Home · Research · The precedent economy

The Record · citation analysis · modified 7 Aug 2026

Which cases actually get cited

We ran a citation network across 484 Ontario human rights decisions about a single institution, then checked the result against two others. The pattern is the same everywhere: the authorities that end cases are cited about five times as often as the authorities that decide them. This page names both sets, explains what each one does, and says which ones a self-represented applicant should read first.

The five most-cited authorities are all gates

Parry Sound, 2003 SCC 42
sends Code claims to grievance arbitration
46
Dabic, 2010 HRTO 1994
the no-reasonable-prospect test
19
Sigrist, 2008 HRTO 34
the reconsideration threshold
19
Miller, 2009 HRTO 1241
the one-year limitation rule
16
Cartier, 2009 HRTO 1670
good-faith delay
11

Not one of them decides whether discrimination happened. Every one decides whether the Tribunal will look.

The five least-cited are the substance

O'Malley, 1985 CanLII 18
the prima facie test — the core of discrimination law
8
Noble, 2010 HRTO 878
the reprisal test
8
McKinnon, 2007 HRTO 4
systemic remedy
7
Moore, 2012 SCC 61
undue hardship
2
Shaw v. Phipps
circumstantial inference in profiling
2
The ratio, stated plainly. Gatekeeping authorities to substantive ones: 4.6 : 1 at Ontario corrections, 5.7 : 1 at Durham, 2.8 : 1 at Thunder Bay. Three unrelated institutions, one direction.

Doctrine presence across 484 decisions

The gates

One-year limitation
54%
Abandonment
33%
Deferral
30%
Jurisdiction
23%
Case Assessment Direction
22%
No reasonable prospect
13%

The substance

Prima facie test
7%
Poisoned work environment
5%
Undue hardship · duty to accommodate
3%
Systemic discrimination
2%
Racial profiling
0.8%
Intersectionality
0%
Zero. In 484 decisions about an institution whose docket is 40% disability and 21% race, intersectional discrimination — discrimination on combined grounds — is discussed in none of them. That is not a claim that no such discrimination occurred. It is a statement about what the published record contains.

The precedent that vanished

McKinnon is the most consequential systemic remedy ever ordered against an Ontario ministry. A 1998 merits decision found a poisoned racial environment at a detention centre. The orders were extraordinary: the ruling read at parade for five consecutive days, attached to every staff pay slip, published in the institutional newsletter, a Commission-approved training programme. The 2002 orders added third-party consultants reporting every six months, with the Deputy Minister personally responsible for implementation. The adjudicator remained seized and was still issuing compliance rulings in 2007 — roughly thirteen years of continuous supervision of a government ministry.

It has been cited by five later decisions outside its own file — in 2001, 2002, 2009, 2010 and 2016. The most far-reaching systemic order in Ontario human rights law was not developed, not distinguished, and has not been cited in a decade.

Part of the reason is structural rather than judicial. McKinnon was possible because, before 2008, the Commission received and carried every individual complaint to a Board of Inquiry as a party — which is why it is captioned Ontario Human Rights Commission v. Ontario (Correctional Services). Direct access ended that routine carriage in June 2008. The Commission still holds s. 35 and s. 37 powers and still uses them, but an individual applicant no longer arrives with a public body already in the room. The two closed doors.

What to read first, if you are representing yourself

Read in this order, because this is the order the record says you will need them.

  1. Miller, 2009 HRTO 1241 — the one-year rule. The single most likely reason a case ends.
  2. Parry Sound, 2003 SCC 42 — only if you are unionised. It explains why your case may be sent to arbitration and what that means.
  3. Dabic, 2010 HRTO 1994 — the no-reasonable-prospect test, so you know what a dismissal letter is actually asking you to answer.
  4. O'Malley, 1985 — the prima facie test. This is what you must eventually prove, and it is lower than most people assume.
  5. Noble, 2010 HRTO 878 — the reprisal test, if anything happened to you after you complained. Reprisal is a free-standing wrong.
  6. Sigrist, 2008 HRTO 34 — only if you have been dismissed and are considering reconsideration.
The practical inversion: conventional legal research starts with the substantive law and treats procedure as housekeeping. In this system that order is backwards. Four of the six cases above decide whether anyone reads your evidence. The gates, explained.

Method

Citation counts are from a network run across the full text of 484 Ontario corrections decisions, 1984–2026. Doctrine presence is a text-presence pass across the same corpus — it records whether a concept is discussed, not how it was resolved. Ratios were then recomputed independently for the Thunder Bay corpus (57 decisions) and the Durham corpus, and hold in the same direction at all three. Coding source: text-verified.

Counts describe published decisions only. Matters that settle, are withdrawn, or are resolved at mediation produce no published decision and are absent from every figure on this page.