Foreign policy beliefs and support for Stephen Harper and the Conservative Party

Abstract

Similar to other recent Canadian elections, foreign policy did not feature prominently in the 2011 federal election campaign. In fact, many doubt Canadian public opinion on international affairs is linked to the actions taken by recent Governments. In this paper, we examine Canadian public opinion toward a range of foreign policy issues and argue that the survey questions measure two latent dimensions —militarism and internationalism. Our survey evidence indicates the existence of an “issue public” which is prepared to endorse military action and is skeptical of human rights and overseas aid programs, and this group is far more supportive of Prime Minister Harper and the Conservative Party than other Canadians. The absence of an elite discussion, either among politicians or between media elites, about the direction of Canadian foreign policy does not prevent the Canadian voter from thinking coherently about questions pertaining to this issue domain and employing these beliefs to support or oppose political parties and their leaders.

De même que dans d'autres campagnes électorales canadiennes, la politique étrangère n'a pas émergé comme thème prédominant dans la campagne électorale du scrutin fédéral de 2011. En fait, nombreux sont ceux qui doutent que l'opinion publique canadienne soit influencée par les actions des gouvernements récents en ce qui concerne les affaires internationales. Dans cet article, nous examinons l'opinion publique canadienne vis-à-vis d'une série de thèmes de politique étrangère et soutenons que les questions de notre enquête permettent de mesurer deux dimensions latentes – le militarisme et l'internationalisme. Les données de notre enquête révèlent l'existence d'un groupe d'intérêt (issue public) qui est préparé à approuver toute action militaire et est sceptique quant aux programmes de défense des droits humains et d'aide internationale, et que ce groupe est bien plus favorable au Premier Ministre Harper et aux conservateurs que le sont les autres Canadiens. L'absence de débats parmi les politiciens ou les élites médiatiques n'empêche pas à l’électeur canadien de réfléchir de manière cohérente à des questions relatives au domaine discuté ici, et de se baser sur ces croyances pour soutenir les partis politiques et leurs leaders, ou pour s'y opposer.

Keywords:

• Canadian public opinion
• foreign policy attitudes
• survey research

Notes

1. See Wittkopf (1986, 1990, 1994) for a discussion of how these dimensions form in the minds of the American public. For public opinion studies employing the dimensions outside of the United States, see, e.g., Reifler et al. (2011) for the United Kingdom and Desposato et al. (2013) for Brazil and China.

2. For example, Bow (2008) notes that Joe Clark's Progressive Conservative Government was hurt by a controversy over the location of Canada's embassy in Israel, and Prime Minister Paul Martin appeared indecisive on the issue of missile defence in the 2006 campaign.

3. The 2004 federal election featured a debate over the appropriateness of a military build-up and the discussion over Harper's advocacy of sending Canadian troops to Iraq. The 2006 campaign featured a minor debate over the status of a “territorial army” (see Shadwick 2011).

4. The survey respondents' declarations of what issue(s) are most important are often taken as a proxy for salience. But there are critiques of this approach – see Johns (2010).

5. See Martin and Fortmann (1995), Munton (2002–2003) and Rioux and Hay (1998–1999). An important exception is Munton and Keating (2001), who performed an exploratory factor analysis of responses to a series of survey questions and found multiple dimensions of Canadian internationalism. This paper is, perhaps, a testament to the rarity of comprehensive batteries of foreign policy questions on Canadian surveys as their analysis had to rely on a survey fielded in 1985.

6. The survey was conducted by YouGov, Plc. YouGov's matching algorithm is designed to create a web-based sample whose characteristics approximate a probability sample, thus permitting model-based inferences (Rivers 2006). An initial sample of 3511 Canadian adults with proportionate representation from all provinces and territories completed an online screening survey. These respondents were matched on demographic factors (gender, age, education and region) to produce the final pre-election sample of 2800 respondents. The survey data were then weighted to the known population marginals for the Canadian adult population. Non-probability samples have been debated in the survey methodology literature. Some research focusing on the American context found differences in the estimates obtained from probability and non-probability survey samples (e.g., Malhotra and Krosnick 2006). However, other research concludes that there are very few differences between inferences made from high-quality online survey data gathered via “active” sampling methods such as those employed in our survey and those gathered by other survey modes (i.e., in-person, telephone or mail-back questionnaires) (see Ansolabehere and Schaffner, 2014). Similarly, analyses in Canada (Stephenson and Crête 2010) and Great Britain (Sanders et al. 2007, Clarke et al. 2008) have found that differences between probability and online data are very small, and substantive conclusions about relationships between variables of theoretical interest are identical across modes. For a recent review, see Baker et al. (2013). Support for CCPES was provided via a grant to Jason Reifler from the National Science Foundation (SES #1003254), and to Thomas J. Scotto from the Canadian High Commission, London and by grant RES-061-25-0405 from the Economic and Social Research Council of the United Kingdom.

7. The idea is that the decision of France to commit to action abroad plays a distinct role in Canadian foreign policy because such actions signal to Quebeckers that Canada's involvement is justifiable.

8. Support for intervention in Quebec, where support for the Conservatives was weak in 2011, is historically low, a stance some scholars have viewed as a likely reaction to a military and political establishment that is largely English speaking and has a historically British lineage (e.g., Granatstein 2004).

9. The bulk of troops were stationed in Afghanistan. In many locations, Canada's commitment involved less than one dozen troops (e.g., the lone troop the nation has as part of the United Nations peacekeeping force in Cyprus). However, the number of missions the Canadian military is partner to might surprise many citizens and, given the results of this study, would likely engender opposition.

10. Possible responses to both questions included: strongly agree, agree, neither agree nor disagree, disagree, strongly disagree and don't know. For ease of presentation in the cross-tabulations, the two agree and two disagree categories were combined. Responses in the “don't know” and “neither agree nor disagree” categories also were pooled. The conclusions drawn from the cross-tabulations when the full six-point scales were analyzed are not substantively different from those drawn when the collapsed scale is analyzed. The multivariate analyses below use a full five-point scale, with “don't know” responses treated as missing at random and multiply imputed. The substantive interpretations that stem from the results are not sensitive to coding decisions.

11. Combat missions involving Canadian troops formally ended in July 2011 (see Government of Canada 2013).

12. In contrast to the division at the public level, elite support for the Libyan mission was quite high with even the Bloc Québécois supporting the government, perhaps noting the leadership France was taking in the operation. Before the dissolution of Parliament, Prime Minister Harper did not seek approval from Parliament. After the election, approval was granted via a 294–1 vote, with the lone Green Party MP, Elizabeth May, dissenting. NDP opposition grew over the summer (Nossal 2013). The difference in the distribution of responses provided by Conservative identifiers and other respondents was significant χ² = 98.31 (df = 2; p < 0.001).

13. The human security agenda was a particular focus of Lloyd Axworthy during his tenure as the Minister for Foreign Affairs for Jean Chrétien in the late 1990s (see Hillmer and Chapnick 2001, Bloomfield and Nossal 2007).

14. In contrast to previous surveys where the peacekeeping question was “Canada should participate in peacekeeping operations abroad even if it means putting the lives of Canadian soldiers at risk”, we designed the 2011 CCPES question to deemphasize the military aspect of peacekeeping operations. Still, an exploratory factor analysis (EFA) suggested that Canadian attitudes towards peacekeeping are manifestations of both beliefs about the use of the military and internationalism. As discussed above, the Canadian mission in Afghanistan evolved to the point where heavy combat operations by the armed forces were the norm, and previous Canadian peacekeeping missions have not been without casualties. Canadians appear to understand that peacekeeping operations require heavily armed soldiers and are not naïve enough to think that such missions are without danger to the lives of soldiers and others on the ground. As shown in Table 2, levels of approval, however, vary considerably.

15. To preserve cases that would otherwise be lost due to item non-response (providing a “don't know” response or refusal to an individual survey item) or survey non-response (i.e., completing the campaign-period survey but not the post-election survey), we impute missing data (Rubin 1987, Allison 2001, Little and Rubin 2002). Multiple imputation was carried out using IVEware version 0.1 for SAS (Raghunathan et al. 2002). Ten multiply imputed datasets were created, the inferences from which were combined using SAS PROC MIANALYZE to produce inferences that account for sampling variation and the uncertainty associated with the imputation process.

16. This composite scale was highly reliable (coefficient α = 0.96).

17. All items were scaled between 0 and 1 before being averaged to produce the composite scores.

18. As noted above, since EFA demonstrated that the peacekeeping item contributed to both the militarism and internationalism factors (i.e., it cross-loaded), it is included in both composite scales. Further, the EFA indicated that attitudes toward Canada-United States relations (closer or looser ties) and agreement with the statement that “Canada should only use its armed forces abroad when it gets approval from the United Nations” did not load on either the militarism or internationalism factor. We had no theoretical expectations regarding the effects of these two variables on attitudes toward the Harper government, and as neither had statistically significant effects, they were excluded from the analyses.

19. Additional analyses of the interaction effect between Conservative party identification and militarism indicate that there is no significant difference between Conservative identifiers and non-identifiers when militarism is above 0.6 (roughly the 75^th percentile). In this regard, one should be mindful of potential “ceiling effects” among Conservative identifiers, whose ratings of the Harper government tend to cluster near the top of the 0 to 10 scale (mean = 7.49, median = 7.71).