Advancing a perspective on the intersections of diversity: challenges for research and social policy.

Wilkinson, Lori

ABSTRACT/RESUME

The influence of intersecting identity markers on the life chances of individuals has become a central concern to policymakers and academics. Governing bodies and various social institutions have a vested interest in accommodating intersectionality to create inclusive spaces for all citizens. Similarly, researchers want to accurately depict the lifestyles of their subjects. Yet when intersections are applied in politics and research, the three identity markers of gender, social class, and ethnicity prevail. Other important identity markers such as Aboriginal status, ability/disability, age, ethnicity, religion, language, immigration, and region are largely neglected despite the recognition that these markers can greatly influence individual outcomes. As a result, researchers and policymakers have not fully engaged in intersections and measuring their combinations in comprehending various social phenomena. This paper opens the discussion on the need for an intersectional theory, a description of what the theory might look like, and an examination of the academic and policy challenges related to such theorising. The anticipated outcome of this work, and the others in this edited volume, is for researchers, policymakers, and concerned stakeholders to begin thinking in terms of intersections in order to enrich our understanding of life and experiences.

Lorsque la recherche et la theorie deviennent plus complexes, l'incidence du croisement des marqueurs de l'identite sur les chances d'epanouissement des personnes est devenue une preoccupation principale. Les organes directeurs et differentes institutions sociales ont aussi un interet personnel a concilier l'intersectionalite afin de creer des espaces inclusifs pour tous les citoyens. Toutefois, lorsque les intersections sont appliquees au domaine des politiques et de la recherche, les trois marqueurs de l'identite du sexe, de la classe sociale et de l'ethnicite predominent. D'autres marqueurs de l'identite importants tels que l'appartenance a une population autochtone, la capacite/l'incapacite, l'age, l'ethnicite, la religion, la langue, l'immigration et la religion sont largement negliges malgre la reconnaissance que ces marqueurs peuvent influer grandement sur les resultats individuels. Par consequent, les chercheurs et les decideurs ne se sont pas entierement engages envers les intersections et a mesurer leurs combinaisons dans la comprehension des divers phenomenes sociaux. Ce document traite du besoin de mener de la recherche dans le domaine des intersections, donne un apercu preliminaire d'une theorie intersectionnelle et souligne ses problemes possibles. En conclusion, ce document fait etat des defis que pose la mise en aeuvre de la theorisation intersectionnelle dans la recherche dans les domaines des politiques sociales et des sciences humaines.

INTRODUCTION

Understanding how the intersection of various identity markers influences individual life chances is a complex task. Policymakers are challenged in the design and implementation of legislation to alleviate the social problems faced by specialised groups while maintaining their commitment to govern society. Researchers are also challenged in the way they conduct their studies by examining several identity markers within their investigations. Rarely do they have the rime to consider the effects of the intersection of these identity markers. Instead, the goal is to isolate the "unique" effects of one variable on particular social phenomena. It has become increasingly apparent, however, that this way of doing research is rather limited in its ability to accurately represent the complexity of social life.

An examination of poverty illustrates the difficulty involved in examining the intersections of diverse identity markers. The poverty rate among Canadians is estimated to be 11.8%, yet the poverty rate for children is slightly higher at 14.3% (Sullivan, 2000; Statistics Canada, 2003). These figures exemplify the simple interaction between two variables, age and poverty rate. Taken alone, this information is useful in suggesting that the rate of poverty is slightly higher among Canadian children than adults. The addition of other identity markers such as family composition, disability status, and other characteristics further refines the understanding of poverty. When two additional identity markers, family composition and gender, are included in this example the results indicate that 56.8% of children who live in female headed, lone-parent families live in poverty (Freiler and Cerny, 1998). When disability status is added to this mix, the statistics change again. Over 37% of all low income families have one child with a disability (Wiegers, 2002). The addition of these three additional identity markers highlights the complexity of poverty. It is only by examining the above sentence outlining the intersection between family composition, age, and disability status that the intricacy of poverty becomes more apparent. For researchers, this finding will help fine-tune theoretical and methodological practices with greater insight into the lived experiences of children with disabilities living in poverty. For policymakers, this more precise information will help governments and service agencies create or revamp legislation and funding programs for the low-income, single-parent families most in need of assistance. The benefit derives from identifying the groups most likely to need assistance and thus being able to direct finite government funds toward those most in need of them. Without this knowledge, policymakers and front-line service workers would be less likely to adequately target the families that require certain types of assistance.

The remaining question, however, is whether this approach of studying multiple intersections is really useful in understanding the intersections of diversity? This paper attempts to address this and four other questions and their influence on various intersections of diversity including: immigration, socio-economic status, disability, language, sexual orientation, race/ethnicity/religion, age, Aboriginal peoples, gender, and region. Thus, in addition to determining the utility of studying multiple intersections, I also begin to address the following questions:

* Is an intersectional perspective needed?

* What would that perspective look like?

* What are the challenges to implementing the results of intersectional research in research and social policy?

* What are some of the implications of an intersectional perspective on policy and research?

While the answers to these questions are preliminary, it is hoped that this paper will generate debates about the way research is conducted, how social problems are theorised, and how social policies are constructed in Canada.

RECOGNISING THE NEED FOR AN INTERSECTIONAL PERSPECTIVE

What is an intersectional perspective? Brewer et al. (2002:3) define intersections as "the intertwined nature of gender, race, class, ability, sexuality, caste and other influences". It is a "condition in which a person simultaneously belongs to two or more social categories or social statuses and the unique consequences that result from that combination" (Deaux, 2001:1). Until very recently, the intersection that mattered most was the "holy trinity" of gender, race, and social class. These were acknowledged as the central characteristics influencing the life chances of individuals (Creese and Stasiulis, 1996) and were arguably the markers most likely at the root of various inequalities in Canadian society. Other identities such as Aboriginal status, ability/ disability, age, sexual orientation, ethnicity, religion, language, immigration, and region were seen as important, but often considered secondary characteristics owing to a perception that their power to shape various social outcomes was not as great as that of the three primary characteristics. As a result, many researchers and policymakers failed to fully engage in the examination of other intersections involving a host of other identity markers and their impact on social inequality in society. A fully intersectional perspective better enables policymakers and front-line service workers to target their services to those most in need. For researchers, intersectional thinking provides richer detail in studies of various social problems.

Understanding intersections is key to comprehending diversity. According to Krane et al. (2000:2), all aspects of identity (especially race, sex, and class) must be used to fully understand systems of oppression such as capitalism, patriarchy, and white supremacy. This framework allows for context-specific understandings of oppression such as those faced by immigrants, Aboriginal peoples, those of various social classes, and people of colour, to name just a few (Ng, 1993; Bakan and Stasiulis, 1997; Razack, 1998). Unfortunately, an intersectional perspective remains largely marginalised in the social sciences. In fact, it is not seriously considered outside the study of women, Aboriginal peoples, immigrants, and people of colour (Brewer et al., 2002; Frideres, 2002). One of the challenges of this endeavour will be to articulate the perspective effectively enough to increase its use throughout a variety of disciplines and in the study of a wide variety of peoples.

Why is an intersectional perspective required? There are three good reasons including: the need for research to be more reflective of the realities that intersections impose on individuals, the need for theory to be inclusive of multiplicity, and the need to guide policymakers and front-line service workers in delivering programs and services to meet the needs of groups facing discrimination and inequality in society.

First, most research methods do not adequately capture intersections of diversity. In quantitative research, studies will often examine the impact of multiple influences on a social event. For example, studies of educational attainment of youth will look at various identity markers such as gender, age, visible minority status, and ethnicity, all of which are known to have an influence on schooling outcomes. When these "variables" are entered into a statistical equation, however, the influence of each identity marker is considered separately. Thus, gender may explain a certain proportion of an educational outcome, and visible minority status may explain another portion. Intersections of identity markers are only considered when two or more identity markers are highly correlated. One solution is to combine these variables into a single measure, effectively examining a statistical intersection. For example, when two identity markers such as visible minority status and immigrant status are highly correlated, statisticians can measure the impact of the intersection of visible minority status and immigrant status on educational outcomes. (1) In other words, they can compare the differences between immigrants who are also members of visible minority groups with immigrants who are non-visible minorities. The problem with this procedure is that the intersections are hot regularly examined and the method is used only to justify the inclusion of two statistically related variables. It is not intended to study intersections. Furthermore, there is a paucity of statistical measures that allow a thorough examination of intersections.

Qualitative methods are more amenable to studying intersections because of their interest in small-scale populations and very detailed topics. This mandate encourages researchers to examine the intersections of various identity markers. Studies on such precise topics as female Filipino domestic workers are not out of the ordinary. However, this method too has its drawbacks. In terms of sample size, it is both costand time-prohibitive to undertake studies of a large enough scale to produce enough data to be "generalisable" to the larger population. Thus, while a study of female Filipino domestic workers may provide precise detail on the lives of these women, the findings cannot be applied to all female domestic workers. This is a real problem as policymakers and front-line service workers look to researchers for large-scale studies that might support the creation of new programs. If the results are not sufficiently generalisable, then the policymaker cannot use the data.

Second, an intersectional perspective is needed to encourage researchers to create models that actively utilise multiple identity markers. For instance, there are several immigration theories relating to the process of societal integration. Mainstream adaptation theory suggests that while most newcomers will have difficulty adjusting to their new lives in Canada, immigrants, for the most part, integrate relatively successfully. Though they may experience problems with employment, language, or cultural differences, over time, most will settle successfully. The underclass absorption perspective, in contrast, argues that the integration process is more difficult, involving various obstacles such as language barriers, discrimination and racism, and the recognition of foreign credentials. In this perspective, immigrants will experience difficulty in adjusting to Canadian society. While these theories appear to contradict one another, they can be complementary as they can both explain the variations in situations faced by immigrants. By focusing on the variations in the intersections of identity markers, some immigrants could best be described using the mainstream adaptation model, while others are best described using the underclass absorption model.

Finally, a perspective on intersections would guide the design of policies and delivery of government programs. As the example above of the intersection between age, disability status, and poverty illustrates, identity markers mingle in infinite combinations (female, single parent family, and disability status) that differentially influence various life circumstances. Policymakers, researchers, and community service workers are aware that certain combinations of identity markers can result in unequal treatment, discrimination, and exclusion (Hicks, 2003:3). The problem, however, is that often policies and programs do not reflect the actual lived experiences of Canadians. That is, the lived experiences of individuals with many different intersections can be considered when legislation is proposed and programs are designed. Currently, government programs are often directed toward alleviating inequalities based on single identity markers. Programs that are designed to address problems faced by women, Aboriginal persons, and persons with disabilities cannot adequately account for more than one identity marker at a time. As a result, programs designed to alleviate poverty among women may not work as well for Aboriginal women given the fact that the intersection of Aboriginal status is hot built into these programs. Mwarigha (2003) argues that government services that target only one identity marker will disadvantage some individuals and benefit others. As a result, poverty programs directed at women may have a slightly greater impact on the lives of white women and may overlook the unique problems that impoverished Aboriginal women may face.

The following example highlights the need for intersectional thinking in both policy and research. Aylward (1999) effectively argues that several intersecting identity markers must be considered in comprehending the power dynamics and interpreting the outcome of any event. (2) She describes the case of an African-Canadian youth charged with assaulting a police officer and resisting arrest while riding his bike in a predominantly lower-class, black neighbourhood in Halifax. The incident occurred as the 13-year-old African-Canadian male encountered a white police officer arresting his cousin. According to the account of the incident provided by the youth, he stopped his bike and asked the officer if his cousin was under arrest. When he received an affirmative answer, he subsequently asked the officer if he should call his aunt. At this point, the young man was arrested.

The incident was recalled much differently by the white police officer. In his version of the events, the police officer claimed that the youth became verbally abusive as he approached, demanding that the officer release his cousin from custody. When the police officer refused, he contended that the 13-year-old physically assaulted him by pushing him with his bicycle.

In the first trial, the youth was acquitted of all charges, including the assault of a police officer. The prosecution, however, appealed the decision, stating that the black, female, Youth Court Judge was biased in her initial ruling. The prosecution successfully argued in the Provincial Appeals Court that the judge's African-Canadian racial identity biased the case because she brought "her life experiences to the adjudication process" (Aylward, 1999:96). Subsequent appeals to the Supreme Court of Canada eventually resulted in a verdict in favour of the youth and the dropping of all charges against him. In their findings, the Supreme Court of Canada noted that Judge Sparks was not biassed and there was, in fact, evidence of "racially motivated overreaction" by the police officer when he arrested the young man (Supreme Court of Canada, 1997; c.f. in Aylward, 1999: 102).

This complicated case can best be understood by using an intersectional perspective. The intersections between various identity markers of the players involved can be used to deconstruct the events that led to the young man's arrest and the subsequent appeals. The most obvious intersection in this case would be age, race, gender, and social class. In the initial incident, the seemingly incompatible intersection of an adult, male, white police officer (assuming the officer can be identified as middle-class) was at odds with the identity markers held by the young, male, black, lower-class youth (Aylward, 1999). The intersection of these different identity markers shows how the situation was differentially defined by the police officer and the young man. But an intersectional perspective can further outline the complexity of the situation. The incident occurred in a primarily lower-class, largely black neighbourhood known to the police as a "hotbed" of crime. Had the event happened in a middle-class neighbourhood, the situation may have developed differently. The subsequent involvement of the black female judge and the intersections of identity she brought to the courtroom intensified the complex web of identity markers that had begun to define the situation. Now there was a three way interaction between the police officer, the judge, and the young man, each person bringing his or her own unique identity markers and their intersections to define the situation. Had age been the only identifier used to describe these events, the explanation would be more simplistic, involving a conflict of a more socially powerful, older police officer versus a less powerful young person, certainly a less holistic interpretation of the events.

This case clearly exemplifies the need to include the influence of the intersections of multiple identity markers in addition to their individual impact on social outcomes. While it is well known that visible minority status is an identity marker that sometimes results in negative social outcomes for some individuals in terms of lower wages, lower health status, and other aspects, the intersection of this identity marker with others is more complex and less articulated in studies.

WHAT WOULD AN INTERSECTIONAL PERSPECTIVE ENTAIL?

There are no easy answers as to what a perspective on intersections would look like. For this reason, this part of the paper provides a perspective on how investigators and policymakers should consider intersections when conducting research, evaluating programs, and implementing new policies. As all stakeholders involved become comfortable in examining all the types of individual diversities, a theory of intersections is likely to follow. For now, a perspective on intersections is intended to encourage all stakeholders to engage in this important debate.

How can an intersectional perspective account for all the possible interconnections of identity markers that make up an individual without being burdened by an endless combination of intersecting characteristics? This is a major problem confronting quantitative research given its propensity toward large equations containing several variables. While qualitative research may be better able to handle multiple identity marker combinations, given its preference for small sample sizes and in-depth data gathering techniques, it too has problems. Qualitative research cannot undertake a study of more than a few intersections atone time and the results cannot be generalised to other groups. It would appear that an intersectional perspective implies the use of triangulation and multiple methods in order to take advantage of the benefits of both methodologies and to be flexible enough to be used in a variety of disciplines.

Quantitative research allows larger, more geographically diverse samples that can be used by policymakers and investigators to draw generalised conclusions about the population. The chief advantage of qualitative research is an in-depth investigation of individuals that garners more detail about the characteristics of the intersections of diversity. The researchers can return to the quantitative information and situate the results of the qualitative study within the general framework of society and as a result, come to a better understanding of how certain life situations and social barriers may prevent equal participation. This multi-method approach is superior in capturing the complexity of various intersections.

An intersectional perspective must overcome these and other challenges to entice researchers to use it. The definitions of identity markers must also change to meet the specifications of an intersectional perspective. Gender, race, ethnicity, and other characteristics have meanings that are socially constructed and highly contested. One example of this problem is the definition of "Metis," an ethnic group with a highly contested membership (Frideres, 1998:38). The Metis National Council has established three criteria that differentiate Metis from other Aboriginal peoples, while other organisations have very different definitions regarding who can legitimately call themselves Metis (Dunn, 1994). In this instance, an intersectional perspective may guide the discussions in this and in other instances where there are many differing opinions. As well, research must account for identity markers with definitions that are more flexible and change throughout the life course of the individual. Age, region, disability status, and socio-economic status may fluctuate during a lifetime. The difficulty that must be overcome is in identifying a method that best captures the flexibility of these more dynamic identity markers.

As Gilmartin (2002:161) laments, "[w]riting about the intersections of identifies is a challenge because it is difficult to simultaneously discuss gender, sexual, racial, ethnic and class identities--and yet these identities are ultimately inseparable". Lester (2002) chronicles several court cases involving the intersection of gender, sexual orientation, and race. According to her assessment, courts have had difficulty in handling cases involving intersectional discrimination. As a result, when plaintiffs go to trial, "they often find that their claims have been subjected to a kind of radical surgery through which the racial aspects of their experience have been cut away from its sexist dimensions" (Lester, 2002:85), or vice versa. As a result, sexism and racism are generally separated for the sake of clarity and simplicity. In Brooms v. Regal Tube (1988), an African woman who experienced both racial and sexual discrimination at work received compensation only for sexual discrimination despite the fact that much of the evidence presented showed black women in graphic scenes of torture. In this case, the court was unable to simultaneously consider the two forms of discrimination. As a result, in these "sex plus" suits, the gender discrimination tends to dominate while the racial discrimination is set aside (Lester, 2002) as racism is more difficult to argue in court. An intersectional perspective would legitimate the simultaneous acts of discrimination in order to ensure justice is served. Until there is a mechanism in place, the courts, as well as researchers and policymakers, will be reluctant to undertake such complex tasks.

Any attempt at an intersectional perspective must therefore avoid binary oppositions. Comparisons of women versus men, Aboriginal versus non-Aboriginal, immigrant versus non-immigrant, persons with disabilities versus persons without, succumb to benchmarking. Benchmarking is the attempt to situate a comparison of individuals who are "different" to those who are "normal". The implicit assumption of these comparisons is that individuals who are different from the norm are not fully integrated or productive members of society. Studies regarding immigrants are a good example of the negative implications of benchmarking. The assumption is that immigrants must change their way of life in order to "become" more like nonimmigrants. As a result:

 [I]t is immigrants and not Canadian society or institutions that are
 seen as needing change. In fact, social changes in metropolitan
 centres that are attributed to immigration are generally not
 interpreted as desirable as they are often viewed as urban problems
 brought about by increases in the immigrant population beyond the
 "absorptive capacity" of Canada. Li, 2003:52

As a result, this type of research implicitly blames various social problems and failed attempts at integration on immigrants, perpetuates negative stereotypes, and may discourage immigrants from maintaining their culture and way of life. Avoiding benchmarking implies a change in research design and execution. Researchers should refrain from interpreting their findings in binary opposition instead of embracing the diversity within a single identity marker. Today, when statisticians look at the effects of racism, for example, they use a binary, visible minority status variable (visible minority status versus non-visible minority status), rather than considering the full range of ethnic origins. This results in a superficial expose of racism rather than a more detailed one.

There are a number of other issues beyond research that must also be considered in an intersectional perspective. An intersectional perspective must be inclusive of all stakeholders. Governments, service agencies, academics, individuals who are personally involved (i.e., Aboriginal peoples, immigrants, persons with disabilities), and the general public are all represented to give voice to all decisions surrounding both research and policy. Negotiations around the new First Nations Governance Act would benefit greatly from the presence of all stakeholders--Aboriginal peoples, governments, community service agencies, and non-Aboriginal persons--as they all have a stake in the new legislation. While these negotiations would increase the length of time required to launch new legislation, the participation of all stakeholders would encourage cooperation and understanding between all sectors of society with acceptance and understanding as an unintended and beneficial result.

While the idea of an intersectional perspective has merit, there are potential pitfalls. Mwarigha (2003:3) warns that intersections can be used to target certain groups for exclusion and to justify stereotypes and legitimate discrimination. In much the same way that binary oppositions can be used to perpetuate stereotypes, these intersections can lead to more targeted discrimination. This is one area where policy has much influence.

POLICY IMPLICATIONS

An intersectional perspective should be applied to policies and programs; otherwise it lacks a basis beyond theoretical utility. Currently, many government policies and programs are designed with the entire population as the target group. As a result, some legislation, like the Employment Insurance program, is designed to aid all workers but may fail to adequately take into account the needs of persons with unique intersections of identity markers. For instance, those of lower socio-economic classes and those less educated are more likely to be dissatisfied with current policies since they are largely marginalised in society and may experience difficulty in accessing resources and programs that help alleviate problems in employment (Mackenzie, 2003:74).

Canada does have programs that are intended to serve smaller, more unique populations. When policies are designed to target a specific population, they usually deal with single identity markers such as gender, immigrant status, and Aboriginal status (Hicks, 2003:5). Thus, legislation such as the Indian Act is designed only to account for the influences of a single identity marker; Aboriginal status. In this instance, these policies have a limited ability to make adjustment for the intersections of two or more identity markers such as the intersection between age, Aboriginal status, and gender. In a sense, these "targeted policies" have the same problems as programs geared toward the "general" population in that "interactive effects that occur in those groups characterized by multiple identity markers [mean] the total problem to be overcome is greater than the sum of its parts" (Burstein, 2003:2). In other words, problems experienced by people who have certain combinations of identity markers are more challenging than others given the influence of intersections. Policies must reflect this uniqueness or remain largely ineffective in solving the problems of marginalised groups.

Another policy recommendation suggests that all citizens become educated about the need for some policies to be directed at specified groups of people. There seems to be a widely held perception in Canadian society that some people are undeserving of special programs and funding. Certain programs are viewed as favouring some groups and not others. Thus, programs such as the Employment Equity Act and the Multiculturalism Act, among others, are considered forms of reverse racism or reverse sexism that unfairly "demonise" the dominant group (Mwarigha, 2003; Nakache, 2003). As a result, the purpose of these programs is often misinterpreted, leading some to believe that certain groups receive privileges or opportunities over others based on ascribed characteristics such as gender, ability, Aboriginal status, or visible minority status. If these programs were discontinued, youth, abused women, people with disabilities, and ethnic minorities would be disadvantaged by a system that clearly tries to create "one-size fits all" policies. Extensive education of the "public" is needed in order to change attitudes toward specialised treatment. This may alleviate much of the criticism and misunderstanding surrounding certain "specialised" policies (such as the right to hunt, fish, and log) that are granted to some Aboriginal peoples but not others.

These suggestions would mean that government and academics must become leaders in implementing an intersectional approach. The first and foremost change requires increased funding for government programs in order to help agencies become more effective in servicing their diverse clientele (Mwarigha, 2003). Funding directed toward intersectional research is also required. The implementation of an intersectional approach both in policy and research will be expensive, as it requires experts from disciplines to meet regularly to learn other perspectives. It is hoped that these initiatives will decrease policy fragmentation and service gaps (Mwarigha, 2003:9; Kligman, 2003). While the initial financial commitment will be expensive, the future returns will help create a society where intersections are a valued part of everyday life.

Along with funding, cooperation and communication between government, service providers, and academics and between all levels of government is required. The issue of the recognition of foreign credentials is one where an intersectional perspective would be valuable. This problem involves many different stakeholders. Governments are concerned with attracting skilled labour to Canada. Professional organisations are interested in creating regulated systems to assess professional credentials and work experience in a fair and consistent manner across occupations. Universities and other educational institutions must consistently evaluate educational credentials and identify gaps in education. Immigrants will benefit from programs that recognise their skills and professional qualifications. Should this level of cooperation be achieved between governments, proCessional organisations, educational institutions, and immigrants, not only will immigrants face fewer barriers to full labour market participation, but productivity in the Canadian economy will also increase.

An important question for governments to consider is whether it is better to create policies that address the intersections of identity markers or whether "policy coordination and delivery by government agencies" should be used to address the influence of intersecting identities on social outcomes (Burstein, 2003:4). If new government policies are the answer, this may entail an entirely new approach to government. Would Canadians be amenable to this? Probably not. Rapid social change is a cause of stress and concern, and any government would be constrained by the time and costs necessary to implement many of these new policies.

Another important question that must be addressed before policy change can take place is addressing the potential problems of policies directed at "special" groups. If legislation is created to address variations in intersections of identity markers, the result may encourage rather than discourage discrimination. In other words, the act of addressing inequality, disability, or visible minority status may encourage differential treatment of persons. Policies designed to account for intersections of diversity must be wary of this unintentional outcome.

CONCLUSION

A perspective on the intersections of identity is intended to encourage a new way of confronting various social inequalities and discrimination in both research and policy. While Canadian society has made tremendous gains in changing attitudes and decreasing tolerance of discrimination, many injustices remain. Despite their long history, legislation such as the Employment Equity Act, the Multiculturalism Act and others have largely failed in their mandate to increase acceptance of diversity and decrease inequality and discrimination. A perspective on the intersections of diversity may encourage all stakeholders to recognise and value the diversity that defines Canadian society. It will also help to increase the ability to alleviate certain social inequalities by targeting programs and services toward the individuals most in need of them. Research that uses an intersectional perspective will result in a more holistic vision of society.

As the previous discussion indicates, however, there are some dangers to the intersectional perspective. Examination of the intersections of diversity may further isolate certain individuals and entrench certain stereotypes given the inherent increase in attention to marginalised groups. It will also be a costly endeavour in terms of the time and finances needed to pursue consultation with all stakeholders and in the costs associated with bringing all the groups together. Nakache (2003) warns, however, that intersectional theorising may not be the best solution to alleviating discrimination:

 Looking at identity markers may not, however, be the most useful way
 to analyse issues linked to justice and the intersection of
 diversity; some argue that we need to primarily phrase the
 discussion in terms of policy areas (access to justice, legal aide,
 language barriers, restorative justice, etc.), rather than on
 markers of identity or intersections of identity. Nakache, 2003:5

In other words, perhaps the focus should be on changing the way society is organised rather than on focusing on individuals and their varying characteristics. This is an important question that stakeholders need to consider before embarking on a large-scale policy change. Despite this potential problem and the need for governments to sensibly consider their options in this endeavour, an intersectional perspective would definitely advance research given its emphasis on multi-disciplinary and multi-methodological approaches. The result will be a more comprehensive understanding of social diversity.

ACKNOWLEDGEMENTS

I would like to thank James S. Frideres and the three anonymous reviewers for their helpful suggestions regarding this paper. The Department of Canadian Heritage, Multiculturalism Branch, is gratefully acknowledged for their financial support. The encouragement of the Association for Canadian Studies and the Metropolis Project is also appreciated.

NOTES

(1.) Combining two variables into one measure is a common occurrence in statistical research. The reason variables are combined is that the measures are observed to be too similar, meaning that the two variables may be measuring the same concept, thus violating the assumption that each variable in the equation must measure a discrete characteristic.

(2.) While Aylward makes the claim that intersections must be taken into account to fully understand the outcome of any event, given her support of Critical Race Theory, she prioritises race as the central identity marker.

REFERENCES

Aylward, Carol A. (1999) Canadian Critical Race Theory: Racism and the Law Halifax: Fernwood.

Bakan, Abigail and Daiva Stasiulis (1997) Not One of the Family: Foreign Domestic Workers in Canada. Toronto: University of Toronto Press.

Burstein, Meyer (2003) "Business Formation, Entrepreneurship, Self-Employment and Capital Markets." Intersections of Diversity Challenge Papers. Ottawa: Metropolis Project, Association for Canadian Studies, and Canadian Heritage.

Brewer, Rose M., Cecilia A. Conrad, and Mary C. King (2002) "The Complexities and Potential of Theorizing Gender, Caste, Race, and Class." Feminist Economics 8(2): 3-18.

Creese, Gillian and Daiva Stasiulis (1996) "Introduction: Intersections of Gender, Race, Class and Sexuality." Studies in Political Economy 51(3): 5-14.

Deaux, Kay (2001) "Social Identity." Pp. 1-9 in Encyclopedia of Women and Gender, Volumes One and Two. New York: Academic Press.

Freiler, Christa and Judy Cerny (1998) Benefiting Canada's Children: Perspectives on Gender and Responsibility. Ottawa: Status of Women Canada.

Frideres, James S. (1998) Aboriginal Peoples in Canada. Don Mills, ON: Oxford University Press.

Frideres, James S. (2002) Immigrants, Integration and the Intersection of Identities. Working Paper. Ottawa: Canadian Heritage.

Gilmartin, Katie (2002) "The Culture of Lesbianism: Intersections of Gender, Ethnicity and Sexuality in the Life of a Chicana Lesbian." Pp. 160-279 in Gender, Nonconformity, Race and Sexuality: Charting the Connections. Toni Lester, ed. Madison, WI: University of Wisconsin Press.

Hicks, Peter (2003) "Social Transfers." Intersections of Diversity Challenge Papers. Ottawa: Metropolis Project, Association for Canadian Studies, and Canadian Heritage.

Kligman, Sarah J. (2003) "Education and (Re)Training and the Intersections of Diversity: Challenge Paper." Intersections of Diversity Challenge Papers. Ottawa: Metropolis Project, Association for Canadian Studies, and Canadian Heritage.

Krane, Julia, Jacqueline Oxman-Martinez, and Kimberley Ducey (2000) "Violence Against Women and Ethnoracial Minority Women: Examining Assumptions About Ethnicity and Race." Canadian Ethnic Studies/Etudes ethniques au Canada 32(3): 1-18.

Lester, Toni (2002) "Race, Sexuality, and the Question of Multiple, Marginalized Identities in the U.S. and European Discrimination Law." Pp. 84-101 in Gender Nonconformity, Race and Sexuality: Charting the Connections. Toni Lester, ed. Madison, WI: University of Wisconsin Press.

Li, Peter S. (2003) Destination Canada: Immigration Debates and Issues. Don Mills, ON: Oxford University Press.

Mackenzie, Patrick (2003) "'Information and Knowledge." Intersections of Diversity Challenge Papers. Ottawa: Metropolis Project, Association for Canadian Studies, and Canadian Heritage.

Mwarigha, M.S. (2003) "Housing." Intersections of Diversity Challenge Papers. Ottawa: Metropolis Project, Association for Canadian Studies, and Canadian Heritage.

Nakache, Delphine (2003) "Justice." Intersections of Diversity Challenge Papers. Ottawa: Metropolis Project, Association for Canadian Studies, and Canadian Heritage.

Ng, Roxana (1993) "Racism, Sexism and Nation Building in Canada." Pp. 50-59 in Race, Identity and Repression in Education. C. McCarthy and W. Crichlow, eds. New York: Routledge.

Razack, S. (1998) Looking White People in the Eye : Gender, Race and Culture in the Courtrooms and Classrooms. Toronto: University of Toronto Press.

Smith, Marilyn (2003) "Culture." Intersections of Diversity Challenge Papers. Ottawa: Metropolis Project, Association for Canadian Studies, and Canadian Heritage.

Statistics Canada (2003) On Poverty and Low Income. Ottawa: Minister of Public Works and Government Services.

Sullivan, Thomas J. (2000) Introduction to Social Problems. Fifth Edition. Boston: Allyn and Bacon.