Conservatism is Changing. But Not in Canada.
Some data and working research on why fusionism endures in Canada
I thought I would share some initial results of the research I’ve been working on for the last few months. Overall, I find that while changing electoral conditions have driven considerable transformations in centre-right coalitions across the west, there remains a striking degree of continuity in Canada. In contrast to other countries, there remains a strong association between social traditionalism and economic liberalism among voters here. The question is: what makes Canada different? I suggest a few theories.
This is the meat of a chapter of my dissertation; Happy to receive any feedback - so please reach out if you’d like to read the manuscript!
‘Populist Entrenchment’ and Changing Party Systems
Global patterns of electoral competition have undoubtedly shifted over the last three decades, as the main roster of competitors and the common battlelines between them have been replaced. Across most western democracies, the old-reliable, post-war pattern of dueling moderate center-right and center-left parties has been upended by “populist” challengers of both the left and the right. This has, as seen below, been gradual and long-term, starting towards the end of the 1990s:
What’s important to note - and part of the reason why we should care - is that this isn’t just about a change in the teams involved. It amounts to a broader paradigmatic shift. Parties are not merely neutral electoral competitors, but also play a part in shaping the broader parameters that political contestation takes place within. Part of what characterized the Post-WW2 party system was the fact that, despite claiming the support of historically-opposed social constituencies, the center-left and center-right promoted similar things: the welfare state, globalization, neoliberal economics, and a managerialist procedure to policy-making. Developments linked to democratic consolidation, economic stability, and a decline in the salience of historical social conflicts had led these parties to grow increasingly moderate, professionalized, and nonideological.
But the entrenchment of those political movements we have grown accustomed to calling ‘populist’ reveals two things. The first is the prescience of what Johnathan Hopkin has labeled as “anti-system politics”: a broader rejection of the ‘political and economic order’ described above. That an overarching systematic critique of the status quo is now the defining ethos (or vibe, as the kids would say) of our time, including both the practical failure of existing procedures to deliver democratically-legitimate and equitable social goods alongside the more normative malaise that liberalism has itself fallen into to.
The second, and deeply interrelated, trend this shows is that the dimensions of social conflict have themselves changed. Competition between political parties continues to move away from conventional social divisions around economic distribution and class towards a set of conflicts that are expressed in more culturally salient terms, focused on the main issues of immigration, denationalism/transnationalism, multiculturalism, welfare chauvinism, and economic protectionism, among others.
Driven by a set of socio-structural changes brought about by post-materialist values, knowledge capitalism, and globalization, this may reflect the emergence of a novel cleavage that - while not necessarily replacing historical class-based divisions – is in the process of realigning the left and the right. Seen this way, the function of populist parties has been to introduce these new issue dimensions into existing party systems.
The detrimental consequences each has had on the mainline centre-left and centre-right has lead, throughout the 2020s, to a considerable degree of rhetorical, programmatic, and electoral experimentation; a blind grasping towards some sort of post-neoliberal order.
For the bruised centre-left, it appears that ongoing electoral viability can only be maintained through an effective, though elusive, balance of the diverging preferences of both its traditional working-class base and its more recent support among well-educated and urban knowledge-economy professionals. While the latter are the most likely to indicate support for immigration, transnational integration, cosmopolitanism, and neoliberal-adjacent economic policies, the former have moved towards protectionism, autarchy, and culturally conservative positions. Given pushback to the ‘woke’ progressivism of the 2010s, the impetus seems to have moved towards articulating an abundance-laden project of fair economic growth and technological progress.
But what this change means for the center-right is much less clear and understood. This is driven not only by the fact that the centre-right is far more contingent and coalition-based than the left, but that a good amount of populist pushback has expressed itself within existing, mainline party organizations. Indeed, the centre and more radical right have a deeply interactive relationship.
But what’s more important is that what we generally agree to be conservatism - the ‘fusionist’ alliance between social traditionalists and economic libertarians from a shared aversion to government overreach - is itself coming apart. These ‘right-wing’ positions on cultural and economic dimensions are not only becoming less covaried, but are increasingly mutually exclusive. It would seem that parties are now forced to choose one over the other. The thrust of the movement has likewise moved away from the Burkean moderation and gradualism advocated by its intellectual forebears towards a more confrontational and disruptive style of politics.
Canadian Exceptionalism
But while these same social structural transformations have likely occurred in Canada, there remains a striking degree of continuity when it comes to party behavior. It has yet to see significant disruption in its federal party system that parallel developments in comparable cases, whether through the entry of insurgent populist challengers or in the internal transformation of the mainline parties.
The Conservative Party of Canada (CPC) has, in particular, shown only minimal change from the programmatic commitments and electoral base of the Harper government (2006-2015). As in the past, rhetoric remains directed towards conventional conservative attacks on bureaucratic inefficiency, excessive government spending, and economic overregulation, evoked primarily to legitimize established neoliberal goals. And, when engaged, cultural or social issues are framed as a matter of individual freedom, and topics with moral or explicitly exclusionary implications – abortion, same-sex marriage, minority rights, immigration – are avoided whenever possible.
Existing explanations of the more moderate nature of the CPC focus on the claim that political conditions in Canada make a more right-wing populist or ideologically conservative program electorally unfeasible. While some analysists argue that the country’s particular conditions necessitate a uniquely moderate approach to politics, others suggest that – even while Canada may now contain a conventional left-right structure of competition – Canadians disproportionally fall onto the left side of the spectrum.
There are good reasons to believe that this could still be the case. The average Canadian certainly exhibits the sort of demographic traits that are now associated with the left, historically being more educated, urban, and cosmopolitan in their values than residents of other comparable advanced democracies. These voters are, furthermore, much more likely to be concentrated in the competitive metropolitan electoral districts that every party needs to win to form government.
Still, the emerging cleavage that has been observed throughout other comparable western democracies suggests that these conditions may have changed.
Although a comprehensive analysis of the probable emergence of a novel cleavage in Canada has yet to be conducted, research has already detected some demographic-based voting changes that may indicate a growing realignment along the same lines seen elsewhere, including economic class, urban/rural differences, religiosity, education, and cultural attitudes.
The accepted wisdom that Canada lacks the necessary conditions for the emergence of a populist or radical right political movement, will still possible, is a claim that can no longer be taken for granted and requires further evidence.
A New Cleavage in Canada?
I collected data from the Canadian Election Study (between 1997-2021, excluding 2006) to assess the strength and structure of a potentially new cleavage, alongside its consequences for the electoral opportunities available for the CPC. I ask whether there are possible electoral changes that, in introducing tension within an existing centre-right coalition, are incentivizing or pressuring the CPC (alongside other centre-right parties in Canada) to change in the way that other centre-right parties are.
Cleavages are understood to be the emergent property of three necessary interrelated elements: (1) a relatively permanent and consequential social division that creates clear and objectively definable groups, (2) corresponding identities or sets of attitudes, values, and interests, and (3) an organizational expression, most often seen through formal support for competing political parties.
Once formed, they remain in place over long periods of time because, even while the precise social fault lines that shaped them may decrease in salience, the broader social identities, values, and behaviors that they produced do not. In order to have confidence that a new cleavage has in fact emerged to reshape the electoral map, we should find evidence in all three of these elements.
There is considerable disagreement about how best to conceptualize the micro foundations and structure of new social divides1. There remains nuances around whether what we’re seeing amounts to either a shift in the nature of economic class status, the emergence of uniquely values-based differences, or an identarian resentment.
To keep things simple, I chose to examine the Canadian case through the GAL-TAN (Green, Alternative, Liberal - Traditional, Authoritarian, Nationalism) framework. GAL-TAN follows the prevailing view in so far as it conceptualizes the ‘new politics’ to be rooted in a novel, non-economic, values-based conflict that operates alongside an existing economic cleavage.
Generally, GAL-TAN cleavages should be rooted in nonmaterial attitudes on “democratic freedoms and rights” that pertain to the postmaterialist and traditionalist divides. While one side should be seen to support “expanded personal freedoms” and “democratic participation”, the other should instead prioritize “order, tradition, and ability” and “believe that the government should be a firm moral authority on social and cultural issues”.
This should be found to effect and structure a range of both cultural and economic issue positions, including (but certainly not limited to) abortion access, LGBTQ+ rights, transnationalism, immigration, multiculturalism, welfare chauvinism, environmentalism, and economic protectionism. The most relevant socio-structural positions found to have strong relationships with these positions elsewhere are education, religiosity, occupation, age, gender, and urban-rural residency.
Fourteen issue position variables are included in the analysis, representing a cross-section of relevant policy attitudes. Measures for relevant socio-structural positions were also derived from demographic variables used in all the datasets: age, gender, education, province, education, religion, and country of birth.2
Findings
Stage One - Exploratory Factor Analysis
I first conducted several stages of exploratory factor analysis to both identify which issues covaried and the latent underlying dimensions behind that covariance. The idea here is that, by sorting covarying issues into the same factors, the technique can determine whether an unobserved attitudinal dimension (whether pertaining to economic or GAL-TAN divides) accounts for patterns in the data.
The process detected just that; through five issue position variables it found clear, coherent, and stable latent factors that align with expectations regarding an economic and cultural attitudinal dimension:
Cultural Dimension: made up of positions on ‘equal rights’, ‘family values’, and “newer lifestyles’ ; it appears that many respondents have grouped their dislike of post-materialist cultural and social movements (as contained under the labels of “equal rights’ and ‘newer lifestyles’) alongside their support of traditional, authoritative, and orderly ‘family values’.
Economic Dimension: made up of positions of income gap reduction and government providing a “decent standard of living”; rooted in competing viewpoints about both the relative role the government ought to have in the distribution of material resources and whether the existing unequal outcomes of that process are problematic.
These findings do not appear to support the claim that a novel attitudinal dimension has emerged but instead suggest that the composition or structure of the existing factors has not significantly changed since 1997.
But while these factor scores demonstrate that a latent cultural dimension exists in the electorate, they do not say that much on how it actually informs or impacts the structure of Canadian party competition. In order to have significance, a latent factor cannot merely inform a relatively narrow or isolated policy area (abortion or same-sex marriage, for example), as much as it has to relate to how individuals organize and relate to their broader political orientation.
It could very well be the case that while this broader attitudinal structure has existed for several decades, it can only be said to have formed into a cleavage in recent years, whether driven by an increased political salience, a stronger alignment with socio-structural differences, or specific party cues.
Stage Two - Issue Position, Demographic, and Vote Choice Regressions
Having established the existence of a cultural factor in the Canadian electorate, I proceeded to extract factor scores from each respondent and, in treating them as a measure for said attitudinal dimension, separately examined its relationship with each of the consistent elements of a cleavage - (1) broader policy orientations, (2) specific socio-structural positioning, and (3) impact on vote choice – through a series of regression models.
To assess the extent to which the cultural attitudinal dimension may inform a broader policy orientation, I used an OLS (ordinary least squares) model to regress factor scores on both the economic dimension and the included issue position variables that did not map onto any clear factor.
As with the factor loadings, they suggest that cultural attitudes do in fact inform a broader policy orientation that has shown little change overtime. Higher traditionalist social attitudes (as measured by the cultural factor) remain associated, as they were in 1997, with support for limited government economic intervention, lower immigration rates, higher defense spending, and skepticism towards bilingualism policy.
Interestingly, while existing findings have shown that social conservative positions are developing negative associations with liberal economic policy, these results show a strong - though qualified - connection between more traditionalist social attitudes and economic liberalism. Ultimately, they depict a set of issue positions that are consistent with an ideal type of what we associate with a broader center-right policy orientation, combining economic liberalization, social traditionalism, greater defensive spending, and immigration skepticism. And, while new issues (like international trade and the conflict between the environment and jobs) seem to have been incorporated, there is no evidence that they have started to unravel it. The cultural attitudinal dimension, while not new, constitutes this element of a cleavage structure.
How does this broader policy orientation relate to demographic variables? I repeat the model, but regress the cultural factor scores on age, gender, education, religion, region, and birthplace.
These variables show clear, though inconsistent, relationships with cultural attitudes in the ways expected. As with issue positions, the prevailing picture appears to be one of continuity. In line with established findings from in and out of Canada, education and religion seem to be the strongest and most constant predictors of cultural attitudes.
While all religious categories are associated with more traditional cultural attitudes, the possession and level of university education (as in the difference between undergraduate and post-graduate degrees) makes one less likely to express concern with newer lifestyles, equal rights, and family values.
Age, albeit less consistent, also appears to be relevant, particularly regarding the differences between younger adults and seniors.
Breaking from existing Canadian scholarly emphasis on regional differences, the analysis does not find any consistent relationships with region.
But the impact of gender and birthplace do seem to have changed in interesting, though predictable, ways. The findings on gender certainly supports existing evidence that men and women are becoming increasingly polarized in their cultural policy orientations.
And the fact that being born in Canada becomes statistically associated with more progressive cultural attitudes after 2015 is consistent with the claim that urban immigrant communities, despite more often voting Liberal, are generally more socially conservative.
Compared to issue positions, socio-structural positions - while significant - play a much less important and consistent role in the shaping and maintenance of cultural attitudes.
I finally examine the impact of both the cultural and economic dimensions on vote choice across the datasets. Given that vote choice is a categorical variable, I perform this through a multinomial logistic regression model that assesses the impact one’s position on the two attitudinal dimensions has on the probability for voting for each major party.
Like the other cleavage components, these too present a consistent and conventional picture: both traditional cultural attitudes and more liberal economic attitudes significantly predict support for centre-right parties (whether the CPC, PCPC, or Reform/Alliance). Note also how close the NDP and Liberals are for both measures. This pattern, furthermore, remains in place throughout all of the datasets covered demonstrating that, while it matters to vote choice, it seems to do so in a way that has not changed over the last several decades.
What Makes Canada Different?
Although Canadians structure many of their political attitudes along a discernable cultural dimension, there is no evidence of change. By and large, cultural traditionalist attitudes pertaining to GAL-TAN divides map quite easily onto the existing conflict dimensions of the Canadian party system, particularly in regards to their relationship with liberal economic attitudes. This suggests that, although the party’s base may be increasingly made up of social traditionalists, the Conservative Party in Canada can continue to base its wide appeal on economic policy alongside a tactful, if increasingly rightward, engagement with salient cultural or social issues.
What, then, accounts for the apparent continuity in the Canadian electoral marketplace, particularly as it regards the centre-right’s coalition? And how will this matter to the party’s behavior moving forward? I finish here by suggesting five possible answers:
Empirical Limitations: it could be the case that the findings are an artifact of the research methodology. Certainly, some limitations caution against a certified answer. This analysis is missing important issue position and demographic variables, and it only looks at the aggregate national picture (ie. how would this look through the lens of 338(ish) electoral districts?). Still, this doesn’t seem like it would really affect my main claim: that economic liberalism and social traditionalism remain tethered in Canada for the time being.
Political Economy: one approach to accounting for Canada’s divergence may be to emphasize specific aspects of its socioeconomic structure. The Canadian economy’s reliance on extractive export commodities, for instance, could mean that the changes brought about by neoliberal policy (ex. deindustrialization, declining union membership) has had less of an overall impact then it did in cases like the United States and Britain; that Canada’s natural resource sector prevented the conditions behind the much-discussed “losers/winners of globalization” or “depths of despair” dynamic that has been observed elsewhere. Certainly, in this present moment, no serious voice on the centre-right really disputes the fact that our needs are not best met by economic liberalization and free trade.
Canadian Conservatism Itself: alternatively, this could be a story about but the way conservative ideology is itself marketed and packaged by Canadian conservative actors, whether elite or otherwise. Quite a lot of ink has been spilled (a good amount of which is accurate) on the way that Canadian Conservatism is inherently different than its American or British cousins. ‘Fusionism’ is itself also a historically contingent development, based not so much on inherent ‘conservative’ attitudes but a coalition built and maintained by mid-century centre-right elites. Perhaps, while it has fallen apart elsewhere, it is maintained in Canada through the legitimacy of existing party elites.
Political Opportunity Structures: another approach would be to suggest that specific aspects of Canada’s institutional environment prevent anti-fusionist insurgencies. It could be the case that although a more radical - or at least ‘anti-fusionist’- right-wing base exists in Canada, the country’s political structures – whether the electoral system or the internal cohesiveness of the CPC itself – prevents their expression.
Reform: Finally, one could point to an important critical juncture in the development of centre-right politics in Canada: the Reform-PC party split. This could inculcate Canadian conservatism from the transformations elsewhere, whether it is because of the fact that contemporary actors fear the electoral consequences of division or that the split has already introduced and incorporated into the political system those socio-structural changes having impacts elsewhere.
Existing labels include libertarian-authoritarian, integration-demarcation, cosmopolitanism-communitarianism, and GAL-TAN
Other relevant variables, such as occupation, union status, income, and rural-urban residency, lacked consistent measures and were therefore unable to be included.