In the past few years alone, there have been several high-profile cases of hate crimes against Canadian Muslims. On January 29, 2017, six Muslim people were shot dead in a Québec City Mosque (Austen & Smith, 2017). On September 12, 2020, a Muslim person standing outside of a mosque near Toronto was stabbed to death (CBC News, 2020). Additionally, on June 6, 2021, four Muslim people were killed by a driver who purposefully hit them with a car in London, Ontario (Westoll & LeBel, 2021). Though these are just a few high-profile incidents highlighted by Canadian media, these events are not isolated. Hate crime trends like those just mentioned have encouraged efforts by the Canadian government to “collect data to contextualize hate crime reports and to conduct needs assessments for impacted communities” (Canada Parliament, House of Commons, Member motion, 2017). These efforts include assessing different sources of information pertaining to hate crimes, including police reports and surveys.

While police-recorded hate crimes represent one source of information on the experiences of the Canadian Muslim community, additional evidence on the occurrence, nature, and impact of hate crimes is provided by self-reporting and media reports. For example, a survey of Canadian Muslims in 2002 by the Canadian Council on American Islamic Relations found that 56% of respondents experienced at least one anti-Muslim incident in the twelve months since 9/11 (CAIR-CAN, 2002). Further, the Canadian Islamic Congress (CIC, 2003) noted a 1600% increase in the incidence of anti-Muslim hate crimes reported to them following the incidents of 9/11. Comparatively, hate crimes against Muslims in the US have been found to be 5 times more common than before the 9/11 attacks (Samari, 2016). Additionally, in a poll recorded after 9/11, it was found that 60% of Americans reported adverse sentiments toward Muslims (Samari, 2016). Further, the Council on American-Islamic Relations reported that after the attacks in Paris, France on November 13, 2015, there was an increase in threats, violence, and discriminatory acts targeting Muslim Americans and those perceived as Muslim (Samari, 2016). In addition, negative media coverage contributes to the perpetuation of Islamophobic views because in social situations, Muslims are seen as the “enemy” (Considine, 2017). Further perpetuating Islamophobia, attacks committed by Muslims had an average of 449% more coverage in the media than attacks committed by non-Muslims (Considine, 2017). Negative media depictions and self-reported increases in Islamophobia post-9/11 may help explain the disproportionate hate crimes experienced by Muslims, not only in Canada but internationally as well.

Islamophobia against Muslim Canadians has been reported in the form of negative attitudes towards this marginalized population (Wilkins-Laflamme, 2018). The consequences of Islamophobia are widespread and impact many facets of well-being. Islamophobia leading to hate crimes can have negative impacts both socially and psychologically. For example, Muslim Canadians that have experienced hate crimes have reported economic burdens due to having to change jobs, having to use a car out of fear of harassment on the public transport system, and even having to change to more secure living locations (Mercier-Dalphond & Helly, 2021).

The economic burden from these hate crimes is not the only consequence of Islamophobia. Psychological distress and decreased well-being are also commonly reported effects of hate crimes (Lefevor et al., 2018; Padela & Heisler, 2011). Psychological distress is defined as non-specific symptoms of stress, anxiety, and depression such as feeling nervous, restless or fidgety, hopeless, sad, and worthless (Kessler et al., 2002; Viertiö et al., 2021). Further, psychological distress can also be defined as specific clinical features, including symptoms of depression, anxiety, and somatic symptoms (such as lack of energy or headaches) (Drapeau et al., 2012). Psychological distress can have far-reaching effects on overall health, including an increased mortality risk (Barry et al., 2020). Therefore, understanding the psychological consequences of Islamophobia and hate crimes is incredibly important.

Experiencing crime of any nature is distressing, but hate crimes are particularly damaging as victims are more likely to experience psychological distress, such as feeling nervous, anxious, or depressed; anger; and safety concerns compared to victims of other types of crime (Altemeyer, 1996; Boeckmann & Turpin-Petrosino, 2002; Gerstenfeld, 2011). For example, findings on the impacts of hate crimes against sexual and religious minorities in America indicate that victims of religiously-motivated hate crimes experience increased psychological distress (Lefevor et al., 2018; Padela & Heisler, 2011). This discrimination is not only associated with psychological distress but can lead to various other health problems like depression, anxiety, substance abuse, and suicide (Assari & Lankarani, 2017). Specifically, Lowe et al. (2018) found that discrimination against Muslim American college students was associated with a high mental health burden, with 42.9% of students being classified as having symptoms of major depression or generalized anxiety disorder.

Similarly, in Canada, there is evidence that the intersection of race and religion contributes to the experience of psychological distress. Specifically, Rousseau et al. (2011) found that psychological distress was significantly more elevated among Arab Canadians that were Muslim. However, these authors did not examine the impact of hate crimes on the psychological distress of these populations. Expanding on previous findings regarding experiences of distress among Canadian Muslims (Rousseau et al., 2011) and the mental health impacts of hate crimes on American Muslims (Lefevor et al., 2019, and Padela & Heisler, 2011), the current study assessed the psychological distress that follows hate crime experiences among Muslims in Canada. Psychological distress can have far-reaching effects on overall health (such as anxiety disorders) (Barry et al., 2020). Therefore, understanding the psychological impact of religion and race-based (such as Islamophobia motivated) hate crimes on Muslims in Canada is imperative to improving the health and well-being of racial and religious minority populations.

To develop a random sample, we used an innovative sampling strategy known as Respondent-Driven Sampling (RDS). RDS is based on Markov chain mathematics and social networking theory, and has been successfully employed by researchers seeking to identify and conduct research with hard-to-reach populations that may be widely dispersed. Briefly, RDS begins by identifying a small number (15 in this case) of “seed” respondents in the target community. These seeds must be motivated individuals who will be asked to distribute the survey to three different people in the identified community, and so on. This method is described in greater detail in Perry and Alvi (2011).

For this survey, six Canadian cities were selected (Calgary, Edmonton, Montreal, Ottawa, Toronto, and Vancouver). The researchers, onsite research assistants, and members of local Muslim organizations facilitated the identification and solicitation of 15 initial seeds per city, emphasizing diverse representation from across Muslim communities. The seeds were preferentially selected based on whether they had a large social network, were respected members of the community, and would be able to recruit others to participate in the study (Heckathorn, 1997; 2002). The seeds were provided with a link for the online survey, as well as contact information for the research team, including community collaborators.

To be eligible to participate, the respondents had to identify as Muslim. To recruit survey respondents through chain referrals, all survey respondents, including the seeds, were asked to recruit three other individuals to complete the survey as well. Information about how to refer someone to the study was included in the survey. To prevent duplicate responses, each person was assigned a unique identification number that they entered in the survey. If participants did not have access to a computer or other device, participants were able to complete the survey at a partner agency site or at the university of one of the research team members.

The survey consisted of 48 questions about experiences related to being Muslim, being the victim of a hate crime, psychological distress, and sociodemographic information. The questions were adapted from a pilot study conducted in Ontario, Canada (Perry & Alvi, 2011). The survey was available in both French and English, and took approximately 45 minutes to complete.

Participants were asked to identify their own ethnicities. These were then regrouped into 7 categories: Asian, Canadian/European, Maghreb, Middle Eastern, Sub-Saharan African, Turk, and Mixed ethnicity. Mixed ethnicity was given to any participant that indicated more than one ethnicity. These regrouped categories were recoded as dummy (binary) variables. For example, for Asian ethnicity ( yes = 1; no = 0). See Table 2 for demographics.

Participants were asked to identify their gender (man = 1, woman = 0, other = NA). There were two cases that identified their gender as other and were therefore considered as “missing” in the analyses as there were too few to consider this third group.

Participants were asked to identify their age group in years (18–24 = 1, 25–30 = 2, 31–34 = 3, 35–40 = 4, 41–44 = 5, 45–50 = 6, 51+= 7).

Participants were asked whether they were born in Canada (yes = 1 or no = 0).

Participants were asked to identify the region they were living in at the time of the survey (Calgary or Edmonton = 1, Ottawa = 2, Toronto = 3, Vancouver = 4, Montréal = 5). For the regression analyses, each of these levels were recoded into dummy variables (for example, lived in Ottawa (yes = 1; no = 0)).

Participants were asked to indicate their highest educational attainment (No certificate, diploma, or degree = 1; High school diploma or equivalent = 2; Trades or apprenticeship certificate or diploma = 3; Certificate or diploma from a college, CEGEP, or other = 4; University certificate or diploma below bachelor’s degree = 5; Bachelor’s university degree = 6; Master’s university degree = 7; Degree in medicine, dentistry, veterinary medicine = 8; Doctorate degree = 9).

Participants were asked if they felt they were easily identifiable as Muslim (yes = 1; no = 0).

Participants were asked a total of 18 questions on how many times in the last year and since their arrival to Canada had they been verbally (for example, “received phone calls, texts, or emails that were humiliating or insulting”) or physically assaulted (for example, “had things purposely thrown at you”). If participants stated that an event occurred greater than 0 times to any of the verbal aggression questions, this was coded as “yes” for being the victim of verbal aggression (yes = 1; no = 0). If the participants answered “yes” to any of the physical assault questions, this was coded as “yes” for being the victim of physical assault (yes = 1; no = 0). See Table 3 for reported frequencies of victimization across the different sociodemographic characteristics.

This measure consisted of 14 items. Participants were asked how they felt after being a victim of verbal aggression or physical assault. They were asked to answer Yes, No, or I don’t know to statements such as, “I felt afraid.” An exploratory factor analysis revealed that five items should be retained to create a factor of psychological distress (for example, the factor loadings were all greater than .5, which is considered fair–good; Tabachnick & Fidell, 2007). These items with frequencies in brackets were: 1) “I felt ashamed/guilty” (18%); 2) “I felt afraid” (39%); 3) “I felt depressed” (36%); 4) “I lost confidence” (30%); and 5) “I was less productive at work” (21%).

A factor score was generated for each participant based on these items (M = −0.03, SD = 1.00). The higher the score, the more distress the participant experienced.

The first set of analyses included descriptive statistics for the demographic characteristics of those that reported being a victim of verbal aggression or physical assault. To address the research questions, three regression models were used. To address the first research questions (1a and 1b), two separate logistic regression models were estimated with verbal aggression and physical assault as the outcome variables. Gender; age group; education; ethnic group; self-identifying as visibly Muslim, Immigrant, or Canadian born; and the geographical region they are in were the independent variables. To address the second research question, a third linear regression model predicting psychological distress included the same independent variables with the addition of verbal aggression and physical assault as independent variables.

The sample consisted of 231 participants². The percentages reported in text are the rounded valid percentages (excluding missing cases). The sample had slightly more women (58%) than men (41%). Two participants reported not identifying with either man or woman (1%), but given the low case count, they were excluded from the regressions. Within the sample, 74% were not born in Canada. The ethnicities in order of decreasing frequency were Asian (27%), Maghreb (19%), Mixed Ethnicity (19%), Middle Eastern (16%), Sub-Saharan African (6%), Turk (8%), and Canadian/European (4%). Most of the participants were younger than 40 years old (72%). The sample had overall high levels of education. For example, 35% reported having a Bachelor’s degree, 32% reported having a Master’s degree, and 9% reported having a Doctoral degree. Approximately 1% of the sample reported not having any degree.

In terms of other characteristics, 64% of the sample reported that they self-identified as visibly Muslim. See Table 2 for more demographic information and how the key variables compare to the population of Muslims in Canada.

In the current sample, 74% reported being a victim of verbal aggression, and 63% reported being a victim of physical assault. To further understand the socio-demographics of those that experienced assaults, cross-tabulations were used (see Table 3). The key observations from the current sample are that higher levels of verbal aggression or physical assaults respectively were observed in women (63%; 61%), those not born in Canada (73%; 73%), persons residing in Montréal (28%; 28%), persons between the ages 18–24 (21%; 22%) and between the ages 35–40 (22%; 21%), those self-identifying as visibly Muslim (66%; 69%), and persons identifying as Asian (26%; 29%). See Table 3 for hate crime frequencies presented by these independent variables.

In the logistic regression model with verbal aggression as the outcome, the model fit was χ² = 27.11, df = 15, p = .028, r² = .15 (Cox & Snell R-Square). There were no significant associations between sociodemographic characteristics and having been a victim of verbal aggression. See Table 4a.

In the logistic regression model with physical assault as the outcome, the model fit was χ² = 31.37, df = 15, p = .008, r² = .17 (Cox & Snell R-Square). The only significant associations with the likelihood of being a victim of physical assault were self-identifying as visibly Muslim and living in Vancouver. Specifically, when the participant self-identified as visibly Muslim, they were three times more likely to have experienced a physical assault compared to those that did not self-identify as visibly Muslim (b = 1.18, OR = 3.27, p = .004). If the participant lived in Vancouver, they were nine times more likely to have experienced a physical assault compared to living in Montréal (b = 2.24, OR = 9.42, p = .009). See Table 4b.

Participants were asked to report on how they felt psychologically after experiencing verbal aggression or physical assaults. The final measure as described above represents a factor score derived from the following feelings: feeling ashamed/guilty, feeling afraid; feeling depressed; feeling a loss of confidence; and feeling less productive at work.

In the regression model with psychological distress as the outcome, the model fit was F(17, 104) = 3.099, df = 17, p <.001, r² = .34. The significant associations with psychological distress were being a woman (coded as zero) (b = −.49, p = .007), self-identifying as visibly Muslim (b = .46, p = .018), residing in Vancouver (b = .84, p = .015), and being a victim of physical assault (b = .70, p < .001). See Table 4c.