says the content audit did most of this work by retiring bad batches, not the generator fix — and the arithmetic agrees: 3,220 of the 3,283 live questions were written before the fix existed, and they are already even. 21 commits, most of them a new browser-level test lane that arrived “six defects late” by its own commit message. | What’s nextThe fix is in and it is not finished. Of the 141 questions generated since it landed, 68.1% sit exactly where that question’s own identifier says they should — and 39.7% still land in slot C, above the 35% gate. Part of that is deliberate: the module sorts four comparable numbers by value instead of permuting them, so a numeric question’s slot is its value rank, and models still bracket the true answer. Separating “working as designed” from “still broken” needs the repository’s own audit script, not a database query, and that is the next job. The content audit still owes one collection that has to be written rather than read. |
| Empowered Compass | Live | alpha |
Where it standsEvery question a voter can answer now carries written guidance underneath it — all 60 of them, checked one by one. The library holds 64 topics; the 60 the open season serves are exactly the 60 that carry guidance. The calibration page was rebuilt around it: it names which option changed and says so in words rather than printing a number, lets a reader pin a lens, and shows what the officials said beside the question. First-time users get an onboarding overlay. Compare was removed from the Compass entirely and that traffic now funnels to Essentials, which is the app built to answer it. | What’s nextSeason 2 stays open, Season 1 stays closed, and Season 3 is still a draft holding the same 17 answers it held on Friday — checked, because a season opening is a launch that no other figure announces. The audit’s tail bands are untouched this cycle. 24 rungs are still known to ask a genuinely different question in Season 2 than in Season 1, and none can be carried across without a ruling. |
| Treasury Tracker | Live | alpha |
Where it standsThe largest expansion by entity count this platform has done: 778 governments in three days. 61 New York cities outside NYC, then all 717 New York school districts, FY1995 to FY2025 — 82,564 budgets across an operating and a revenue scope. Every one of them landed with a real provenance grade rather than an unknown: the “unknown audit grade” pile this page named last cycle sits unmoved at 38,491 while the database grew by 82,564 budgets. The sunburst and icicle charts got labels, real font metrics, and a phone treatment that stops treating a picture as a control. | What’s nextThe oldest debt on this page is now four editions old and still unpaid: nobody has timed the real page, five runs before and after, so three cycles of performance work remain unproven. A faster call is not a faster page until the page is timed. Beyond that: school-district budgets are a different shape from a city’s, and the comparison gate that refuses to set cash beside GAAP has not yet been asked whether it should refuse to set a school district beside a city. |
| Empowered Essentials | Live Alpha | alpha |
Where it standsZIP-to-representatives with funding, votes and stances. One commit this cycle, and it is a labelling honesty fix: outside spending now says whether it supports or opposes a candidate, instead of showing a total that silently mixed the two. That is the reader-facing half of a change written down as an architecture decision on the platform side the same day. 12,226 curated officials and candidates are served, beside a 77,708-record discovery pool still in triage. | What’s nextStill owed, and now unchanged for a fourth cycle: portraits for the 70 Puerto Rico alcaldes without one. That figure is exact — 78 municipios carry a mayor, 70 of them have no default image, and the island’s other 83 seats all do. Compass is about to send it traffic, now that compare has moved here. |
| Read & Rank | In Development | build |
Where it standsA blind taste test for candidates: the question, not the race, is the unit of comparison. No commits this cycle, after 48 last cycle rebuilt the app around its redesign. This is the first window that contains the new landing page rather than ending just before it. | What’s nextThe number that had been stuck moved by one. 9 people revealed a ballot in the quarter against 20 race starts, where it was 8 on 19. It remains a start problem rather than a finish problem, and it is still gated on a stance pool that gained nothing this cycle. One quiet window is not a verdict on the redesign; the next pull is the first that could be. |
| Fallacy Finders | In Development | build |
Where it standsv1.1 is 4 of 6 phases done and built the Fallacy Sleuth: curation-time AI proposing timecoded fallacy tags, with names stripped before the model sees a word and every clip re-run with the speakers swapped. Quiet for a seventh cycle: no commits at all. It is still the one repository on this page with no remote — so its state can only be read from the machine this report is written on, while everything else here is read from a published branch. Last edition called giving it one “the smallest open task on this list”. It was not done. | What’s nextIts dependency is quiet too this cycle: On the Record shipped no code, though three more meetings were ingested. Clip sourcing at volume still depends on transcripts the quality gate will accept. |
Connected accounts, one human one voice. Parked during the election push; resumes after November.
| Feature | Status | Journey | Where it stands | What's next |
|---|---|---|---|---|
| Civic Spaces | Live | alpha |
Where it standsGeo-scoped communities, neighborhood to federal. No commits this cycle, after three consecutive busy weeks in a parked pillar that produced the shared banner library, the compact location banner and the dashboard. A parked pillar going quiet is the plan working, not a stall. | What’s nextIt stays parked until after November. The groundwork laid early is the thing v4.0 will be built on. |
| Empowered Quests | Live | alpha |
Where it standsFact-verification quests with consensus scoring; verified facts already feed Essentials. Red gem quests unlock at VR 90+. No commits this cycle. | What’s nextv1.6 planning. |
| Focused Communities | Paused | build |
Where it standsv1.0 shipped April 17: per-topic forums wired to the Compass. Waiting on the post-election Connect pivot. | What’s nextv1.1: badge ratification and argument mapping. |
| Empowered Listening | Paused | build |
Where it standsStructured debate infrastructure; 5 of 6 phases done — timers, transcription, audience notes. | What’s nextPhase 6: voting and badges. Then Common Grounds can build on it. |
| Emparks | Design | design |
Where it standsPlatform-wide annotation system, five mark types. Full design doc; no build yet. | What’s nextAwaits Connect-pillar capacity. |
| Common Grounds | Design | design |
Where it standsAdversarial-collaboration mediation. Full design doc; depends on Empowered Listening and Fallacy Finders arbiters. | What’s nextBuilds after its dependencies land. |
The shared spine every app plugs into.
Auth, tiers, XP and gems, politician database, TIGER geofencing, elections. 27 commits, and most of them are an apology to the campaign-finance data. A sweep through the California filings detached committees that had been attached to the wrong person by a surname match: 599 foreign committees let go and 354 kept on the starts-with-surname shape alone, then 204 off 44 live politicians, 103 off seven more, 63 off inactive records, and 9 re-attached to their real owners. 2,260 federal politicians now carry finance records. One of those corrections broke four live pages — an appended note made the stored JSON unparseable and the campaign-finance tab returned a server error until it was fixed the same day. The LegiScan importer was also rebuilt to read a dataset a slice at a time rather than a whole session at once, after last cycle’s 8.4 million-vote load showed it could not do it twice.
A version number has been dropped from this heading. Previous editions printed “v2.23”; nothing in the repository produces that string, and a label this page cannot derive does not belong on it.
Speaker-attributed, searchable transcripts of public meetings, synced to the original video. No commits this cycle, and the dataset grew anyway — which is the normal shape for an ingest pipeline between code changes. It reads 180 meetings, 21,755 transcript segments and 1,442 identified speakers, including LA City Council with 7,200 recorded votes: three new meetings, 39 new segments and 10 new speakers since Friday. The standing caution is repeated rather than retired: segments carry no creation timestamp, so dated meetings are the evidence here — 20 of them in the last 30 days — and the two drops earlier this year are still unexplained. Built by volunteer Chris Andrews.
The invisible work. This is what "evidence-based" costs, and we paid it.
Nothing in this table moved, and the table reproducing exactly is the point of running it. It partitions stored rows rather than served positions, so it reconciles against the row count: 38,311 rows across 4,228 politicians and 64 topics; 1,846 of the 2026 candidates have some. 32,000 of them resolve to a position a voter is served, held by 4,208 politicians at 7.6 topics each, with 1,629 standing withdrawn. One figure keeps its date said out loud: 2,722 sources were independently re-checked in a sweep that finished in April 2026, and the number has not moved since, because every row in it was written then. It is a dated achievement, not a running total.
| Coverage area | Politicians | Stances |
|---|---|---|
| US Congress + national | 556 | 9,694 |
| 2026 candidates & discovered names (no seat) | 931 | 6,206 |
| California (state + local) | 373 | 5,048 |
| Massachusetts (state + local) | 302 | 3,887 |
| Texas (state + local) | 270 | 2,412 |
| Maryland (state + local) | 188 | 2,316 |
| Maine (state + local) | 186 | 1,796 |
| Virginia (state + local) | 166 | 1,009 |
| Oregon (state + local + county) | 154 | 826 |
| Utah (state + city + county) | 57 | 682 |
| Wisconsin (state + county + local) | 208 | 629 |
| Washington (state + county + city) | 168 | 502 |
| Arizona (state + local) | 129 | 438 |
| North Carolina (state + local) | 128 | 411 |
| Tennessee (state + local) | 119 | 231 |
| 38 more jurisdictions (incl. Florida and DC) | 293 | 2,224 |
All sixteen rows came back identical, and both totals still agree at 4,228 politicians and 38,311 stances. That is the strongest form of this check: when every named row reproduces, the resolution order that assigns a politician to a jurisdiction has not shifted under us. It is also the weakest possible week — the last stance row was written on October 9, so the table had nothing to absorb. The check and the drought are the same observation read two ways.
Nothing was added to the certified corpus either. One stratum of one evidence class remains the only measured part of it — 57 state-level, record-based stances, certified October 3 at 0.830 and 0.932 with no severe disagreements, and a leakage control that drops to 54 items and still certifies. News-sourced and statement-sourced evidence are not certified, nor is the local tier.
Treasury Tracker’s database holds budget data for 8,925 government entities across all 50 states plus DC, spanning fiscal years 1976 to 2028:
It is worth saying why 717 of these are a good use of a weekend, because they are not on anybody’s ballot in the way a mayor is. A New York school district levies its own tax, adopts its own budget, and in most of the state puts that budget to a public vote every May. It is one of the few governments an ordinary person is asked to approve a number for directly — and the number is, for most households, the largest local line on the bill. A platform that says it helps people follow the money cannot hold every city in New York and none of the districts that spend more of it.
What this does not do is make them comparable to a city, and the app does not yet know that. A district’s budget is mostly instructional salaries and state aid; a city’s is police, fire, roads and sanitation. The comparison gate already refuses to set a cash-basis budget beside a GAAP one, which is the same class of problem one layer down, and it has not been asked this question yet. Until it is, the honest framing is that we hold these budgets rather than that we compare them. That is named here so it is on the record before somebody charts one against the other.
The third leg, alongside votes and stances. The platform holds 29.1 million individual campaign-finance contributions totaling $8.6 billion, covering 2,260 federal politicians and candidates across 24 election cycles (1980 to 2026). Every record traces to its original FEC filing, and superseded lines from amended filings are retired individually rather than counted twice. This cycle’s work was subtraction: the California committee sweep described above removed roughly 970 wrong attachments and kept the several hundred that turned out to be real. A money page that credits a contribution to the wrong person is worse than one that is missing it, which is why that work counts as progress even though every figure in this paragraph either held still or barely moved.
Every state has at least one researched official, and all 50 congressional delegations are covered nationally. Shading shows the depth of state and local research. No tile changed colour this cycle, and none should have — a tile shades only when an answer is published, and no answer was published. The map below is identical to Friday’s. Puerto Rico is still an outlined tile: 161 seats seeded, no stances yet.
Hover any state for its exact count. The seeded tier is mostly governors and statewide officers; deeper tiers add legislatures, counties and cities.
Who actually uses what we build — measured, not guessed.
724 people used the platform in the last 90 days, across every app, over 1,554 sessions and 3,741 page views. Sessions average five minutes thirty-two and 67% do not bounce. 105 of the 344 who reached the main site clicked through into an app — 31% in the quarter, the fifth pull running at that share.
Visitors rose and the other two framing numbers fell: 698 to 724 people, but 1,574 to 1,554 sessions and 3,786 to 3,741 views, with the average visit twenty-seven seconds shorter. More people, each doing slightly less. Read every figure here as an order of magnitude, not a trend line — at this volume a couple of people either side is the whole swing, and the window slides. Over nine pulls the main site reads 306, 316, 312, 297, 300, 314, 337, 340, 344. The hostname map was re-run first and found no new host.
Everyone who arrived, against the people who did the one thing that app exists for. Bars are the 90-day rate; counts are people.
The two rows this page had called stuck both moved by exactly one person. Civic Trivia went from twelve finishers on nineteen starts to 13 on 20, after three identical pulls; Read & Rank went from eight ballots on nineteen starts to 9 on 20. One is not a recovery and we are not reporting it as one — but a number that had refused to move at all has moved, and that is worth the sentence. The Compass stays our healthiest funnel: 118 of its 137 visitors opened the quiz, and 37 finished.
Unique visitors per app, 30 days and the full quarter. Each app is counted once across every hostname that serves it — Treasury drew traffic on four this quarter, the main site on two.
Treasury Tracker rose furthest for the third cycle running — 190 to 206 in the quarter, 95 to 110 in the month — and it is now within reach of the main site in the 30-day window, 110 against 151. Civic Trivia is flat at 112 for the quarter and still falling in the month, 20 to 19. Civic Spaces is absent from this table, and that is a measurement gap rather than a result: it serves no traffic our analytics can see.
An earlier edition published a rate here — seven people in a quarter, as 5% of 152 visitors — and withdrew it. It stays withdrawn. The events were sound; the denominator was not, because Treasury pushes its in-app navigation into the address bar without recording a page view, so every visitor is filed against the bare domain. The direction stays wrong too: this quarter those few people produced 166 category drills, 101 entity selections, 101 line-item views, 56 year changes, 19 searches and 18 chart changes.
82 people reached a budget address and 24 fired any past-the-front-page event, and the 82 splits the way it has every cycle: 60 landed on our own organisation’s budget, one click from the donate page, which selects nothing and so fires nothing. A note on how the other half is counted, because the published series is not quite what it looked like. Earlier editions reported it by subtraction — 82 minus our own — and printed 16, then 19, then 23. Asked directly, 25 people this quarter opened some other government’s budget; asked by subtraction the same window gives 22. The gap is three people who looked at both. Neither reading changes the conclusion: the donate-driven half is still more than twice the other, and the day that reverses is the day this section gets a real denominator. It is not that day.
How much further we have to go, honestly.
Restart the stance work. It is the one thing on this list that did not happen this cycle, and with three weeks to the election it is the thing a voter will notice missing. The audit’s tail bands are untouched; 24 rungs still ask a different question in Season 2 than in Season 1 and need a ruling before anything carries across. Then put the 8.4 million votes to work — 3,861 legislators carry at least one recorded vote, and that record feeds exactly the evidence class the one certified stratum rests on. Finish the 2026 candidate waves, settle whether the trivia generator’s remaining slot-C share is design or defect, write the one collection an audit cannot fix, and time the Treasury page properly. Success in November is one thing: any voter in a served area can look up every name on their ballot and follow the money behind it.
Reactivate the parked pillar: Empowered Listening phase 6, Focused Communities v1.1, Civic Spaces v4.0 — whose shell, shared theme, banner library and dashboard are groundwork done early, and which went deliberately quiet this cycle. This is where the platform stops being a reference tool and becomes a community, and where our sustaining-donor model lives.
Designed but not built: the Gems stipend economy, Awareness Exchange, Badges, Symposium tournaments, Emparks, Common Grounds, and tools for citizens to run for office on the strength of their solutions. Roughly a third of the founding vision, waiting on the first two pillars.
A volunteer-built platform, visible in the commit history.
An interactive Purpose mind map walks the whole vision — the thesis, and the Why, How, Who and What — with a short explainer on every concept and pillar feature.