What should IA practitioners applying culturally relevant GBA Plus do when local disaggregated data about the population(s) of interest are not available?

Disaggregating data is the process of breaking down data into sub-categories, which are grouped together based on shared characteristics. Examples of categories include gender, sex, geographic location, race, ability, Indigenous status, level of education, etc. The collection of data that can be disaggregated according to these and/or relevant categories helps proponents better understand how certain segments of a population might be impacted differently by a resource project and thus provide appropriate mitigations. 

When disaggregated data are not available, proponents can reach out to members of a community, and search for other relevant research, to understand what disaggregated data would likely uncover.  For example, in the case of the Kudz de Keyah project, BMC Minerals referenced the Government of Canada’s MMIWG report and community input from Indigenous women-led organizations to help establish conditions for Indigenous women’s safety due to the unavailability of disaggregated data about Indigenous women. Proponents should make similar efforts to learn about necessary conditions for ensuring the safety and inclusion of other groups, such as people with disabilities, lone mothers, and others.

Additional detail

When disaggregated data for a population of interest is not available, IA practitioners can start by engaging with that group (e.g., women, girls, Two-Spirit and gender-diverse peoples, Elders, people with disabilities and others) by either inviting them to participate in the conversation or consulting with local representative organizations. If the community of interest is willing to engage, proponents should move forward by using data collection parameters established by or with the community. This means that that proponents should only enter community spaces and regions as permitted by the community. Another approach for data collection is reaching out to local non-profits and organizations that support the community, as they may collect their own data about the individuals who access their services. They may also be able to offer a nuanced perspective on the social, political, economic and environmental challenges facing the community. 

If the communities affected by the project do not want to provide data, IA practitioners can rely on relevant secondary data. To ensure meaningful insights are drawn from secondary data, practitioners can still engage directly with the community or groups within the community. If initial attempts to gather data are met with reluctance, practitioners should assess whether the engagement approach or relationship-building efforts need improvement. Working with trusted local organizations can facilitate this process, but when individuals are the focus, their hesitancy might stem from insufficient trust or poor engagement strategies. Taking time to establish genuine relationships, listen to concerns, and ensure culturally appropriate and inclusive outreach methods can help increase participation. However, when engaging with marginalized communities, it is crucial to recognize that they may choose not to share information to avoid jeopardizing their safety. In such situations, secondary information sources can provide a valuable and reliable alternative. These sources, which often represent an established body of evidence, can help fill knowledge gaps while ensuring the protection of marginalized groups. Prioritizing their safety over direct data collection reflects a thoughtful and ethical approach to gathering information. When secondary data are needed, proponents should prioritize sources from individuals or organizations that share geographic, gender, or racial relevance to the community, ensuring the data resonates as much as possible with local lived experiences. It is important that culturally relevant GBA Plus practitioners and researchers provide a clear rationale to explain the decision to use regional or extrapolated data so that the analysis and assumptions made about the population under study are transparent.  

Examples

Including data collection as part of the project

The Woodfibre LNG Gender and Cultural Safety Plan report (2024) notes that while data on gender-based violence and harassment have been collected in the areas of other industrial projects in the province, data do not exist for the Squamish community. The plan therefore includes funding research to collect Squamish-specific data about gender and sexual violence and harassment during the construction phase of the project. The report explains that “the findings will provide insight to better understand trends and the relationship between industrial development projects and violence against Indigenous women, girls and 2SLGBTQIA+ People” (p. 53). 

Supplementing with interviews

The Cariboo Gold project GBA Plus used BC statistics to include information at the provincal level on police-reported sexual assaults, including higher rates for Indigenous women and women with disabilities. It also provided some specific details about rates of intimate partner violence in Quesnel (the largest town in the region where most services were located). It also included some contextual information about services in the area for “women and diverse groups” that includes information on the women’s resource centre, women’s clinic, domestic violence shelters (among others). While limited, the report notes that it conducted a phone interview with the women’s centre and obtained some supplemental information on existing use of the centre and the capacity for an increase in use. While the data in this report are limited, this highlights that key informant interviews can provide important community-level information that could be useful in an analysis, even if it is not officially published data. 

Imperfect proxies

The African Nova Scotia Prosperity Index Report explains that there are little data specifically about the African Nova Scotian population. Where available, it draws on statistics that include self-identification as Black; however, this is not the same as the African Nova Scotian population. In some instances, they were able to cross tabulate to include people who self-identify as Black and as third generation Canadian or more. In the report, they use this where possible as a proxy (albeit imperfect) for the historic African Nova Scotian population. This provides the ability to do some level of analysis specifically on African Nova Scotian populations while also highlighting the need for more specific data collection. 

Citations and useful resources

• Beyers, D. W. (1998). Causal Inference in Environmental Impact Studies. Journal of the North American Benthological Society, 17(3), 367–373. Available at : https://doi.org/10.2307/1468339.
• Nakaizumi, T. (2022). Role of Causal Inference in IA. Contributions to Economics, 105–110. Available at : https://doi.org/10.1007/978-981-19-5494-8_8.
• Vitus, K. (2008). The Agonistic Approach: Reframing Resistance in Qualitative Research. Qualitative Inquiry, 14(3), 466–488. Available at : https://doi.org/10.1177/1077800407309331.
• Woodfibre LNG. (2024). Gender and Cultural Safety Plan. Available at : https://projects.eao.gov.bc.ca/api/public/document/6644e69624b6ce0022ff52b9/download/Woodfibre%20LNG%20Gender%20and%20Cultural%20Safety%20Plan%20v8%20May13%20final%20clean.pdf.
• Yukon Environmental and Socio-economic Board. (2024). Decision Document – Kuz ze Keyah Project. Available at : https://yesabregistry.ca/projects/5942a72b-b77d-403d-83d6-bc2ffffc0c7b/activity.