Indigenous Data Gaps: The Number Nobody Could Source

Kishawna Peck

5 days ago

6 min read

Yesterday, on Truth and Reconciliation Day, I woke up wanting to do something that would actually help Indigenous communities. A graphic or a reshared quote wasn't it. I wanted to use what I have: I'm a business owner who makes decisions, and a founder with an audience.

I'm a Black woman and a non-Indigenous founder living and working in Toronto on the territory of many nations, including the Mississaugas of the Credit, the Anishinaabeg, the Chippewa, the Haudenosaunee, and the Wendat peoples, covered by Treaty 13 and the Dish With One Spoon Wampum Covenant.

I went back to the talks from our past Indigenous conference speakers to listen and learn. I expected answers. I came away with more questions and an uncomfortable look at how incomplete, broken, and missing the data is.

Kimberly Murray, Mohawk of Kanehsatà:ke, spoke on our stage about the impact of missing data on Indigenous communities. She was Canada's Independent Special Interlocutor for Missing Children and Unmarked Graves at the time. She told a room of data and AI professionals about her years at the Truth and Reconciliation Commission, when she asked Canada and the churches where the number 150,000 came from.

No one could tell her.

That's the number most of us know for how many First Nations, Métis, and Inuit children attended residential schools. Kimberly believes the real count is higher. Without the data lineage and methodology, the data is incomplete, and so is the story.

How do you read a number?

Every number was counted for a purpose, using a definition. Who counted, what they counted, and what they left out changes the story.

Gwen Phillips, a citizen of the Ktunaxa Nation and BC First Nations Data Governance Champion, spoke on Indigenous data sovereignty. She sits on the board of the First Nations Information Governance Centre. She told us what gets collected about First Nations is what's easiest to collect: diabetes rates, addiction rates, suicide rates. "Those are sickness indicators," she said. The Ktunaxa want to measure how healthy their governance is, how healthy their partnerships are, and how healthy their people are.

When a community leads the counting, the picture changes. The First Nations Regional Health Survey is led and governed by First Nations. In its third phase, 80.6% of First Nations adults and 76.6% of youth surveyed on reserve and in northern communities rated their sense of community belonging as somewhat or very strong.

I picked up Catherine D'Ignazio and Lauren Klein's Data Feminism again while thinking this through. Data isn't neutral, and what gets counted counts.

A low number can make a problem look small, and resources don't show up. A high number can get read as a stereotype, and nobody asks what produced it.

The 2021 census counted 1,807,250 Indigenous people, 5% of the population of Canada. First Nations make up 1,048,405, Métis 624,220, and Inuit 70,545. There are more than 600 First Nations, 50 Inuit communities, and over 70 Indigenous languages. Most data sources put all of that in one bucket.

"The world needs to know we're Ktunaxa," Gwen said. Indigenous peoples are not a monolith, and the data should say which nation, which community, which people.

The census also has gaps. It couldn't finish counting in 63 of the 1,026 census areas classified as First Nations reserves. Kahnawake, Akwesasne, and Six Nations didn't take part in 2021, so people living on reserve in those communities weren't counted.

Residential schools were federal policy, run with churches. The TRC called it cultural genocide. About 140 schools are recognized under the 2007 Indian Residential Schools Settlement Agreement, operating between 1831 and 1997. 9,471 people asked Canada to recognize an additional 1,531 institutions. Canada applied a two-part test: was the child placed there under federal authority for education, and was Canada responsible for running it? Ten institutions met it. Day schools and residences run only by provinces or churches weren't recognized.

In 2015, the TRC confirmed 3,200 deaths of children at residential schools. No name was recorded for 1 in 3. No cause of death was recorded for half. The National Centre for Truth and Reconciliation memorial register now lists over 4,100 children, and former TRC Chair Murray Sinclair said he believed the real number was higher.

200,000 Indian Affairs files were destroyed between 1936 and 1944. The surviving names sit scattered across admission forms, medical files, census records, and letters. There is no master list.

Communities are searching school by school with ground-penetrating radar. Radar finds disturbed soil, and confirming takes more work. Kimberly told us Canada has very few experts who can analyze this data. She asked our room to get involved and think about how AI could help, like telling burials apart from rocks, pipes, and roots.

The RCMP counted 1,181 Indigenous women and girls murdered or missing between 1980 and 2012: 1,017 homicide victims and 164 missing. Indigenous women were 4.3% of women in Canada and 16% of women killed.

The Native Women's Association of Canada ran Sisters in Spirit. It drew on public records and family submissions and documented 582 cases by March 2010. Federal funding ended that year, and NWAC says the real number is higher.

Maryanne Pearce, a federal civil servant, built a database for her PhD in law at the University of Ottawa. Her sources were police data, legal databases, media, and academic work. It lists 3,329 missing or murdered women and identifies 824 as Indigenous.

In 2016, federal Indigenous Affairs minister Carolyn Bennett said the true number was "way bigger than 1,200." Her office said that came from survivors and families, not new data. The National Inquiry couldn't give a final number.

Four sources, none of them final.

Police records use binary gender categories, so Two-Spirit people are invisible in the data. Officers also fail to record Indigenous identity when it relies on their own perception. There is no national database of missing and murdered First Nations, Métis, Inuit, Two-Spirit, and gender-diverse people. In May 2025, four Indigenous organizations (Pauktuutit Inuit Women of Canada, Les Femmes Michif Otipemisiwak, Ontario Native Women's Association, and 2 Spirits in Motion) announced they are building one.

Indigenous children are 7.7% of children aged 14 and under in Canada and 53.8% of children in foster care. Indigenous people are about 5% of the population and 32% of people in federal prisons. Indigenous women make up 50% of women in federal custody. In 2016, the Canadian Human Rights Tribunal found federal child welfare funding created an incentive to remove First Nations children as a first resort.

In 2015, the TRC called on governments to eliminate the overrepresentation of Indigenous people in custody within ten years. Eight years later, the Indigenous share of federal prisoners had grown by over 30%. The Yellowhead Institute's Twenty-Five Years of Gladue found the Supreme Court's 1999 direction to weigh Indigenous circumstances in sentencing hasn't worked. Over those 25 years, the Indigenous share of federal inmates grew from 12% to 33%. Yellowhead tracked the TRC's 94 Calls to Action every year until 2023, when they found zero completed that year and 81 still incomplete. They stopped writing the reports.

National race-based data on police stops doesn't exist. What we have comes from individual cities. In Toronto in 2020, Indigenous people faced 1.5 times more police enforcement than their share of the population.

I couldn't find a published count of Indigenous people working in data and AI.

What Gwen and Kimberly asked of data and AI professionals

Kimberly asked four baseline questions of any data on Indigenous peoples: who owns it, who controls it, who has access to it, and who possesses it.

She asked data and AI professionals to consider how our work can support survivors and communities leading searches for missing children. She named specific areas: cross-referencing records to build complete lists of taken children, spatial mapping of school grounds and nearby cemeteries, analyzing radar data, looking for burial patterns, and tracking personnel across institutions. She set conditions. Any work must be trauma-informed, survivor-led, and Indigenous-governed.

Gwen asked us to learn about the peoples whose territories we work on. She named the Mississaugas of the Credit and asked what we know about their governance and their vision for the future.

What Woven will do

Woven has had Indigenous speakers in two of our nine years. We offered free tickets to Indigenous attendees, and repeatedly heard travel costs kept people from coming.

Here is what we're doing:

Speak directly with Indigenous community leaders about what supports their communities need to participate in Woven programming, and act on what we hear.
Invite and include more Indigenous voices at Woven events every year, reporting how many we invited and how many joined.
Build a fund covering travel and tickets for Indigenous participants.
Intentionally include Indigenous-owned businesses in vendor searches, reporting how many we invited to quote and how many we hired.
Report on all of this every September 30.