Demographic diversity, perceived workplace discrimination and workers' well-being: context matters

Abstract

Purpose

The main objective of this article is to contribute to the advancement of scientific knowledge on the determinants of perceived workplace discrimination and its consequences on workers' well-being in Canada.

Design/methodology/approach

The authors used a representative sample of 7,706 workers aged 18 to 65 based on data from the 2016 General Social Survey of Canadians at Work and Home to conduct logistic regression models.

Findings

Women and visible minorities are at greater risk of perceiving that they have experienced workplace discrimination, but immigrants' perceived workplace discrimination risk is no different from that of non-immigrants. This risk is higher in public administration than in other industries and varies between provinces. Perceived workplace discrimination increases stress and is associated with a lower level of self-reported mental health.

Practical implications

Since perceived discrimination has a detrimental effect on workers' well-being, organizations should pay special attention to their employees’ perceptions. Relying only on official complaints of discrimination can lead organizations to underestimate this issue because many employees are not inclined to file an official complaint, even if they believe they have been discriminated against.

Originality/value

The authors findings are original because they suggest that visible socio-demographic characteristics (gender and visible minority) affect perceived workplace discrimination, which is not the case for invisible socio-demographic characteristics (immigrant). They point out that the province of residence is an element of the context to be considered and they indicate that workers in the public sector are more likely to perceive discrimination than those in other industries. These empirical contributions highlight that, despite anti-discrimination laws and government efforts to promote equity, diversity and inclusion, perceived workplace discrimination persists in Canada, particularly among women and visible minorities and it has tangible impacts on the workers' well-being.

Citation

Boulet, M., Lachapelle, M. and Keiff, S. (2023), "Demographic diversity, perceived workplace discrimination and workers' well-being: context matters", Evidence-based HRM, Vol. 11 No. 1, pp. 35-51. https://doi.org/10.1108/EBHRM-08-2021-0179

Introduction

In recent decades, workforce demographic diversity has increased (Shore et al., 2018). In 2016, half of the jobs were held by women, a quarter by immigrants and just over a fifth by visible minorities in Canada (Government of Canada, 2020). Past literature has shown that these groups are at a higher risk of being discriminated against in the workplace (Banerjee et al., 2018; Bobbitt-Zeher, 2011; Stainback et al., 2011; Zschirnt and Ruedin, 2016) and this growth in diversity creates challenges for organizations (Bove and Elia, 2017). In this context, providing an inclusive work environment has become an important responsibility for organizations (Lee et al., 2021).

In addition to corporate social responsibility, promoting the well-being of the population, gender equality, decent work and reducing inequalities are part of the Sustainable Development Goals of the United Nations adopted by Canada (Government of Canada, 2018; United Nations, 2016). Consequently, equity, diversity and inclusion are among the priority issues of sustainable development in Canada. With respect to these goals, Canada is committed to adopting an inclusive approach by integrating historically marginalized groups, which include indigenous peoples, women, immigrant populations, people with disabilities and people who identify with the LGBTQ2 community (Government of Canada, 2018). Therefore, studying perceived workplace discrimination and its effect on the well-being of workers is relevant regarding these objectives.

Furthermore, Alteri (2020) examined the relationship between changes in the representation of different groups and complaints of discrimination in the federal public service of the United States. She found that the rising ratios of minority employees and women led to higher rates of complaints of racial and gender discrimination. In addition, according to Rubin and Alteri (2019), less than 1% of federal government employees filed a discrimination complaint in the United States each year between 2006 and 2014. They also showed that ethnicity and gender are the two motives most often invoked in discrimination complaints. However, although effective, the focus on officially filed complaints can lead to underestimation of the problem. By comparison, in the 2019 Public Service Employee Survey , 8% of public servants in Canada perceived they had experienced discrimination in their job in the past year and the most frequent source of discrimination perceived by public servants was a person having authority over them (Government of Canada, 2020). Furthermore, Du Mont and Forte (2016) showed that 15% of Canadians perceived to have experienced discrimination over the past 5 years, which suggests that individuals who perceive they are experiencing discrimination do not all report a complaint.

Lee (2022) mentioned three main reasons for focusing on perceived discrimination. First, employees who are pressured by organizational authorities may not file a discrimination complaint, even if they have legitimate reasons to do so, because of fear of reprisal or a lack of confidence in the complaint system. Second, examining perceived discrimination can help managers improve diversity management practices by addressing an underestimated problem. Third, empirical evidence shows that perceived discrimination in the workplace has a concrete impact on employee well-being (Lee, 2022; Triana et al., 2015).

Most research on perceived discrimination focuses on individual characteristics, such as sex and race, as explanatory variables (Avery et al., 2008; Ayalon, 2014; McCord et al., 2018), but some studies have shown that national context and laws can affect workers' perceived discrimination (Dhanani et al., 2018; Triana et al., 2015). Using a meta-analysis, Triana et al. (2019) found that the association between perceived gender discrimination and employees' well-being is stronger in countries with labor policies that value gender equity. This finding led them to conclude that organizations must consider the national context in organizational decisions to prevent gender discrimination.

Based on this evidence, our study focuses on factors related to both demographic diversity (gender, immigration status and visible minority) and the external context of organizations—namely, Canadian province and industry. Given that the workforce is increasingly diverse and that demographic characteristics affect perceived discrimination, we used the 2016 Canadian General Social Survey database, which is representative of the population, to pursue our two main objectives.

First, we aimed to identify the demographic characteristics and contextual factors that affect perceived workplace discrimination in Canada. Drawing a portrait of the demographic characteristics leading to perceived workplace discrimination will help to identify at-risk groups and to establish whether anti-discrimination laws succeed in eliminating perceived discrimination among workers of the target groups. The use of representative data to examine the discrimination perceived by the target groups (women, immigrants and visible minorities) is an undeniable advantage of this study, since it allows capturing the social trends. Examining provincial differences is relevant since it will show if perceived workplace discrimination is less common in some provinces. If so, future research should attempt to investigate why these provinces are more inclusive to find avenues for improvement in the other provinces. To our knowledge, our study is the first to examine the effect of the Canadian provinces on perceived workplace discrimination. This is an important contribution because organizations are not hermetic systems and their employees are influenced by the external context (national culture, social values, public policies and labor laws), which varies by province. Examining perceived discrimination in public administrations is relevant since, as employers and legislators, they must behave in an exemplary manner, especially since equity, diversity and inclusion are stated values of these organizations.

Second, we attempted to measure the impact of perceived workplace discrimination on worker's well-being. Measuring this association is important to ascertain if perceptions, whether accurate or biased, have a concrete impact on the well-being of workers in Canada. Such a finding could convince some employers to focus on this perception to take a step further towards promoting an inclusive workplace.

Diversity and perceived workplace discrimination

Van Knippenberg et al. (2004) defined diversity as “differences between individuals on any attribute that may lead to the perception that another person is different from self” (p. 1008). This definition is based on self-categorization theory (Tajfel, 1982; Turner et al., 1987), which states that individuals categorize themselves at different levels, such as the interpersonal level, where the self is defined as a unique individual and is compared to others. Accordingly, people categorize themselves and others in social groups as in-group or out-group based on characteristics such as sex, age, or ethnic origin. Hence, out-group individuals are viewed less favorably, which can explain discrimination against them.

Referring to Allport (1954), Triana et al. (2015) defined discrimination as “denying equal treatment to individuals because of their group membership” (p. 491). Workplace discrimination can be objective or subjective (Lee, 2022). Objective discrimination is perceived by an observer based on existing criteria, while subjective, or perceived, discrimination occurs when an individual judges their situation to be discriminatory based on their perception and criteria (Lee, 2022; Hopkins, 1980). Perceived discrimination refers to “an individual's perception of receiving (negative) differential treatment based on some characteristics (e.g. age, gender, race/ethnicity)” (Xu and Chopik, 2020, p. 1). More specifically, Dhanani et al. (2018, p. 148) defined perceived workplace discrimination as “an employee or job applicant's perception of unfair or negative treatment based on membership in a particular social group (Chung, 2001).”

Regarding the relevance of focussing on perceived workplace discrimination, Naff (1995) suggested that subjective (or perceived) discrimination can be as damaging to women's careers than objective discrimination. Previous studies have shown that perceived discrimination has concrete impacts on employee attitudes at work (Triana et al., 2015, 2019). In addition, Lee (2022) suggested that perceived discrimination is a relevant measure because it includes both reported and unreported discrimination due to fear of reprisal or lack of trust in the complaint mechanism. Banerjee (2008) also revealed that objective discrimination is associated with perceived workplace discrimination.

In a meta-analysis of objective discrimination, Zschirnt and Ruedin (2016) showed that ethnic discrimination is still common in hiring decisions. They found that comparable minority applicants must send approximately 50% more applications to be invited for an interview than applicants from the majority group. Banerjee (2008) found that immigrants were more likely to perceive discrimination in the workplace than natives. Moreover, despite their disadvantage in the Canadian labor market (Boudarbat and Boulet, 2007), new immigrants are less likely to perceive discrimination than long-term immigrants, since the latter may have higher expectations in terms of fair treatment (Banerjee, 2008). Banerjee (2008) also found that education increases the perception of discrimination among immigrants, which she also explained by the higher expectations for equity among educated immigrants.

Women and minority groups are at a higher risk of perceiving workplace discrimination (Hirsh and Lyons, 2010; Triana et al., 2015). Avery et al. (2008) found that perceived gender discrimination in the workplace is more prevalent among women than men, while perceived ethnic discrimination in the workplace is more prevalent among Black and Hispanic employees than White employees. McCord et al. (2018) showed that women perceive more gender-related workplace mistreatment than men, but women and men report comparable perceptions of all other forms of mistreatment. Similarly, racial minorities perceive more workplace racial mistreatment than Whites (McCord et al., 2018). In the same vein, Bae et al. (2017) found that gender diversity reduces the perception of inclusion in organizations and Foley et al. (2015) revealed that women perceive more gender discrimination because they make gender comparisons and perceive biases against women as a group.

Therefore, relying on self-categorization theory and the empirical evidence presented below, we postulate that:

H1.

Women, immigrants and visible minorities are more likely to perceive that they have experienced discrimination in the workplace.

Perceived workplace discrimination in context

Recent literature has highlighted the role of social and organizational context as determinants of perceived workplace discrimination (Kartolo and Kwantes, 2019; Triana et al., 2015, 2019). Kartolo and Kwantes (2019) indicated that the demographic composition of an organization is not the only factor influencing employee perception of workplace discrimination; societal and organizational culture also plays important roles. Their findings suggested that employees perceived more workplace discrimination when organizations promoted a culture of competition than when they focused on collective goals.

Furthermore, a meta-analysis by Triana et al. (2019) found that labor laws and cultural norms, as contextual factors, moderate the relationship between perceived gender discrimination in the workplace and employees' psychological health. The correlations between perceived gender discrimination in the workplace and employees' outcomes were stronger in countries with more broadly integrated labor policies and tightly enforced labor practices focused on promoting gender equality. The correlations were also stronger in countries where cultural practices were more gender-equal.

To study the effect of diversity management practices in public sector organizations, McGrandle (2017) used contingency theory, which indicates that organizational practices must align with organizational culture and the external environment. He indicated that, according to this theory, there is no “one best way” to manage organizations. Rather, the internal and external environment of each organization creates a unique context requiring distinctive practices to achieve optimal functioning (McGrandle, 2017). Indeed, the contingent approach to human resources management (HRM) states that the effects of HR practices depend on the context (Delery and Doty, 1996; Knies et al., 2017) and that these practices should be chosen to fit the specific context of each organization (Clinton and Guest, 2013).

Since some recent studies have noted that the organizational context influences perceived discrimination in the workplace, we believe it is relevant to ascertain whether other contextual factors affect this perception. For example, the Canadian province is likely to do so for several reasons. First, French is the first official language of Quebec, while the other provinces are predominantly English-speaking. Second, Quebec's welfare regime is more similar to the social-democratic model, which emphasizes equality and gives considerable role to the state, while that of other Canadian provinces is more similar to the liberal model, which insists on individual freedom (Bernard and Saint-Arnaud, 2004). For example, Quebec's parental insurance plan is more generous and has a universal childcare policy, which increases the participation of women in the labor market (Beaujot et al., 2013). Third, while the federal grid is applied in the other Canadian provinces, Quebec is the only one to have its own selection grid for immigrants received as skilled workers (Boudarbat and Boulet, 2010), showing a desire to have decision-making power in the choice of newcomers. In addition, Quebec, Ontario and British Columbia are the three main provinces of economic immigration and the three provinces with the most ethnocultural diversity according to the 2016 Census (Immigration, Refugees and Citizenship Canada, 2020; Statistics Canada, 2017).

Given these differences, the contingency theory and Quebec's social-democratic inclination, we hypothesize that:

H2.

The risk of perceived workplace discrimination is lower in Quebec than in Ontario, British Columbia and other provinces.

The distinction between public and private organizations is another contingency likely to affect perceived workplace discrimination. Representative bureaucracy theory postulates that the more demographically similar the public workforce becomes to the general population (i.e. passive representation), the better the public administration will serve citizens by reflecting the interests of various groups in its decision-making processes (i.e. active representation) (Kingsley, 1944; Lee, 2022; Mosher, 1968). Therefore, this theory assumes that public sector employees stand up for citizens with similar demographics because of their shared experiences, attitudes and values (Lee, 2022). According to Lee (2022), these principles, which concern the public sector employee–client relationship, can also be applied to the employee–supervisor relationship. Supervisors of disadvantaged groups, such as minorities and women, tend to support the interests of employees in these groups (see, for example, Grissom and Keiser, 2011; Marvel, 2015).

In Canada, the Employment Equity Act (SC, 1995, c. 44), which applies to public and private organizations under federal jurisdiction, aims to facilitate access to employment and the organizational representativeness of women, Aboriginal peoples, persons with disabilities and members of visible minorities. In Quebec, public organizations diverged from other industries, since they are subject to the Act Respecting Equal Access to Employment in Public Bodies (CQLR c A-2.01), which gives them specific obligations in terms of the representativeness of the same target groups. In Quebec, private organizations are therefore not subject to this law.

Villadsen and Wulff (2018) noted three arguments suggesting that ethnic discrimination in employment is more prevalent in the private than in the public sector. First, public organizations are funded by taxpayers, which increases public scrutiny (media and accountability) compared to private organizations. In such an environment, illegal discrimination is likely to be exposed. Second, public organizations are traditionally more bureaucratic and have more red tape (Rainey et al., 1995). A greater prevalence of rules and standards could reduce workplace discrimination because, in the presence of a highly formalized hiring process, it is more difficult to disregard a candidate meeting the selection criteria based on their ethnic origin. Third, public managers are more focused on the public interest than private-sector managers (Perry, 2000). Using testing, that is, assessing the hiring practices of organizations by submitting fake resumes in response to job postings in Denmark, Villadsen and Wulff (2018) found little evidence that public employers are fairer in their hiring decisions. Their findings suggest that there is no discriminatory difference between the public and private sectors.

Banerjee et al. (2018) sent 12,910 fake resumes in response to 3,225 job postings in Canada and found that Asian-named applicants received fewer calls for an interview than English-named applicants from organizations of all sizes. However, their results showed that large organizations discriminated against these applicants less than those of smaller size. Banerjee et al. (2018) suggested that large organizations discriminate less because they devote more resources to candidates' evaluation, have a more professional recruitment process informed by HRM knowledge and have more experience with diversity by having a larger workforce.

The results obtained by Hirsh and Lyons (2010) suggest that workers with a greater sense of entitlement (professional authority, promotion experience, union membership) and knowledge of legal rights (level of education, age) are more likely to perceive workplace racial discrimination. They also found that perceived discrimination increases with organizational size but that employees in the public sector are not less likely to perceive discrimination than those in the private sector. The researchers insisted on the importance of the work context in understanding how individuals determine that they have been unfairly treated.

Despite the representative bureaucracy theory and the arguments advanced by Villadsen and Wulff (2018), given the empirical evidence noting no difference in workplace discrimination between the public and private sectors, we formulate the following hypothesis:

H3.

The risk of perceived workplace discrimination is not different in public administration than in other industries.

Perceived workplace discrimination and well-being

The World Health Organization (WHO) (1946) defined health as “a state of complete physical, mental and social well-being and not merely the absence of disease or infirmity” (p. 100). Health is apprehended by both the absence of negative states and the presence of positive states (Gilbert et al., 2011). Psychological well-being has two facets: eudemonic well-being focuses on self-realization and defines well-being in terms of optimal functioning, while hedonic well-being centers on happiness and construes well-being as the pursuit of pleasure and the avoidance of pain (Ryan and Deci, 2001). Therefore, well-being can be assessed using both negative indicators (depression, anxiety and stress) and positive ones (self-esteem, job satisfaction and psychological health).

In a meta-analysis, Pascoe and Richman (2009) found that perceived discrimination reduces mental health and that this negative relationship is present in all ethnic groups and in both sexes. Similarly, Schmitt et al. (2014) performed a meta-analysis to examine the relationship between discrimination and psychological well-being. Their results revealed that the negative relationship between these variables was weaker for positive measures of well-being, such as self-esteem or positive affects, than for negative measures, such as depression and anxiety. Furthermore, Berger and Sarnyai (2015) explained the physiological mechanisms linking perceived discrimination to mental health outcomes by looking at the stress hormones produced in the context of racial discrimination. Although this evidence does not specifically focus on workplace discrimination, it does show that discrimination harms the well-being of individuals in general.

Regarding workplace mistreatment, Rospenda et al. (2009) found that workplace harassment and discrimination are associated with poor mental health and problems with alcohol use in the USA workforce. The perception of workplace discrimination decreases both job satisfaction and performance (Choi and Rainey, 2014; Dhanani et al., 2018; Di Marco et al., 2016; Sloan, 2012). In their meta-analysis, Triana et al. (2015) found that perceived racial discrimination in the workplace is negatively related to psychological health and that the effect of perceived racial discrimination was stronger when minorities were more represented in the samples, which suggests that they are more likely to perceive discrimination and/or react more strongly to perceived discrimination. More recently, Triana et al. (2019) conducted a meta-analysis on perceived gender discrimination in the workplace and their results showed that perceived gender discrimination is negatively linked to psychological health. Sloan (2012) revealed that workers with strong support from colleagues are protected from the negative effects of mistreatment and Ragins and Cornwell (2001) found that organizational practices supporting equity, diversity and inclusion reduce the perception of discrimination among gay employees and attenuate its deleterious effects on work attitudes.

To explain the negative relationship between perceived workplace discrimination and workers' well-being, we relied on the job demand-resource (JD-R) model of Demerouti et al. (2001). The JD-R model divides working conditions into two categories: demands—dimensions of the job requiring physical or psychological efforts causing costs for the employee and resources—aspects of the job facilitating the achievement of work objectives by reducing constraints and their costs (Bliese et al., 2017; Bakker and Demerouti, 2017). Accordingly, perceived workplace discrimination is conceptualized as a job demand of the working environment that requires a psychological effort of adaptation from workers, increasing their stress and reducing their well-being.

Based on the JDR model, as well as the empirical evidence presented above, we postulate that:

H4.

Perceived workplace discrimination is negatively associated with workers' well-being.

Method

Data and sample

To test our hypotheses, data from cycle 30 of the 2016 General Social Survey (GSS) on Canadians at Work and Home were used. The GSS is a representative survey of the Canadian population that explores individuals' perspectives on their work, home, leisure and well-being. The target population for the survey includes individuals aged 15 and over who reside in the 10 Canadian provinces and do not live in an institution. Since our study focuses on workplace discrimination, we selected paid workers aged 18 to 65. The final sample contained 7,706 workers, of which 47.9% were women, 22.1% were visible minorities and 20.3% were immigrants.

Measures

Perceived workplace discrimination is sometimes a dependent variable and sometimes an independent variable in our models. It was measured with the following question: “In the past 12 months, have you experienced unfair treatment or discrimination while at work?” This question was completed by the following definition of discrimination to ensure respondents' understanding: “Discrimination means treating people differently, negatively, or adversely because of their race, age, religion, sex, or anything else.” This variable dichotomized respondents into two groups and was coded as 1 = yes and 0 = no.

Self-rated mental health is a dependent variable in our study and was measured with the following question: “In general, would you say your mental health is … ?” Possible answers were excellent, very good, good, fair and poor. We recorded the participants into two categories: 1 = those with good mental health or better and 0 = those with fair or poor mental health.

The stress level is a dependent variable measured with the following question: “Thinking of the amount of stress in your life, would you say that most days are … ?” A five-point Likert scale was used: not at all stressful; not very stressful; a bit stressful; quite stressful; and extremely stressful. Participants' responses were recoded into two categories: 1 = quite stressful and extremely stressful and 0 = not at all stressful; not very stressful; and a bit stressful.

Demographic diversity was measured using three independent variables. The first is gender, for which females were coded 1, while males were coded 0. To measure the second variable, immigration background, the participants were asked the following question: “Are you now, or have you ever been a landed immigrant in Canada?” The GSS defines “immigrant” as “a person who has been granted the right to live in Canada permanently by immigration authorities.” Respondents matching this definition were coded 1 and non-immigrant were coded 0. The third variable, visible minority identification was measured using the following question: “You may belong to one or more racial or cultural groups on the following list. Are you … ?” The respondent options were: White; South Asian; Chinese; Black; Latin American; Arab; Southeast Asian; West Asian; Korean; Japanese; and Other. White respondents were coded 0 = not identifying with a visible minority, while the others were coded 1 = identifying with a visible minority.

Two more variables concerning the external organizational context were conceptualized as independent variables in our study. The first was the respondents' provinces of residence, determined by Statistics Canada from their postal code and classified into ten options: Newfoundland and Labrador; Prince Edward Island; Nova Scotia; New Brunswick; Quebec; Ontario; Manitoba; Saskatchewan; Alberta; and British Columbia. We recorded the respondents' provinces to form four groups: 1 = Quebec; 2 = Ontario; 3 = British Columbia; and 4 = rest of Canada (ROC). The second variable, industry, was measured by Statistics Canada using the North American Industry Classification System 2012, which is a common classification system developed by Canada, the United States and Mexico. The two-digit classification was used to divide industries into 20 categories. For the purposes of our study, which aimed to identify respondents from the public sector, we recorded the respondents' industry into four categories: 1 = public administration; 2 = educational services; 3 = health care and social assistance; and 4 = others.

The literature has identified other factors affecting perceived workplace discrimination or workers' well-being. We therefore used these factors as controls in our analysis to prevent them from biasing the association between our variables of interest. The education level was measured using the following categories: 1 = less than high school diploma; 2 = high school diploma; 3 = trade diploma; 4 = college and other non-university diploma; 5 = university diploma below the bachelor's level; 6 = bachelor's degree; and 7 = university diploma above the bachelor's degree. The variable age was assessed from three age groups: 1 = 18–34; 2 = 35–54; and 3 = 55–65. Statistics Canada also asked respondents what their occupation and main work tasks were. With this information, the 4-digit occupation code was determined based on the 2016 Canadian National Occupational Classification (NOC). For the purposes of our study, the single-digit classification was used, which divides Canadian occupations into ten main categories: 1 = management occupations; 2 = business, finance and administration occupations; 3 = natural and applied sciences and related occupations; 4 = health occupations; 5 = occupations in education, law and social, community and government services; 6 = occupations in art, culture, recreation and sport; 7 = sales and service occupations; 8 = trades, transport and equipment operators and related occupations; 9 = natural resources, agriculture and related production occupations; and 10 = occupations in manufacturing and utilities. Union type was measured using the following categories: 1 = single; 2 = married; and 3 = cohabiting. Parental status was evaluated based on the number of children, while child age was measured from the youngest child's age and divided into four categories: 1 = no child; 2 = younger than 5; 3 = 5–12 years old; 4 = 13–14 years old.

Analytical strategy

Statistical analyses were performed using Stata Special Edition 15.1. Since the dependent variables of the study are dichotomous, logistic regression models were estimated. Table 1 presents the proportion of perceived discrimination in the workplace according to demographic diversity. Table 2 examines the associations between demographic diversity (gender, immigrant, visible minority), context (province, industry) and perceived workplace discrimination. Table 3 illustrates the relationship between the perception of discrimination in the workplace and workers' well-being (mental health and stress). All analyses were weighted using individual weights. The significance threshold used is p < 0.05.

Results

Descriptive results

Table 1 indicates that 8.9% of workers perceived themselves to have been discriminated against in their workplace. This table shows that the proportion of women who perceived that they had experienced workplace discrimination (10.9%) is higher than that of men (7.0%). The proportion of immigrants who perceived that they had been discriminated against is higher than that of non-immigrants (12.3 versus 8.0%). Finally, the proportion of visible minorities who perceived that they had experienced workplace discrimination (13.1%) exceeds that of workers who did not identify as a visible minority (7.7%).

Multivariate results

Table 2 displays the effects of variables related to demographic diversity and context on perceived workplace discrimination. Regarding demographic diversity, women are 2.1 times more likely to perceive having experienced discrimination in their workplace than men. Visible minorities are 48% more likely to perceive that they are discriminated against than workers who are not visible minorities. However, other things being equal, immigrants are no more likely than non-immigrants to perceive that they have been discriminated against. These results partially confirm Hypothesis 1, since immigration background does not affect the perception of workplace discrimination as it does for gender and identification as a visible minority.

Regarding the factors linked to the worker's context, the results of Table 2 show that workers in Ontario have a 45% higher risk of perceiving that they have experienced discrimination in their workplace than workers in Quebec. This risk is also 44% higher among workers in the rest of Canada. However, the risk of perceived discrimination in the workplace is not different between workers in Quebec and British Columbia. These results confirm Hypothesis 2 for Ontario and the rest of Canada but not for British Columbia.

Table 2 also reveals that workers in public administrations have a 64% higher risk of perceiving that they have been discriminated against in their workplace than those in other industries. The risk of perceived discrimination in educational services and health care and social assistance, in which there is a significant proportion of public jobs, is not different from that in other industries. This invalidates Hypothesis 3, which assumed that the risk of perceived workplace discrimination is not different in public administration than in other industries.

Among the control variables (results available on request), the risk of perceived workplace discrimination varies by occupation. Compared to managers, this risk is higher among workers in sales and service occupations, in trades, transport and equipment operators and related occupations and in manufacturing and utilities occupations. Married workers are less likely to perceive discrimination than those who are single and workers with two children are less likely to perceive discrimination than those without children.

In Table 3, we observe that perceived workplace discrimination reduces the chances of being in good psychological health by 70% and increases the risk of being stressed by 2.36 times. These results confirm Hypothesis 4.

Regarding mental health, Table 3 shows that visible minorities are 67% more likely to be in good mental health than White workers, while women and men, as well as immigrants and non-immigrants, are as likely to be in good mental health. Compared to Quebeckers, workers in other provinces are approximately half less likely to report being in good mental health. Workers in public administration are more likely to be in good mental health than those in other industries.

Regarding stress, immigrants are 25% less likely to be stressed than non-immigrants, while visible minorities are 30% less likely to be stressed than White workers. Men and women do not have a different stress risk. Contrary to their mental health advantage, workers in Quebec are more stressed than those in other provinces. The likelihood of workers being stressed does not differ by industry.

Discussion

Our descriptive results showed that 8.9% of workers in Canada believe that they have experienced discrimination in their workplace. This rate is worrying, especially since we found that perceived discrimination reduces the well-being of workers. The descriptive analyses also indicated that women, immigrants and visible minorities are more likely to perceive that they have experienced discrimination in their workplace.

However, when other factors that may influence perceived workplace discrimination are considered, we found that women and visible minorities are at greater risk of perceiving they have experienced workplace discrimination, but this risk does not differ between immigrants and non-immigrants. This empirical fact is an important contribution that would be relevant to explore in future work. Thus, this suggests that visible demographic characteristics (sex, skin color) further increase perceived discrimination in the workplace than invisible ones (immigration). Visible differences therefore appear to be more relevant for individuals in the process of social categorization than those that are not visible, possibly because they are more difficult to hide. Despite government efforts to promote gender and ethnic equity in Canada, it turns out that women and visible minorities are more at risk of perceiving that they have experienced discrimination in the workplace. These findings suggest that additional initiatives are needed to improve HRM practices related to equity, diversity and inclusion in organizations to reduce the perception of discrimination among these groups. The results also point out that the existing anti-discrimination laws that target gender and visible minorities fails to completely eradicate perceived workplace discrimination.

In immigration policies, the governments of Quebec and Canada have put much effort into facilitating the integration of immigrants into the labor market, given the difficulties that have been noted in this regard. However, our results add that immigrants are not at greater risk of perceiving discrimination in the workplace, which is positive, but may be due to their lower sense of entitlement (Hirsh and Lyons, 2010) or their lower expectations in terms of fair treatment, especially for recent immigrants (Banerjee, 2008). An important avenue of future research would be to adopt an intersectional approach by examining the interaction between gender, immigration background and visible minority identity on perceived workplace discrimination. Such a study would help to refine our understanding of this issue and to ascertain whether certain groups of immigrants are more at risk than others.

For organizations, tackling employees' perceptions of discrimination is relevant since our results obtained using a representative survey of the Canadian population agree with those of previous studies (see, for example, Choi and Rainey, 2014; Dhanani et al., 2018; Di Marco et al., 2016; Sloan, 2012; Triana et al., 2015; Triana et al., 2019), which shows that this perception is negatively linked to workers' well-being. This finding is important since Allen (2019) argues that despite legal reforms prohibiting discrimination, it persists, posing an additional health risk for historically marginalised groups and maintaining social inequalities. Thus, a workplace that promotes the well-being of employees must address the issue of perceived discrimination not only to eradicate discrimination, which is illegal, but also to prevent its perverse effects on workers' psychological health. Since the single focus on formally filed complaints does not provide a complete picture of discrimination in the workplace (Lee, 2022), organizations should include questions about perceived discrimination and feelings of inclusion in their employee surveys, if any. For HRM professionals, this implies that more efforts must be made to implement practices that promote an inclusive work environment (Shore et al., 2018).

Another empirical contribution of our study is that we found that workers in Quebec are less inclined to perceive that they have been discriminated against in their workplace than those in Ontario and the rest of Canada. This finding supports those of previous works (Kartolo and Kwantes, 2019; Triana et al., 2015, 2019) showing that social context impacts perceived workplace discrimination. It suggests that the Quebec model is better to promote workers' feeling of inclusion. . This result can be explained by the more generous employment and family policies in Quebec and/or the more social-democratic and egalitarian Quebec's welfare state model compared to that of other Canadian provinces (Beaujot et al., 2013; Bernard and Saint-Arnaud, 2004). Nevertheless, this could also result from Quebeckers' weaker sense of entitlement or other external factors that are not controlled for in this study such as social values, organizational culture, or organizations' diversity management practices.

Regarding the difference between workers in public administration and those in other industries, our results contradict Hypothesis 3, which is also an important contribution of this article. Despite the greater accountability and bureaucracy of public organizations, the red tape forcing them to adopt highly formalized recruitment and selection processes and the greater motivation of public managers to protect the public interest (Rainey et al., 1995; Perry, 2000; Villadsen and Wulff, 2018), workers in public administration are actually more likely to perceive that they have been discriminated against than those in other sectors. Public administrations would therefore be less inclusive employers. However, this finding could be explained by a more pronounced feeling of entitlement (Hirsh and Lyons, 2010) and higher expectations in terms of fair treatment (Banerjee, 2008) of workers in public administrations, especially due to the exemplarity expected from this employer given its role as legislator. This finding could also derive from the higher representativeness of public organizations in terms of the groups targeted by legislation on employment equity. Indeed, Alteri (2020) found that the rise in the proportions of women and minorities among public employees increased complaints of racial and gender discrimination. Moreover, Ashikali et al. (2020) showed that inclusive leadership is necessary to ensure inclusion in diverse teams. In any case, further research is needed to better understand the greater propensity of public administration employees to perceive discrimination in their workplaces. This empirical contribution has important implications since it suggests that, despite Canada's commitments towards the promotion of equity, diversity and inclusion and the Sustainable Development Goals (Government of Canada, 2018; United Nations, 2016), Canadian public administrations cannot conclude that they are exemplary as employers. Even if this result is due to the greater representativeness of the target groups within public administrations, it remains that additional efforts and HRM practices that promote inclusion are necessary in the public sector to improve the employees' feeling of inclusion.

Finally, the impact of the province and the industry on perceived discrimination in the workplace, two elements of the organization's external context, underlines the relevance of contingency theory when investigating workers' perceptions. Organizations are permeable systems that are subject to external influences not only through national culture, social values and public policies, but also through the province, even possibly the city in which the organization is located. The context of public administration also appears to be distinct from that of other industries. Nevertheless, we cannot know whether this is due to the internal context of these organizations, such as the organizational culture or employees' motivations, or the external context, such as the distinct economic conditions of the industries or differences in the respective missions of organizations—public administrations aim to serve the population, while private organizations aim to maximize their profits. Nonetheless, this result agrees with studies that have mobilized the theory of contingent HRM (Delery and Doty, 1996; Knies et al., 2017; McGrandle, 2017) and suggests that organizations should pay particular attention to their internal and external context to put in place equity, diversity and inclusion management practices that effectively fit their employees to reduce their perception of discrimination. A future study comparing the culture of inclusion in public administrations and private companies could provide interesting explanatory leads.

This study is not without methodological limits. First, the secondary data used did not allow us to consider several factors likely to influence the discrimination perceived by workers, such as social values, organizational culture, diversity management practices in their organization, their motivation for public service, or their personality traits. These variables could have improved our study by providing more explanatory power. However, the GSS is a representative survey of the Canadian population that provides access to a large pool of working women, immigrants and visible minorities, a crucial point when examining the perceived discrimination of these groups.

Second, the data used are cross-sectional, which allows the measurement of associations between the variables of interest but not to distinguish cause from consequence. Thus, although we find a negative relationship between perceived workplace discrimination and workers' well-being, it is not clear whether the perceived discrimination leads to a lower level of well-being or vice versa.

Third, another limitation of our study is that it focuses on the perceived workplace discrimination based on a single item. This could lead to overestimating discrimination (Allen, 2019). However, as mentioned previously, this measure includes discriminatory events that is not reported by workers for fear of reprisal or lack of confidence in the complaint mechanism, which are relevant when looking at the consequences on their well-being (Lee, 2022). According to the results of the 2019 Public Service Employee Survey, the most frequent source of perceived workplace discrimination by public servants in Canada is a person in authority over them (Government of Canada, 2020), which could discourage complaints.

Finally, our data came from a single source—the individuals who answered the questionnaire. Therefore, our study is subject to common source bias, which is likely to inflate the correlations. This bias is more likely to affect the associations noted between perceptual variables, such as perceived discrimination, stress and self-rated mental health. On the other hand, it is unlikely that the variables linked to demographic diversity (gender, immigration, visible minority), which are factual, will change with the source of the responses. In addition, because the province of residence was determined by Statistics Canada from the postal code of the respondents and the industry was coded using a national classification, the risk that the correlations between these variables and the perception of discrimination are artificially inflated is low.

Despite these limitations, our results show that visible demographics (sex and skin color) increase perceived workplace discrimination, which is not the case for the invisible one (immigration). The risk of perceived discrimination in the workplace is lower in Quebec than in other Canadian provinces and higher in public administration. These findings suggest that, despite anti-discrimination laws targeting women and visible minorities and government efforts to promote equity, diversity and inclusion, perceived workplace discrimination persists and it has tangible impact on the well-being of workers, which highlights that there is still room for improvements in equity, diversity and inclusion in the workplace in Canada.

A proactive law requiring employers to eliminate workplace discrimination without a complaint mechanism similar to Quebec's Pay Equity Act could be developed. Moreover, in their desire to be recognized as socially responsible and employers of choice, both private and public organizations must also be proactive in eliminating discrimination. To promote inclusive workplaces, Shore et al. (2018) suggested that a focus on preventing exclusion, in which managers commit to complying with anti-discrimination laws, is the foundation of an inclusive organization. However, they indicated that if this is the only way the organization demonstrates its commitment to diversity, employees who are members of historically marginalized social identity groups will not feel included. HRM professionals must go further by implementing practices that promote psychological safety, the feeling of being respected and valued, participation in decisions and they must recognize, honor and encourage advancement of diversity (Shore et al., 2018).

Table 1

Perceived workplace discrimination across demographic characteristics

	Discrimination (%)	CI 95%
Gender
Women	10.9*	[9.63–12.36]
Men	7.0	[5.92–8.19]
Immigrant
Yes	12.3*	[10.00–15.02]
No	8.0	[7.14–8.94]
Visible minority
Yes	13.1*	[10.57–15.86]
No	7.7	[6.83–8.57]
Total	8.9	−
N	7,706	7,706

Source(s): Statistics Canada, 2016 General Social Survey *p < 0.05

Table 2

Perceived workplace discrimination determinants (odds ratio)

Fixed part
Constant	0.03***
Independent variables
Women (ref.: Men)	2.11***
Immigrant (ref.: Non-immigrant)	1.27
Visible minority (ref.: White)	1.48*
Province (ref.: Quebec)
Ontario	1.45*
British Columbia	1.35
ROC	1.44*
Industry (ref.: Other industries)
Public administration	1.64**
Educational services	1.05
Health care and social assistance	0.86
Fit
Wald test	134.80
Df	(35)***
Pseudo R2	0.052
N	7,706

Note(s): The following variables were controlled in model 2: education level, occupation, union type, parental status, child age and age (unstandardized coefficients)

Source(s): Statistics Canada, 2016 General Social Survey ***p ≤ 0.001, **p ≤ 0.01, *p ≤ 0.05, †p ≤ 0.10

Table 3

Perceived workplace discrimination impact on well-being (odds ratio)

	Mental health	Stress
Fixed part Constant	10.00***	0.49**
Independent variables
Women (ref.: Men)	1.00	1.11
Immigrant (ref.: Non-immigrant)	1.31	0.75*
Visible minority (ref.: White)	1.67*	0.70*
Perceived workplace discrimination (ref.: No)	0.30***	2.36***
Province (ref.: Quebec)
Ontario	0.45***	0.69***
British Columbia	0.53**	0.57***
ROC	0.46***	0.57***
Industry (ref.: Other industries)
Public administration	1.70*	0.86
Educational services	1.03	0.99
Health care and social assistance	1.30	0.82
Fit
Wald test	201.31	239.17
Df	(36)***	(36)***
Pseudo R2	0.067	0.055
N	7,706	7,706

Note(s): The following variables were controlled in all models: education level, occupation, union type, parental status, child age and age (unstandardized coefficients)

Source(s): Statistics Canada, 2016 General Social Survey ***p ≤ 0.001, **p ≤ 0.01, *p ≤ 0.05, †p ≤ 0.10