Linking Canadian public datasets keeps revealing what no single file shows: weak pollution enforcement, hollowed-out rental markets and concentrated public money

Asked:

“What are documented cases where combining Canadian open datasets revealed something new that a single dataset could not show, such as fraud detection, spending patterns, environmental violations, or housing insights?”

Eleven documented Canadian cases, each from an independent source, spanning roughly 2019–2026 across three domains: environment (1), housing (5), and public integrity and spending (5). “Open” is read practically here — several cases rest on public registries or records obtained through access-to-information requests rather than open-licensed downloads.

How each case links data to a new finding

Input datasets flow through a linkage step — matching organizations or owners across inconsistent names, adding context, or connecting money across systems — into the case that produced the finding. Ribbon colour marks source quality; hover any case or input for details. Pick a domain:

The strongest penalty signal is national: 2,217 air-pollution enforcement actions over 2000–2020 produced just $16.28M in fines and no jail sentences, with weak responses for 32% of repeat offenders — online.ucpress.edu

Linking non-resident ownership to census incomes shows correlations of 0.92 (Metro Vancouver) and 0.81 (Toronto) between foreign ownership and the decoupling of house prices from local incomes — doi.org

Resolving corporate ownership across 8,536 defence suppliers shows foreign-owned firms are about 17% of the supplier base yet hold about 55% ($55.9B) of observed procurement value — blog.govconweekly.ca

A self-published AI screen of 1.28M federal grant records flagged 846 entities critical and 6,931 high risk — flags are leads for review, not adjudicated wrongdoing — and found $71.5B in large grants with no documented purpose — nateglubish.substack.com

All eleven cases

Click a column header to sort. Evidence type is read from the publishing context: peer-reviewed journal, official statistical program, investigative or civil-society research, or self-published current tool. Risk scores in the self-published grant screen are leads for review, not findings of fraud.

Case	Domain	Evidence type	What linkage revealed	Date	Source

Shortlist of 11 documented Canadian data-linkage cases, one row per case, compiled 2026 from 11 independent hosts (no URL supports more than one row). Dollar figures are as reported by each source in CAD; risk flags count entities scored for review, not confirmed wrongdoing. Full dataset lists and finding texts were shortened for space; each source link carries the complete record. No coordinates supplied, so no map is drawn.

This report was generated automatically by Keenable SELECT at a user's request, from publicly available web sources linked herein. Keenable does not review, verify, or endorse its contents and makes no representation as to accuracy, completeness, or timeliness; AI-based extraction may contain errors. Nothing in this report is investment, legal, financial, or other professional advice. All trademarks and referenced content remain the property of their respective owners; no affiliation or endorsement is implied. To report an error, rights concern, or request removal: legal@keenable.ai.