Sources

Indigenous Services Canada · First Nation community

How it joins

band_name_within_province

Quality method

ISC operational tracking of advisories in effect >12 months; not a survey. Point-in-time snapshot (as_of).

Load decision

ingested: 38 active long-term advisories; 37 matched to a band, 1 unmatched (flagged). ISC page offers no CSV; extracted as-published.

Publisher pageMethodology

Indigenous Services Canada · First Nation community

How it joins

band_name

Quality method

ISC operational tracking, <12 months. BC advisories excluded (held by First Nations Health Authority).

Load decision

ingested: 48 short-term advisory systems as published (HTML; no CSV). BWA/DNC/DNU in advisory_kind.

Publisher pageMethodology

Agriculture and Agri-Food Canada (from Statistics Canada Business Register) · CSD

How it joins

csd_uid (SGC)

Quality method

Business Register (administrative), not a survey; counts of active business locations. Individual businesses suppressed by StatCan; no census non-response applies.

Load decision

ingested: Row per (CSD SGC, NAICS code, employment-size class ESR9); value = count of business locations (FREQ).

Publisher pageMethodology

Statistics Canada · CSD

How it joins

csd_uid == 2016 GEO_CODE (POR)

Quality method

2016 Census. Per-geography GNR (short-form) and GNR_LF (long-form) captured in statcan_csd_quality at census_year=2016. Note 2016 used the long-form CENSUS (unlike the 2011 voluntary NHS), so it is directly comparable to 2021. Suppression tokens kept as-published.

Load decision

ingested: ~2,247 characteristics per CSD; Total count stored as value, male/female kept in dimensions. Second time-point for the multi-year census series (with 2021 anchor). Reserves whose CSD code changed 2016->2021 do not match and are simply absent for 2016 (flag-not-guess).

Publisher pageMethodology

Statistics Canada · CSD

How it joins

csd_dguid

Quality method

Comprehensive Census Profile. Carries per-geography TNR_SF (short-form) and TNR_LF (long-form) Total Non-Response rates and a 5-digit data-quality flag; captured per community in statcan_csd_quality. Anchor source covering the ~2600 census characteristics (population, income, labour, education incl. field of study, housing, language, Indigenous identity, Registered/Treaty status) per community. Also serves as the community-level source for email items 2/5/12/14/18 whose standalone cross-tab cubes are not separately materialized.

Load decision

ingested: ~2,631 characteristics per CSD. Total count stored as value; men+/women+/rate columns kept in dimensions. Used as the anchor per-community census dataset and as the TNR response-rate source.

Publisher pageMethodology

Indigenous Services Canada · CSD

How it joins

csd_uid->band_number

Quality method

Composite of census income/education/housing/labour; blank where suppressed or below population threshold. Inherits census non-response.

Load decision

ingested: Per-community well-being score (0-100) + 4 components x census year.

Publisher pageMethodology

Statistics Canada · CSD-to-CSD

How it joins

csd_dguid (residence)

Quality method

2021 Census long-form journey-to-work. Per-CSD TNR in statcan_csd_quality; small flows suppressed/rounded as published.

Load decision

ingested: Origin (residence) CSD -> destination place-of-work flows, by gender. Only non-zero Total-Gender flows stored.

Publisher pageMethodology

Statistics Canada · CSD-to-CSD

How it joins

csd_uid (residence)

Quality method

2016 Census long-form journey-to-work; per-CSD GNR in source. Small flows rounded/suppressed as published.

Load decision

ingested: 2016 companion to statcan_98100459 (2021). Origin (residence) CSD -> place-of-work flows by sex; non-zero flows only. Makes commuting multi-year (2016 + 2021).

Publisher pageMethodology

Statistics Canada · n/a (catalogue of datasets)

How it joins

none

Quality method

A catalogue/index of GDI datasets, not itself a dataset. Individual member tables would be assessed and ingested separately if selected.

Load decision

reference_catalogue: Registered as a pointer; no single dataset to load.

Publisher page

Public Health Agency of Canada / Pan-Canadian Public Health Network · Health region / national aggregate (by equity stratifier)

How it joins

none

Quality method

Interactive tool reporting health inequalities by equity stratifiers (including Indigenous identity) at aggregate/health-region level; not published per First Nation community. Response/coverage varies by underlying survey.

Load decision

reference_tool: Registered; aggregate-only, not community-joinable. Specific extracts possible on request.

Publisher pageMethodology

Statistics Canada · CSD

How it joins

csd_dguid

Quality method

Experimental modelled index combining census and administrative data; no direct survey response rate. Underlying census inputs subject to per-CSD TNR (see statcan_csd_quality).

Load decision

ingested: Housing + transportation cost index per CSD (experimental estimates).

Publisher pageMethodology

Statistics Canada · CSD

How it joins

csd_dguid

Quality method

2021 Census long-form (25% sample) and 100% data. Per-CSD Global Non-Response Rate captured in statcan_csd_quality; on-reserve/small populations subject to StatCan suppression (symbols preserved) and the 2021 incompletely-enumerated-reserve list.

Load decision

ingested: Income sources and taxes (32) x Income statistics (8) per CSD. 253,952 cells for 992 linked FN IRI CSDs; ~82k with published numeric values, remainder suppressed/blank as published.

Publisher pageMethodology

Statistics Canada · Reserve/settlement (name + province; no codes published)

How it joins

name match to fn_census_subdivisions.csd_name (flag-not-guess)

Quality method

THE low/zero-response signal: reserves/settlements where the 2021 Census could not complete enumeration (COVID restrictions, evacuations, access refusal). 61 CSDs flagged in statcan_csd_quality.incompletely_enumerated (57 exact/relaxed name matches + both CSD parts of Six Nations (Part) 40 and Fort Albany (Part) 67). NOT flaggable: Kanesatake (QC), Pakuashipi (QC), Granville Lake (MB) — Indian settlements (S-E), outside the IRI-only fn_census_subdivisions crosswalk; recorded here rather than guessed.

Load decision

ingested: Prior-census populations stored on matched rows (prev_population_2016). StatCan page publishes names only, no CSD codes; matching is derived and evidence-noted.

Publisher pageMethodology

Indigenous Services Canada · Community (project point)

How it joins

Community Number->band_number

Quality method

Administrative project register; investment shown only for completed projects.

Load decision

ingested: ISC-funded infrastructure projects, First Nation communities only.

Publisher pageMethodology

Statistics Canada · CSD

How it joins

csd_dguid

Quality method

2021 Census. 2945160 published value cells loaded for our CSDs; 1875960 cells suppressed as published (recorded per community in __cell_coverage_summary__). Per-CSD TNR in statcan_csd_quality.

Load decision

ingested: Registered/Treaty status(3) x Gender(3) x Statistics(3) x Age(20) x Indigenous identity(9) per CSD. Published cells: 2945160; suppressed: 1875960.

Publisher pageMethodology

Statistics Canada · CSD

How it joins

csd_dguid

Quality method

2021 Census. Per-CSD Total Non-Response captured in statcan_csd_quality; on-reserve/small counts suppressed as published; see 2021 incompletely-enumerated-reserve list.

Load decision

ingested: Full detail: gender(3) x age(8) x statistics(5) x Indigenous identity(9) per CSD.

Publisher pageMethodology

Statistics Canada · CSD / Indigenous community (per-geography pages)

How it joins

per-DGUID page retrieval only

Quality method

2021 Census long-form Indigenous Population Profile (25% sample). response_rate_value=25 records the long-form sample fraction, not a universal response rate. Per-CSD TNR_SF/TNR_LF and incompletely-enumerated flags are in statcan_csd_quality; all 992 IPP CSD availability records join there. Indigenous-identity subpopulation cells are smaller than whole-community cells, so suppression and TNR caution bite harder. The CSV product carries as-published cell flags; no separate per-geography IPP quality flag was found.

Load decision

ingested: Per-geography pulls from the stable StatCan CSV endpoint replaced the still-nonexistent bulk file. Completed 2026-07-19: 605 available CSDs, 387 explicit absent CSDs, 0 errors, and 15,501,915 as-published cells. Consumers must join statcan_csd_quality before interpretation.

Publisher pageMethodology

Statistics Canada · CSD

How it joins

csd_dguid (available at source)

Quality method

Available to CSD but the largest cube (259M datapoints; 3x15x3x40x12 per geography), overwhelmingly suppressed for small reserves. Field-of-study and education headline figures are captured via the Census Profile anchor.

Load decision

skipped_available_at_source: DELIBERATELY SKIPPED (JME decision 2026-07-09). The full field-of-study cross-tab (field x education level x age x sex) is ~15M published cells for our CSDs, overwhelmingly suppressed for small reserves. The 2021 and 2016 Census Profile anchors already carry field-of-study SUMMARY characteristics per community. Full detail remains available at source (WDS getFullTableDownloadCSV/98100393) and can be loaded on demand via pipeline/load_big_cubes.py 98100393. The 6.84M-row partial load from the 2026-07-08 overnight crash (Google Drive unmount) has been cleared. Companion cubes 98100266 and 98100456 were loaded in full.

Publisher pageMethodology

Statistics Canada · Province/Territory (with 610-category First Nation/band membership breakdown)

How it joins

membership name -> band_number

Quality method

Geography stops at province/territory; the 610-member "Membership in a First Nation or Indian band" field is a StatCan person-level classification, not a community geography, and StatCan publishes no band_number crosswalk for it. Community-level labour-force status is available per CSD via the Census Profile anchor.

Load decision

ingested: Band-keyed useful slice loaded 2026-07-09 into statcan_fn_labour_force: 610 membership groups x on/off-reserve residence x gender x 8 labour-force metrics (counts + participation/employment/unemployment rates), national totals, all ages, Count statistic = 43,920 rows. 475 named groups matched to a band by name (flag-not-guess). The only source reporting labour force by BAND MEMBERSHIP rather than geography, so it captures members living OFF reserve. Full cube (~67M cells: all provinces x ages x registered-status) available at source, not materialized.

Publisher pageMethodology

Statistics Canada · CSD (2016 vintage)

How it joins

csd_uid (stable code, 2016 DGUID)

Quality method

Administrative vehicle-registration counts. 919 of our reserves appear in the table; 236 with a published value, the rest suppressed as published. Small-count suppression symbols kept.

Load decision

ingested: CORRECTION: earlier not_community_level/2-reserve claim was an error from an (IRI)-tag filter. This IS a CSD-level, multi-year (2017-2026) table; joined on the stable CSD code across the 2016/2021 vintage gap. Headline total-fuel x total-vehicle margin stored per community per period.

Publisher pageMethodology

Statistics Canada · CSD

How it joins

csd_dguid

Quality method

2021 Census. 6299370 published value cells loaded for our CSDs; 4012470 cells suppressed as published (recorded per community in __cell_coverage_summary__). Per-CSD TNR in statcan_csd_quality.

Load decision

ingested: Place of work(5) x Industry NAICS(21) x Occupation NOC(11) x Gender(3) x Statistics(3) per CSD. Published cells: 6299370; suppressed: 4012470.

Publisher pageMethodology

Indigenous Services Canada · Registry group (band)

How it joins

registry_group_number~band_number

Quality method

Indian Register snapshot; groups <40 and cells <10 suppressed. Earlier years (2014-2022) available as separate yearly datasets, not yet stacked.

Load decision

ingested: Registered (Indian Act) population by band x year, on/off-reserve x gender. Machine-readable per-band CSV exists ONLY for 2022, 2023, 2024 (all loaded, ~646 groups/year). Earlier years 2014-2021 were published as PDF reports only (per-band tables inside the PDFs; no CSV/open-data record) - would require table extraction/OCR. registry_group_number = ISC band number for the vast majority of groups.

Publisher pageMethodology

Statistics Canada · CSD

How it joins

csd_dguid

Quality method

Building Permits Survey (administrative permit data). Reference years 2018-2022. Cube archived.

Load decision

no_reserve_coverage: Residential building permit counts and values per CSD, by type, by year. COVERAGE FINDING: 0 First Nation reserve CSDs appear in this table — the Rural Canada Housing Profiles derive from property-assessment/permit (CHSP) data that excludes on-reserve land. Non-reserve CSDs exist at source; reserves are absent.

Publisher pageMethodology

Statistics Canada · CSD

How it joins

csd_dguid

Quality method

Derived from the Canadian Housing Statistics Program (administrative property assessment/registry data), not a survey; no census non-response. Reference years 2019-2022. StatCan cube marked archived.

Load decision

no_reserve_coverage: Counts of residential properties, assessed values, living area per CSD, by year. COVERAGE FINDING: 0 First Nation reserve CSDs appear in this table — the Rural Canada Housing Profiles derive from property-assessment/permit (CHSP) data that excludes on-reserve land. Non-reserve CSDs exist at source; reserves are absent.

Publisher pageMethodology

Statistics Canada · CSD

How it joins

csd_dguid

Quality method

Investment in Building Construction program (administrative/building-permit based). Reference years 2018-2023. Cube archived.

Load decision

sparse_reserve_coverage: Values of investment in residential building construction per CSD, by year. COVERAGE FINDING: only 1 First Nation reserve CSD present (assessment/permit-based source largely excludes on-reserve land).

Publisher pageMethodology

Statistics Canada · CSD

How it joins

csd_dguid

Quality method

Modelled estimates from administrative + survey sources (not a per-community census). Reference years 2019-2023.

Load decision

ingested: Active non-profit counts, revenues and employment per CSD, by year.

Publisher pageMethodology

Indigenous Services Canada · First Nation community (varies by dataset)

How it joins

band_number / community (varies)

Quality method

ISC portal linking multiple community-level administrative datasets (e.g., drinking-water advisories, community infrastructure, band governance). Administrative data (no census non-response). Each dataset would be ingested individually with its own provenance.

Load decision

reference_portal: Registered as a pointer; high-value community datasets behind it (water advisories, infrastructure) can be ingested on request.

Publisher pageMethodology