Half of Canada’s child-care communities have exactly one option

Where the programmes are

Licensed programmes per 1,000 children aged 0–4 is the cleanest cross-province supply signal our data supports, but it must be read with the registry differences in mind. British Columbia and New Brunswick license small family child cares individually, so each home appears in the registry; Ontario licenses home child care through agencies, so thousands of individual homes sit behind 165 agency records. The rate below measures licensed points of access a parent can look up, not spaces.

On that measure New Brunswick now leads the country at 27.7 programmes per 1,000 children under five, a shade above British Columbia’s 26.3. New Brunswick is a small province with a wide licensing net: 947 facilities for 34,230 young children, 135 of them licensed family homes. Ontario sits last at 6.1 for the opposite reason: its home-care sector is licensed at the agency level and never surfaces as individual records. Read the gap as a difference in how provinces write down their supply as much as a difference in supply.

The provider mix itself tells a policy story. Saskatchewan’s licensed system is 66% family child-care homes, the highest share in the country, while Ontario’s registry is 96% centres. Quebec’s 3,774 installations include no individually licensed homes at all: home-based care there runs through recognized coordinating offices, another reminder that provincial registries measure different things.

The one-option communities

49.4% is just under half. In these places there is no comparing, no waitlist hedging, no choosing between a centre and a home: one licence serves the community, or the family drives. Saskatchewan has the starkest geography: 127 of its 176 communities (72%) are one-option, and 75.9% of all its licensed programmes sit in just five cities. Ontario (271) and Quebec (297) cover the most communities of any province and also carry the largest absolute counts of one-option towns.

How far is the next licensed programme?

The obvious objection to the headline is the strongest one, so here is the measurement rather than an argument. For every one-option community we took its single listing and found the nearest other visible licensed programme in a different community. Because a postal-district centroid cannot answer a distance question, we ran it only where the coordinates are street addresses: British Columbia, Alberta, Nova Scotia, New Brunswick. That is 310 one-option communities, of which 309 were computable — one British Columbia community’s listing sits on a shared centroid point and was dropped.

160 of 309 (52%) have another licensed programme within 10 km, and 260 (84%) within 25 km. So for about half of them, “one option” describes the registry rather than the drive. The other half is the finding that survives contact with a map: roughly 48% of these communities have no second licensed programme within 10 km in any direction, and 16% have none within 25.

One more number belongs beside the headline. The 1,074 above is the raw grouping of registry city names. The named list of these communities publishes 1,031, because 43 of the 1,074 are not communities at all — they are spelling or punctuation variants of a city that already holds other licensed programmes (“Ottawa,” “bunaby,” “toronto  ontario”), and each is itemised there with the evidence that disqualified it. On the cleaner definition the share is 1,031 of 2,132 (48.4%) rather than 49.4%. Both are published, here and there, and the finding is the same either way: just under half. The raw grouping is the honest count of what the registries wrote down; the named list is the honest count of what a parent could go and visit.

What parents actually pay in 2026

“$10-a-day child care” is a federal promise implemented eight different ways. Ottawa counts eight provinces and territories delivering regulated care “for an average of $10 a day or less” across thirteen jurisdictions²⁴; counting published provincial rules across the eight provinces in this dataset, three qualify. Both statements are true — they use different denominators, and the difference is the reason this section exists.

The chart below is drawn so that unlike things look unlike. A dot is a single published rate. A bar is a published band. Nova Scotia gets no marker at all, because it publishes no single provincial figure — inventing one to fill the row would be the exact error the chart exists to prevent.

Three of those rows deserve a sentence they did not get in the last edition. Quebec is leaving the group. The $9.65/day reduced contribution is current and regulated, but the ministry’s own 2026-2027 budgetary rules state that it holds only until December 31, 2026 and rises to $10.60/day on January 1, 2027 — a 9.8% increase, published in the province’s own document.⁶ Anyone quoting “only Quebec, Manitoba and Saskatchewan” after New Year’s Day will be quoting a finding that has expired.

Ontario’s two numbers come from two different promises. The province capped CWELCC fees at $22/day on January 1, 2025 and states that this brought the provincial average to $19/day.¹¹ Separately, the federal one-year extension provides funding “to help maintain eligible child care fees at their current average of $19 per day until December 31, 2026” and runs the agreement to March 31, 2027.¹² The commitment with the end date is to the average, not to the cap; the previous edition of this report merged them, and that was wrong.

Manitoba publishes more than we credited it with. An earlier version of this work told Manitoba parents that nothing published tells them which facilities are funded. The province’s own page says the opposite: funded or unfunded status, and the fees a specific facility is allowed to charge, are published per facility in Manitoba Child Care Search.²⁰ The finding that survives is narrower and still worth knowing — the bulk provider list Manitoba publishes for download carries no funding column, so the split cannot be counted across the province, only checked one facility at a time. Manitoba states that over 1,000 licensed facilities must comply with the regulated $10-a-day maximum,²¹ against the 1,139 licensed Manitoba programmes in this dataset.

One figure has been removed rather than repeated. Previous editions gave Nova Scotia’s largest reduction as $23/day for centre infant care; that row could not be found on the source page as fetched, so this edition states only the reductions the published table shows — $18.25/day for full-day centre toddler and preschool care, $17.00/day for family-home infant care.¹⁴ A British Columbia upper bound of $45/day has gone the same way: the province publishes the $200/month floor and no ceiling,¹³ so the chart shows a floor and no ceiling.

The transparency gap

Whether a parent can see a facility’s inspection history depends almost entirely on their postal code. Six of the eight provinces publish inspection or enforcement findings in some form, and as of this edition all six are cited to a government page we opened — the previous edition counted Ontario and British Columbia among them without a source for either, which was the weakest link on this page. We can confirm bulk, machine-readable data in three — Alberta, Nova Scotia, Quebec. Those three counts need reading carefully, because each is a dated extract of a file that has since moved, not the live record.

As extracted on 2026-07-05, Alberta’s published Child Care Information dataset carried 15,566 inspection records covering 2,874 facilities between July 2024 and December 2025 — 11,770 routine inspections, 2,274 follow-ups and 1,522 complaint-triggered visits. The province publishes a rolling 18-month window refreshed quarterly, so today's file will differ.^3†

As extracted on 2026-07-11, Nova Scotia’s Child Care Directory carried 645 inspection records across 339 facilities (328 annual, 317 unannounced), the most recent dated August 2025. The directory is updated bi-annually; the directory was refreshed on 2026-07-19 and is now current to June 30, 2026, so today's file will differ.^4†

Quebec does both things at once, which the previous edition got wrong. As extracted on 2026-08-10, its quarterly registers of administrative penalties carried 730 penalties totalling $699,500 issued to 321 facilities between January 2023 and September 2025.¹⁵ Those registers are published quarterly; two further registers (October–December 2025 and January–March 2026) have appeared since, so today's registers cover a longer window.^26† Quebec also publishes inspection results for individual facilities on the ministry’s interactive childcare map,²⁷ so the honest axis is not enforcement versus inspection — it is bulk versus one facility at a time.

Ontario, British Columbia and New Brunswick publish findings through per-facility lookups, and the three lookups are not the same thing. New Brunswick’s Early Childhood Services Portal carries past and recent inspection reports under each facility’s “More Details” link, giving the public a brief compliance history one centre at a time.²⁸ (The province states no retention window; an earlier edition of this report said five years, which no published page supports, and the claim has been dropped.) British Columbia licenses child care through the regional health authorities, and the province points families at those five authorities’ own inspection pages, where you can “view summary reports of inspections and substantiated complaints”²⁹ — summary reports of inspections and substantiated complaints, held by the five regional health authorities rather than by the province, so there is no single provincial file to read.

Ontario is the narrowest of the three, and this edition says so plainly. Ontario publishes confirmed violations and offences only — compliance orders, protection orders, administrative penalties, restraining orders and prosecutions. A routine Ontario inspection that ends in compliance is never published, and the provincial file of assessment results is withheld. The child care violations registry covers licensed providers back to August 2015 and unlicensed providers back to January 2012.³⁰ That is a real disclosure and it belongs in the count of six. It is also not what a parent thinks they are getting: an Ontario centre with a spotless registry entry may simply never have been ordered to fix anything, and there is no way from the outside to tell that apart from never having been inspected.

The Ontario question the last edition left open is now closed. That edition could not reach data.ontario.ca and therefore refused to state whether Ontario publishes compliance findings in bulk. We reached it on 2026-08-11. The dataset exists, it is titled “Compliance of licensed child care facilities with licensing requirements,” it describes exactly the file this report wanted — the results of the assessments that check every licensed facility against the Child Care and Early Years Act — and it is withheld: “This data is not available right now. We are reviewing the data to determine if it can be made open to the public.”³¹ Its last validated date is February 2017. The sibling dataset of serious incidents in licensed child care facilities carries the identical notice.³² Two withheld files rather than one absent file: Ontario’s bulk position is a decision, not a gap, and the report now records it as a finding instead of as an open question.

Saskatchewan and Manitoba: no bulk inspection or enforcement dataset found on either province's open-data portal or child-care ministry pages, searched 2026-08-11. A negative like that needs a documented search rather than a citation, so the search and its date are stated instead of a link. No province publishes data that would let anyone honestly compute a comparative “safety score,” which is why MapleCub doesn’t: we show dated inspection facts where they exist and say plainly where they don’t. “No violations on record” in a province that publishes no records is not the same thing as a clean history.

The same gap applies to capacity. Only Quebec (237,634 spaces²), Alberta (222,932, including school-age care^3†) and Nova Scotia (19,216, covering 334 of its 348 listed programmes^4†) state licensed capacity in their open data — 479,782 spaces in total. British Columbia, Ontario, Saskatchewan, Manitoba and New Brunswick publish programme lists without machine-readable capacity, which means nobody outside those governments can compute how many licensed spaces exist per child. British Columbia’s own map data carries a per-facility CCFRI authorization flag but does not identify designated $10-a-Day sites,^25† so even the province’s cheapest programme cannot be counted from the outside. For a national programme spending billions to create “new spaces,” the count of existing ones is surprisingly hard to audit.

Source: MapleCub Canadian Childcare Dataset — Aug 2026 · computed 2026-08-10 · N = 20,516 · CSV · inspection and capacity figures are dated extracts; extract dates and refresh cycles stated inline

The visibility problem — and how much of it is ours

This section is about a MapleCub field before it is about Canada, and the previous edition had that the wrong way round. Google ratings reach our dataset through a matching step, and we have matched only 9,252 of the 19,569 programmes in the seven enriched provinces — 47%. Of the ones we did match, 5,991 (65%) carry a rating, and 132,162 parent reviews sit behind them. The national 31.8% “with a Google rating” is therefore a floor: it is what we have found, not what exists.

The distinction is not pedantic, because in three provinces the two numbers are the same number. We hold a Google place record for 18.1% of British Columbia’s listings, 6.5% of Saskatchewan’s and 20.5% of Manitoba’s — and essentially every listing we matched carries a rating (98.5% in British Columbia). Publishing “18% of British Columbia’s programmes have a Google rating” would be publishing our own queue depth with a province’s name on it. The previous edition set eight provinces side by side on that column and drew a conclusion from the spread. The spread was mostly ours. This edition does not do that.

Websites behave differently and need saying differently. A website can reach us from a registry field or from our own harvest, so it is not capped by the Google match: 2,613 listings across the seven provinces carry a website on a row we never matched. It is still conditioned by it. Read live on 2026-08-11, 62.2% of matched listings carried a website against 25.3% of unmatched ones. So the published 7,747 (40%) is a floor as well, and it is a frozen one: it is the quarterly snapshot figure, while the live column stood at 8,369 at that read and keeps rising as the harvest runs. The snapshot is what we cite; the gap between them is disclosed rather than quietly closed.

New Brunswick is excluded from every figure in this section. Its 947 programmes entered the dataset on 2026-08-09 and have not been matched to Google presence data yet, so the denominator here is the 19,569 programmes in the other seven provinces. Reporting New Brunswick as 0% website coverage would describe the state of our matching queue rather than the state of the province — and the two paragraphs above are that same sentence, extended to every province it applies to.

Source: MapleCub Canadian Childcare Dataset — Aug 2026 · computed 2026-08-10 · N = 19,569 · CSV · presence figures cover the 19,569 programs in the seven previously-enriched provinces

Province snapshots

Quebec deserves one more number: 2,788 of its 3,774 listed installations (74%) are subsidized (subventionné), the reduced-contribution network where parents pay $9.65/day in 2026⁵. The remaining quarter operate at market rates offset by Quebec’s refundable tax credit. New Brunswick has an equivalent split: 641 of its 947 facilities (68%) are designated under the provincial Parent Fee Grid¹⁰. Explore any province in depth: British Columbia · Ontario · Quebec · Alberta · Saskatchewan · Manitoba · New Brunswick · Nova Scotia

Methodology & data

Programme counts, provider types, locations, capacity and inspection figures were computed on 2026-08-10 from MapleCub’s dataset, which is built from the eight provincial licensing registries, carried through with the registry identifier each record was ingested with, and deduplicated (290 duplicate records identified and merged in the July 2026 dedup pass, excluded from every figure in this report). A community means a normalized city name: lower-cased, trimmed, non-empty, so that case variants in the Manitoba and Saskatchewan registries do not count twice. Google presence signals (ratings, review counts, websites) were matched conservatively, requiring name agreement, so they undercount rather than overcount — and because they undercount by different amounts in different provinces, they are published as match coverage rather than as a province ranking; New Brunswick has not been through that matching step and is excluded from those figures. Child population denominators are Statistics Canada, Table 17-10-0005-01 (July 1, 2025 estimates)¹. The full build is documented on the Data Lab methodology page.

Sourcing rule. Every external claim on this page — every dollar figure, every inspection count, every published rule — is cited to the government page it came from, with the date we fetched it, in the Sources list below. Figures computed from MapleCub’s own dataset are not cited to anyone else: they carry the “Source: MapleCub Canadian Childcare Dataset” line and the CSV instead, and they are ours to defend. Inspection and enforcement counts are published as dated extracts with the publisher’s own refresh cycle stated, because re-downloading those files today will not reproduce our numbers. Where a source has moved on since our figure was computed, or where we could not open the page ourselves, the citation carries a dagger and the Sources entry says why. Sources were last audited 2026-08-11.

What “geocoded” means here. 20,481 of 20,516 records (99.8%) carry a coordinate. That is the easy half of the sentence. Of those, 9,400 (45.9%) are address-level and 11,081 (54.1%) are the centroid of a postal district or a city. We call a point a centroid when more than 3 visible listings on one identical lat/lng, resolving to more than 1 distinct street address; one address with several licences on it is a building, several addresses on one point are not. Coordinates are address-level in British Columbia, Nova Scotia, New Brunswick, Alberta — Alberta except in its two biggest cities — and a postal-district centroid in Ontario, Quebec, Saskatchewan, Manitoba. Saskatchewan's S0K resolves 57 listings in 27 separate towns to one point; Alberta pins 314 Calgary listings and 256 Edmonton listings to one point each. An earlier edition said only that “a small number of Manitoba records geocode to a municipal centroid,” which understated the number by four orders of magnitude and named the wrong province.

No count in this report is computed from coordinates, so nothing above moves. The one place coordinates are load-bearing is the nearest-alternative finding in the one-option section, and that analysis is restricted to the four provinces where the coordinate is an address.

Known limitations: registries differ in scope (individually licensed homes vs. agency-based home care), so cross-province programme counts are not space counts; Alberta’s published capacity includes school-age care while Quebec’s is early-childhood only, so we do not compute cross-province coverage ratios; Nova Scotia’s published capacity covers 334 of our 348 NS listings, not all of them; Manitoba is the only province with listings carrying no coordinate at all (1,104 of 1,139) — which is a different and much smaller problem than the centroid split described above, and the earlier editions’ habit of naming only Manitoba here read as though the other seven were fine; eight Saskatchewan and 47 Manitoba rows carry no city value and sit outside the community counts, and 35 of those 47 Manitoba rows carry no coordinates either. A 2026-08-10 repair merged registry abbreviations such as “Wpg” into their proper communities wherever the community was determinable; the 12 that stayed unrecoverable were set to null and are part of the 47.

Two further limitations belong here rather than in a footnote. The website count is frozen to the quarterly snapshot (7,747); MapleCub’s website discovery keeps running between editions, so the live directory shows more — 8,369 when it was read on 2026-08-11 — and the frozen figure is the one to cite. And the per-1,000 rate counts programmes of every type against a 0–4 denominator; the school-age caveat, and what happens to the ranking if the exclusion is applied, is stated in full beside the chart above.

The province-level dataset behind this report is free to download and reuse with attribution (CC BY 4.0). Province-level counts of licensed child-care programs, provider types, geographic coverage, published licensed capacity, Google-presence signals, and inspection-data availability for BC, ON, QC, AB, SK, MB, NB and NS. Derived from provincial licensing registries, deduplicated and verified.

Machine-readable JSON for developers: /data/state-of-child-care-2026.json — same figures as the CSV, same licence.

<img src="https://www.maplecub.ca/charts/programs-per-1000-by-province-2026.png"
     alt="Licensed child-care programmes per 1,000 children aged 0-4, by province, 2026"
     width="760" loading="lazy">
<p>Source: <a href="https://www.maplecub.ca/research/state-of-child-care-2026">MapleCub, The State of Licensed Child Care in Canada 2026</a> (CC BY 4.0)</p>