Counted but Not Accounted For

Introducing

Counted but Not Accounted For

A First Look Beyond the 2022 Canadian Survey on Disability Headline

A UN CRPD Article 31 Disability-Data Accountability Tool

Created by ABILITYscaled

This plain language first look introduces Counted but Not Accounted For, an Article 31 accountability tool developed by ABILITYscaled. Article 31 is the legal obligation on statistics and data collection under the United Nations Convention on the Rights of Persons with Disabilities (CRPD), to which Canada is a State Party. The tool asks whether disability data is protected, disaggregated, accessible and used to identify barriers, assign responsibility and support remedies. Grounded in Canada, the method can be adapted for CRPD State Parties with Westminster-style public law structures.

The core message is simple: being counted is not the same as being accounted for.

The Canadian signal cards below are the first application of the tool. In this document, the Canadian Survey on Disability (CSD) refers to Statistics Canada’s national survey for estimating disability among people aged 15 and over in Canada. Its 2022 headline reported that 27% of Canadians aged 15 and over — about 8.0 million people — had one or more disabilities that limited daily activities.

That headline is essential, but it is not a complete accountability map. Some people at high rights risk are outside the CSD’s age or household coverage; others appear only in benefit, school, hospital, shelter, long-term care, correctional or other administrative systems. Visibility in a dataset does not necessarily mean protection, support, accommodation or remedy.

Counted but Not Accounted For helps reviewers, advocates, public bodies, parliaments, national human rights institutions and monitoring bodies move from prevalence reporting to accountability. It asks who is missing, which barriers are not measured, who is responsible, which rights are affected and what remedy or corrective action should follow.

Data sources vary by country, but the accountability test does not. A CRPD State Party must be able to show whether disability data is protected, disaggregated, accessible and used to identify barriers, assess rights implementation and support corrective action.

People at highest rights risk may appear in surveys, benefit files, school, hospital, shelter, long-term care or correctional records and still be denied accommodation, support, safety or remedy. When children under 15, institutionalized people, people under correctional control, people experiencing homelessness, people blocked by benefit barriers or people facing violence are missing, under-measured or visible only through fragmented records, data does not prove inclusion; it can hide exclusion.

1. How to read this first look

The full tool applies an Article 31 accountability test to disability data. It asks whether the data is protected, disaggregated and accessible, and whether it is used to identify barriers, assign responsibility and support remedies or corrective action. This Canadian first look previews that test through one practical question: who is missing, under-measured or visible only through fragmented systems behind the headline numbers?

Read the figures as estimates, floors and signals, not as a final national count. The signal cards use available public numbers to show where disability data still fails to account for people at high rights risk.

Note: In other CRPD State Parties with Westminster-style public law structures, the same signal-card approach can be rebuilt using local disability survey, census, benefit, institutional, justice, corrections, housing, health, education, safety and violence data.

2. Accountability signals behind the Canadian Survey on Disability (CSD) headline

The Canadian Survey on Disability (CSD) headline is an essential starting point, but not a complete accountability map. The signal cards below translate that gap into estimates, floors, proxies and warning signs. Each card identifies a population, system or rights-risk area where people with disabilities may be outside the adult survey frame, under-measured in collective or administrative systems, or visible only through fragmented records. Together, they show why Article 31 requires more than prevalence: it requires data that connects people to barriers, responsible actors, safeguards and remedies.

The figures are not additive. They come from different systems and are used to show accountability signals, not to produce one combined total.

2.a Populations outside or below the adult CSD headline

1 · EXCLUSION SIGNAL · POPULATION OUTSIDE CSD

400,000 Children under 15

Children under 15 with disabilities or disability-related functional limitations who are outside the adult CSD headline.

Source logic: The adult CSD headline covers persons aged 15+; child-disability and functional-limitation evidence is used as a first-look estimate.

Rights risk: Child supports, school accommodations, family supports, safeguards and remedies may be invisible when adult prevalence reporting is treated as the whole disability picture.

Reference anchor: Statistics Canada’s 2022 CSD materials establish the adult survey frame and 15+ coverage; Statistics Canada child disability/functional-limitation sources support the separate first-look child signal.

Calculation note: The 400,000 figure is used as a first-look estimate for children under 15 with disabilities or disability-related functional limitations; the full calculation note will accompany the technical release.

Authorities: Statistics Canada, Canadian Survey on Disability, 2017 to 2022; Statistics Canada, Canadian Survey on Disability, 2022: Concepts and Methods Guide; Statistics Canada child disability and functional-limitation sources; full citations and technical notes will accompany the technical release.

2 · INSTITUTIONAL FLOOR · COLLECTIVE SETTINGS

160,000+ Collective settings and long-term care

Disabled people aged 65+ in collective dwellings, including long-term care and other institutional or collective settings.

Source logic: Collective-dwelling and age-based public data are used as a conservative institutional floor.

Rights risk: The full institutional and long-term-care population may face hidden support, consent, accommodation, complaint and remedy gaps.

Reference anchor: Statistics Canada collective-dwelling, age and disability-prevalence sources support this institutional and long-term-care floor.

Calculation note: The 160,000+ figure is a conservative floor for disabled people aged 65+ in collective dwellings, not the full institutional or long-term-care disability population.

Authorities: Statistics Canada collective-dwelling and disability-prevalence materials, including CSD and census-related sources; full citations will accompany the technical release.

3 · GOVERNANCE SIGNAL · FIRST NATIONS, INUIT AND MÉTIS DATA GOVERNANCE

300,000+ First Nations, Inuit and Métis data governance

Estimated First Nations, Inuit and Métis persons with disabilities or disability-related functional limitations, used as a governance-sensitive signal rather than a prevalence count.

Source logic: Public Indigenous population and disability-prevalence evidence is used as a cautious first-look signal, not a final count.

Rights risk: Accountability must respect Indigenous data sovereignty and community control over collection, interpretation, sharing and use.

Reference anchor: Statistics Canada Indigenous population and disability materials support the signal; OCAP®, Inuit research guidance and the UN Declaration Act support the governance requirement.

Calculation note: The 300,000+ figure is a cautious first-look signal and does not replace distinctions-based data governance.

Authorities: Statistics Canada Indigenous population and disability sources; First Nations Information Governance Centre, The First Nations Principles of OCAP®; Inuit Tapiriit Kanatami, National Inuit Strategy on Research; United Nations Declaration on the Rights of Indigenous Peoples Act.

2.b Correctional and state-control data gaps

4 · CUSTODY ESTIMATE · ADULT CUSTODY

9,100 Adult custody

Estimated disabled adults in custody: about 3,400 federal and 5,700 provincial/territorial.

Source logic: Public correctional average counts are combined with the 2022 CSD adult disability prevalence rate.

Rights risk: Baseline only; it does not directly count mental health, cognitive disability, accommodation needs, remand, youth custody, supervision or release supports.

Reference anchor: Statistics Canada and federal correctional data support the custody-count base.

Calculation note: The 9,100 estimate applies the 2022 CSD adult disability prevalence rate to adult custody counts; it is not a direct disability count.

Authorities: Statistics Canada, Adult and Youth Correctional Statistics, 2022/2023; correctional data tables; CSD 2017 to 2022; Public Safety Canada, Corrections and Conditional Release Statistical Overview 2022.

5 · DATA GAP · CORRECTIONAL CONTROL

No public denominator Correctional disability data gap

No single public disability denominator spans custody, remand, youth custody, community supervision, accommodations, complaints and release supports.

Source logic: Correctional data is dispersed across custody, supervision, accommodation, complaint and release systems.

Rights risk: Disability, accommodation, complaint, release-support and remedy needs cannot be directly counted from public data.

Calculation note: No national figure is calculated; the signal is the absence of a public rights-based disability denominator.

Authorities: Statistics Canada correctional statistics and data tables; Public Safety Canada; Correctional Service Canada and oversight materials.

6 · STATE-CONTROL SIGNAL · COMMUNITY SUPERVISION

25,000+ Community supervision

Likely disabled adults under probation, parole, statutory release or conditional sentences; likely higher because disability is not directly counted.

Source logic: Public community-supervision counts are combined with the 2022 CSD adult disability prevalence rate.

Rights risk: State control outside custody can affect breach risk, supervision conditions, supports, accommodations and remedies.

Calculation note: The figure is a state-control signal, not a direct disability count.

Authorities: Statistics Canada correctional services data; Public Safety Canada; Statistics Canada CSD materials.

7 · ACCOMMODATION-RISK SIGNAL · FEDERAL CUSTODY

70% / 79.2% Federal custody mental-health and accommodation risk

CSC evidence reports mental disorders among over 70% of federally sentenced men and 79.2% of federally sentenced women in the sample.

Source logic: CSC mental-disorder prevalence evidence is used as an accommodation-risk signal, not a custody count.

Rights risk: Adult custody estimates may understate mental health, cognitive disability and accommodation needs.

Calculation note: No separate population total is calculated; the percentages show accommodation-risk pressure.

Authorities: Correctional Service Canada mental-disorder prevalence studies and correctional oversight materials.

2.c Crisis, access and safety signals

8 · CRISIS SIGNAL · HOMELESSNESS

40,713 / nearly 30,000 Homelessness and hospital crisis

40,713 visibly homeless people counted; nearly 30,000 hospitalizations recorded homelessness as a crisis factor.

Source logic: Visible homelessness counts and hospital-use data are crisis signals, not full disability counts.

Rights risk: Hidden homelessness, inaccessible shelters, unmet health needs and missing supports may remain outside headline data.

Calculation note: The figures come from different systems, are not added together and are not direct disability counts.

Authorities: Housing, Infrastructure and Communities Canada, Everyone Counts 2024; Canadian Institute for Health Information homelessness and hospital-use materials.

9 · ACCESS SIGNAL · BENEFIT ACCESS

6.7 million Benefit access

Up to 6.7 million people form a benefit non-use and access-barrier signal, where formal uptake may understate disability need.

Source logic: Benefit uptake and non-use evidence are used as an access signal.

Rights risk: Benefit gates, eligibility rules and administrative barriers can hide people who need income support, accommodation or remedies.

Calculation note: The figure is a first-look access signal, not a direct count of unmet benefit need, and should not be added to other signals.

Authorities: Statistics Canada disability income profile and CSD data tables; Employment and Social Development Canada materials.

10 · EMPLOYMENT BARRIER SIGNAL · WORK AND SUPPORTS

1 million+ Work and supports

Disabled people not working but able to work with the right supports, accommodations or barrier removal.

Source logic: Labour-market and support-needs data are used as an employment barrier signal.

Rights risk: Accommodation, support, transport and discrimination barriers may keep people out of work despite capacity and willingness.

Calculation note: The figure is a first-look employment barrier signal, not a complete count of labour-market exclusion.

Authorities: Employment and Social Development Canada employment strategy progress report; Statistics Canada disability employment and income profile; CSD data tables.

11 · SAFETY SIGNAL · VIOLENCE AND REMEDIES

55% Women with disabilities and intimate-partner violence

Women with disabilities reporting lifetime intimate-partner violence, used as a safety, access-to-services and remedies signal.

Source logic: Violence-prevalence evidence is used as a safety and remedy signal.

Rights risk: Prevalence data alone does not show access to shelters, services, complaints, protection, accommodations or remedies.

Calculation note: The figure is a safety and remedies signal, not a complete measure of violence, service access or protection outcomes.

Authorities: Laura Savage, Intimate Partner Violence: Experiences of Women with Disabilities in Canada, 2018, Statistics Canada, Juristat; ESDC CRPD Article 16 indicators.

3. Data assembly protocol: crosswalk and matrix

These signal cards identify missing, under-measured or rights-risk populations behind the adult CSD headline. Cards are included where public data points to exclusion, under-measurement, fragmented visibility, a missing rights-based disability denominator or high rights risk.

The figures were assembled through a crosswalk-and-matrix process. The crosswalk asks where disability-related data is held or missing across systems and, where First Nations, Inuit or Métis people or data are concerned, who governs that data. It covers surveys, censuses, benefits, institutions, justice, corrections, housing, health, education, safety and violence data. The matrix identifies who is responsible for acting on what the data reveals.

Each figure is treated by evidentiary role: direct count, estimate, conservative floor, proxy or warning signal. For correctional systems, this means separating four signals rather than treating correctional control as one number: adult-custody disability estimate, correctional data gap, federal mental-health and accommodation-risk evidence, and community supervision outside custody counts.

The crosswalk keeps the cards from becoming a loose list of statistics by showing which system holds the information and where gaps appear. The matrix keeps the analysis from stopping at measurement by linking each data gap to responsibility, implementation duties, oversight and corrective action. Together, they show how each figure is used and what accountability response should follow.

The figures are not added together because the purpose is not one grand total. The purpose is to show where public data fails to account for people at rights risk and to identify actors who should explain, remedy or monitor that failure. Under Article 31, a missing rights-based disability denominator is not a technical footnote; it is part of the accountability gap.

4. A safeguarded approach to data

This tool does not call for more data at any cost. It rejects surveillance, forced disclosure, Indigenous data extraction, immigration-enforcement misuse, outing, retaliation, stigma and service loss.

Rights-based disability data must be protected, purposeful, accessible and connected to remedies. Sensitive data should be collected, linked, disaggregated or reported only where safeguards are in place: privacy protection, safe disaggregation, Indigenous data governance, immigration-status protections, anti-outing protections, accessible complaint pathways and remedies for misuse.

In Canada, this includes respecting First Nations, Inuit and Métis data sovereignty and distinctions-based governance, including community control over how Indigenous disability data is collected, interpreted, shared and used.

5. What makes this tool different

Counted but Not Accounted For is designed to connect disability data to accountability, not simply to describe prevalence. It combines:

• a plain-language thesis: being counted is not the same as being accounted for;
• an Article 31 accountability test for protected, disaggregated, accessible and rights-linked disability data;
• a rights-risk population audit that identifies who is missing, under-measured or visible only through fragmented systems;
• a dataset crosswalk and actor-by-actor matrix showing where information is held, where gaps appear, who is responsible and what action should follow;
• findings, requested remedies and implementation details for public bodies, advocates, reviewers and monitoring institutions; and
• a concise CRPD Articles 31 and 33 summary for use in reporting, monitoring and implementation.

It can be used in Canada as a CRPD shadow-report annex, policy audit framework, monitoring tool, briefing document or implementation checklist, and adapted for other CRPD State Parties with Westminster-style public law structures.

The difference is that the tool asks not only how many people are counted, but whether the data shows who is responsible, what rights are affected and what corrective action follows.

6. The ask

Governments should publish and maintain an Article 31 disability-data accountability framework. In Canada, and in other CRPD State Parties with Westminster-style public law structures, that framework should show who is excluded or under-measured, identify barriers, name responsible authorities, protect sensitive data, measure rights implementation and outcomes, and report corrective action.

The answer is not a national disability registry, mass disclosure or a demand for perfect data. It is a rights-based public account of whether persons with disabilities are not only counted, but protected, supported, accommodated and remedied in practice.

7. Early release and next steps

This plain-language first look previews the framing, signal-card method and possible adaptations of the larger Counted but Not Accounted For tool. The complete tool, related reports and supporting materials are planned for release later this summer. Questions, feedback and connections are welcome via LinkedIn or by email at Lance@Abilityscaled.com.

Planned next steps include the complete tool release, related plain-language reports, French-language materials, additional language versions as capacity allows, jurisdiction-specific adaptations, and selected guidance or implementation support.

This document may be shared under the licence terms. Adaptation requires permission where the licence does not allow derivatives. The aim is accessible, rights-based disability-data accountability while protecting the integrity of the main tool and plain-language framing.

8. Source note

The signal cards include source logic, calculation notes and authority anchors. The authorities listed in the cards are first-look anchors, not full citations. Full citations, technical notes and the CRPD Articles 31 and 33 framework will accompany the complete tool release.

Counted but Not Accounted For is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International Licence, unless otherwise stated.

Original cover text

ABILITYscaled

COUNTED BUT NOT ACCOUNTED FOR

An Article 31 Disability Data Accountability Tool

Developed by ABILITYscaled

Previewing a broader Article 31 Disability Data Accountability Framework

CANADIAN FIRST LOOK

Who Is Missing from the CSD Headline?

Children under 15 outside the adult CSD headline	~400,000
Institutional & long-term care populations	160,000+
Custody & incarceration (adults)	9,100
Community supervision (probation, parole, release)	Tens of thousands
Homelessness signals (shelters & rough sleeping)	40,713 + 30,000
Benefit-access gaps (missing from administrative data)	Up to 6.7 million
Work-support barriers & accommodation gaps	1 million+
Violence & safety risks (especially for women & girls)	55% higher risk of IPV

PART OF A LARGER FRAMEWORK

This short-form document demonstrates one example output from the broader Counted but Not Accounted For Article 31 Accountability Framework.

FRAMEWORK COMPONENTS

• Data Systems
• Rights Monitoring
• Safeguards
• Responsible Authorities
• Remedies & Corrective Action

THIS DOCUMENT

A first look at who is missing from the Canadian Survey on Disability headline.

Developed by ABILITYscaled

Building disability data systems that are inclusive, rights-based, transparent and accountable.