The Empty Ballot

Ontario is about to elect thousands of people almost nobody can read up on — the flagship analysis of the 2026 municipal field. Field truth as of August 10–11, 2026; nominations close August 21, 2 p.m.

On October 26 — the fixed date of Ontario's Municipal Elections Act — the province will hold roughly 2,300 elections in a single day: 2,289 races across every one of its 414 directly-electing municipalities and their school boards, choosing the councils, mayors and trustees who run local government — roads, water, policing, transit, planning, schools [receipt §1]. We tried to do the thing every civics teacher says a voter should do: look the candidates up. All of them.

Here is what a voter can actually find.

Of 4,743 researched candidates — every single one checked individually, to a dated conclusion — 85.9% have published nothing a voter can hold them to [AGGREGATES_v21]. For most of them the word is literal: 3,461 candidates (73.0%) have no findable published platform at all. Another 613 (12.9%) have a website with no platform on it — real writing, often, a bio or a list of priorities, but nothing specific enough to check later [AGGREGATES_v21]. And no, the rest aren't simply on social media instead: counting ANY findable web presence, site or social, the number reachable at all rises only to 33.0% — 1,566 of 4,743 [receipt §3, §11]. Just 416 candidates in the entire province — 8.8% — publish a platform containing even one commitment concrete enough to check later: an action with a number, a mechanism, or a date attached [AGGREGATES_v21].

Around that silence sits a quieter fact about the ballot itself. 316 races currently have no registered candidates at all, and 956 more are on acclamation track — no more candidates than seats. Counted at the seat level (many races elect several members), that puts 1,457 of Ontario's 3,180 listed seats — 45.8% — on course to be filled without a contest if nothing changes by the August 21 filing deadline [receipt §2]. And where the no-contest map and the no-platform map overlap is the darkest, cleanest number we found: of 948 fully-researched acclamation-track races, 854 — 90.1% — have no registrant who has published a platform anywhere we could find [receipt §5]. Well over a thousand seats — 1,309 — are on track to be filled this fall without a contest and without any registrant publishing, anywhere we could find, what they intend to do [receipt §5].

This is a structural story, not a lazy-candidates story

Say the fairness line first, because it is true: absence of a published platform is a fact about the public record, not about a candidate's effort, character, or worth. Much of municipal Ontario campaigns door to door, and a candidate knocking on hundreds of doors a week in a township of 3,000 may be doing more direct democracy than any website ever could. The claim in this piece is narrower and, we think, more damning of the system than of any person in it: a voter who goes looking can't read anything — and nobody designed it that way, and everybody tolerates it.

Ontario municipal politics runs without parties: the Municipal Elections Act has no municipal party system — unlike the federal and provincial tiers, and unlike the civic-party machines of Vancouver or Montréal. Nobody hands an Ontario municipal candidate a platform template, staffs their war room, or writes their communications; and in hundreds of races, no opponent forces the issue either. The result is exactly what you would predict, and now it is measured: website coverage falls from 64.0% of candidates in Toronto to 7.4% in the North, and from 52.6% in municipalities over 100,000 people to 3.4% in municipalities under 5,000 (every regional and office figure in this piece is the August-10 snapshot cut [receipt §3]; the §11 corrections removed 13 wrong-attribution website values province-wide, corrected toplines carry the §11 cite, and the full cuts republish post-correction at registrar-final) — the places where, simultaneously, 24.4% of races have nobody running at all [receipt §2–3]. The smaller the government, the closer it sits to your basement flooding and your kid's school bus — and the less there is to read.

The school boards are the blind spot inside the blind spot. Trustee races — the offices that govern Ontario's school-board budgets — run 47.0% acclamation-track, 29.1% zero-candidate, and just 15.0% of trustee candidates have any campaign website [receipt §2–3].

Incumbency does not rescue the record: of 307 candidates on file flagged as incumbents, only 55 (17.9%) have a platform or policy page on record [receipt §3]. Many sitting officials do publish something real — 150 candidates province-wide, 29 of them incumbents, publish delivery records of what they did in office, and a delivery record is a genuine public document [AGGREGATES_v21]. But a record is what you did, not what you propose to do, which is why this instrument counts it separately rather than scoring it as a platform. In the province-wide comparison we ran, challengers as a class out-publish incumbents [receipt §3].

What the platforms that do exist reveal

We read all of them — every captured candidate site in the province, individually, under a public, viewpoint-neutral instrument, with a fifth of the reads independently blind double-rated — 90.8% class agreement after adjudication, 83.1% raw, both published — and every disagreement adjudicated on the record [RELIABILITY_v21]. Three findings carry the piece:

1. Candidates know their own job. The folk theory that municipal candidates promise federal miracles is false. Jurisdictional fit is the strongest dimension we measure (4.22/5): people running for council overwhelmingly campaign inside council's actual powers [AGGREGATES_v21]. The raw material of honest local politics is already there.

2. Almost nobody will tell you when — or with whose money. Time-boundedness is the weakest dimension in the province (1.30/5), as it has been in every edition we've run; fiscal grounding sits at 2.58/5 [AGGREGATES_v21]. A dated, costed commitment — "we will do X by year Y for about Z" — is the rarest object in Ontario's entire platform corpus. Not because it's hard: because nothing in the system asks for it.

3. There is no copy-paste democracy — because there is almost nothing to copy. We compared all 484 assessable platform texts against each other for borrowed language and found zero confirmed cases of platform planks copied between unrelated campaigns [receipt §7]. The consultant-boilerplate scandal we went looking for does not exist. The vacuum is worse than plagiarism: plagiarism would at least mean the words existed.

And one service finding, published in aggregate only because publishing detail would be a target list: 47.0% of 998 checked campaign sites have no DMARC record — meaning anyone on earth can send email that appears to come from those campaigns [receipt §8]. Each affected candidate gets that finding privately and courteously; none is named.

The good news is genuinely good

Of the 80 candidates whose platforms clear the instrument's top scoring band [AGGREGATES_v21], 74 — from big-city wards to small-town mayoralties — also cleared the two gates we add before any name is printed: an independent second reading, and a verbatim, checkable commitment quoted from their own text [EXCEPTIONAL_SHORTLIST]. They are named, praise-only, with their own words, at the exceptional platforms page. They prove the bar is passable by anyone — several of them are first-time candidates with a free website template and a weekend's writing.

Be clear about what is and isn't fixable by August 21. The gap this piece measures is structural — no party infrastructure, no staff, a capacity gradient that tracks town size — and no ten-day window repairs that. What IS true, and worth saying just as plainly, is that the door itself was never the barrier: filing costs as little as $100 ($200 for head of council) and an afternoon. In the 316 zero-candidate races, a single filing would put the first name on the ballot, and that filer's platform would be, literally, the only thing to read; in the hundreds of acclamation-track races where no registrant has published — 90.1% of them [receipt §5] — a second filer with a written plan would create the race's entire public record. So we ship first steps, not remedies: a free platform template designed to take a candidate from nothing to written-costed-dated in a weekend, per-race pages for every one of the 2,289 races — each carrying the same fairness line this piece opens with — so voters can see what's on offer where they live, the full method for anyone who wants to replicate this anywhere, and a standing promise: publish, tell us, and we re-measure gladly. This baseline exists to be beaten.

An election where there is nothing to read is not an election a voter can take part in with their eyes open. Ontario didn't choose that — it defaulted into it. A default is not a destiny: the habit of published, checkable commitments starts with one race, one platform, one afternoon — and the measuring stick is now public. The deadline to be on a ballot is August 21.