Published: December 2004
(Please note: The views and opinions expressed by the author are their own and do not necessarily reflect those of the Ontario Human Rights Commission.)
by Carol Agocs
Carol Agocs is Professor and Director of the Local Government Program in the Department of Political Science, University of Western Ontario. Her research addresses race and gender discrimination in employment, and equality policy and its implementation. Her publications include Employment Equity: Cooperative Strategies for Organizational Change (1992), co-authored with Catherine Burr and Felicity Somerset, and Workplace Equality: International Perspectives on Legislation, Policy and Practice (2002).
Few complaints of systemic racial discrimination in employment have been heard and substantiated by human rights tribunals and boards of inquiry, in part because of the challenge of gathering persuasive evidence. A three-dimensional diagnostic framework is proposed, highlighting numerical representation and distribution, employment systems, and organizational culture. For each dimension, examples of types of qualitative and/or qualitative evidence are suggested, using research approaches borrowed from the social sciences, employment equity and pay equity.
The Canadian human rights community is grappling with the need for effective approaches to challenging racism that is institutionalized within our workplaces. Few race-based employment-related complaints are referred to tribunals, and of those, few have addressed systemic discrimination and/or entailed systemic remedies. Our study of 135 cases concerning complaints of racial discrimination in employment heard by boards of inquiry or tribunals from 1980 through 1998 across Canada found that less than half of decisions favoured the complainant. Race complaints are said to be difficult to prove because evidence of direct discrimination is rarely present, forcing reliance on circumstantial evidence; clearly this situation characterizes systemic complaints.
Most race-based cases dealt with by human rights commissions are individual complaints that, if substantiated, result in individualized remedies such as compensation for lost pay or small monetary awards for pain and humiliation. Yet individual stories are situated in an organizational context and suggest a need for a systemic analysis of the social and cultural environment of the workplace in which these events occurred. How widespread was the pattern of racism, over what length of time, and what forms did it take? What policies, practices and behaviours encouraged, permitted and condoned racism? How did the workplace culture encourage and validate the expression of racist assumptions and attitudes?
My working definition of systemic or institutional discrimination is that it consists of patterns of behaviour that are part of the social and administrative structures of the workplace, and that create or perpetuate a position of relative disadvantage for some groups and privilege for other groups, or for individuals on account of their group identity. Systemic discrimination is multi-dimensional and may entail separate and cumulative impacts on grounds of race, gender, disability and other identities. Our study of race-based cases found that 80 percent of complainants were men. Assessing systemic discrimination means considering the interacting and cumulative impacts of discrimination on multiple grounds. Some diagnostic approaches may address this issue better than others.
The purpose of the discussion that follows is to introduce conceptual lenses useful in seeing systemic racism, and that may assist in identifying credible evidence of systemic race discrimination in the workplace. Three analytical dimensions of the workplace can be a starting point:
- The numerical representation and distribution of members of racialized minorities;
- Employment systems – the policies and decision making processes and practices that affect all aspects of people’s careers in organizations; and
- Organizational culture -- patterns of organizational behaviour involving communication, informal social relations, decision-making behaviour, norms, and employers’ response to concerns about discrimination and/or harassment.
This three-dimensional framework was developed with employment equity diagnosis and planning in mind. Employment equity entails the analysis of numerical representation and employment systems using approaches that may also be useful in human rights systemic cases. Numbers continue to be the primary criterion for assessing results under employment equity, and the diagnosis of discrimination in organizational culture remains part of the unfinished business of employment equity.
However, the analysis of systemic racism should address all three dimensions of organizational behaviour: systems and culture as well as numbers. Examining numerical representation and distribution alone is insufficient because it deals with symptoms, not causes and mechanisms of systemic racism: but it is an important beginning.
Quantitative Approaches: Numerical Representation and Distribution
An analysis of systemic discrimination generally includes evidence concerning the access of minorities to employment, and once hired, their representation across the vertical ranks and horizontal job categories. Such “stock” data showing a cross-sectional snapshot is accompanied by “flow” data showing minority representation among those hired, promoted and terminated during the time period under study.
Employment equity analysis compares representation and distribution in a workplace with availability of workers in relevant labour markets in order to investigate under-representation. Internal availability data may show whether qualified and interested members of under-represented groups are present in the organization but have not been promoted. External availability data may disclose the presence of qualified persons in labour markets from which the employer might reasonably draw in hiring. In a human rights context, availability analysis demonstrates whether the organization could have made staffing decisions that would produce a more representative workforce, but did not.
The choice of a comparator group is important in establishing a prima facie case of adverse impact or systemic discrimination. The comparator group must be relevant to the specific complaint and to the period under study, and should be at an appropriate level of detail. Availability data should report minority representation in the specific labour market from which applicants would be drawn.
An issue arises in specifying a standard for determining when under-representation is serious enough to be deemed evidence of systemic discrimination, unless there are zeroes, or near-zeroes, of minority group members in the category in question. If there is some representation, then it is necessary to present an argument as to why the existing level of representation should be considered discriminatory. In the United States, an 80 percent rule has long been used although this standard has been contested. It suggests that there is prima facie evidence of discrimination if the hire or promotion rate of the plaintiff’s group is less than 80 percent of the rate for the successful comparison group. The 80 percent rule was used by the Canadian Human Rights Tribunal in National Capital Alliance on Race Relations (NCARR) v. Canada Health and Welfare (1997).
There is also an important role for statistical analysis in determining appropriate remedies when systemic discrimination is substantiated. What is the appropriate level of representation to be sought and attained? The answer has generally been based on availability data, as in the systemic cases of Action Travail des Femmes (1987) and NCARR (1997). This imposes a conservative standard, since the representation of racialized minorities in comparator groups may itself reflect systemic discrimination.
Delays in the human rights process also introduce bias, as in the NCARR case, in which court-ordered remedies to be implemented by 2003 were based on 1993 data on under-representation.
Data deficiencies often pose constraints in using statistical evidence. Relevant organizational data may not be available. Since the repeal of the Ontario Employment Equity Act in 1995, Ontario employers have not been required to collect such data, and the repeal legislation was interpreted as requiring the destruction of data collected under the Act. There may also be a lack of relevant external availability data, which are typically drawn from the census. Timeliness is an issue, and since data related to race are based on samples, they may not be readily available in sufficient detail, for example, disaggregated to a local labour market or to a particular occupational category.
Employers covered by federal employment equity legislation, who are required to maintain workforce data and make them publicly available, are limited in number. Even if such data are available, revisions to the federal Employment Equity Act and the Canadian Human Rights Act in 1995 disallowed the use of data collected under the Act as the basis for a human rights complaint. Moreover, the use of statistical evidence as the sole basis for a complaint under the Canadian Human Rights Act is not permitted. Prior to these revisions the Canadian Human Rights Commission used these data as a basis for negotiating voluntary agreements with employers that contained significant equity-related provisions. Perhaps it remains for human rights adjudication to clarify circumstances in which workforce data generated under employment equity requirements may be used in human rights proceedings. It seems clear that a systemic complaint based on statistical under-representation alone will not succeed: analyses of employment systems and/or workplace culture are also needed.
Another example of quantitative analysis relevant to systemic cases is found in applications of regression analysis in pay equity contexts and in analyses of employment discrimination. Multiple regression analysis of data on hiring and promotion can help to determine whether under-representation occurs for reasons that are not explained when the effects of seniority, qualifications and other legitimate factors are taken into account. Multiple regression can also reveal whether salary differences between groups can be substantially accounted for by characteristics such as education or seniority, or whether a salary gap remains that is not explained. The gap may be attributed to discrimination, provided that the job-related factors expected to determine salary have been included in the analysis. This kind of analysis uses the organization’s own compensation data without the need for external comparative data. Multiple regression analysis may assist in identifying discrimination on more that one ground, for example both gender and race. However a lack of data on the racial identity of the workforce may be an impediment.
Since the mid-seventies there has been a substantial increase in the complexity and rigour of statistical evidence and expertise used in employment discrimination cases. Yet it would be unfortunate to refrain from pursuing such cases because of these challenges. It is important to creatively exploit the possibilities for using statistical evidence when it is available, even under less than ideal conditions, and to seek to advance case law by pursuing systemic complaints.
Identifying Systemic Racism in Employment Systems
Identifying and proving systemic race discrimination in human rights cases entails documenting the job barriers contained in organizational policies and/or practices. The three-step guide to analysis set out in the Meorin decision of the Supreme Court of Canada [British Columbia Public Service Employee Relations Commission v. BCGSEU, 1999] now serves as a guide to identifying systemic barriers and their adverse impacts. The NCARR case also provides analyses of how employment policies and practices, both formal and informal, create barriers in hiring, job assignment, promotion and terms and conditions of employment. Evidence in these cases came from the insider knowledge of employment practices provided by advocacy organizations and unions that represented the complainants as well as employees who participated in surveys and interviews.
A review of employment systems, a core requirement of employment equity implementation, depends on the employer to scrutinize its own practices and uncover its own job barriers, in consultation with employees. This approach has the advantage of tapping organizational knowledge and minimizing employers’ resistance. But self-monitoring has also contributed to the widespread failure of employers to implement employment equity requirements, as seen in the delays and perfunctory effort identified in compliance audits by the Canadian Human Rights Commission.
Field trials using paired actors have been used in human rights as well as research contexts, such as the well-known 1985 Toronto study by Frances Henry and Effie Ginzberg, to identify direct but covert race discrimination in hiring. Employment testing compares hiring outcomes for minority applicants with outcomes for majority group applicants when the two groups of testers are given matching qualifications and apply for the same job vacancies. Useful evidence may include interview records and testers’ narratives in addition to the success ratios for the two groups. The U.S. Equal Employment Opportunity Commission and the Office of Federal Contract Compliance have authorized use of testing to uncover discrimination.
Although employment testing is not often used in litigation, it can provide evidence that is seen as objective and has persuasive power. It can lend itself to experimental designs involving more than one ground of discrimination, such as both gender and race. Testing could assist in cases where data on under-representation point to the need for evidence of discrimination, or where there have been several complaints against an employer but evidence adequate for developing a case is lacking. Though the use of testing may be limited to hiring, and to lower-ranking positions, in these contexts testing could be useful. Discrimination in hiring is difficult to substantiate because applicants may not be aware of discrimination against them, or may lack insider information allowing them to challenge it.
The analysis of employment systems and the use of field trials can reveal mechanisms of discrimination by showing how decision-making processes that may appear neutral actually result in adverse impact. However it may be difficult or impossible to gain access to knowledge about how some employment decisions are made, especially for senior level positions, and employees rather than applicants. Insiders who know the organization’s actual employment practices may suffer retaliation if they disclose information useful in a human rights context. Human rights cases typically do not have access to the kind of evidence used in the Texaco case in the United States, where tape recordings captured senior executives making racist comments.
Identifying Systemic Racism in Organizational Culture: Qualitative Approaches
Organizational culture encompasses shared patterns of informal behaviour, such as communication, decision making and social interaction, which are the observable evidence of deeply held and largely unconscious values, assumptions and norms. Prevailing organizational values and norms are socially constructed by dominant groups and typically reflect the standpoint of members who are white, male, heterosexual and able-bodied. In workplaces where whiteness is constructed as normative, informal social behaviour may harass, exclude or marginalize members of racialized minorities and/or Aboriginal people. The culture of an organization or department may render minority group members invisible, create a glass ceiling, or perpetuate a poisoned environment that is intimidating, abusive, hostile, humiliating or offensive. Evidence of discrimination in organizational culture in human rights cases usually comes from individual testimony and occasionally from corporate documents and employee surveys, as in the NCARR case.
Systemic racial discrimination in organizational culture takes different forms depending upon the groups that are represented. Issues may differ for Aboriginal people, Blacks, and people of Asian and South Asian ancestry, as well as for immigrants as compared with Canadian-born persons, and women as compared with men. With research-based tools to assist in the diagnosis of organizational culture, some of these differences can be identified and examined. Perhaps systematic assessment of organizational culture could provide useful evidence in human rights contexts, and could supplement individual testimony, which may not reveal patterns.
In research designed to develop an assessment tool useful in diagnosing systemic racism in the culture of the workplace, we reviewed employment-related human rights cases to identify specific behaviours that had been found discriminatory on the basis of race. To gain an understanding of everyday discrimination we also conducted focus groups with 106 working people who identified themselves as Black, Aboriginal, South Asian or Asian. Men and women were interviewed in separate focus groups, as were each of the four identity groups, by trained facilitators of similar background. Using an interview protocol, the groups discussed systemic racial discrimination that participants had experienced or observed according to several dimensions of workplace culture, including communication, informal social behaviour, decision making, norms, and employers’ response to concerns about racism or sexism.
Analysis of the interviews and cases yielded an inventory of specific behaviours that members of each identity group and gender had identified as discriminatory. These behaviours that had occurred in workplace culture were used as the basis for developing a diagnostic questionnaire that was field-tested on a small scale in a public sector workplace. A statistically significant difference was found between white and racial minority respondents in the degree to which they had observed or experienced racism in their workplace on each of the dimensions of organizational culture.
Perhaps such a diagnostic tool, or at least the methodology used to produce it, might have a place in a human rights context where there are allegations of systemic racism in organizational culture but evidence is difficult to acquire. It might also be used to monitor a workplace over time to identify impacts of remedies or anti-racism interventions.
Individual human rights complaints often refer to racial slurs, jokes or other behaviours that constitute racial harassment, sometimes in combination with sexual harassment. Less frequently, allegations of poisoned environment harassment are made. Some individual complaints contain descriptions of behaviours that reflect a racist workplace culture as well as adverse impacts of employment systems, and under-representation of racial minorities. Yet often the dots are not connected and an underlying pattern of systemic discrimination is not made explicit. Perhaps if assessment tools helpful in identifying systemic racism and sexism in organizational culture were available, more vigorous and successful efforts could be made to advance systemic complaints.
Conclusions: Using Social Science Approaches in Systemic Discrimination Cases
In pursuing systemic human rights cases alleging race discrimination in employment, a broad and strategic approach encompassing numbers, systems and culture is necessary to advance the case law. Successful pursuit of systemic cases requires that racist patterns of organizational behaviour be brought to the level of awareness of decision-makers and adjudicators so that their existence cannot be denied or illegitimately explained away. Obviously the collection, analysis, interpretation and presentation of information that contributes to a diagnosis of systemic race discrimination need to be rigorous and to employ credible methodologies. The development of diagnostic methods and applications is a work in progress and will evolve as they are tested and revised in the context of human rights cases. Such an evolution presupposes that all parties to the human rights decision-making context – investigators, lawyers, subject matter experts, judicial decision-makers, arbitrators and mediators, advocates for equality-seekers, and others who make the human rights system work, are aware and knowledgeable about the use of methods and concepts borrowed from the social sciences. Training and professional development will be needed to support this.
The spread of knowledge about successful systemic cases beyond human rights specialists to the general public would help to build knowledge about systemic discrimination, as well as understanding and support for the human rights process. Perhaps most important, learning from successful cases would build awareness that systemic discrimination can be identified and addressed in practical ways that improve the workplace for everyone.