Académique Documents
Professionnel Documents
Culture Documents
JOEL S. FETZER
Keywords: Realistic group conflict, British Columbia, First Nations, treaties, ecological inference
Résumé
Bien que les peuples autochtones constituent l’un des groupes les plus socialement, économique-
ment et politiquement marginalisés au monde, relativement peu de scientifiques sociaux quanti-
tatifs ont étudié au niveau mondial ou au Canada l’opinion publique sur cette ethnie. Cette note
de recherche cherche les racines de ces attitudes en étudiant pourquoi les citoyens de la Colombie-
Britannique : 1. ont décidé de participer aux élections et 2. ont voté «oui» sur le controversé
Référendum sur la négociation des traités autochtones de l’an 2002. L’essai examine des théories
scientifiques générales sur les attitudes à l’égard des minorités ethniques. Cette révision théorique
indique que le soutien pour la mesure pourrait être le résultat des conflits réels entre les Premières
Nations et les Canadiens d’origine européenne. Pour évaluer cette théorie, la note de recherche
examine ensuite les données du sondage Reid Express BC d’Avril 2002 ainsi que du recensement
et des chiffres officiels du vote. Les tableaux croisés de l’analyse du sondage Reid et l’analyse
écologique des statistiques électorales selon la méthode de Gary King indiquent que les différences
régionales et ethniques, mais pas les circonstances économiques individuelles, semblent avoir
motivé une grande partie du conflit sur cette mesure.
Mots clés : Théorie des conflits réels, Colombie-Britannique, Premières Nations, traités, l’analyse
écologique
INTRODUCTION
Although Indigenous peoples around the world suffer from poor economic outcomes
and widespread discrimination and prejudice (World Bank n.d), the global literature
on Indigenous rights has mainly focused on the antecedents of current policies (e.g.,
Anaya 2009; Attwood and Markus 1999; Barker 2005) rather than the causes of anti-
Indigenous public attitudes per se. Outside of Canada, a small group of scholars such
as Fetzer and Soper (2011), Pedersen et al. (2004), and Saldierna and Muñiz (2012)
has used advanced statistical methods to estimate the effects of various cultural, psy-
chological, and demographic variables on mass-level views of Indigenous rights in
general samples. Among the more sociological regressors from these publications,
advanced education, youth, and residence in a region with little Indigenous popula-
tion appear to have reduced self-reported hostility to native peoples.
Within Canada, scholars have similarly examined the history, culture, and poli-
tics of First Nations (Flanagan, Alcantara, and Le Dressay 2010; Havemann 1999), and
several investigators have used quantitative analysis of public-opinion polls to study
general attitudes toward Indigenous groups. These publications conclude that
Canadians’ views of First Nations arise from birth cohort (Ponting 2000), region
(Gibbins and Ponting 1978; Martin and Adams 2000), the proportion of natives in
the local population (Kalin and Berry 1982), gender (Delic 2008), unemployment
(Delic 2008), support for limited government (White et al. 2015), previous contact
with Indigenous individuals (Morrison, Morrison, and Borsa 2014; but see Denis
2015), education (Delic 2008), ethnic prejudice, economic conservatism, perceived
group conflict (Langford and Ponting 1992; see also White et al. 2015), and Catholic
identity (Delic 2008; but see Morrison, Morrison, and Borsa 2014). While such inves-
tigations give us the general outline of a theory of attitudes toward Indigenous
Canadians, these reports typically base their findings on abstract questions from pub-
lic-opinion surveys instead of on exit-polls or actual voting returns from a relevant
ballot measure or election. Ordinary respondents might not be particularly well-
informed about or interested in Indigenous issues unless the question has somehow
become highly contentious or politicized at the federal, provincial, or local level. The
somewhat unthinking responses interviewees might give during a generic social sur-
vey may differ from those they would proffer in the heat of an extended political cam-
paign over Indigenous matters. And until we are sure of the underlying roots of
public opposition to First Nations’ interests, we remain ill-prepared to suggest effec-
tive educational measures or political strategies to advance Indigenous rights.
Joel S. Fetzer | 151
This research note therefore focuses on this gap in the literature by examining
the factors that help explain ordinary citizens’ views of Indigenous rights during one
critical case of high political mobilization around this question, the 2002 Aboriginal
Treaty Negotiations Referendum in British Columbia. In particular, I investigate why
ordinary British Columbians decided 1. to vote in; and 2. to vote “yes” on two repre-
sentative and substantively important items from the controversial measure (i.e.,
whether one agreed that “Private property should not be expropriated for treaty set-
tlements” in question 1 and whether one affirmed that “Aboriginal self-government
should have the characteristics of local government, with powers delegated from
Canada and British Columbia” in question 6). Although only two short, qualitative
research notes (Fetzer 2016; Kajlich 2002) and two unpublished MA theses (Lozanski
2002; Rickey 2002) have looked at this referendum in any depth, both tabulations
from survey data and actual electoral returns are readily available for analysis.
This study primarily derives its empirical hypotheses from “realistic group conflict”
(Bobo 1983; Campbell 1965; Sherif et al. 1961). This prominent theory of ethnic rela-
tions holds that people’s political behavior generally follows from their economic self-
interest (Downs 1957) and/or “subjectively appreciated threat and challenges to group
status” (Bobo 1983, 1196). During the illustrative Robbers Cave Experiment (Sherif et
al. 1961, 151), for example, inter-group tension “was produced systematically through
the introduction of conditions of rivalry and frustration perceived by the subjects as
stemming from the other group.” The theory further states that conflict will exist
unless an individual believes that cooperation with the other group is good for him or
her personally. Or the two subsets of people may reach harmony if they together
embrace “superordinate goals . . . the attainment of which is compelling but which
cannot be achieved by the efforts of one group alone” (Sherif et al. 1961, 183).
In many ways, this theory seems well-suited to explain voting on the 2002
Aboriginal Treaty Referendum. First Nations and European Canadians constitute
two relatively clearly demarcated groups in British Columbia, and each party has
experienced many frustrations attributed rightly or wrongly to the other. Both
groups are competing for limited material resources such as land as well as less tan-
gible assets such as prestige in the eyes of most Canadians. Historically, the two enti-
ties have all-too-often shown little sign of cooperating or embracing common goals;
rather, the European-Canadian-dominated federal and provincial governments have
usually neglected if not actively suppressed the interests of Indigenous North
Americans. We might therefore translate Bobo’s (1983, 1198) U.S.-based formula-
152 | Canadian Ethnic Studies/Études ethniques au Canada
tion of conflict theory into Canadian terms by contending that “[Canadian] social
organization allows and fosters in whites the belief that [First Nations], insofar as
they demand changes in the racial status quo, are a threat to their life-styles, as well
as to other valued resources and accepted practices.”
In the case at hand, higher-income British Columbians in 2002 might oppose
government payments or land transfers to First Nations out of fear that their taxes
would thereby increase. The unemployed might similarly be more concerned about
any economic harm from a pro-First Nations treaty. Fearing a decline in their local
political influence and access to previously public natural resources, European-ori-
gin rural residents surrounded by Crown lands and a greater proportion of First
Nations individuals may be more opposed to pro-Indigenous treaty provisions than
would apartment dwellers or other urbanites in Vancouver (see Plant 2013). Visible
Minorities, on the contrary, should in the aggregate be more willing to accommo-
date the interests of other minorities such as First Nations since these respondents
are presumably less opposed to shifting political power, social status, and perhaps
even some economic resources1 away from the European-Canadian majority.
Although the above factors seem most relevant for testing group-conflict theory,
this research note also briefly examines the possible effects of such other demo-
graphic or control variables as education, age, gender, and religious practice. The first
three controls are standard for social-scientific analyses, and I included the fourth
regressor because of the well-documented anti-referendum mobilization among
many of British Columbia’s religious groups in 2002. In particular, provincial leaders
of the Anglican Church (Crawley et al. 2002), United Church (BC Conference 2002),
and Jewish Congress (Maclean’s 2002), among others, denounced the exercise and
instead urged solidarity with the region’s Indigenous peoples (see also Morales 2013).
To evaluate these hypotheses about the causes of public support for the Treaty
Referendum, this essay relies on two data sources. The first is the BC Reid Express
poll conducted by Ipsos-Reid (2002) during 2–9 April 2002. Interviewers spoke to
the 800 adult BC residents by telephone, and the firm then weighted the results by
region, age, and sex to match the province’s demographics according to the 1996
census. Unfortunately, bivariate crosstabs are the only form in which data from this
survey now seem to exist (Braid 2013). The second data resource is the official report
of the BC Chief Elections Officer (Elections BC 2002), which contains the actual
numbers of registered voters and valid “yes” and “no” ballots for the six questions of
the referendum in each of the 79 provincial electoral districts. I later merged these
electoral results with demographics from the 2001 census (BCStats 2001).
Joel S. Fetzer | 153
RESULTS
How well does realistic group conflict predict reported attitudes toward the referen-
dum? Table 1 suggests that most strictly economic forms of the theory did not find
support in these data. Although a few trends appeared, the results for neither income
nor employment status reached statistical significance for reported turnout or antic-
ipated voting on either question (because First Nations and their advocates were
against the wording of the questions and even holding such a referendum in the first
154 | Canadian Ethnic Studies/Études ethniques au Canada
Source: Ipsos-Reid (2002). * p < .05. Only valid respondents included in analysis (i.e., not those answering “don’t
know”).
Joel S. Fetzer | 155
place, I treated both “no” voting and complete abstention as measures of opposition).
Regional differences did appear, however. Residents of metropolitan Vancouver
seemed especially unlikely to participate in the referendum, while substantially more
respondents from the rural “rest of British Columbia” and partly rural Vancouver
Island said they intended to mail in their ballot. Among the controls, higher educa-
tion appears to have suppressed “yes” votes on both questions, and advanced age sim-
ilarly seems to have discouraged support for item 1 (private property).
TABLE 2. Percentage of Yes Voters and Turnout for Question 1 of Treaty Referendum
TABLE 3. Percentage of Yes Voters and Turnout for Question 6 of Treaty Referendum
Table 2 presents the results of the ecological and parallel analyses for question 1.
Here, apartment dwellers were much less likely to return their ballots or vote “yes” on
the measure. Rural residents in general exhibited a higher turnout rate than did their
urban or suburban peers, and voters from the country were also slightly more willing
to affirm the item on private property. Visible Minorities showed opposition to
European Canadians’ interests but solidarity with First Nations by turning out in
much-reduced numbers and voting “yes” significantly less often. As predicted, the con-
trol variable of religiosity correlated positively with abstaining and “no” voting. And
education appears to have also decreased willingness to participate in the referendum.
Joel S. Fetzer | 157
The final table largely confirms these findings. People living in apartments once
again disproportionately failed to participate or else massively voted “no.” Citizens
registered in rural constituencies were a little more likely both to mail back their bal-
lots and to vote “yes.” Visible Minorities, in contrast, tended disproportionately to
have either disagreed with the exercise in principle or voted “no” on question 6
(Aboriginal sovereignty). Among the control variables, religiosity was associated with
less “yes” voting, and a university degree made one almost certain not to participate.
CONCLUSION
Overall, analysis of both survey and ecological data suggests that several manifesta-
tions of realistic group conflict can help us understand why ordinary British
Columbians supported or opposed the 2002 Aboriginal Treaty Negotiations
Referendum. In particular, regional differences pitting urbanites living on privately
owned land in Vancouver against the rural population residing on or near Crown
lands seem to have motivated much of the conflict over this controversial poll (cf.
Gibbins and Ponting 1978). In addition, the less tangible competition for status
between the ethnically European majority and Indigenous or other racial minorities
might have accentuated political discord in 2002 (also following Langford and
Ponting 1992). The typically null results for income and employment status, in con-
trast to findings by Delic (2008) but parallel to those by Martin and Adams (2000),
do not lend credence to more narrowly self-interested forms of realistic group con-
flict. Thus, purely individual economic circumstances do not seem to have played
much of a role in determining one’s views of indigenous rights, perhaps because
British Columbians are not mobilizing politically around these poles. Rather, a more
group-based calculus seems to be at work in provincial politics.
Of course, King’s ecological inference method more easily yields estimates of
bivariate relationships than multivariate ones. In the datasets of official electoral and
census figures, for instance, Visible Minorities appear less supportive of the election
itself and of the “yes” side (see Tables 2 and 3). Because this correlation is only bivari-
ate, however, we cannot be sure that some third factor (e.g., urbanicity) is not actu-
ally the root cause; perhaps Visible Minorities merely appear to be pro-Indigenous
rights because they live disproportionately in major metropolises such as Vancouver.
Their ethnicity would not then be driving their treaty-related attitudes, but rather
their urban environment would. We should probably therefore remain cautious
about concluding that ethnicity is necessarily an important determinant of views on
Indigenous rights and instead recommend that future researchers in this area con-
tinue to test related hypotheses, especially using multivariate methods.
The significant results for region, in contrast, are undeniable. Standard χ2 analy-
158 | Canadian Ethnic Studies/Études ethniques au Canada
sis of Reid’s individual-level survey data in Table 1 demonstrates that rural residents
were more likely to take part in the referendum. Ecological inference of electoral and
census numbers in Tables 2 and 3 likewise reveals both greater participation and
increased “yes” voting among those living outside of urban areas. And official, pop-
ulation statistics from BC’s electoral and demographic authorities also document
conclusively that rural residents were more enthusiastic about the referendum itself
and more in favor of the “yes” side (cf. Rural vs. non-Rural rows in Tables 2 and 3
based on returns by constituency).
These findings about the rural/urban divide, at least, thus appear strong enough
to support certain policy recommendations. In particular, activists in the field might
consider the best way to rebut arguments claiming that generous treaty settlements
or other pro-First Nations policies will be a zero-sum game in which non-
Indigenous rural residents will permanently lose political influence, social status,
and access to Crown lands and other valuable resources. Instead, proponents might
try to paint equitable approaches to native issues as win-win solutions in which
European Canadians living in the countryside also stand to gain economically, polit-
ically, and socially through collaboration with First Nations.
After the Tla’amin Nation concluded its treaty with Canada and British
Columbia, for instance, the group’s Chief Negotiator described the effect on politi-
cal and economic relations among local ethnic groups as follows:
So far, in our part of the world, up in Powell River, the completed treaty is good news.
There’s been a whole lot of new relationships built with the City of Powell River; for
example, the regional district, the local chambers of commerce, [and] the development
corporation boards are cross-pollinating . . . [in a] very positive [way]. I think that dis-
cussion [about the treaty] at the grocery store in Powell River is good. (Francis 2013)
For this First Nation in British Columbia, therefore, justice for Indigenous peoples
may have coincided with something like the welfare of all.
NOTES
1 Of course, if one views the acquisition of land and other property as a zero-sum game, one might argue that
Visible Minorities should oppose First Nations’ exclusive claims to Crown lands since such demands threaten Visible
Minorities’ own access to those areas. Because the Province’s Visible Minorities tend to live in cities such as Vancouver
that contain relatively little Crown land, however, these individuals might see Indigenous land claims in relative rather
than absolute terms (i.e., the proportion of land held by the majority versus minorities is more important than the
absolute amount of property each group has access to) and perceive the question as more academic than concrete
because urban Visible Minorities are unlikely to use Crown lands frequently whoever controls them.
2. As King details in his book-length explanation of this technique (1997, 28-140), the maximum-likelihood rou-
tine estimates two major parameters. The first, labeled βb, could be the proportion of some subset of all citizens (e.g.,
African Americans), who turn out to vote in a particular election. The second, labeled βw, is the proportion of all remain-
ing citizens (i.e., not members of that subset) who turned out to vote. EzI also provides standard errors for both of these
estimates, which allow researchers to calculate the precision of EzI’s estimates of βb and βw. At least under ideal circum-
Joel S. Fetzer | 159
stances, MLE triangulates in on the most plausible values for these two parameters by searching across a (for the bivari-
ate case) two-dimensional plane that represents all possible combinations of βb and βw and plots all observed pairs of the
independent and dependent variables by district (e.g., the proportion African-Americans in a given county versus the
proportion of citizens who turned out to vote in that same area). EzI represents this plane as an MRI-like tomography
plot that highlights the most likely combination of these two statistics in the underlying population of individuals. The
EI2 routine in the program extends this bivariate approach to more than one dimension, allowing for limited multivari-
ate analysis (e.g., by calculating the proportion of African Americans who voted Democratic versus Republican given
that they turned out in the first place).
REFERENCES
———, and J. Christopher Soper. 2011. The Determinants of Public Attitudes toward the Rights of
Indigenous Peoples in Taiwan. Taiwan Journal of Democracy 7.1: 95–114.
Flanagan, Tom, Christopher Alcantara, and André Le Dressay. 2010. Beyond the Indian Act: Restoring
Aboriginal Property Rights. Montreal: McGill-Queen’s University Press.
Gibbins, Roger, and J. Rick Ponting. 1978. Canadians’ Opinions and Attitudes Towards Indians and Indian
Issues: Findings of a National Study. Ottawa: Department of Indian and Northern Affairs.
Havemann, Paul, ed. 1999. Indigenous Peoples’ Rights in Australia, Canada & New Zealand. Auckland:
Oxford University Press.
Ipsos-Reid. 2002. BC Reid Express – April 2002 (0066-00558844-7777): Detailed Tables. Electronic file.
Vancouver: IPSOS-Reid.
Kajlich, Helena. 2002. The British Columbia treaty referendum: An appropriate democratic exercise?
Indigenous Law Bulletin 5.17: 11–14.
Kalin, Rudolf, and John W. Berry. 1982. The social ecology of ethnic attitudes in Canada. Canadian
Journal of Behavioural Science 14: 97–109.
King, Gary. 1997. A Solution to the Ecological Inference Problem: Reconstructing Individual Behavior from
Aggregate Data. Princeton: Princeton University Press.
Langford, Tom, and J. Rick Ponting. 1992. Canadians’ Responses to Aboriginal Issues: The Roles of
Prejudice, Perceived Group Conflict, and Economic Conservatism. Canadian Review of Sociology and
Anthropology 29.2: 140–166.
Lozanski, Kristin Elizabeth. 2002. Public Space, Democracy, and Colonialism: British Columbia’s
Referendum on Treaty Principles. MA thesis, University of Victoria.
Maclean’s. 2002. BC Referendum Controversy. April 13. http://www.thecanadianencyclopedia.ca/en/arti-
cle/bc-referendum-controversy/ [22 July 2016].
Martin, David, and Chris Adams. 2000. Canadian Public Opinion Regarding Aboriginal Self-
Government: Diverging Viewpoints as Found in National Survey Results. American Review of
Canadian Studies 30.1: 79–88.
Morrison, Todd G., Melanie A. Morrison, and Tomas Borsa. 2014. A Legacy of Derogation: Prejudice
toward Aboriginal Persons in Canada. Psychology 5: 1001–1010.
Palmer, Vaughn. 2001. Keeping a promise as innocuously as possible. Vancouver Sun, 4 December, A10.
Pedersen, Anne, et al. 2004. Attitudes toward Indigenous Australians: The Role of Empathy and Guilt.
Journal of Community & Applied Social Psychology 14.4: 233–249.
Ponting, J. Rick. 2000. Public Opinion on Canadian Aboriginal Issues, 1976–98: Persistence, Change, and
Cohort Analysis. Canadian Ethnic Studies/Études ethniques au Canada 32.3: 44–75.
———. 2006. The Nisga’a treaty: polling dynamics and political communication in comparative context.
Peterborough, ON: Broadview Press.
Rickey, Pamela Mary. 2002. Talking to the Wall: Public Input into the Wording of the British Columbia
Treaty Referendum Questions. MA thesis, University of British Columbia.
Robinson, William S. 1950. Ecological Correlation and the Behavior of Individuals. American Political
Science Review 15: 351–357.
Saldierna, Alma Rosa, and Carlos Muñiz. 2012. Impacto del consumo mediático sobre el apoyo a las medi-
das políticas dirigidas hacia indígenas. Evaluación del posicionamiento de la población mexicana.
Estudos em Comunicação 11(May): 67–88.
Sherif, Muzafer, et al. 1961. Intergroup Conflict and Cooperation: The Robbers Cave Experiment. Norman:
Institute of Group Relations, University of Oklahoma.
University of British Columbia. 2001. British Columbia Provincial Election 2001.
http://esm.ubc.ca/BC01/results.html [30 July 2014].
White, Stephen, et al. 2015. Public Policies toward Aboriginal Peoples: Attitudinal Obstacles and Uphill
Battles. Canadian Journal of Political Science/Revue canadienne de science politique 48.2: 281–304.
World Bank. n.d. Indigenous Peoples: Still Among the Poorest of the Poor. Policy Brief of the World Bank.
http://siteresources.worldbank.org/ EXTINDPEOPLE/ Resources/407801-1271860301656/HDNEN_
indigenous_clean_0421.pdf [26 July 2014].
Joel S. Fetzer | 161