# Sources and claim checks

Accessed 4 November 2026. Raw inputs in `raw/` are the 2025-filtered Eurostat JSON-stat responses, fetched directly from the official dissemination API. Both report `updated=2026-10-20T23:00:00+0200`.

| Claim ID | Published claim | Primary source and edition | Inputs and formula | Expected display | Limit or caveat |
| --- | --- | --- | --- | --- | --- |
| C01 | Share of children receiving any formal childcare or education | Eurostat `ilc_caindform25b`, 2025, unit `PC`; [Data Browser](https://ec.europa.eu/eurostat/databrowser/bookmark/e3d108b7-79b2-4e1e-b230-35069be22544?lang=en); [JSON-stat API](https://ec.europa.eu/eurostat/api/dissemination/statistics/1.0/data/ilc_caindform25b?time=2025&lang=EN) | `freq=A`, `yn_arope=TOTAL`, `age=Y_LT3`, `unit=PC`, `geo=DK/NL`, duration `H0`; `any=100−H0`, rounded to one decimal | Denmark 67.2%; Netherlands 66.5%; gap 0.7 percentage points | Population share by EU-SILC, not count of children or available places. Values are 2025 estimates.
| C02 | Share receiving at least 25 hours a week | Same source/filter, duration `H_GE25` | Read the published cell directly; do not infer it from rounded bands | Denmark 62.0%; Netherlands 11.5%; gap 50.5 percentage points | 25-hour Barcelona-target indicator. No causal interpretation or statistical-significance claim.
| C03 | Denmark–Netherlands has the widest ≥25h gap among near-equal participation pairs | Same source; all 27 EU countries, age `Y_LT3`, `yn_arope=TOTAL`, 2025 | Compute all 351 country pairs; keep pairs with absolute any-care difference ≤3.0 pp; rank by absolute `H_GE25` difference | 34 pairs meet the screen; Denmark–Netherlands is largest at 50.5 pp | Three-point screen is an exploratory selection rule, not a published Eurostat threshold or statistical test. It follows the earlier 30-hour research lead; do not describe this as a global-first discovery.
| C04 | Romania and EU comparison under age three | Same source/filter, `geo=RO` and `EU27_2020` | `any=100−H0`; long-hours share is direct `H_GE25` | Romania: 11.1% any, 7.4% ≥25h; EU: 40.6% any, 29.4% ≥25h | Descriptive rates. They do not establish demand, capacity, price or why families use formal care.
| C05 | Age-band labels and duration definitions | Eurostat `ilc_caindform25b` dimension metadata and [2026 Key Figures, p.43](https://ec.europa.eu/eurostat/documents/15216629/24279737/KS-01-26-036-EN-N.pdf/6546c110-cf94-2b55-e64a-39f68170325d) | Preserve source codes: `Y_LT3`, `Y1-2`, `Y3-CSA`, `CSA1-Y12`; durations `H0`, `H1-24`, `H_GE25` | Zero hours, 1–24 hours, 25 hours or more; under 3, ages 1–2, ages 3 to minimum compulsory school age, and one year after compulsory school age to age 12 | Under 3 includes infants and is not the same group as ages 1–2. Compulsory school age differs between countries.
| C06 | Definition of formal childcare or education; 25-hour threshold | Eurostat, [2025 indicator explanation](https://ec.europa.eu/eurostat/web/products-eurostat-news/w/ddn-20250930-2) and [glossary](https://ec.europa.eu/eurostat/statistics-explained/index.php?title=Glossary:Children_in_formal_childcare_or_education) | Quote/paraphrase the official definition: institutionalised, intentional, planned and formal education programmes provided through public organisations and recognised private bodies; indicator threshold is 25 hours/week | Plain-language scope note beside the chart | The old 30-hour series remains published but is a distinct indicator; never relabel it as the 25-hour Barcelona-target series.
| C07 | Observation flags | Eurostat JSON-stat `status` field; [EU-SILC metadata](https://ec.europa.eu/eurostat/cache/metadata/en/ilc_sieusilc.htm) | Preserve exact source code(s) for each country/age/duration cell. Render `b`, `e`, `p`, `u` using official definitions | `b` break in series; `e` estimated; `p` provisional; `u` low reliability | A missing flag means no status code was attached; it is not proof of zero sampling error. Denmark/Netherlands `Y_LT3`, total cells are unflagged.
| C08 | Dataset version and source reuse | Eurostat API metadata (`updated` field); [Eurostat reuse and citation terms](https://ec.europa.eu/eurostat/help/copyright-notice) | Snapshot downloaded 4 November 2026, filtered to 2025. Keep raw bytes, SHA-256, API query and transformations. State source and that values were adapted/translated. | `Source: Eurostat, ilc_caindform25b, accessed 4 November 2026. Calculations and translations by Marius Comper.` | Reuse of statistical data is allowed with source acknowledgement. Because data are transformed and copy translated, include the Eurostat non-responsibility notice in both language editions.

## Saved input hashes

- `raw/ilc_caindform25-2025.json` — SHA-256 `a36a1e767fea9bc7a466f1107367146c4764f00531ae5b479d2a29acd9d67894`
- `raw/ilc_caindform25b-2025.json` — SHA-256 `3035bc787f4b88d4d1d9bfde38b4e6c365cb2855c09224cb46292363d0bf096c`

The `TOTAL` cells are independently cross-checked between the two API tables before build. The separate CIRCABC folder Eurostat says contains variances and confidence intervals could not be downloaded: the browser connector returned a cache miss and direct HTTP redirected to `sorry.ec.europa.eu` with a 406 page. Therefore the page reports descriptive point estimates and does not claim significance.

## Attribution and disclaimer

Source: Eurostat, datasets `ilc_caindform25` and `ilc_caindform25b`, accessed 4 November 2026. Data are modified and the explanatory copy is translated by Marius Comper. Neither the European Commission nor any person acting on behalf of the Commission is responsible for the use that might be made of this information.