Continentalism in the Canadian Context

For much of Canada's history, continentalism has been portrayed as a largely passive ever-present force stemming from the country's geopolitical and economic positionality. Echoing a realist understanding of the notion (see Inwood, Reference Inwood1997), many argue that continentalism has reflected the rational self-interest of Canada given its geographic, economic and political proximity with the United States, as well as the latter's international primacy (Granatstein, Reference Granatstein2007). In these accounts, continentalism is understood as an acknowledgement of positive outcomes associated with linkages between Canadian and American societies, cultures and economies (Clarkson, Reference Clarkson2006) rather than being a fully-fledged, coherent and purposeful policy strategy or analytical framework (Massie and Roussel, Reference Massie, Roussel, Smith and Turenne-Sjolander2013).

Many historical studies focus on the economic and security realms of continentalism. On the former, authors like Duquette (Reference Duquette, Randall and Konrad1995) and Martin (Reference Martin1995b) have focused on continentalism in trade, treating it as the motivating force behind policies designed to maximize economic benefits for Canada by broadening available markets. Economic continentalism can be traced back to the nineteenth century, when support for a commercial union with the United States emerged in Canada, largely as a result of practical and self-interested considerations (Pennington, Reference Pennington2007). Support for closer economic ties and integration further grew in the twentieth century concurrently with the United States’ increasing economic hegemony. Several studies of economic continentalism have emphasized this period, highlighting the establishment of notable treaties such as the US-Canada Free Trade Agreement in 1988 and the North American Free Trade Agreement in 1994 (Clarkson, Reference Clarkson2001; Pennington, Reference Pennington2007).

In the defence realm, continentalism has been treated as a practical necessity given Canada's geopolitical positioning (Bland and Young, Reference Bland and Young1988; Roussel, Reference Roussel2004). It has frequently been studied as a series of defence policies with an emphasis on the rational benefits stemming from cooperation and integration with the United States (Paquin, Reference Paquin2009; Roussel, Reference Roussel2004). Canada's integration into US defence systems, as institutionalized by the North American Air Defense Command (NORAD) of 1958 and the Defense Production Sharing Arrangement of 1959, has constituted a significant portion of discussions on security continentalism (Clarkson, Reference Clarkson2006; Roussel, Reference Roussel2004; Bercuson, Reference Bercuson, Carment, Hampson and Hillmer2003). It is interesting to note that Canadian leaders actively refrained from framing continentalist defence policies as being representative of a broader foreign policy strategy or framework. Instead, such actions were showcased as being consistent with a broader internationalist foreign policy, further restraining the emergence of continentalism as a fully-fledged framework, whether in an analytical or policy sense (Roussel, Reference Roussel2004).

From the above, we can conclude what continentalism, as historically conceptualized, is not. As argued by Massie and Roussel (Reference Massie, Roussel, Smith and Turenne-Sjolander2013), it has, in the past, not constituted a coherent foreign policy strategy in contrast to established dominant ideas and frameworks such as liberal internationalism (Smith and Turenne-Sjolander, Reference Smith, Turenne-Sjolander, Smith and Turenne-Sjolander2013). It has also historically not constituted an active political project or been part of a partisan ideology (Granatstein and Hillmer, Reference Granatstein and Hillmer1991). This is consistent with studies having shown that few partisan differences in overall diplomatic standing toward the United States existed during the late twentieth century (Vucetic and Ramadanovic, Reference Vucetic and Ramadanovic2020). Relatedly, Bow and Black (Reference Bow and Black2008) have shown that Canada's guiding strategies in navigating its relations with the United States have remained relatively stable between alternating partisan governments.

However, recent studies have shown a shift in expressions of Canadian continentalism over the past decade, challenging the previously prevailing realist conception of the concept. As a result, an increasing number of scholars view continentalism as a coherent analytical framework and a potentially leading paradigm in Canadian foreign policy (Smith and Turenne-Sjolander, Reference Smith, Turenne-Sjolander, Smith and Turenne-Sjolander2013; Massie and Roussel, Reference Massie, Roussel, Smith and Turenne-Sjolander2013). In particular, Massie and Roussel introduced the notion of “neocontinentalism” as an emerging dominant idea in 2013. In contrast to the notion of continentalism traditionally embraced by most scholars, neocontinentalism is seen as a political project that actively promotes the idea that Canada can gain influence and secure its interests on the world stage by nurturing its already close relationship with the United States (see Hart, Reference Hart2008). In addition, it has been linked with partisan and ideological orientations, with Massie and Roussel (Reference Massie, Roussel, Smith and Turenne-Sjolander2013) showing that it was partially motivated by conservative values and ideals under the Harper government. These studies suggest an evolved form of continentalism, increasingly seen as a coherent analytical framework and political project. Notably, it is increasingly recognized as a dominant idea in Canadian foreign policy, gaining prevalence alongside long-established ones such as liberal internationalism (Roussel and Robichaud, Reference Roussel and Robichaud2004; Dorion-Soulié and Roussel, Reference Dorion-Soulié and Roussel2014).

This shift has prompted new directions among researchers, with an increasing number of studies testing whether official policy has indeed become more continentalist through analyses of foreign policy and official rhetoric (Dorion-Soulié and Roussel, Reference Dorion-Soulié and Roussel2014; Paquin and Beauregard, Reference Paquin and Beauregard2013; Vucetic and Ramadanovic, Reference Vucetic and Ramadanovic2020). Moreover, in line with the move away from a realist and interest-driven view of continentalism, authors have increasingly incorporated ideational elements into their studies. As demonstrated by recent literature, if we focus on the ideational realm, continentalism can be associated with a particular set of political ideologies (Massie and Roussel, Reference Massie, Roussel, Smith and Turenne-Sjolander2013), specific beliefs or worldviews (McLaughlin, Reference McLaughlin2017), a strategic culture (Massie, Reference Massie2009), and issue-specific values and judgments (Boucher, Reference Boucher, Juneau, Lagassé and Vucetic2020). In a broader sense, ideational continentalism in these different realms can be summarized as stemming from perceptions of shared values, norms, and standards of behaviour with the United States. This rising acceptance of ideational and normative elements of continentalism closely mirrors the constructivist turn in IR, which has emphasized the inability of material factors and exogenously assumed interests to fully explain international outcomes (Wendt, Reference Wendt1992).

Despite the renewed interest, few studies have assessed the extent to which this evolved form of continentalism has gained widespread acceptance in Canadian society beyond narrow government elites. In contrast, such analyses have been undertaken regarding competing frameworks such as liberal internationalism (see Munton and Keating, Reference Munton and Keating2001; Paris, Reference Paris2014). As shown above, much of the recent work on continentalism focuses on concrete policy actions or the official level (Vucetic and Ramdanovic, Reference Vucetic and Ramadanovic2020; Dorion-Soulié and Roussel, Reference Dorion-Soulié and Roussel2014; Paquin and Beauregard, Reference Paquin and Beauregard2013). Several studies have focused on the ideational perspective and studied continentalism as a Canadian strategic culture in recent years (Massie Reference Massie2009; Nossal et al., Reference Nossal, Roussel and Paquin2011). However, if adopting Snyder's (Reference Snyder1977) definition of strategic culture, such analyses remain centred on political elites and decision-makers. Whether a novel form of continentalism has gained widespread acceptance among broader Canadian actors including the general public remains an understudied question.

Given the shift away from the realist understanding of continentalism and recent studies that have shown its development into a broader and more coherent political project and an emerging dominant idea since the Harper government (Massie and Roussel, Reference Massie, Roussel, Smith and Turenne-Sjolander2013), this can be seen as a significant gap in understanding Canada-US relations and Canadian foreign policy more broadly. Indeed, following Nossal et al.'s (Reference Nossal, Roussel and Paquin2011: 135) definition of dominant ideas as being shared by “a majority of members” within society, understanding whether a new form of continentalism has truly emerged as a dominant idea in Canada requires investigating the extent to which it has gained acceptance among a broader array of domestic actors. More specifically, given that many studies have empirically demonstrated that national publics hold predictable, principled and stable preferences on matters of international relations (Herrmann et al., Reference Herrmann, Isernia and Segatti2009; Rathbun, Reference Rathbun2007; Kertzer and Zeitzoff, Reference Kertzer and Zeitzoff2017; Boucher, Reference Boucher, Juneau, Lagassé and Vucetic2020) and that public opinion holds the potential to significantly constrain Canada's actions on the world stage (Boucher and Nossal, Reference Boucher, Richard Nossal, Hampson and Saideman2015), understanding whether continentalism has become broadly accepted among the general public is key to determining whether it has indeed become a dominant idea and a fully-fledged analytical framework suitable to the analysis of Canadian foreign policy.

We argue that ideational elements of continentalism, which have increasingly been incorporated into existing studies on the question (McLaughlin, Reference McLaughlin2017; McKay, Reference McKay2018), constitute an especially interesting avenue of research since they represent guiding forces in shaping preferences (Nossal et al., Reference Nossal, Roussel and Paquin2011) and can be differentiated from simple pragmatic and self-interested preferences for a closer relationship with the United States. Indeed, preferences for closer economic and security ties could simply be attributed to perceived interests associated with the latter's international power, thereby reflecting the traditional realist conception of continentalism. In contrast, our conceptualization of ideational continentalism sees it as stemming from Canadians’ perception of shared values, norms and standards of behaviour with the United States. Such views, whether expressed in cultural, ideological, strategic or identarian realms, are likely more resilient and less contextual in nature (see Parsons, Reference Parsons1951; Nossal et al., Reference Nossal, Roussel and Paquin2011). Especially if they become embedded in individuals’ national identities through the processes outlined in constructivist literature, such elements can become embedded in citizens’ national identities and represent strong guiding principles for their foreign policy attitudes and preferences. This link will be further explained in the following section.

Transnational Norms, Identity, and Foreign Policy

Based on the above, continentalism has increasingly been recognized as a coherent paradigm and dominant idea in Canadian foreign policy in recent years. As part of such a trend, studies have tied the notion to ideational elements such as ideologies, values and culture (Dumas, Reference Dumas2015). Under the Harper government, scholars have argued that ideational elements of neocontinentalism were tied to perceptions that cooperation and alignment with the United States were in line with the conservative values espoused by the governing majority (Massie and Roussel, Reference Massie, Roussel, Smith and Turenne-Sjolander2013). However, the newly emergent expressions of continentalism seem to have survived the end of the Harper government and continued to be present under Trudeau (Vucetic and Ramadanovic, Reference Vucetic and Ramadanovic2020). Moreover, no empirical research has conclusively established a link between present-day continentalism and conservative ideologies in a stable and context-independent manner. Therefore, we adopt the wider view that continentalism in the ideational realm is manifested by perceptions of shared norms, values and identities by respective Canadian and American actors. Interestingly, available literature in International Relations has extensively demonstrated that such ideational elements and perceptions often become embedded in the identities of national actors, after which they play a structuring role in shaping foreign policy attitudes and preferences.

The study of national identity has greatly expanded in IR since the early 1990s when the rise of social constructivism took place. If accepting Anderson's (Reference Anderson1983) conception of national identities as “imagined communities,” it becomes clear that such a notion is inherently malleable and affected by social processes of interaction, including beyond national boundaries. This forms the basis for Wendt's (Reference Wendt1992) claim that national identity is constantly reshaped by practice and interactions with other states. Several authors have pointed out that membership in treaties or international organizations plays a role in the international diffusion of norms, values and standards of practice (see Finnemore and Sikkink, Reference Finnemore and Sikkink1998). Through this mechanism, states sharing membership in the same international organizations can incorporate shared normative notions in their respective identities. Risse-Kappen (Reference Risse-Kappen1997) and Adler and Barnett (Reference Adler and Barnett1998) specifically showcase how membership in security alliances such as NATO and the OSCE has led to the rise of shared identities, values, and practices among their members. The above studies are particularly relevant to the study of continentalism in the Canadian context. Indeed, authors have shown that dominant ideas such as internationalism, through its promotion of Canada's membership in numerous international bodies and organizations, have become deeply embedded in Canadians’ sense of national self (Turenne-Sjolander, Reference Turenne-Sjolander, Smith and Turenne-Sjolander2013). More specifically on the question of continentalism, the significant extent of institutionalized defence integration and cooperation between the United States and Canada through organizations such as NORAD and NATO make the above mechanism highly relevant.

The same underlying process can be discerned in the economic realm through an examination of classical liberal literature in IR. Indeed, following the findings of authors such as Russett and Oneal (Reference Russett and Oneal2001), international commercial ties lead to additional interactions between the domestic actors of different countries and can thereby contribute to the development of shared transnational norms, values and identities. In the North American context, Rankin (Reference Rankin2004) found that debates surrounding economic agreements such as NAFTA engender collective values concerning economic integration among the Canadian and American public. On a broader level, authors such as Bow and Santa-Cruz (Reference Bow, Santa-Cruz and J2012) have argued that diplomatic interactions and the process of negotiating regional economic agreements involving Canada, the United States, and Mexico have created a closer and more aligned identity between Canada and its southern neighbour. Numerous studies focused on economic cooperation between the two have linked these interactions to common norms and ideas (see Macdonald, Reference Macdonald2020).

Through the above, a view emerges in which both the economic and security ties and integration that have been attained between the United States and Canada can be seen as leading to the development of identarian and ideational shared traits (or perceptions thereof) among actors from both countries; a phenomenon that can lead to the existence of ideational continentalism. It is clear from existing studies that its emergence relies at least partially on preexisting ties in material realms. This further justifies our primary focus on ideational continentalism, as it can be considered to be a more deeply embedded form of such a phenomenon which can be differentiated from preferences or policies of alignment that are more transient and contextual in nature yet rely on the existence of continentalist orientations in more material realms.

While this article focuses on an analysis of public attitudes and does not seek to investigate policy outcomes, it is important to note that numerous scholars have pointed out the constraining effects of national identity on countries’ foreign policy orientations. Indeed, we treat continentalism in its ideational realm as an analytical framework helpful for the analysis of the guiding forces behind Canadians’ foreign policy attitudes and preferences rather than a policy framework expressed through concrete government actions. However, continentalism in the ideational realm can still have important implications on policy. Authors such as Herrmann et al. (Reference Herrmann, Isernia and Segatti2009) argue that national identity has become especially important since the rise of mass politics on questions of foreign policy and demonstrate how it is associated with changes in support levels for international cooperation or military intervention. In the Canadian context, many have investigated different aspects of identity as driving forces shaping Canadians’ foreign policy views (see Therien and Mace, Reference Therien and Mace2013; Turenne-Sjolander and Cornut, Reference Turenne-Sjolander and Cornut2016). Others have also shown that identarian attachment can heavily shape Canadians’ perceptions and understandings of foreign policy (Boucher and Nossal, Reference Boucher, Richard Nossal, Hampson and Saideman2015; McDonough, Reference McDonough2013). When also considering studies having shown that the domestic public holds predictable, principled and stable preferences on matters of foreign policy (Herrmann et al., Reference Herrmann, Isernia and Segatti2009; Rathbun, Reference Rathbun2007; Kertzer and Zeitzoff, Reference Kertzer and Zeitzoff2017; Boucher, Reference Boucher, Juneau, Lagassé and Vucetic2020) and that public opinion holds the potential to significantly constrain states’ foreign policies (Gries and Turcsanyi, Reference Gries and Turcsanyi2022; Boucher and Nossal, Reference Boucher, Richard Nossal, Hampson and Saideman2015), it becomes apparent that the identities of Canadians play a part in shaping and constraining the country's foreign policy. This makes our findings particularly relevant to further the understanding of how ideational continentalism can shape Canada's ties with the United States and foreign policy more broadly.

Hypotheses

To examine whether ideational continentalism has become a force shaping Canadians’ foreign policy attitudes and preferences in the current-day context, we formulate two hypotheses. The focus on ideational continentalism is justified, given that scholars of Canadian foreign policy increasingly see continentalism as having evolved into a coherent political project and dominant idea partially comprised of ideational elements. In addition, constructivist literature in IR elucidates that mechanisms supporting the development of ideational continentalism ground themselves on the existence of material ties in economic and defence realms without, however, being guaranteed by them. As a result, this analytical choice corresponds to the toughest possible test to assess whether a new form of continentalism has emerged as a meaningful analytical framework for explaining Canadian foreign policy.

In addition, we seek to differentiate continentalist attitudes deeply embedded in the views of domestic Canadian actors from more transient and pragmatic desires to align with the United States to further certain contextual interests. While findings limited to the latter would not necessarily negate the existence of continentalism in Canada, they would lend greater support to the traditional realist conception of continentalism and would not indicate its emergence as a coherent analytical framework, a purposeful political project or a dominant idea. Importantly, perceived ideological affinities and shared values with the United States become embedded in citizens’ national identity through a long socialization and localization process (see Finnemore and Sikkink, Reference Finnemore and Sikkink1998) and thus are stable and slowly shifting. This distinction helps to differentiate between interest-based transient preferences for alignment and ones associated with a deeper set of continentalist attitudes and beliefs more easily.

Hypothesis 1 hence aims to examine the direct link between the strength of Canadians’ national identity and their foreign policy preferences with the United States. This first step is necessary for establishing whether there is indeed any significant linkage between Canadians’ self-proclaimed national identity and their foreign policy attitudes. Indeed, national identity is an inherently complex notion, with the possibility of it not holding unidirectional effects on alignment preferences with the United States representing a real possibility. The demonstrated heterogeneity in national identities among francophone and anglophone Canadians, and between ones residing in different provincial settings, further raises such a possibility (Breton, Reference Breton2015; Nossal et al., Reference Nossal, Roussel and Paquin2011). Despite this, recent developments and evidence lead us to expect a positive tie between the two. Indeed, authors have pointed out that increasing continentalist integration between the United States and Canada in trade, commerce and defence has countered historical expressions of Canadian nationalism which would have led us to expect an effect in the opposite direction (Clarkson, Reference Clarkson2006). Studies have concurrently shown that Canadian identity has become more similar to that of Americans over time (Raney, Reference Raney2011). Empirically, survey evidence since the turn of the century has consistently shown that Canadians’ and Americans’ political and cultural attitudes are significantly closer to each other than with any other state's population (Dalton, Reference Dalton and J2019). Finally, the emergence of neocontinentalism under the Harper government has been argued to have represented the birth of a political project actively pushing for closer ties and alignment with the United States, one which had been largely absent throughout Canada's recent history (Massie and Roussel, Reference Massie, Roussel, Smith and Turenne-Sjolander2013). As a result, we hypothesize that a stronger sense of self-reported national identity is tied to a more aligned foreign policy preference with the United States among the Canadian public.

Hypothesis 2 aims to further test whether any linkage found in line with Hypothesis 1 can be explained through ideational continentalism embedded in national identity. The two hypotheses combined point to a state of affairs that would suggest that continentalism has, in recent years, emerged as a dominant idea and has led to the perception of shared norms, values and identities with the United States among Canadian domestic actors. Following this view, continentalism has embedded itself in Canadian national identity and plays a role in shaping Canadians’ foreign policy preferences and attitudes toward closer alignment with the United States.

Notably, the focus on alignment in our hypotheses is consistent with historical Canada-US relations. It is important to acknowledge that different levels of bilateral engagement can exist in foreign policy, ranging from simple ties to institutionalized policy integration. Notably, these levels differ in terms of the extent to which they constrain Canada's sovereign policy orientations. While ties involve little such sacrifice, actual institutional integration significantly does so. We take the view that, historically, Canada-US relations have involved ties in numerous policy realms, alignment in some, and integration in a few. Using alignment preferences as our dependent variable constitutes a tougher test for the significance of ideational elements of continentalism, as it presupposes an extensive degree of ties and involves some relinquishing of policy autonomy. In contrast, using support for ties only would involve little trade-off and would hence risk being seen in overwhelmingly favourable terms by survey respondents. Furthermore, using support for integration would risk misrepresenting the reality of Canada-US relations, as many expressions of cooperation between the two states do not necessarily require institutionalized integration (see Chatsko, Reference Chatsko2012).

It is important to note that failing to provide evidence in favour of Hypothesis 1 would not necessarily imply the lack of narrow interest-based continentalist views among Canadians, especially in some material realms. However, it would make the conclusive differentiation between contextual preferences for integration or alignment with the United States based on perceived interests, which have been present throughout much of Canada's history, and more lasting and embedded continentalist orientations consistent with the emergence of continentalism as a dominant idea and coherent political project difficult. On the other hand, only finding evidence for Hypothesis 1 (while failing to establish the validity of Hypothesis 2) would make it difficult to support the relevance of ideational continentalism and would similarly place doubts on whether continentalism has truly evolved away from its traditional realist conception. This is because Canadian national identity may lead to foreign policy alignment through other mechanisms than continentalist attitudes, which lie beyond the scope of this analysis.

Finding evidence to support both hypotheses would not imply that Canada's national identity has become fully continental or that elements of competing frameworks do not play concurrent roles in shaping Canadians’ foreign policy views. Many scholars have indeed suggested other frameworks to be significant in shaping Canada's foreign policy orientation. These include but are not limited to, liberal internationalism (Paris, Reference Paris2014), multilateralism (Brown and Olender, Reference Brown, Olender, Besada and Kindornay2013) and Atlanticism (Wolfe, Reference Wolfe1991). National identities represent highly complex notions that are shaped by a multitude of factors and are often considered to be in constant evolution. In light of this, combined evidence in favour of both hypotheses should simply be taken to indicate that continentalist orientations have, to an extent, embedded themselves in Canadians’ national identities and represent one factor among many that shape their foreign policy attitudes and preferences. In line with the many studies that have shown that the mechanisms through which ideational continentalism develops are dependent on pre-established economic and security linkages (see Risse-Kappen, Reference Risse-Kappen1997; Adler and Barnett, Reference Adler and Barnett1998; Russett and Oneal, Reference Russett and Oneal2001), evidence in favour of our two hypotheses would likely be indicative of existing continentalism in material realms such as the economy and defence in the Canadian context.

Data and Research Design

This article utilizes Canadian responses to the Sinophone Borderlands survey project conducted between August and October 2022 (Turcsanyi et al., Reference Turcsanyi, Kironska, Gerstl, Arnez, Cenek, Chen, Chubb, Gries and Wang2022). The survey, conducted in English and French, contains data from 1,512 respondents across Canada. To ensure the national representativeness of our sample, we applied survey weights provided by the Sinophone Borderlands project to the dataset.Footnote ¹ The survey is selected as it represents one of the latest and most comprehensive projects targeting the understanding of individual foreign policy preferences designed specifically under the context of the current US-China rivalry.

It is important to note that the timing of this survey may raise concerns that it captures a period of unusually friendly orientations toward the United States as a result of the ongoing Ukraine war and the recent end of the Trump presidency. However, we believe that, while short-term contextual factors like these continuously play a role in shaping the public's attitudes and preferences, responses to survey items indicating ideational continentalist orientations embedded in respondents’ sense of national identity represent longer lasting and more stable views (see Finnemore and Sikkink, Reference Finnemore and Sikkink1998). Moreover, our analysis examines variations between respondents within the same period. As a result, the potential effects of the Ukraine war and the American domestic political context should apply evenly across units and can hence be controlled for. Additionally, our model includes multiple political attitude controls to absorb potential confounders in the above effect. The specific timing of the survey thus cannot be considered to represent a major limitation for our analysis and its results.

Measuring Foreign Policy Alignment

As the dependent variable intending to capture respondents’ preferences on the extent to which Canada's foreign policy should be aligned with the United States, the corresponding survey question was used. This question asks “How closely should Canada's foreign policy align with the United States?” and records responses on an ascending scale, with zero corresponding to a response of “not at all” and ten corresponding to one of “completely.” This variable is chosen as it captures the overall preference for alignment with the United States on the foreign policy front without specifying any specific issue or referring to any geopolitical context. Recognizing that foreign policy alignment encompasses a wide range of areas from diplomatic stances and cultural exchanges to defence and economic policies, we intend to capture attitudes toward alignment vis-à-vis the United States on a wide spectrum, rather than on specific policies. The exclusion of more specific issues helps eliminate the potential bias introduced by issue framing and contextually related factors and renders the selected variable appropriate for detecting general attitudes toward foreign policy. However, we recognize that this generalization may also result in heterogeneity in respondents’ interpretation of alignment and thus include various socio-demographic and political controls in our analysis. The distribution of responses is shown in Figure 1. As illustrated, most Canadians prefer relatively high levels of alignment with the United States.

Figure 1. Distribution of Respondents’ Preferred Level of Foreign Policy Alignment with the United States

Measuring Canadian National Identity

The key independent variable of the study is the strength of respondents’ self-assessed national identity. To capture this, we construct an index based on several survey questions. The index takes the average of respondents’ level of agreement with the following statements: “I love my country,” “I am glad to be Canadian”; “I feel a bond with Canadian people”; and “The fact that I am Canadian is an important part of my identity.” While the last two statements intend to directly measure the individual perceived level of resonance with being Canadian, the first two are included to infer the feeling of Canadian exceptionalism within the broader expression of national identity. Indeed, as established in existing literature, feelings of exceptionalism constitute an inherent aspect of states’ national identities (Simonsen, Reference Simonsen2022). It is also particularly relevant in the Canadian context in which a strong desire for differentiation vis-à-vis the United States has been argued by some to be tied to such exceptionalism (Doran and Sewell, Reference Doran and Sewell1988).

We recognize that the constructed index represents a relatively reductionist and simplistic measure of identity. However, relying on it is justified given that this article's focus lies not in expanding knowledge regarding the complexities of Canadian national identity, but in offering a targeted analysis of its links with Canadians’ foreign alignment preferences with the United States. Furthermore, this choice is consistent with the many existing studies in the field relying on similar measures in their instrumentalization of identity (see Huddy and Khatib, Reference Huddy and Khatib2007; Pehrson et al., Reference Pehrson, Vignoles and Brown2009). To ensure the validity of the index, the internal consistency of the four statements used to construct it was tested statistically. The resulting Cronbach's Alpha score of over 0.90 indicates a high degree of internal consistency and thus validates the usage of the index. As shown in Figure 2, Canadians exhibit a strong sense of national identity, with responses of five or above (on a scale of seven) representing the most popular categories.

Figure 2. Distribution of Respondents’ Perceived Level of Canadian Identity

Measuring Ideational Continentalism

A pivotal variable in our study is the one measuring ideational continentalism embedded in respondents’ national identity. To capture it, an index is constructed using the weighted average based on the Principal Component Analysis (PCA) factor loading scores of four variables: perceived similarity of cultural values between the United States and Canada, the cultural attractiveness of the United States, favourability toward US political values, and favourability toward Americans.

To identify the appropriate variables for the index, we perform an exploratory PCA on eight variables. These include seven variables related to the ideational dimensions which reflect respondents’ opinions about American values, norms and attitudes, and one variable indicating the perceived importance of the United States to Canada's economy. The material variable serves as a conceptual comparative point, representing one of the main traditional material sources of continentalism having been identified in historical literature. The PCA reveals high factor loadings for four variables. Consequently, these four variables are selected to construct the ideational continentalism index. The economic importance variable is found to have only a moderate loading score (0.51). This implies that it contributes much less to the principal component intending to capture ideational continentalism relative to included variables. This finding aligns with our conceptual framework in suggesting that, while expressions of the newly emergent form of continentalism in the ideational realm may have material roots, they remain conceptually distinct from simple material self-interest. Details of the exploratory PCA can be found in Appendix 5.

The single-component and no-rotation PCA results stemming from the final four variables (standardized for index construction) are shown in Table 1. The component achieves an eigenvalue of 2.40, accounting for 60 per cent of the total variance. This lies above the common greater-than-one threshold for standardized data, indicating its significance. All four variables have factor loadings higher than the common thresholds for index construction (Pituch and Stevens, Reference Pituch and Stevens1992; Field, Reference Field2005). This demonstrates the variables’ strong and positive contribution to the final index of ideational continentalism.

Table 1. Ideational Continentalism Index ─ Factor Loadings and EigenvalueFootnote ²

The index captures ideational continentalism mainly from the perspective of cultural and political values. This renders it an appropriate measure of ideational continentalism which, as laid out in the previous sections, largely relies on the development of shared values, norms, and standards of behaviour between Canadian and American domestic actors or at least on the perception of such proximity by the public (Raney, Reference Raney2011). It furthermore relies on the accepted notion that respondents consequently view values closer to their own through more favourable lenses (Roccas et al., Reference Roccas, Sagiv, Oppenheim-Weller, Elster and Gal2013; Rokeach, Reference Rokeach1973), thereby justifying the inclusion of questions pertaining to both favourability and similarity.

Control Variables

A set of independent variables aimed at controlling for various characteristics of respondents that could hold an effect on both the strength of their national identity and their alignment preferences with the United States are included in the model. These pertain to respondents’ demographic characteristics, socio-economic backgrounds and political attitudes. Details of the question wordings and their descriptive statistics can be found in Appendix 1 and 2.

Regarding socio-demographic controls, we include age, gender, education and urban/rural residence status. Age is obtained directly through respondents’ self-provided answers. Gender is represented through a categorical variable with three levels: female, male and other. Education level also contains three categories: primary (elementary school), secondary (high school) and tertiary (university/college). For urban/rural residency, four categories are obtained, ranging from “village” to “big city.” As political attitude controls, we include respondents’ level of political interest, their self-identification on a left-right ideological scale and their degree of support for capitalism. Political interest, measured on a 10-point ascending scale, is included as it is recognized as an important factor in shaping political understanding and behaviour (Luskin, Reference Luskin1990). The political ideology variable measured on a left-right scale takes a value from 1 to 7, with 1 representing “far left” and 7 “far right.” Regarding this variable, scholars have highlighted the importance of considering political beliefs and ideology when studying policy attitudes (Jacoby, Reference Jacoby1991; Gromet et al., Reference Gromet, Kunreuther and Larrick2013). Particularly in the case of Canada-US relations, empirical evidence suggests that the large divide on political and social spectrums concerning issues such as welfare, gun control and abortion could affect Canadians’ willingness to align with the United States (Parkin, Reference Parkin2022). This is also why the level of support for capitalism is included, which is measured on an ascending continuous scale. The inclusion of the latter control also stems from the observation that the general impression of the United States is empirically associated with favourability toward capitalism (Abdelfatah and Arablouei, Reference Abdelfatah and Arablouei2021). Given that these ideological traits are likely to also be linked with Canadians’ sense of national identity, it is important to include them in the model.

Furthermore, we include variables indicating respondents’ region of residence. Regions are coded as respective binary variables with a value of 1 indicating respondents’ residency in the region in question. This is included to account for the potential that significant heterogeneity in Canadian national identity exists across regional clusters, leading to different provincial baselines. Indeed, many studies have established different degrees of national and provincial belonging based on individuals’ province of residence (Statistics Canada, 2015). In some settings such as Quebec, national identity has been shown to take on significantly different expressions than in other regions holding less pronounced regional identities (Raney, Reference Raney2009). Specifically, the “two solitudes” model has been argued to affect various aspects of Canadians’ perceptions and attitudes (Gagnon, Reference Gagnon2013). In terms of identities, many studies have found Quebecers to hold strong provincial identities that differ from the sense of Canadian national identity that is prevalent in other provinces (Brie and Mathieu, Reference Brie and Mathieu2021). This distinctiveness also manifests in perceptions of certain ideational notions such as liberalism (Turgeon et al., Reference Turgeon, Bilodeau, White and Henderson2019), individualism (Rousseau and Côté, Reference Rousseau and Côté2017) and support for economic welfare (Johnston et al., Reference Johnston, Banting, Kymlicka and Soroka2010). On the foreign policy front, Quebecers have also shown significant differences in their level of support for militarism (Massie and Boucher, Reference Massie and Boucher2014) and North American free trade (Martin, Reference Martin1995b) among other international issues. Given the existence of strong provincial heterogeneity in attitudes, especially in Quebec, it is important to include these provincial controls to obtain a complete view of the relationship of interest.

Specifying a Model of Foreign Policy Alignment

To test Hypothesis 1, the study employs an OLS regression. While ordinal regression strategies were, in the past, preferred for the estimation of coefficients derived from variables consisting of Likert scale survey responses, recent studies have questioned the benefit derived from such a choice and have instead privileged standard linear models whose results are easier to interpret and contextualize (Robitzsch, Reference Robitzsch2020). We therefore decide to privilege a linear regression strategy for the analysis. The regression specification is shown below in Equation 1.

Equation 1. Regression Model Specification

$$FP\;alignment_i = a + \beta _1 \times lvl\;of\;CanID_i + \mathop \sum \limits_j \delta _jX_{\,ji} + \mathop \sum \limits_g \theta _gGender_{gi} + \mathop \sum \limits_e \varphi _eEducation_{ei} + \mathop \sum \limits_u \sigma _uUrban/Rural_{ui} + \left({\mathop \sum \limits_k \gamma_kprovince_{ki}} \right) + \varepsilon _i$$

“FP alignment” is the preferred level of foreign policy alignment with the United States for individual i. The level of “CanID” is the indicated strength of person i's self-assessed strength of Canadian identity. X represents the set of continuous control variables including age, political interest, political ideology and support for capitalism. The next three terms are three categorical socio-demographic controls: gender, education and urban/rural. The last term represents the regional dummies for all Canadian regions, which we will include in certain iterations of the model. Several different model iterations were run as a test for robustness, with the above equation representing the iteration in which all mentioned control variables are included.

Devising a Mediation Analysis

To test the mediating role of ideational continentalism as outlined in Hypothesis 2, we employ a mediation analysis. This allows us to conclusively determine how much (if any) of the demonstrated link found between the strength of respondents’ national identities and their alignment preferences with the United States is attributable to elements of ideational continentalism embedded within their national identities.

The pathways linking our index of ideational continentalism, the strength of respondents’ national identity and their resulting foreign policy alignment preferences with the United States are shown in Figure 3. The mediation analysis aims at untangling any demonstrated effect between Canadian national identity and alignment preferences with the United States into distinct direct effects (c’) and effects mediated through ideational continentalism. The latter would require significant correlational pathways between both Canadian national identity and ideational continentalism (a) and between the latter and alignment preferences with the United States (b). The mediation analysis result will help to reveal whether any correlation between Canadian national identity and alignment preferences with the United States found in the first regression outlined above is, in fact, due to some embedded notions of ideational continentalism and not some other components of Canadians’ national psyche that lie beyond the interest of this article.

Figure 3. Mediation Analysis Diagram

Findings

We ran several model iterations with different combinations of controls relating to demographic characteristics, political attitudes, and region of residence. All model iterations show robust and consistent results. The full table of results can be found in Appendix 6. Figure 4 presents the output of the full model.Footnote ³ As shown, we find a significant and positive correlation between respondents’ sense of Canadian identity and preferences for a greater level of foreign policy alignment with the United States. In other words, Canadians with a strong sense of national identity tend to hold more favourable foreign policy preferences vis-à-vis their southern neighbour, providing support for Hypothesis 1.

Figure 4. Full Regression Model Coefficient Plot

All demographic controls are found to be insignificant while the political attitude controls, such as favourability toward capitalism and political interest, are found to be positively correlated with more pro-US foreign policy stances. Moreover, respondents expressing support for more right-wing political ideologies are found to favour closer alignment with the United States (p = 0.04). As many existing empirical studies have identified linkages between nationalistic sentiments and different political ideologies (Thompson, Reference Thompson2023), we further tested the potential moderating effect of political ideology on the relationship of interest and found no significant results (See Appendix 7). This suggests that a stronger level of Canadian national identity would hold similar effects on alignment preferences with the United States across the left-right political spectrum.

Our primary model includes regional dummy variables in recognition of the existing literature having established the importance of accounting for regional differences in studying public opinion and policy preferences in the Canadian context (Cochrane and Perrella, Reference Cochrane and Perrella2012). Figure 4 illustrates such differences using Ontario (n = 586) as the baseline. The results showcase little variation in the effects of interest at the regional level, with only Quebec (n = 340) showing significant differences. This finding is consistent with existing literature highlighting the distinctiveness of Quebecers in holding attitudes significantly different from other provinces’ populations (Fiorino, Reference Fiorino2015). This difference can be linked to Quebec's distinctive culture and provincial identity (Mendelsohn, Reference Mendelsohn2003; Gagnon, Reference Gagnon2013). On the foreign policy front, although Quebecers have historically demonstrated different foreign policy attitudes on defence and trade issues (see Massie and Boucher, Reference Massie and Boucher2014; Martin, Reference Martin1995a), recent studies have found that they do not diverge significantly from anglophone Canadians (Boucher Reference Boucher2021). Our findings suggest that, on the issue of foreign policy alignment with the US, Quebecers hold more favourable attitudes relative to other Canadians.

To understand whether the key independent variables included in the model operate differently in Quebec and the rest of Canada, we conducted further statistical tests. Firstly, Wald tests reveal no significant evidence that any of our explanatory variables operate differently in shaping alignment preferences between Quebec and Canada (see Appendix 4). Figure 5 presents models conducted using subsets of Quebec residents and non-Quebec residents. While the result suggests that the effect of our main variable of interest is insignificant when considering only residents of Quebec, the overlapping confidence intervals shown in Figure 5, alongside non-significant Wald test results, make it impossible to establish the existence of significant differences in effects. Hence, our results provide support for the existence of some stable and unified orientations in Canadian national identity across different provincial settings, in line with the recent findings of authors such as Bilodeau et al. (Reference Bilodeau, White, Turgeon and Henderson2020). However, we acknowledge that drawing straightforward conclusions on these questions on the basis of our findings is difficult, partially as a result of the relatively small sample size of Quebec respondents. Hence, further investigation of these effects represents a welcome avenue for future studies.

Figure 5. Differences in Effects between Quebec and Other Provinces

Turning to Hypothesis 2, Figure 6 presents the result of the mediation analysis using the constructed index of ideational continentalism as a mediator. The analysis was performed using the “mediation” package in R (Tingley et al., Reference Tingley, Yamamoto, Hirose, Keele, Imai, Trinh and Wong2014). This is necessary to establish whether the above-demonstrated link between Canadian national identity and preferred alignment with the United States is at least partially mediated by elements of embedded ideational continentalism. The mediation analysis was run multiple times using weighted survey data with different combinations of control variables. It is particularly important to address any individual-level differences as both of our key variables attempt to measure identity and related ideational elements (see Cislaghi and Heise, Reference Cislaghi and Heise2019). Both the naïve and the full models of the mediation analysis yield consistent results, with the detailed output shown in Appendix 8. Figure 6 presents the result of our full mediation analysis, including the full set of demographic and political attitude controls, as well as province-level heterogeneity.

Figure 6. Mediation Analysis Result (Full Model)

The results reveal a significant and positive linkage between levels of Canadian identity and ideational continentalism, suggesting that people with a stronger sense of Canadian identity are indeed more likely to have a higher level of embedded continentalist beliefs, providing sound justification for the mediator choice. However, although the previous regression results reveal a positive and significant link between Canadian national identity and alignment with the United States, the direct coefficient becomes insignificant after the inclusion of the mediating variable. This indicates that general feelings of Canadian identification hold no significant direct effect on foreign policy preferences vis-à-vis the United States beyond the continentalist attitudes embedded within it. On the other hand, the indirect pathway through ideational continentalism is highly significant. This suggests that a stronger sense of Canadian national identity is associated with more US-aligned foreign policy preferences through the channel of embedded ideational continentalism. This is further confirmed by the highly significant Average Mediation Effect (see Appendix 8). Overall, the result provides strong support for Hypothesis 2.