2. Attitudes toward Aboriginals and Aboriginal Policies

While private polling companies in Canada have regularly released research on Aboriginal issues (for recent examples see Environics, 2010; Ipsos Reid, 2013), academic studies are less numerous. J. R. Ponting and his various co-authors (Langford and Ponting, Reference Langford and Rick Ponting1992; Ponting Reference Ponting and Satzewich1998, Reference Ponting2000; Ponting and Gibbins, Reference Ponting and Gibbins1980, Reference Ponting and Gibbins1981) have made the largest systematic contribution to our knowledge. Relying on four national surveys administered in 1976, 1986, 1994 and 1998 that dealt exclusively with attitudes towards Aboriginal peoples and attitudes towards Aboriginal policy, Ponting and his colleagues found, among other things, that Canadians give little thought to these matters. Early surveys showed that attitudes were somewhat inchoate and inconsistent but not uniformly negative. Ponting and Gibbins (1980: 92) described them as dominated by “indifference and moderation.” Years later Ponting (Reference Ponting2000) found little change in levels of familiarity with Aboriginal issues and slight declines in support for issues like land claims, although a solid majority of Canadians still expressed what he described as pro-Aboriginal attitudes. The exception was any hint of “special status.” Ponting found Canadians continue to be highly resistant to the idea that equality meant anything other than treating everyone equally. The treaties conferred no special rights in the minds of Canadians and the concept of cultural protection for Aboriginals barely achieved majority support in the various surveys.

In Canada there has been only a modest amount of research on the structuring of attitudes toward Aboriginals. In one of the most important studies, Langford and Ponting (Reference Langford and Rick Ponting1992) found three critical determinants of Aboriginal policy attitudes: levels of prejudice towards Aboriginal peoples, group conflict perception and economic conservatism. First, Canadians who held unfavourable attitudes toward Aboriginal peoples (prejudice) were less likely to support policies giving greater autonomy or special status to Aboriginal peoples and were less likely to see Aboriginal issues as a priority. Second, those who felt Aboriginal people receive favourable treatment from governments while non-Aboriginal needs are neglected (group conflict) were also less likely to support Aboriginal policies or see Aboriginal issues as a priority. Of particular importance was the interaction between group conflict and prejudice; underlying prejudice became important for attitudes only when group conflict was prominent in the eyes of respondents. Finally, economic conservatism was associated with lower levels of support for policies giving greater autonomy or special status to Aboriginal peoples. In other words, some Canadians oppose special treatment for Aboriginals on broad ideological grounds, independent of whatever prejudice they may (or may not) harbour.

Though there have been many important developments in the nearly three decades since their data were collected (1986), Langford and Ponting's Reference Langford and Rick Ponting1992 study is the most recent to examine the sources of Canadian attitudes towards Aboriginal policies. Many Canadians may well have been unfamiliar with Aboriginal issues in the mid-1980s; the contemporary indigenous rights movement in Canada was still relatively new, and the idea of self-government had only recently entered political discourse (Papillon, Reference Papillon, Turgeon, Papillon, Wallner and White2014). The subsequent years saw the failures of the Meech Lake and Charlottetown accords, the Oka crisis, the Royal Commission on Aboriginal Peoples, the Nisga'a Treaty and other agreements, and the emergence of the Idle No More movement. These numerous developments in the intervening years, combined with strenuous demands for justice and reconciliation (for example, Saul, Reference Saul2014), suggest that a new examination of the determinants of public attitudes toward Aboriginal policies is warranted.

To provide some perspective on these attitudes, it is useful to begin by considering the more extensive body of research on attitudes concerning policies directed at a different minority group in another setting: African Americans in the United States. There are two clear perspectives on the sources of public opinion about racial policy in the United States (see Hochschild, Reference Hochschild, Sears, Sidanius and Bobo2000). Sidanius and colleagues sum up the first, and certainly more prominent view: “Group identity and social dominance values are among the most important factors driving the racial policy attitudes of white Americans” (Sidanius et al., Reference Sidanius, Singh, Hetts, Frederico, Sears, Sidanius and Bobo2000: 228). Although there are many different strands of thought within this perspective—some focus on the link between racial antipathy and a variety of core values (Kinder and Sanders, Reference Kinder and Sanders1996; Sears et al., Reference Sears, van Laar, Carrillo and Kosterman1997), others on group interests, identities, and beliefs about group status (Bobo et al., Reference Bobo, Kluegel, Smith, Tuch and Martin1997) —the common thread linking them is their emphasis on the crucial role of racial antipathy in shaping opposition to policies aimed at addressing racial differences. Racism no longer takes the form of racial hatred or doctrines of racial superiority, but rather “symbolic” racism or “racial resentment,” which is characterized by more subtle “legitimizing myths” that allow respondents to express their racial attitudes through their opposition to special favours or their belief that discrimination no longer exists.

Paul Sniderman and his colleagues maintain, in contrast, that there is no single “race” issue but a variety of issues with different political dynamics. Central to their account are white Americans' ideological predispositions, and the ways in which arguments about policies aimed at helping achieve racial equality are framed (Sniderman and Piazza, Reference Sniderman and Piazza1993; Sniderman et al., Reference Sniderman, Crosby, Howell, Sears, Sidanius and Bobo2000). White Americans who support efforts to improve the socioeconomic conditions of African Americans, for example, are not necessarily the same people who support affirmative action policies (Sniderman and Piazza, Reference Sniderman and Piazza1993). Among the various sources of support for and resistance to policies that would assist African Americans, white Americans' political ideology is as likely a contender as racism. For example, while there is evidence of white resentment toward such initiatives as state sponsored scholarships for blacks, those who define themselves as “conservatives” are generally opposed to this kind of policy regardless of the racial group that would benefit (Feldman and Huddy, Reference Feldman and Huddy2005).

The influence of ideology on policy attitudes comes with an important qualification. Ideological beliefs do not shape policy attitudes equally for everyone. Political “sophisticates” —those with higher levels of education or political interest—are more capable than others of connecting policy positions to political principles (Sniderman et al., Reference Sniderman, Brody and Tetlock1991). When it comes to the politics of race in the United States, the influence of ideological orientations on policy attitudes clearly hinges on respondents' levels of education; differences in the policy preferences of “liberals” and “conservatives” are sharper among the college-educated and more muted among those with lower levels of education (Sniderman and Piazza, Reference Sniderman and Piazza1993: 123; Sniderman et al., Reference Sniderman, Crosby, Howell, Sears, Sidanius and Bobo2000: 264).

This brief overview of pertinent attitudinal research suggests three tentative lines of inquiry that relate to Aboriginal public policy. First, attitudes toward policies supportive of Aboriginal peoples might depend on exactly what type of public policy is being contemplated. Research on attitudes towards race policies in the United States suggests that support for the amelioration of hardship will likely induce a different response from policies framed as recognizing special needs. Second, to the extent that antipathy towards Aboriginal peoples underlies attitudes toward Aboriginal policies, it is much more likely to be expressed in subtle forms than in blatant assertions of superiority. For these reasons, our approach to prejudice will be to focus on generalized attitudes toward outgroups, rather than antipathy toward Aboriginal peoples specifically. The social psychology literature strongly suggests that prejudice is generalized: people who dislike one outgroup are inclined to dislike many others (Allport, Reference Allport1954; Altemeyer, Reference Altemeyer1998).

Third, it is by no means certain that prejudice is the most important determinant of attitudes toward particular public policies. Government interventions to address particularistic demands are vulnerable to many critiques, including ideological views regarding the appropriate role of the state. In this regard, and drawing on the work of Sniderman and his colleagues (Reference Sniderman, Crosby, Howell, Sears, Sidanius and Bobo2000) with respect to race policy in the United States, we expect views about the role of government to determine policy positions among the most educated and politically attentive. Our working hypothesis is that the more educated our respondents, the more likely they will have the necessary degree of political sophistication to recognize which policy position is most closely aligned with their own political principles.

3. Data and Measurement

The Saskatchewan Election Study (SKES) and the Taking the Pulse of Saskatchewan (TTP) Survey were random sample surveys conducted by the University of Saskatchewan's Social Sciences Research Laboratories (SSRL) from November 8, 2011, to November 21, 2011, and March 5, 2012, to March 19, 2012, respectively. Both were deployed as a telephone survey using WinCATI software. The SKES includes responses from 1,099 Saskatchewan residents 18 years of age and older (a 23.6% response rate), and the TTP survey includes responses from 1,750 Saskatchewan residents 18 years of age and older (a 34.3% response rate). 76 and 104 respondents in the SKES and TTP survey, respectively, self-identified as Aboriginal. Our dependent variables are attitudes toward Aboriginal public policies, and we employ two distinct sets of items as measures. These are described in detail in the next section.

The primary explanatory variables are prejudice and ideology. Our measure of prejudice focuses not on Aboriginal peoples, but on a different outgroup, namely, immigrants. Research has shown, with respect to racial policy attitudes in the United States, that negative sentiments toward African Americans are difficult to disentangle from policy considerations (Shuman, Reference Shuman, Sears, Sidanius and Bobo2000; Carmines et al., Reference Carmines, Sniderman and Easter2011); indeed, for many people, resentment over policies may well drive negative sentiments toward outgroups (Sniderman and Piazza, Reference Sniderman and Piazza1993). To avoid this entanglement, respondents were asked about foreign-born residents, an outgroup for which we are also able to assess prejudiced attitudes. We use respondents' level of disagreement with two statements as our TTP survey measure (Cronbach's alpha = .47, scale = 0–1, mean = .30): “New immigrants make a valuable contribution to Saskatchewan,” and “New immigrants to Saskatchewan should be encouraged to retain their cultural heritage.”Footnote ² The SKES measure is level of agreement with the statement, “Too many recent immigrants just don't want to fit into Canadian society” (scale = 0–1, mean = .54).

Although these measures do not precisely capture the full effects of prejudice—there are other reasons people might agree or disagree with these statements—they have considerable advantages over alternative measures. Alternative indicators of prejudice, especially those that focus solely on Aboriginal peoples, might be more strongly related to policy attitudes, but their “independent” effects would be highly uncertain. This approach mitigates the problem of determining causality by tapping a broader propensity toward prejudice. People rarely differentiate between outgroups when it comes to prejudice (Allport Reference Allport1954; Altemeyer Reference Altemeyer1998), so to the extent that sentiments towards a different outgroups (immigrants) are related to Aboriginal policy, we can be more confident about the causal direction.

To analyze the effects of ideology, we focus on beliefs about the role of government, measured by a “limited government index” consisting of responses to three questions in the TTP survey (alpha = .44, scale = 0–1, mean = .35), and four questions in the SKES (alpha = .62, scale = 0–1, mean = .41). Those who score higher on the measure in the TTP survey are more likely to “disagree” or “strongly disagree” with these three statements: “The Saskatchewan government should raise the income tax rate on high income earners in order to fund programs to bring low income individuals above the poverty level,” “I am willing to pay additional taxes so that prescription drugs are affordable for everyone in Saskatchewan,” and “The Saskatchewan government should not reduce workers' collective bargaining rights.” Those who score the highest on the SKES measure favour “having some private hospitals in Saskatchewan,” believe “the best way to deal with major economic problems” is to “leave it to the private sector,” believe “much less” should be done “to reduce the gap between the rich and the poor in Saskatchewan,” and are more likely to “agree” or “strongly agree” that “Saskatchewan should privatize some of its Crown corporations.” These indices measure the extent to which a respondent desires that government intervene in the economy and society to reduce economic inequality, enforce workers' rights, and increase public ownership.

The last set of measures taps into respondents' levels of political sophistication. Researchers often employ level of education or interest in politics as stand-ins for this concept. Fortunately, the same measure of education level, an 11-point scale ranging from “no schooling” to “a professional degree or doctorate,” is available in both surveys. The SKES also includes a measure of political interest that asks respondents, “Using a scale from 1 to 10, where 1 means not at all interested and 10 means very interested, how interested were you in the recent provincial election campaign?” Both variables have been re-scaled to 0–1.

4. Should Aboriginals Receive Special Treatment?

In some ways this is an empty question. Unlike African Americans, Aboriginal Canadians can claim an embedded constitutional status that, while unclear in its consequences, embodies a robust conception of group rights. The question of whether Aboriginal peoples should be assisted to increase their educational attainment, for example, has been answered by treaties and by the subsidies provided to those who are part of the Indian Register. Educational subsidies, mostly through the Post-Secondary Student Support Program, provide financial support for “status” Indians or Inuit to attend post-secondary institutions. The government of Canada can change the conditions under which these subsidies are obtained, or cap the amount received, but the principle that First Nations students are eligible for subsidies has long been accepted. Similarly, section 87 of the Indian Act exempts the “personal property of an Indian or a band situated on a reserve” from taxation, including income earned on reserve. Exemptions from other taxes, such as the GST/HST, may also apply, although the intention is that they be restricted to on-reserve activities. The list of special arrangements is a long one, and the rationale behind them is traceable to constitutional guarantees and treaty entitlements.

What do non-Aboriginal residents of Saskatchewan think of these sorts of arrangements? Do they distinguish between different kinds of policies? And how do their views differ from those of Aboriginals? We begin with the responses to questions in the TTP survey. Two patterns are apparent in the data presented in Table 1.Footnote ³ First, there are clear differences in opinion between the non-Aboriginal majority and Aboriginal minority in Saskatchewan: non-Aboriginals are far less likely than Aboriginal respondents to express positive views about special arrangements and targeted investments for Aboriginal communities. Far fewer non-Aboriginal than Aboriginal respondents agree that “Aboriginal self-government is important to the future of Saskatchewan,” “Public investments in Aboriginal education pay off in the long run” or that “Governments need to ensure that on-reserve housing is properly funded.” By the same token, non-Aboriginals are more inclined than Aboriginal respondents to agree that “Aboriginal people do not pay enough taxes in Saskatchewan” and they are also slightly less comfortable with the influence that Aboriginal communities exert. Indeed, despite the relatively small number of Aboriginal respondents in the sample, the differences between non-Aboriginal and Aboriginal responses to each of these five statements are large enough that we can be quite confident they are not due to random error.

However, if we focus exclusively on the opinions of non-Aboriginal respondents, there is second discernible pattern in the data presented in Table 1. Non-Aboriginals are more likely to express positive views about targeted investments that involve the amelioration of hardships than about other, more permanent special arrangements for Aboriginal communities. Solid majorities agree that spending on Aboriginal education is effective and that on-reserve housing should be properly funded. At the same time, most agree that Aboriginal communities have too much influence over government policy and that Aboriginals do not pay their fair share of taxes. They are also divided on the importance of Aboriginal self-government, with more expressing strong opposition than strong support.

Although Saskatchewan residents support the proposition that, broadly speaking, Aboriginal peoples require assistance, attitudes reveal traces of resentment that hinge on the belief that Aboriginal peoples have been, and continue to be, afforded unfair special advantages. Lingering misunderstandings regarding treaty rights and the significance of First Nations' tax treatment probably contribute to the belief that Aboriginal people do not pay enough in taxes. Confusion about treaty rights and agreements may also extend to the question of whether Aboriginal communities have undue influence in matters such as resource development. The idea that a minority group ought to have the right to self-government is also a potential source of resentment, yet Aboriginal self-government is a cornerstone of the aspirations of many Aboriginal peoples, as documented in the Royal Commission on Aboriginal Peoples (Canada, 1996). Controversy in this area is fed by differences of opinion regarding what Aboriginal self-government can, and should, consist of, as well as ignorance regarding what Aboriginal governments are currently responsible for. Public attitudes reflect policy controversy. The meaning of self-government, the best way to achieve it, and indeed what its benefits are, have been a source of considerable debate among federal and provincial politicians as well as Aboriginal leaders and communities (Belanger and Newhouse, Reference Belanger, Newhouse and Belanger2008).

The responses to questions about Aboriginal policy in the SKES survey also demonstrate the complexity of these issues. In this survey, the emphasis shifts from constitutional-legal issues to core policy questions, that is, questions that are not automatically justiciable. Once again, however, we are probing questions of special status, specifically whether it is sound practice to have special post-secondary institutions for Aboriginal peoples, whether Aboriginal peoples deserve a separate share of natural resource revenues, and more generally whether Aboriginal people should get by without any “special favours.” The questions about post-secondary institutions and natural resource politics are not hypothetical. In Saskatchewan, First Nations and Métis peoples manage the Saskatchewan Indian Institute of Technology, First Nations University and the Gabriel Dumont Institute. And although post-secondary institutions were not a major topic of debate during the 2011 provincial election, natural resource policies were. In response to the idea that natural resource revenues should be shared with Aboriginal peoples, the Saskatchewan party made clear its opposition, while the NDP declared its support.

The differences of opinion between non-Aboriginal and Aboriginal respondents, presented in Table 2, are again readily apparent. Non-Aboriginal respondents are less likely than Aboriginal respondents to believe that Aboriginal peoples should have their own post-secondary institutions or a separate share of Saskatchewan's natural resource royalties. Likewise, they are less inclined than Aboriginals to agree that governments should do more for Aboriginal people, or that longstanding discrimination is a source of economic disadvantage for Aboriginal people. Non-Aboriginals are also more likely to believe Aboriginals should get by “without any special favours.”

Once again, however, comparing the distributions of non-Aboriginal responses to each of these statements reveals considerable differences. There is widespread opposition to separate post-secondary institutions and a separate share of natural resource royalties. Whatever the reasons, it seems fair to conclude that the non-Aboriginal population in Saskatchewan is opposed to parallel institutional arrangements that would provide separate access to resources revenues and to education. Perhaps it is not surprising, then, that three quarters also reject “special favours” for Aboriginal peoples. Nevertheless, it seems easier for non-Aboriginals to contemplate governments doing more in general terms for Aboriginal peoples than providing specific and special arrangements. Approximately two in five non-Aboriginal respondents agreed that a general effort on the part of government is a good idea. And three in five see discrimination as a barrier to Aboriginal success.

What are we to make of these findings? The distinct differences of opinion between Aboriginal and non-Aboriginal respondents are perhaps not surprising, but they are noteworthy. Understanding the sources of opinion among those in the non-Aboriginal majority takes on greater importance precisely because the sharp differences in their views, compared to those of the minority Aboriginal population in Saskatchewan, is likely to give pause to policy makers, particularly at the political level.

The differences in aggregate levels of support for various policies among non-Aboriginal respondents warrant greater scrutiny. Arguably all of the questions contained in tables 1 and 2 involve public policy, but they are not the same kinds of policies. Those that speak to entrenched rights and political power and ask how much tolerance there is for special arrangements for Aboriginal peoples, receive relatively modest support; others that speak to the idea that more help is required to ameliorate hardships receive greater support. These distinct responses suggest there might be two different issue agendas in the Saskatchewan public: a “social welfare” agenda that enjoys considerable support and a “special arrangements” agenda that does not.

But if there were two different agendas, then we would expect to see attitudinal differences that go beyond aggregate levels of support. Specifically, there should also be some evidence of two distinct clusters in the individual-level responses to the TTP and SKES questions, with strong relationships among questions that focus on amelioration of hardship and strong relationships among those that emphasize special arrangements but weaker relationships between responses across the two clusters.

Tables 3 and 4 present the relationships between non-Aboriginal responses to questions in the TTP and SKES surveys, respectively. The direction of relationships in both tables is coherent. In Table 3, attitudes towards Aboriginal self-government and investments in Aboriginal education and on-reserve housing are positively associated with one another; in turn, these attitudes are negatively associated with perceptions of undue influence and unfair tax treatment. In Table 4, support for greater government effort for Aboriginal peoples, separate universities and colleges, a separate share of resource revenues and the perception that discrimination is an obstacle to Aboriginal economic success are all positively associated with one another; each of these is negatively associated with the view that Aboriginal people should “work their way up…without any special favours.” However, the results show no evidence of distinctive clusters of attitudes; all of the correlations are statistically significant and the differences between them are not especially dramatic.

The implication is that Saskatchewan residents do not distinguish “social welfare” Aboriginal policies from “special treatment” Aboriginal policies. All policies form a single Aboriginal policy agenda in the minds of the non-Aboriginal majority in Saskatchewan. Those who support targeted efforts to improve the social welfare of Aboriginal peoples are generally more inclined to hold favourable views toward special arrangements for Aboriginal peoples. Broadly speaking, what we see is acceptance or rejection of any arrangements, special or otherwise, aimed at Aboriginal peoples. Factor analyses of non-Aboriginal responses to these questions, presented in Table 5, reinforce the conclusion that there is a single dimension to Aboriginal policy attitudes. In both the TTP and SKES data, all five Aboriginal policy items load highly on the first factor. Subsequent factors do not generate enough additional information to warrant consideration (that is, their eigenvalues are well below zero).

Aboriginal policy attitudes are closely linked in both surveys, then, and both sets of items form reliable indexes (Cronbach's alpha equals .72 and .79 in the TTP and SKES surveys, respectively). For these reasons, in the analyses that follow we employ these questions in additive indexes that measure support for Aboriginal policy initiatives. Both of these measures range from zero (low support) to one (high support). The mean scores on the TTP and SKES indexes are .50 and .35 respectively.

The findings presented thus far raise the question of how much of the negativity towards policies that favour Aboriginals is rooted in prejudice and how much in an antipathy toward government interventions of any kind. All of these policy questions imply some level of government intervention to support special claims or particular needs. Regardless of the merits of these claims, or the reality of, for example, Aboriginal post-secondary institutions, there will be some respondents who are unwilling to accept that governments have agreed to devote resources to recognizing these claims. Can we isolate a group of “principled” objectors, those whose responses are rooted in something other than prejudice? If so, who are they and how important are they as obstacles to Aboriginal public policy? These are the questions to which we now turn.

5. Sources of Support for (and Opposition to) Aboriginal Policies

Our goal in this section is to examine the impact on Aboriginal policy attitudes of two variables that were introduced in section 2: prejudice, and beliefs about the role of government. For ease of interpretation, all variables in this analysis are re-scaled to range from zero to one. Although our measures of these variables—as well as those of policy attitudes—differ in the two surveys we analyze, the expectation is that the underlying relationships will be similar. We expect prejudice to be a source of opposition to policies that address the conditions of Aboriginal people simply because individuals who express negative sentiments toward outgroups are unwilling to consider policies aimed at helping them or recognizing their rights. We expect those who believe government should not be playing a significant role in alleviating social conditions in general to be negatively disposed toward special treatment of any group, including Aboriginals. But not everyone who believes in small (or large) government can be counted on to use this ideological principle to inform other topics. Beliefs about abstract ideological principles are likely to be salient determinants of policy attitudes only among those who are capable of recognizing the link between their own principles and particular policies.

In order to isolate the effects of ideological principles and prejudice and to explore the conditional effects of education and interest, we employ OLS regression and generate two models for each dataset, with analysis limited to the non-Aboriginal respondents. The results appear in Table 6. Models 1 and 3 show the effects of prejudice and the unconditioned effects of belief in limited government in the TTP and SKES data, respectively, controlling for levels of political sophistication. As expected, high scores on each of these variables results in reduced support for Aboriginal policies. Scores on the zero to one Aboriginal policy support index decline by nearly one-third (b = −.30) and one-fifth (b = −.18) in the TTP and the SKES surveys respectively, as prejudice increases from its lowest to its highest level, holding all other variables in the model constant. By the same token, in both models policy support also declines by about one-fifth of the scale as beliefs about the role of government shift from strong support for government intervention to strong support for limited government. In the TTP survey data, more education is associated with greater support for Aboriginal policies. Neither political interest nor education level has any discernible independent effects in the SKES data.

What about the moderating effects of political sophistication? These are presented in models 2 and 4 of Table 6. We assess the conditional effects of political sophistication on the relationship between beliefs about the role of government and policy attitudes with two multiplicative interaction terms: “limited government × education” and “limited government × political interest.” It turns out that all interaction terms are negative and statistically significant, suggesting that a strong belief in limited government powerfully reduces support for dedicated Aboriginal policies among those who are most capable of recognizing the connection between their ideological predispositions and these particular Aboriginal policies.

These effects are easier to discern in figures 1, 2 and 3, which show the effects of beliefs in limited government at each level of political education and interest. Figures 1 and 2 present the marginal effects of beliefs in limited government on support for Aboriginal policies along with their associated confidence intervals, at each level of education. These effects are calculated from the estimates in models 2 and 4 of Table 6, holding the effects of prejudice and political interest constant. The solid downward-sloping lines illustrate how the negative effect of beliefs in limited government on support for Aboriginal policies increases with educational attainment. In the TTP data (figure 1), greater support for limited government has virtually no effect when education is at its lowest level, but dramatically reduces support for Aboriginal policies when education is at its highest level. The SKES data (figure 2) also indicate greater support for limited government has a statistically insignificant effect at the lowest levels of education but a strong and statistically significant negative effect on support for Aboriginal policies at higher levels of education. Figure 3 presents the marginal effects of beliefs in limited government at each level of political interest in the SKES data, with the effects of prejudice and education held constant. It, too, shows that beliefs in limited government progressively reduce support for Aboriginal policies as levels of political interest increase. Using either education or interest as a measure, it is clear that the influence of ideology on Aboriginal policy attitudes depends on what we are calling political sophistication.

Note: Calculated from regression results presented in Table 6; broken lines represent the upper and lower bounds of the 95% confidence interval. Source: Taking the Pulse Survey, 2012.

Figure 1. Marginal Effect of Support for Limited Government on Support for Aboriginal Policies, by Education (TTP Survey)

Note: Calculated from regression results presented in Table 6; broken lines represent the upper and lower bounds of the 95% confidence interval. Source: Saskatchewan Election Study, 2011.

Figure 2. Marginal Effect of Support for Limited Government on Support for Aboriginal Policies, by Education (SKES)

Note: Calculated from regression results presented in Table 6; broken lines represent the upper and lower bounds of the 95% confidence interval. Source: Saskatchewan Election Study, 2011.

Figure 3. Marginal Effect of Support for Limited Government on Support for Aboriginal Policies, by Political Interest (SKES)

We have shown in this section that while questions about Aboriginal policy draw different responses depending on how those policies are framed, these responses share common sources. Indeed, the results were remarkably consistent when we performed multivariate analyses on each item in both policy attitude indexes, using the same sets of independent variables (results not shown). In every case, both prejudice and support for limited government generated statistically significant negative responses to policies. Nine of the fifteen interactions between beliefs about the role of government and political sophistication generated statistically significant results, and all of these were in the expected direction (that is, the effects of beliefs about the role of government are relatively stronger among those who are more educated and politically interested). The evidence clearly suggests that resistance to Aboriginal policy initiatives is governed not only by negative outgroup sentiments but also by ideological principles. Whether non-Aboriginals in Saskatchewan support or oppose policies targeted at helping Aboriginal people is driven in part by their level of prejudice, but general views about the role of government are also salient for those equipped with the cognitive resources to link ideological principles to policies. Not surprisingly, people who generally do not believe government intervention is desirable oppose policies that involve government effort.

The robustness of these findings is noteworthy. Although both surveys draw samples from the same population and both allow for measurement of the same core concepts, they were conducted five months apart—with one taking place in the context of a provincial election—and they employ different indicators of those core concepts. The results using both surveys are remarkably similar despite differences in timing and context, the particularities of survey indicators, and the chance effects of sampling.