Cooperation During Minority Governments

What does cooperation mean in the 2022 supply-and-confidence agreement between the LPC and NDP? The NDP promised to support any budget or confidence vote until the end of the legislature in 2025 (PMO, 2022). In addition, both parties have agreed to coordinate on the legislative process, with monthly meetings between party leaders and briefing sessions before bills are introduced to the House of Commons. The agreement therefore creates a formal voting coalition between the two parties (Godbout and Cochrane, Reference Godbout, Cochrane, Field and Martin2022). In addition, the agreement includes a mutually approved policy program aimed at expanding social policies. Shared policy goals include healthcare (dental insurance, universal drug insurance [Pharmacare], national norms on long-term health services), housing (involving construction and regulation, as well as housing benefits), family (entailing a national daycare program), work (including paid sick leave and anti-strikebreaking legislation) and Indigenous reconciliation initiatives (PMO, 2022).Footnote ¹ As will be argued later, this programmatic component and the shared policy objectives are fundamental to understanding the agreement.

Uncommon in Canada, supply-and-confidence agreements are a type of support agreement between the government and an opposition party, in which the opposition party agrees to vote with the government on any confidence vote in parliament in exchange for concessions. Scholars typically classify this support as a party “in” government without being “of” government (Dunphy and Bale, Reference Dunphy and Bale2011; Miller and Curtin, Reference Miller and Curtin2011). This implies that the party formally remains in the opposition but actively contributes to the implementation of key policies, participates in decision-making and claims credit for ensuring a stable and efficient government (Miller and Curtin, Reference Miller and Curtin2011).

In Canada, Godbout and Cochrane (Reference Godbout, Cochrane, Field and Martin2022) identified several factors explaining why federal parties have an aversion to such legislative agreement. The informal convention that the party with the most seats should form government and govern alone fosters a perception of single-party government as the default mode of governance (Stewart, Reference Stewart1980). Institutions further reinforce this convention, including the strong position of the executive, high party discipline and the likelihood of re-election for minority governments (Godbout and Cochrane, Reference Godbout, Cochrane, Field and Martin2022). Moreover, there exists a deeply embedded norm to seek compromise on more consensual proposals and refrain from opposing the government during the first confidence vote (the Speech of the Throne) following an election. The structure of the federal party system also favours an ad-hoc style of majority building in the House of Commons. In the case of minority governments, alliances can form along the left-right axis or the centre-periphery axis (related to federalism) to pass legislation, due to strong regionalist parties or the NDP often holding the balance of power (Cody, Reference Cody2008). Thus, interparty cooperation almost never takes the form of formal agreements during minority governments in the federal parliament.

We can nonetheless identify two cases of legislative agreements in Canadian federal politics: the 1972 LPC-NDP agreement and the failed attempt to form a tripartite coalition LPC-NDP-Bloc Québécois in 2008 to displace the Conservative government of Stephen Harper. In 1972, factors such as ideological proximity, poor party finance and the imperative to avoid an early election that could have led to the return of the Progressive Conservatives were pivotal in the emergence of the agreement (Stewart, Reference Stewart1980), but the NDP withdrew from the agreement to cause an election in 1974, which led to a Liberal majority (Godbout and Høyland, Reference Godbout and Høyland2011). A similar situation occurred in 2008, when the LPC, the NDP and the Bloc Québécois negotiated a coalition agreement in which the LPC and the NDP would share government positions with the parliamentary support of the BQ (Godbout and Cochrane, Reference Godbout, Cochrane, Field and Martin2022). This time again, the agreement was meant to avoid a Conservative government and was fostered by a shared criticism of its handling of the 2008 financial crisis, but Harper prorogued the parliament to escape from a confidence vote, which derailed the LPC-NDP agreement over partisan disputes, particularly cooperation with a Québec autonomist party (Godbout and Cochrane, Reference Godbout, Cochrane, Field and Martin2022). In both cases, opposing the Conservative government using parliamentary leverage and proximity in policy ideas were important factors leading to interparty cooperation.

At the provincial level, we can also find similar factors, such as the 1985 Liberals-NDP agreement in Ontario. Although the Ontario Progressive-Conservative (PC) party arrived first in the 1985 election, the Liberals and New Democrats wished to end the reign of the PC, which had governed the province since 1943, and negotiated a legislative agreement (Cody, Reference Cody2008). Cody (Reference Cody2008) found “unusually similar” electoral platforms and promises between the Liberals and the New Democrats during the 1985 election, which helped both parties to cooperate with the Liberals forming the government and the NDP supporting it. Yet, the Liberals withdrew from the agreement in 1987 for electoral reasons in the face of advantageous polls (Cody, Reference Cody2008). In 2017, the British Columbian NDP and the Green Party of British Columbia entered into a supply-and-confidence agreement over their conjoint majority of one seat over the BC Liberals, after their 12-year-long reign (McElroy, Reference McElroy2017). Ideological proximity was also detrimental to the forming of their agreement, which contained policy proposals related to electoral reform, extensive climate policies, Pharmacare and investment in healthcare, childcare, poverty reduction and a basic income pilot, among others (BC NDP Caucus, 2017). Yet, the NDP withdrew early from the four-year agreement in 2020 to bring about a snap election, in which the party won a majority. More recently, a supply-and-confidence agreement occurred in Yukon between the Liberals and the NDP in 2021 and was renewed in 2023 until the next Yukon general election in 2025 (CBC, 2023b), which is the only case of a non-terminated legislative agreement ever observed in Canada. This agreement clearly states that “The Yukon Liberals and the Yukon NDP campaigned on policy proposals that included points of agreement across multiple themes” (Yukon Liberal Caucus, 2021), revealing the role of electoral platforms and promises in the forming of the agreement. Other than these rare agreements either at the federal or provincial level, however, minority parliaments lead to an ad-hoc majority strategy, as explained by Godbout and Cochrane (Reference Godbout, Cochrane, Field and Martin2022).

In fact, minority governments are often understood by parties and political pundits as elections in waiting (Cody, Reference Cody2008; Pammett and Dornan, Reference Dornan, Pammett and Dornan2022), and yet the 2021 federal election resulted in a supply-and-confidence agreement between the LPC and the NDP. Since 2019, the LPC has been at the helm of a minority government, having lost its majority due to a negative campaign and political scandals that eroded Prime Minister Justin Trudeau's popularity (Brooke, Reference Brooke, Pammett and Dornan2022). The NDP has since held the balance of power and has pressured the LPC government to adopt a more left-leaning stance on social policy (Béland et al., Reference Béland, Dinan, Rocco and Waddan2022; see Maioni, Reference Maioni1998 for an historical analysis of LPC-NDP interactions regarding health care policy). During the COVID-19 crisis, the Liberal government implemented large, expansive —yet temporary—social policy measures, strongly supported and pushed by the NDP (Béland et al., Reference Béland, Lecours, Paquet and Tombe2020). Prime Minister Trudeau triggered an early election in August 2021, believing that a “majority [government] was within reach” (Dornan, Reference Dornan, Pammett and Dornan2022). The 2021 federal election yielded results remarkably similar to those of the 2019 election (although 40 ridings changed seat), with the LPC winning three seats and maintaining a minority government, and the NDP securing one seat (Clarke et al., Reference Clarke, Scotto, Stewart, Pammett and Dornan2022). Unlike the situation in 2019, this time a supply-and-confidence agreement was within reach, despite the majority formed by the LPC and the NDP in the House of Commons remaining almost unchanged.

Nevertheless, cooperation is a constant dilemma for political parties. In Muller and Strøm's (Reference Muller and Strøm1999) traditional framework, political parties typically seek offices, policies or votes, and their choices involve trade-offs between these three goals. In the context of interparty cooperation during minority governments, particularly in Europe and other Westminster systems, the governing party aims to remain in office and implement its program to claim credit in the next election. Meanwhile, support parties seek proximity to decision-making (and offices in the case of a coalition), policy concessions and public prominence (Miller and Curtin, Reference Miller and Curtin2011; Artès and Bustos, Reference Artès and Bustos2008; Dunphy and Bale, Reference Dunphy and Bale2011). The bargaining power of each party stems from its strategic position after the election, especially for smaller parties. For example, in the German parliament, Abed and Siaroff (Reference Abedi and Siaroff2011) differentiate between an influential small party that holds a majority with the larger party but is not the sole influential party, a crucial party that holds a majority alone with the larger party, and a pivotal party that holds a majority with either large party—a kingmaker.Footnote ²

Building on this literature, Godbout and Cochrane (Reference Godbout, Cochrane, Field and Martin2022) outlined three factors that lead parties to cooperate in ad-hoc coalitions during minority government in Canada: ideological proximity along the left-right axis, alignment on regionalization-centralization and electoral incentives. Parties with similar policy preferences tend to cooperate more, and the history of minority governments suggests that the LPC and the NDP naturally align as voting coalition partners (Stewart, Reference Stewart1980; Cody, Reference Cody2008). Electoral incentives influence coalition building, with unpopular parties or those facing leadership challenges or financial issues aiming to avoid an early election. Simultaneously, strong parties seek policy gains by using their parliamentary influence to obtain concessions from the governing party. But cooperation entails strategic risks for supporting parties, as opponents can negatively frame their voting alliance with the government.

Why did the LPC and the NDP decide to cooperate in a formal supply-and-confidence agreement after the 2021 election? In the next section, we turn to the role of ideas in politics as they relate to programmatic beliefs, strategic learning, frames and large-scale crises. The ideational literature provides analytical tools to explain how parties navigated the COVID-19 pandemic and the 2019-2021 election sequence, resulting in a unique scenario where a formal agreement emerged in the House of Commons. Thus, our conclusions align with both ideological (policy ideas) and strategic (electoral gains) explanations for coalition building during minority governments.

Ideas, Strategic Learning, and Large-Scale Crises

Over the last three decades, much has been written about the role of ideas in politics and public policy (for example, Béland, Reference Béland2019; Béland and Cox, Reference Béland and Cox2011; Hall, Reference Hall1993; Berman, Reference Berman1998; Blyth, Reference Blyth2002; Campbell, Reference Campbell2004; Mehta, Reference Mehta, Béland and Cox2011; Parsons, Reference Parsons2007; Schmidt, Reference Schmidt2008). According to Craig Parsons (Reference Parsons2007), this scholarship emphasizes “the logic of interpretation,” focusing on how particular political and social actors interpret the world around them in ways that can influence their behaviour, alone or in conjunction with other factors. This means that the ideational scholarship we draw on operates from the assumption that ideas are not mere epiphenomena purely derived from the institutional and/or material position of actors (Berman, Reference Berman1998; Campbell, Reference Campbell2004). Rather, the ideational approach recognizes the interpretive capacities of parties in relation to the political context they confront (Mudge, Reference Mudge2015a). This leads to a more sophisticated analysis of parties’ motivations and strategies that deepens the traditional framework of “offices, policies, and votes” (Muller and Strøm Reference Muller and Strøm1999). We thus follow Béland and Cox (Reference Béland and Cox2016) in arguing that ideas act as “coalition magnets” for policymakers.

As Sheri Berman (Reference Berman1998) reminds us, scholars turning to the role of ideas in their empirical work must define precisely what these ideas are. Simultaneously, as highlighted by Daniel Béland and Alex Waddan (Reference Béland and Waddan2015), these scholars need to break down the concept of ideas to study specific ideational processes. This is crucial because the term “ideas” encompasses various ideational processes that must be distinguished from one another (Béland and Waddan, Reference Béland and Waddan2015; see also Campbell, Reference Campbell2004). In this article, we focus on one main type of idea—the “programmatic beliefs” of political parties. We study these beliefs in relation to both framing and strategic learning during the 2019-2021 election sequence and the COVID-19 crisis.

In her book The Social Democratic Moment, Berman (Reference Berman1998) turns to the role of ideas to explain the behaviour of Social Democratic parties in Germany and Sweden between the two World Wars. Specially, she turns to what she calls “programmatic beliefs,” situated “between ideologies and policy positions” (Berman, Reference Berman1998: 21). In contrast to ideologies as broad world views, programmatic beliefs “are directly relevant only to particular categories of human action” (Berman, Reference Berman1998: 21). In that, “programmatic beliefs provide guidelines for practical activity and for the formulation of solutions to everyday problems,” which can be aggregated in specific political programs and party platforms (Berman, Reference Berman1998: 21). In other words, programmatic beliefs “provide a relatively clear and distinctive connection between theory and praxis” (Berman, Reference Berman1998: 21; see also Mudge, Reference Mudge2015a), making them practical ideas that enable political parties to align their broad ideological and strategic orientations with concrete policy prescriptions featured prominently in their political platforms.

In this article, we claim that the strategic beliefs of political parties evolve over time through a process known as “strategic learning” (Hay, Reference Hay, Marsh and Stoker1995: 201–202; Béland and Marier, Reference Béland and Marier2006). In contrast to policy learning, which involves experts, politicians and other actors drawing lessons from existing policies to improve or discard them over time (Hall, Reference Hall1993; Heclo, Reference Heclo1974), “strategic learning takes place when elected officials constantly learn about political risks and opportunities related to timing, institutional structures, and the anticipated behaviour of other social and political actors” (Béland and Marier, Reference Béland and Marier2006: 300). More specifically, within the context of this article, strategic learning revolves around extracting electoral and political lessons from changing circumstances that may reshape the programmatic beliefs of political parties and the frames they articulate in public discourse to support or oppose specific policy proposals (Campbell, Reference Campbell2004).

Intrinsically strategic, “frames” are widely studied within the political science and public policy literature concerning the role of ideas, although they are sometimes conflated with the concept of discourse (Schmidt, Reference Schmidt2008). According to John L. Campbell (Reference Campbell2004: 94), frames are “symbols and concepts” that political actors deploy to justify their policy proposals and promote themselves among voters and/or specific interest groups. Framing is an essential ideational process during the policy cycle because it contributes to the definition of the policy problem, the interpretation of the larger social and political context and the prescription and justification of proposed policy solutions (Béland, Reference Béland2019). Frames enable political parties to actively tap into existing cultural and ideological categories when formulating public statements about the policy proposals they support or oppose in close relationship with the discourse and strategies of their political opponents and competitors (Béland, Reference Béland2005). Thus, frames serve as essential analytical tools for understanding the interrelation between programmatic beliefs and strategy-building in the partisan arena, which is especially useful during the coding of qualitative data (see the online Appendix).

Programmatic beliefs, strategic learning and frames are analytical tools to make sense of how political parties interpret changing social contexts, notably in situations of uncertainty (see Mudge, Reference Mudge2015b). Uncertainty is a fundamental aspect of political reality, rendering outcomes challenging to predict for actors and underscoring the important role of ideas in explaining how parties navigate the opportunities and constraints within political equilibriums and elections, serving as mechanisms for change in these equilibriums (Blyth, Reference Blyth2011). Large-scale economic, political and/or public health crises rank among the circumstances most likely to induce acute uncertainty, prompting strategic learning and the revision of programmatic beliefs and framing tactics (Pontusson and Raess, Reference Pontusson and Raess2012; Hall, Reference Hall1993; Bremer and McDaniel, Reference Bremer and McDaniel2020; Farrell and Quiggin, Reference Farrell and Quiggin2017; Hutter et al., Reference Hutter, Kriesi and Vidal2018; Roberts, Reference Roberts2017). According to Mark Blyth (Reference Blyth2002: 37), large-scale crises and their acute uncertainty create space for new ideas which, in turn, help lower the uncertainty faced and “[narrowing] possible interpretations of the crisis, and hence courses of action, to a significant degree.” This article demonstrates how, through strategic learning, political parties can alter their programmatic beliefs in the context of the COVID-19 crisis and the political uncertainty surrounding the 2019-2021 election sequence. Simultaneously, we acknowledge that strategic learning and changes in programmatic beliefs and framing tactics can occur both during and outside large-scale crises, although such changes are more likely to emerge during them.

Empirical EvidenceFootnote ³

To study programmatic beliefs, framing tactics and strategic learning, we compared the party manifestos of the four major parties in the House of Commons (LPC, the Conservative Party of Canada [CPC], NDP, and the Bloc Québécois [BQ]) for the 2015, 2019 and 2021 elections. We chose to focus on the LPC, CPC, NDP and BQ, as they collectively hold nearly all the seats in the House of Commons during the period studied (336 of 338 in 2021). Party manifestos serve as reliable indicators of the concessions, policy ideas and strategies that parties bring to the table during coalition or support agreement negotiationsFootnote ⁴ (Artès and Bustos, Reference Artès and Bustos2008; Matthews, Reference Matthews2011). Moreover, electoral platforms are valuable data sources for analyzing the language through which parties articulate their programmatic beliefs. According to Stephanie Mudge (Reference Mudge2015a), this language comprises policy prescriptions, the framing of policy programs and the social categories targeted by political parties. In our analysis, we utilize these dimensions to understand the beliefs of the four major parties in the House of Commons during this election sequence and the interpretation of the COVID-19 pandemic as a large-scale crisis that required a social policy response.

Additionally, we traced the legislative process that led to emergency policymaking during the first two years of the COVID-19 pandemic. We selected a total of ten bills introduced in the House of Commons between March 2020 and November 2021 for analysis. Specifically, our focus was on examining the votes and “major speeches”Footnote ⁵ during the debates on the second reading of these bill proposals. The choice of the second reading is deliberate, as it relates to debates on the principles and substance of the policy proposals. These debates give each party the opportunity (and time) to formulate a statement detailing their position on the proposed policies. Given the increasing significance of procedural speeches in the House of Commons in recent years (Vallée-Dubois et al., Reference Vallée-Dubois, Godbout and Cochrane2021), analyzing the second reading allows us to focus on the discursive efforts of parties regarding policymaking and the expression of programmatic beliefs. These two data sources—party manifestos and the speeches at the House of Commons—assist us in understanding the ideas and strategies of the four major parties (and their relations with opponents or allies in the party system) since 2015, with a particular focus on the COVID crisis.

In Table 1, we present the ten bills selected along with the corresponding policies proposed and the adoption process. The sequence begins with C-13, which implemented the Canada Emergency Response Benefit and ends with C-2 (after the 2021 election), enacting the final emergency measure, the Canada Worker Lockdown Benefit. These selected bills comprise the most extensive emergency social policy measures either implemented or modified by the federal government during the COVID-19 crisis. Of the ten bills, seven were adopted without formal opposition to the Liberal government. The remaining three underwent debate in the House from April 2021, over a year after the first wave of the virus in Canada. In these non-consensual bills, the CPC opposed all, the NDP voted against the last proposal (C-2), and the BQ consistently supported the government. However, votes, especially in the context of emergency parliamentarism, characterized by accelerated procedures, heightened control from party leaders and a reduced number of MPs and sitting days in the House, may not provide a comprehensive understanding of the programmatic beliefs of the four parties (Malloy, Reference Malloy2020; Turnbull and Bernier, Reference Turnbull and Bernier2022; Lozano et al., Reference Lozano, Atkinson and Mou2021; Rayment and VandenBeukel, Reference Rayment and VandenBeukel2020).

Table 1. Bill Proposals and Social Policy in Canada during the Pandemic

A study of policy ideas and (policy and strategic) learning processes would typically benefit from interviews with key political actors (Campbell and Pedersen, Reference Campbell and Pedersen2014). In our case, party staff and MPs that negotiated in the backstage of the supply-and-confidence agreement could inform our analysis and help confirm or disconfirm our conclusions. Yet, the LPC-NDP deal was still in effect and a contentious issue on the federal political scene at the time of the research, limiting our access to key informants. Considering this, to stay as close as possible to the actors at hand and start mapping the political terrain of the supply-and-confidence agreement, we decided to adopt a different research strategy and analyze documents and speeches produced by parties and MPs. This lack of interviews with the main brokers of the agreement does limit our detailed understanding of the LPC and the NDP during the period we cover. Yet, our analysis identifies ideational and strategic processes that could be further deepened through future interviews.

In the next section, we turn to party manifestos and speeches in the House of Commons to explore programmatic beliefs, framing tactics and strategic learning from 2015 onward and throughout the pandemic. Specifically, we coded the policy proposals, the frames deployed by parties to foster political support, the targeted constituencies and the political lessons learned by parties. Analyzing political lessons can be difficult as they pertain to the internal decision-making processes of actors. In the case of political parties, complex social networks and power dynamics among MPs, party experts, intellectuals, and others influence how parties interpret evolving political contexts (Mudge, Reference Mudge2015a). By focusing on parties’ official positions in the political arena (during elections and in parliament), we use frames to infer how parties draw lessons from the political context and strategically deploy them. Frames typically relate to the broader economic and political context and the party system, serving as useful tools for analyzing the dynamics of parliamentary and electoral competition or cooperation.

Explaining the 2022 Supply-and-Confidence Agreement

In this section, we outline our ideational argument to explain why the supply-and-confidence agreement took shape in the LPC and NDP response to the COVID-19 crisis and the 2019-2021 election sequence. First, we present the programmatic language of the four parties in their manifestos from 2015 to 2021 (see Table 2). Then, we ask whether ideas explain the formation of the LPC-NDP alliance. In response, we find that strategic learning from the 2019-2021 election sequence serves as a mediating factor, linking programmatic beliefs and political strategies.

Table 2. Programmatic Language in Party Manifesto, 2015-2021

Note:* Main expenses and proposals presented in the manifesto

¹ The distinction between social framing and economic framing derived from the coding, see the online Appendix.