Rural crime rates by Indigenous nation in Canada (2019 to 2023)

I often emphasize that sample size is crucial in geography. If the sample size isn't appropriate, the data becomes too vague. This is an issue in how people discuss crime statistics, where the sample sizes are usually too large. In one of my previous articles, I compared citywide crime rates with neighborhood-level rates within the same cities. You'll find that cities are too large of a sample size to provide meaningful crime statistics.

Similarly, when discussing Australia, Canada, the United States, or any country colonized by Europeans, Indigenous populations are often treated as homogeneous: with the same culture, history and issues to solve. This assumption is flawed: it is an overly large sample size.

I wanted to explore violence in rural Indigenous communities in Canada. Indigenous reservations are often generalized to all be the same, assuming that a rural community on the Pacific Coast is the same as one in Northern Saskatchewan, Southern Ontario, or Baffin Island. Even distinctions between First Nations and Inuit can be too broad. For instance, crime rates can vary significantly between different Inuit regions or among different groups of Cree.

While I looked for similar data for the United States and Australia, it wasn't available in the same detailed manner as it was for Canada.

The table above shows crime statistics per 100,000 residents, the standard metric for understanding the frequency of certain crimes relative to population size.

The crime data spans from 2019 to 2023, offering five years of information. 

It's important to note that crimes don't always auto-correlate well year-to-year. This means looking at just one year of crime data can be misleading, as it might represent an unusually low or high year. Averaging multiple years of data provides a more accurate picture of crime trends than focusing on individual annual data.

The violence score uses the same calculation and formula as my article on neighborhood crime (linked in the first paragraph).

The map below shows the general territory area for each Indigenous nation and it is color coated to match the violence score. It uses the same color gradient scale that the neighborhood crime rates charts have. 


Note on Robbery

One interesting thing to note is the low robbery rates in rural areas. Robbery seems to be an urban crime. In rural areas, fewer people and stores mean fewer opportunities for robbery, and the close-knit nature of smaller towns makes it harder to get away with it. Notice how low the robbery rates are in these rural communities compared to other violent crimes like sexual assault and major assaults, which are much higher. In contrast, urban neighborhoods have very different major assault-to-robbery ratios.

In the future, I would like to do an analysis of robbery in urban areas, suburban areas and rural areas to then see how much more frequently robbery is in urban areas when normalizing for general violence (major assaults). For example, if an urban area has a major assault rate of say, 1,000, what can we expect its robbery rate to be? And the same question for suburban areas and rural areas. 

Property crimes are also available for review, but I did not conduct additional analysis on them.


Additional Notes

1. Not all rural areas have detailed crime data from StatCan, especially for specific smaller communities. Many police agencies cluster towns together in their data, often combining Indigenous and non-indigenous towns. This practice presents challenges, particularly in provinces like Alberta and British Columbia. 

For instance, despite British Columbia having around 180,000 First Nations people, many First Nations settlements were clustered with non-Indigenous towns by police agencies, which left me with only about 12,000 people in useable data. In contrast, Manitoba, with a First Nations population of approximately 135,000, had useable data for about 55,000 people due to less clustering, or different clustering approaches.


2. Due to clustering, not all areas are entirely composed of a single nation or ethnicity. For instance, a small portion of the Plains Cree population includes some Nakoda people because the File Hills First Nations Police serves mostly Plains Cree communities but also serves some Nakoda communities. Similarly, a small percentage of the Saulteaux demographic is Sioux people because the Manitoba First Nations Police Force data serves some Sioux and Dakota communities: about 10% of the people it serves are in Sioux and Dakota communities, which is about 900 people.


3. I used the term "Michif" instead of "Métis" because Métis can refer to the Red River Métis who speak Michif or to someone of mixed European (of any type) and Indigenous ancestry (of any type). If a community predominantly consists of Métis and the second most spoken language is Michif, it is a Red River Métis community, a distinct ethnicity with its own language and culture, rather than just a racial classification.

Most Métis people live in cities or live in predominantly European towns. The only predominantly Métis towns, with crime data on StatCan, were in Saskatchewan.


4. Major assault includes: assault 2, assault 3, attempted murder and use of firearms (discharge or  pointing). The reason for this is because it is equivalent to the "Aggravated Assault" definition that the USA uses.


Comments

Popular posts from this blog

Crime rates at the neighborhood level: American, Canadian neighborhoods

Homicide rates: are they more effected by violence in general, or just gun violence?