Statistics Canada yesterday released a report on violence against women that consists mostly of previously released data from various sources. It also includes some new data on women living in the three northern territories.
The GSS [General Social Survey] showed that women in the territories report higher rates of spousal violence than those living in the provinces. Police statistics also indicate that women in the territories also experience higher levels of sexual assault and homicide.
According to the GSS, spousal violence in the territories came to the attention of the police more often than violence occurring in other parts of Canada. In addition, the use of shelters in the territories is the highest in the country.
Overall, 12% of adult residents of the three territories who had been married at some point in their life, or who had lived in a common-law relationship, had experienced violence by a spousal partner in the five years prior to the survey. This compares with 7% in the provinces.
Police data show that rates of homicide in the territories are the highest in the country and that rates of sexual assault are also higher than in the provinces, although rates of sexual assault have declined in all three territories in recent years.
The report combines victimisation data gathered via social surveys and data on crimes reported to the police.
Since this study has been covered in various Canadian news sources (see, e.g., CBC and Globe and Mail), the rest of this post is devoted to a very serious shortcoming in the analysis.
The report fails to distinguish women who are living with a husband to whom they are legally married, women who are in a common-law relationship, and women who are separated from a spouse. Studies in many different countries and localities show that women in these different situations face widely varying levels of risk of violence from their spouse/partner. Statistics Canada, however, provided almost no related statistical analysis in a 97-page document devoted to analysis of violence against Canadian women.
I find this, frankly, baffling. My bewilderment is compounded by Statistics Canada's disclosure in this same report that women who are separated or living common-law are indeed at significantly greater risk of spousal violence. From the section of the Executive Summary headed "Risk factors", p. 14 of the full study (pdf):
Risk Factors
- Young women experience the highest rates of violence.
- Women experience higher rates than men of sexual assault, stalking, serious spousal assaults and spousal homicide.
- Partners' use of psychological or emotional abuse, and frequent heavy drinking by partners, raise the risk of violence against women in spousal relationships.
- Women in common-law relationships and those who are separated report rates of spousal violence and homicide that are disproportionate to their representation in the population.
- Stalking by ex-partners raises the risk of ex-partner violence.
Nowhere in the statistical tables and charts that make up the bulk of the report is the bolded statement corroborated. And here I thought the purpose of an executive summary was to summarise findings documented in the full report. Not this time, apparently.
The report does contain analysis of spousal homicide victims by detailed accused-victim relationship (Fig. 25, p. 38), but the statement in the Executive Summary refers more broadly to spousal violence.
Finally, there's this sentence in the conclusion of the report:
Women living in common-law unions are at higher risk of assault and homicide by their partners than married women.
The report substantiates the conclusion with respect to homicide, but not assault—a much more prevalent form of violence.
Clearly, Statistics Canada has spousal-violence data of sufficient quantity and quality to differentiate married women from those separated or living common-law, else the statement in the Executive Summary could not be made. Why, then, does the report itself not provide statistical documentation for its own executive summary? Why did Statistics Canada not publish relevant and important analysis that it clearly has the capability and opportunity to undertake—indeed, in this case, that it implicitly admits having already done?
Sources:
Statistics Canada, 2006. Measuring Violence Against Women: Statistical Trends, 2006. Statistics Canada catalogue no. 85-570-XIE.http://www.statcan.ca/english/research/85-570-XIE/85-570-XIE2006001.pdf (accessed 2 October 2006)
Statistics Canada, 2006. "Violence Against Women: Statistical Trends." The Daily, 2 October. Statistics Canada catalogue no. 11-001-XIE.
http://www.statcan.ca/Daily/English/061002/d061002a.htm (accessed 2 October 2006).
Previous related post: A primer on crime statistics