Introduction
Much has been written about “the new politics of the welfare state” and Paul Pierson's seminal contributions to that literature (Pierson, Reference Pierson1994, Reference Pierson1996; Myles and Pierson, Reference Myles, Pierson and Pierson2000). Pierson argues that changing fiscal, economic and demographic conditions meant the “old politics” of social program expansion that characterized the post–World War II era was replaced by a “new politics” of social policy centred on attempts to control spending. In this new environment, program clienteles replaced organized labour as the key societal drivers of welfare state programs, and blame-avoiding replaced credit-claiming as the dominant motivation of politicians’ actions with respect to the welfare state. In general, he argued, welfare state programs have been resilient, with some of them undergoing significant retrenchment but seldom dismantling.
Since Pierson's Dismantling the Welfare State? appeared, each of the new politics arguments have undergone considerable scrutiny and gained substantial empirical support (Jensen et al., Reference Jensen, Wenzelburger and Zohlnhöfer2019). Even John Stephens—one of the initial proponents of power resources theory, which emphasizes the role of labour parties and trade unions as the driving force of social politics (Esping-Andersen, Reference Esping-Andersen1990; Korpi, Reference Korpi1983)—claims the main arguments Pierson (Reference Pierson1996) put forward about retrenchment “have withstood the test of time” (Stephens, Reference Stephens2015: 274).
Recent scholarship has also challenged some of Pierson's arguments, however. Some scholars have claimed that intertwined factors of increased labour market inequality and new social risks have increased tensions within traditional welfare state constituencies but also created opportunities to address welfare state challenges through new coalitional strategies and new programmatic “social investment policies” focused on job creation, facilitation of labour market participation and social inclusion rather than passive “compensatory” benefit receipt (Armingeon and Bonoli, Reference Armingeon and Bonoli2007; Häusermann, Reference Häusermann2018; Häusermann et al., Reference Häusermann, Kemmerling and Rueda2020; White, Reference White2012). These developments are sometimes seen as providing politicians with new opportunities for credit-claiming in welfare state reform (Giger and Nelson, Reference Giger and Nelson2011; Bonoli, Reference Bonoli, Bonoli and Natali2012). While this latter literature has focused mostly on developments in Western Europe, it has also been applied to Canada and other so-called liberal welfare states (Jenson, Reference Jenson2004; Banting, Reference Banting2005b).
The central claim in this article is that neither Pierson's key “new politics” nor the arguments about the “newer” politics of social risk are adequate to understand recent developments in social policy. Instead, both sets of arguments are best viewed as more likely to hold under some conditions than others. The article's central contribution is to develop both theory and empirical evidence on the conditions under which Pierson's arguments and later modifications are likely to hold in a stronger or softer form—or not at all. We draw on Pierson's own writings and on cross-national evidence that has developed in the theoretical and empirical literature in the more than 25 years since Dismantling the Welfare State? was published, as well as on evidence from recent developments in the Canadian welfare state.
Canada is generally characterized as a “liberal” welfare state, like the United States and United Kingdom (see White, Reference White2012, Reference White2017), but it has always been an uneasy fit within that category: it has a comprehensive public health care system based on the principle of universalism (Tuohy, Reference Tuohy2018) and, as discussed further below, a public pension system that combines a substantial benefit floor, a contributory earnings-related component and an income-tested supplement that lifts most seniors out of poverty. Canada is a striking case of shifting trajectories in welfare state reform in recent years: at the federal level, the politics of expansion has dominated Canadian social policy debates and actions over the last two decades, in strong contrast to the situation prevailing during the Mulroney (1984–1993) and early Chrétien (1993–1997) years, which featured the politics of retrenchment prominently (Bashevkin, Reference Bashevkin2000; Bernier, Reference Bernier2003; Olsen, Reference Olsen2002). Empirically, this article examines the “new politics of federal social policy expansion” during the Harper (2006–2015) and Justin Trudeau (2015–) years, with a focus on three policy areas: family policy, specifically child benefits and childcare transfers; public pensions; and Employment Insurance (EI). These policy areas were selected because they represent a diverse set of constituencies and because they are the primary areas of federally delivered cash benefits in Canada. Each of these sectors has experienced selective expansion: the creation and/or merging of social programs (the Universal Child Care Benefit and, later, the Canada Child Benefit) in the case of child benefits; the expansion of an existing program (the Canada Pension Plan) in the case of pensions; and a mixture of expansion and retrenchment in the case of EI.
After this introduction, the first section addresses Pierson's key “new politics” arguments, as well as critiques, extensions and more recent evidence provided by recent scholarship regarding each of those arguments. The critiques help to highlight general conditions under which these arguments are likely to hold in a stronger or more muted form. Those general “facilitating and limiting conditions” are compared to Canadian conditions to develop a set of theoretically informed expectations about when, and in what form, selective welfare state expansion will occur. We argue that the interaction of five conditional factors in Canada—the availability of fiscal resources, the presence of new social risks and strategic responses to them, the intensity and nature of partisan competition, differing policy preferences of the main political parties, and the role of political institutions—have facilitated deviations from the patterns predicted by Pierson's analysis of “the new politics of the welfare state.” This is the case because, under some conditions, the interaction among these factors can make retrenchment, but also expansion, occur. The second section offers an empirical analysis of the expansion of federal child benefits in Canada from the mid-2000s to the late 2010s. The third section explores the politics of Canada Pension Plan (CPP) expansion from the 2008 financial crisis to the 2016 reform of the program and, more briefly, policy shifts in the quasi-universal Old Age Security (OAS) program. The fourth section discusses the more complex patterns of change in EI. The final section compares the three cases to assess the conditional nature of the seven central arguments in Pierson's new politics framework. The article concludes with reflections on the new politics of social policy expansion in Canada and in other wealthy democracies, along with suggestions for further research.
From Welfare State Expansion to Retrenchment and Back
Pierson's new politics framework can be broken down into seven specific empirical and causal arguments, which we address in turn. Our thesis is that Pierson's analytical framework is likely to predict actual social policy patterns more accurately under some conditions than others. Our analysis outlines those conditions and compares them with the conditions that have characterized the Canadian social policy environment over the last two decades. When facilitating conditions are present in a sharp and persistent form in a particular societal context, the direction of social policy change should be strongly consistent with those suggested by Pierson's analytical framework; when those conditions are mixed, intermittent or absent, or when limiting conditions dominate, Pierson's arguments should hold only in a more muted form or not at all.
1. The politics of permanent austerity
A first argument is that a politics of permanent austerity has come to dominate social policy reform in advanced industrial societies. Pierson cited several bases for intense retrenchment pressures on the welfare state, including slower economic growth and demographic aging. These pressures are by no means uncontested, however. As Pierson (Reference Pierson2002: 370) puts it: “The contemporary politics of the welfare state takes shape against a backdrop of both intense pressures for austerity and enduring popularity.” More recent literature has noted that external pressures, such as demands for austerity placed by international financial institutions in exchange for financial bailouts, or accession criteria for countries seeking to join the European Union, can add to austerity pressures (Stepan and Anderson, Reference Stepan and Anderson2014). Yet there may also be greater opportunities for social policy expansion when the fiscal situation of the state is relatively favourable. This positive reality can take different forms: budget surpluses, an auspicious debt-to-GDP ratio and, in the case of social insurance programs, long-term actuarial balance or surplus.
Canadian experience features several factors relevant to a conditional understanding of retrenchment forces or expansionary forces likely to dominate. Canada's debt-to-GDP ratio reached a peak in the mid-1990s and fell dramatically between 1997 and 2007, before rapidly increasing during the global financial crisis. These conditions suggest that “permanent austerity” pressures are likely to be relatively muted in Canada in comparison to many wealthy democracies. Equally important, those fiscal pressures will be uneven over time, with selective expansionary initiatives appearing on the Canadian agenda after the improvement in Canada's fiscal situation in the new century.
2. Welfare state resilience
A second Pierson argument is that in the face of these challenges, welfare state programs have proven largely resilient. He did not claim, however, that all welfare state programs were equally resilient, and he discusses a number of factors that may make programs more vulnerable, such as whether a social insurance program was nascent or “mature” and paying out benefits to a large clientele (Reference Pierson1994: 56), and variation across programs in opportunities to obfuscate loss-imposition.
In the Canadian context, welfare state programs vary substantially on several conditional factors affecting program resilience. For example, while pensions and health care serve broad and politically popular constituencies, the constituencies are narrower for EI, childcare and social assistance, suggesting greater opportunities for retrenchment. While the Canada/Quebec Pension Plan (CPP/QPP) and Old Age Security (OAS) and the Guaranteed Income Supplement (GIS) are mature and thus presumably more resilient, federal transfers for childcare are not—and therefore more vulnerable.
3. New social risks
The importance of new social risks is another theme highlighted by Pierson, especially in his later work (Reference Pierson2011). Along with other scholars (for example, Esping-Andersen, Reference Esping-Andersen1999; Häusserman, Reference Häusermann2010), Pierson has highlighted deindustrialization, heightened wage inequality, increased family instability and increased female labour market participation as factors not adequately addressed by “industrial social policies” targeted primarily on male breadwinners. Although the constituencies affected by new social risk generally have weak mobilizing potential (Bonoli, Reference Bonoli2005), policies directed at these risks, such as government financing of childcare and active labour market policies, nevertheless offer significant credit-claiming potential in an era where loss-imposition has been dominant and core constituencies of left parties are shrinking. Opportunities to address new social risks have been taken up to varying degrees by wealthy democracies of different social policy regime types, even after the onset of the Great Recession (Bonoli, Reference Bonoli2010; Morgan, Reference Morgan2013; White, Reference White2012; van Kersbergen et al., Reference van Kersbergen, Vis and Hemerijck2014). According to Huber and Stephens (Reference Stephens2015: 269), “The main innovations and the only welfare state expansions since the 1990 have occurred in activation and work/family reconciliation policies.” At least two other types of selective expansions are at least in theory possible, however: selective enrichment of benefits for existing clienteles (which may, in part, be to make them less likely to oppose retrenchment) and rollbacks of previously enacted retrenchment in response to political pressure.
Canada has shared in the social transformations that create the new social risks (Jenson, Reference Jenson2004). Evidence suggests that liberal welfare states have generally been later to adopt policies targeted at new social risks and devoted fewer resources to them (Bonoli, Reference Bonoli2007). While generally characterized as a liberal welfare state, Canada has always been an uneasy fit within that category. Thus, it is plausible to expect that initiatives to address new social risks will be on the Canadian agenda and that some will be enacted.
4. Self-reinforcing policy feedbacks
A fourth argument stressed by Pierson is that in understanding this new politics—and, in particular, program resilience—a critical role is played by self-reinforcing policy feedbacks and especially the large constituencies that develop around programs created decades ago. For Pierson (Reference Pierson1996), these groups are the product of welfare state development itself. During the postwar era, new social programs—and the enlargement of existing schemes—created genuine social constituencies containing strong interest groups and public expectations.
A developing literature suggests that welfare state programs may have self-undermining, as well as self-reinforcing, policy feedbacks (Jacobs and Weaver, Reference Jacobs and Kent Weaver2015), which may lead to endogenously derived policy reform and even substantial restructuring. Programs perceived to have uncontrollable deficits, for example, may mobilize powerful opponents in national finance ministries (Weaver, Reference Weaver2010). In a related vein, even in the absence of explicit retrenchment, “policy drift”—failure of policy to respond to changing socio-economic conditions—may expose citizens to increased risks and lead to calls for a policy response (Hacker, Reference Hacker2004).
In the Canadian context, the relative strength of self-reinforcing and self-undermining policy feedbacks varies across programs, depending on their maturity and cost trends as well as ideological divisions over programs. In pensions, while many conservatives would prefer a stronger role for private sector provision, the popularity of current federal programs provides a bulwark against retrenchment and a potential opening for program expansion under favourable electoral and fiscal conditions. Policy drift was an important factor in both pensions and childcare. A decline in private pension plan coverage and shift from defined benefit to defined contribution plans in the private sector led to increased concerns about seniors’ income security. Stagnating incomes and steady increases in female labour force participation amplified concerns over affordability of childcare. A combination of fiscal concerns and policy drift can be expected to help push policy reforms, including expansionary ones, onto the policy agenda.
5. Partisanship and labour power versus organized constituencies
Pierson (Reference Pierson1996: 144) argues that “because retrenchment is a distinctive process, it is unlikely to follow the same rules of development that operated during the long phase of welfare state expansion” (for a review of the literature on retrenchment, see Starke, Reference Starke2006). Pierson is particularly skeptical about the continued relevance of power resources theory in an era of permanent austerity. Power resources theory (Esping-Andersen, Reference Esping-Andersen1985; Hicks, Reference Hicks1999; Korpi, Reference Korpi1983; O'Connor and Olsen, Reference O'Connor and Olsen1998; Stephens, Reference Stephens1979) focuses on the capacity of the working class to use democratic means—notably labour parties and trade unions—to gain more power and improve labour conditions and social benefits.
According to Pierson (Reference Pierson1996: 151), the power resources approach fails to capture the dynamics of “the new politics of the welfare state” because it focuses too much on the labour movement at the expense of other constituencies, especially those that emerged as a consequence of welfare state development itself. However, numerous authors have argued that partisan/ideological differences in governing parties continue to make a difference in the form and degree of welfare state retrenchment, even if that effect is not as strong as in the period of welfare state expansion (Allan and Scruggs, Reference Allan and Scruggs2004), is mediated by the degree of partisan polarization in a political system (Finseraas and Vernby, Reference Finseraas and Vernby2011) and is likely to be stronger in societies with less generous welfare states than more generous ones (Starke et al., Reference Starke, Kaasch and van Hooren2014).
Canada features a mixed picture on the continued relevance of organized labour and partisanship. Canada has higher union density than the United States but lower than in many Western European countries. Canada's main social democratic party, the New Democratic party (NDP), and before it the Co-operative Commonwealth Federation (CCF), has relatively weak ties to organized labour. Although the CCF/NDP have never formed government at the federal level, they have done so in many provinces. They have influenced federal politics as opposition parties, especially in minority parliament situations when the larger Liberal party depended on NDP support to stay in power (Maioni, Reference Maioni1998; Olsen, Reference Olsen2002). The two left-of-centre parties are especially likely to engage in social program “bidding-up” when they are more evenly matched and competing for dominance, as was the case during the 2015 election.
Overall, these patterns suggest that political parties, the electoral system, partisan competition and the partisan composition of governments remain important in the development of the Canadian welfare state and that welfare state expansion will be concentrated in periods when the Liberals are in power, especially if the NDP has strong leverage over them.
6. Credit-claiming versus blame-avoiding
Pierson (Reference Pierson1996: 144) argued that “political goals of policymakers are different” in retrenchment-dominated politics—a politician's credit-claiming motivations and strategies were likely to be largely supplanted by blame-avoiding objectives and strategies—and that this was likely to militate against retrenchment (for a review, see Vis, Reference Vis.2016). However, governing parties that engage in retrenchment do not always pay an electoral price for doing so, and some political parties, particularly religious and market-focused parties, may actually be able to claim credit for retrenchment initiatives (Giger and Nelson, Reference Giger and Nelson2011, Reference Giger and Nelson2013). These findings lead to an expectation that in Canada and elsewhere, we should see (1) the left-of-centre parties seeking electoral advantage through credit-claiming opportunities by advocating for selective welfare state expansions when fiscal conditions allow and (2) conservative parties doing so only rarely and framing them in a way that allowed them to claim credit with their core constituencies.
7. Consensus versus unilateral reform packages
Finally, Pierson (Reference Pierson1996: 156) argues that given the unpopularity of welfare state retrenchment, “governments generally seek to negotiate consensus packages rather than to impose reforms unilaterally, which further diminishes the potential for radical reform.” Critiques of this argument have noted that opposition acquiescence rather than consensus has characterized many welfare states’ reforms, especially in the pension sector (Hering, Reference Hering2008; Bonoli, Reference Bonoli, Bonoli and Natali2012). Moreover, multiparty consensus is a policy-making model that is rarely found in Westminster parliamentary systems, where the parliamentary oppositions generally dispute and dissent.
A need for consensus to enact reform is especially likely to characterize political systems where multiple veto points, powerful trade unions and/or consensus traditions constrain unilateral decision making. They are therefore likely to hold unevenly over time and across programs for Canadian welfare state policy making. There are minimal veto points in purely federal-jurisdiction programs when single-party majority government exists, making unilateral action more possible. However, the Canada/Quebec Pension Plan is a classic example of a potential joint decision trap requiring consensus policy making because supermajority provincial assent is needed for changes—though it is consensus among governments, rather than social partners, that is required. In purely federal programs such as EI, there are strong veto points favouring the status quo only when a minority government faces an ideologically hostile majority. When a minority Liberal government is supported by a more ideologically outlying party, the NDP, pressure is likely to favour welfare state enrichment.
This review suggests that a conditional approach to “new politics” arguments can add important insights to understanding welfare state developments in the twenty-first century. Our analysis further suggests that many of the policy dynamics underlying the “new politics of the welfare state” are likely to appear in a somewhat muted form in Canada. Five factors that mediate the effects of relationships hypothesized by Pierson are especially important in understanding recent Canadian social policy developments.
First, regarding Pierson's “permanent austerity” argument, variations over time in the fiscal situations of the federal and (for shared-cost programs) provincial governments are likely to influence whether retrenchment or enrichment initiatives get on the agenda and whether those initiatives are enacted. Simultaneously, as we have seen, these conditions varied widely over time in Canada.
Second, Canada, like other advanced industrial countries, has faced and adapted to new social risks related to changing social and economic patterns. In Canada, in addition to factors identified in the international literature—such as labour market restructuring, diversification of family patterns, population aging, and more diverse life courses—immigration and Indigenous realities have also contributed to the emergence of new social risks that feed concrete calls for social policy reform (Jenson, Reference Jenson2004).
Third, heightened electoral partisan competition, especially in the context of minority parliaments, is likely to fuel social policy expansion as politicians seek to secure additional support from key electoral constituencies (Abou-Chadi and Immergut, Reference Abou-Chadi and Immergut2019; Savage, Reference Savage2019). The nature of the party system shapes party competition and its potential influence on social policy expansion. In the case of Canada, the influence of left-wing third parties nationally—the CCF historically and, more recently, the NDP—have altered the policy agenda and intensify pressures on the two dominant federal parties, especially the Liberals, to up the ante as far as welfare state development is concerned (Maioni, Reference Maioni1998).
Fourth, the degree and nature of partisan polarization is likely to shape the content of the short-term proposals that parties put forward. Parties adapt their policy proposals in response to the electoral incentives they face, but they also feature long-term commitments to certain principles such as income redistribution or traditional family values, which are likely to affect the proposals they include in their platforms. In the Canadian context, each of the two main federal parties, the Conservatives and Liberals, is likely to support policy proposals consistent with its partisan identity and need to appeal to its base.
These positions are not static over time, however. Particularly notable in this regard is the evolution of Canadian conservatism: the 1990s split in the Progressive Conservative party largely along regional lines and reunification of the right in 2003 under the Conservative party label meant that “the Canadian right was more clearly defined by an intermixing of or coalition between neoliberals who emphasize freeing the market from government interference, alongside populists and moral conservatives” (Farney and Rayside, Reference Farney and Rayside2013: 7). In the Harper governments (2006 to 2015), neoliberals and moral conservatives were ascendant.
Finally, political institutions directly influence electoral competition as well as the policy preferences and strategies of actors, a reality discussed extensively in the historical institutionalist literature, which focuses on both formal political institutions and existing policy legacies (Lecours, Reference Lecours2005; Orloff, Reference Orloff1993; Skocpol, Reference Skocpol1992; Tuohy, Reference Tuohy2018). In Canada, federalism and intergovernmental relations are particularly crucial institutional factors that have long influenced the development of both federal and provincial social policies (Banting, Reference Banting1987, Reference Banting, Obinger, Leibfried and Castles2005a; Maioni, Reference Maioni1998). During periods of majority government, governments can make changes quickly and decisively if they are willing to pay potential political costs; during periods of minority government, they may be susceptible to credible threats by other parties to enact policies they do not support or risk losing office.
These factors are closely related, and their interaction matters as it can help explain both retrenchment and expansion, which we define with respect to eligibility and benefits for individual benefit programs and in dollar terms for programs of intergovernmental transfers such as the Early Learning and Child Care (ELCC) programs. Child benefits as they interact with childcare, public pensions for older people, and EI were selected as case studies because they represent the major federal (or shared, in the case of the CPP/QPP) cash benefit programs. They also represent distinct demographic constituencies and different levels of program maturity that allow more nuanced testing of the Pierson framework and its relationship to the new social risk literature. Each of these programs experienced retrenchment in the 1980s and 1990s and significant, if selective, expansion in the last two decades. The analysis seeks to explain such decisions in relationship to the analytical framework and, more specifically, the five factors identified above and how they interact over time. To limit the length of the empirical sections, citations are predominantly to secondary sources, but the underlying research included extensive review of primary materials and interviews with policy-maker participants.
Child Benefits and Childcare Supports
Canada has a complex history regarding the distinct yet interconnected programs of child cash benefits and childcare supports in family policy. In the time period under study here, child benefits and childcare became politically linked and traded-off through sharp partisan preferences and electoral competition. On child benefits, the federal government legislated a universal program of family allowances in 1944, which was revised numerous times. It ultimately was combined with other programs and replaced by more targeted tax-based benefit programs in 1992 and again in 1997 (Blake, Reference Blake2009). Regarding childcare, the federal government provided support to the provinces on a shared-cost basis for childcare expenses of low-income families through the Canada Assistance Plan enacted in 1966 and terminated in 1996. However, efforts to create a Canada-wide framework repeatedly failed on the shoals of Quebec opposition to federal intrusion in a sector generally seen as under provincial jurisdiction; rising fiscal concerns; and, from 1984 to 1993, a Progressive Conservative government (Mahon, Reference Mahon2000; Brennan and Mahon, Reference Brennan and Mahon2011). Later in the 1990s, renewed interest in a National Children's Agenda, followed by an easing budget situation and political pressure to respond to the Quebec government's 1997 creation of a publicly subsidized childcare system (Jenson, Reference Jenson, Michel and Mahon2002), culminated in the minority Liberal federal government of Paul Martin (2004–2006) entering into ELCC agreements with all the provinces and territories (Mahon, Reference Mahon2009; Mahon and Brennan, Reference Mahon and Brennan2013; White, Reference White2017). This involved $4 billion of transfer payments over a five-year period to create 625,000 new spaces.
Reflecting increased polarization in Canadian politics, childcare and child cash benefits and the ELCC agreements became prominent flash points in the 2006 Canadian general election. Conservative party leader Stephen Harper ridiculed the Liberal government's agreements with the provinces as impractical and as bureaucratically restrictive, declaring that the Liberal plan was for “politicians to pay other politicians.” They criticized the Liberal plans as representing huge costly promises, strongly implying the public could not count on those promises coming to fruition.
Governing as a minority government and facing strong electoral competition from the Conservatives, Prime Minister Martin responded to Harper's platform by announcing that the Liberals would extend the ELCC agreements from 5 years to 10 years, adding a further $6 billion for a total investment of $10 billion, which presumably would create more than 1 million new childcare spaces by 2015. This move, which was fiscally feasible given the favourable budgetary position of the federal government, was intended as a sharp ideological demarcation with the Conservatives over the future of childcare and the role of the federal government in this sector.
The ELCC agreements permitted any government to end their participation with a year's notice, and when Harper's Conservatives came to power as a minority government in early 2006, they served notice that the ELCC agreements would be cancelled; a Child Care Spaces Initiative with a private sector focus was to take their place. In addition, the Canada Child Tax Benefit was to be folded into a new program, named the Universal Child Care Benefit (UCCB) (Canada, Reference Canada2011). UCCB payments were unconditional, fitting the conservative belief in “freedom of choice” that enables parents to “choose what's best for their children.” Introduced in July 2006 by the Harper government, the UCCB was the first new federal universal income program in more than forty years. Pierson (Reference Pierson1994: 101–2) suggests conservative governments prefer targeted and means-tested programs over universal ones. Conservative policy in this circumstance, however, should be seen as fitting with preferences of Harper's party and his government's agenda (Prince and Teghtsoonian, Reference Prince, Teghtsoonian and Bruce Doern2007) and the ascendance of neoliberal and moral conservative tendencies. Termination of the ELCC agreements and introduction of the UCCB marked a deliberate shift in resources away from provincial and territorial governments and public providers toward individuals and families directly from the federal government, and hence a credit-claiming opportunity for the federal governing party. Political institutions were no barrier to this swift act of reformation of childcare and child benefit policy at the federal level: while the Conservatives only had a minority government, the opposition Liberals supported the UCCB in the House of Commons in order to avoid bringing on an election they were likely to lose (Gidengil et al., Reference Gidengil, Blais, Everitt, Fournier and Nevitte2012: xxi).
The strategic context was different in the run-up to the October 2015 federal election, which featured three-way party competition, with both the NDP and the Liberals hoping to defeat the Conservatives and form a government. The Liberals campaigned on running deficits for the first three years of a government, with major spending initiatives on public infrastructure and social programs. The Harper government announced an increase to the UCCB benefit, prompting Liberal leader Justin Trudeau to signal his party's commitment to restructuring federal child benefit programs (Mahon, with Prince, Reference Mahon, Prince, Béland and P2019). “When it comes to child benefits,” Trudeau announced, “fair doesn't mean giving everyone the same thing, it means giving people what they need” (Bryden, Reference Bryden2015). Upon winning that election and forming a majority government, the Liberals introduced the new Canada Child Benefit (CCB) in their first budget, billing it as “the most significant social policy innovation in a generation” (Morneau, Reference Morneau2016: 57).
Reflecting their distinct partisan policy preferences, the Liberals adopted a generous and progressive selectivity for child benefits. The CCB targets benefits to low- and middle-income families while reducing or removing benefits previously provided to families with higher incomes (generally over $150,000). It is much more generous in benefit payments to eligible families than the previous mix of selective and universal programs. Whereas the UCCB was a taxable and non-indexed benefit, the CBB is both tax-free and indexed to the rate of inflation. Trudeau's CCB entails a modest narrowing of what a child benefit should be in eligibility and coverage along with a substantial deepening of its vertical redistributive effects. The CCB moved beyond traditional debates of selective versus universal social programming by recognizing the popularity of targeted universalism—nominally universal programs designed to be effectively progressive (Béland et al., Reference Béland, Marchildon and Prince2019).
Across Liberal, Conservative, and Liberal governments, early learning and childcare experienced an expansion of federal commitments, then retrenchment, and then again, a renewed federal role and expansion. In 2017 the Trudeau government announced the Multilateral Early Learning and Child Care Framework, which offered to increase federal investment by $1.7 billion over the next 10 years. Major investments in the 2021 federal budget for a Canada-wide early learning and childhood development system guarantee continued partisan (and likely intergovernmental) consideration of childcare supports in relationship to child cash benefits. That budget pledged new federal investments of $33.9 billion for early learning and childcare to 2025–2026 and then more than $8 billion ongoing. Over the same period, the CCB will transfer to families $82.7 billion and then at least $27 billion ongoing. The connection is made not only through partisan debates but also through policy analysis informed by intersectionality that fits together the interplay of cash benefits and childcare infrastructure in tackling inequities (Friendly et al., Reference Friendly, Feltham, Mohamed, Nguyen, Vickerson and Forer2020). Both address the personal financial costs of raising children; both are regarded as social investments in families, workforce participation and economic development; and both are presented by the Trudeau Liberals as enhancing opportunities and growing the middle class.
Overall, recent experience in these linked policy areas confirms the conditional nature of Pierson's arguments and the intertwined nature of their effects. Austerity pressures were strong drivers of policy under both Liberal and Conservative governments, but selective expansion was possible when these financial constraints eased. Both partisan competition and sharply divergent policy preferences between the two major political parties influenced policy choices. Policies that deviated from the preferences of new governing parties after a change in party control in Ottawa, such as the Martin government's ELCC agreements, were most likely to experience retrenchment, especially when those governments had a majority or believed that the opposition would not dare to bring them down. Nascent policies, and intergovernmental ones at that, such as ELCC programs, were especially susceptible to retrenchment but later reinvestment.
Public Pensions for Older People
Canada has a complex set of public benefits for senior citizens that include a quasi-universal Old Age Security program integrated with an income-tested program (Guaranteed Income Supplement and Allowance); the earnings-related Canada Pension Plan (and the Quebec Pension Plan for that province only); and a variety of tax credits. Taken together, this multilayered pension system is relatively affordable (Canada spends less than the OECD average on public pensions, at about 5 per cent of GDP) and quite effective in fighting poverty, in part because the Guaranteed Income Supplement targets low-income seniors. Yet, as in other liberal welfare states such as the United States, changes in private, voluntary and fiscally subsidized Registered Pension Plans (RPPs), especially the decline in coverage and a shift from defined benefit to defined contribution schemes, have transferred financial risks from employers to workers (Hacker, Reference Hacker2004). Although the debate over the inadequacies of RPPs emerged before the 2008 financial crisis, the additional problems it created for existing RPPs gave further impetus to a push to expand public pensions in order to compensate for the obvious income security limitations of private pensions and voluntary savings in Canada.
Not long after the 2008 financial crisis occurred, the NDP and the labour movement began advocating once again for the expansion of CPP, a program with a modest replacement rate of 25 per cent of covered income. The reason CPP expansion could remerge to the centre of the federal policy agenda after the 2008 financial crisis is that the program's long-term fiscal sustainability issues had been addressed a decade earlier, through a reform that combined indirect benefit cuts and, especially, an increase in the combined contribution rate from 5.6 per cent in 1996 to 9.9 per cent by 2003 (Little, Reference Little2008). Through modest retrenchment and substantial payroll tax increases, policy makers left an improved fiscal situation for CPP that made expansion conceivable. Yet, as suggested below, instead of being driven by the federal Department of Finance, as in the mid to late 1990s, the politics of CPP reform that emerged a decade later featured a much more central role for the provinces, especially Ontario.
The 2008 financial crisis and the Great Recession occurred in a context of minority parliament in which the Conservative government of Stephen Harper needed opposition support to stay in power. Because of this electoral and parliamentary context, the Conservatives explored the idea of CPP expansion in 2010 while still in a minority position in the House. In the end, however, Conservatives followed the traditional policy preferences of the party, which aligned with small-business opposition to increased payroll taxes needed to pay for higher CPP benefits. The NDP continued to advocate CPP expansion, with a call for a gradual doubling of CPP retirement benefits in their 2011 election campaign, while the Liberals endorsed only an undefined increase in benefits and the option of voluntary additional CPP contributions (Townson, Reference Townson2011). By 2015, however, the political situation had changed: Harper's government was very unpopular, and both the NDP and the Liberals were viable contenders to take power but only if the vote was not evenly split between them. In a context of heightened electoral competition, both parties promised a CPP expansion in 2015, and even the Conservatives half-heartedly pledged to allow individuals to make voluntary additional CPP contributions in exchange for higher retirement benefits—a position endorsed by its small-business constituency (Curry and Chase, Reference Curry and Chase2015; Flavelle, Reference Flavelle2015).
Federalism also played a central role in setting the table for CPP expansion in the first place, as reform requires support from at least two-thirds of the provinces representing at least two-thirds of the country's population. The intergovernmental dynamics leading to CPP expansion did not involve the usual competitive turf-claiming between Ottawa and Quebec; indeed Quebec, which had already been forced in 2011 to raise QPP tax rates above those for the CPP because of its less favourable demographic situation, was reluctant to simply copy CPP expansion. It was rather the Ontario Liberal government of Kathleen Wynne that passed legislation creating a mandatory, supplemental Ontario Retirement Pension Plan (ORPP) for workers who did not have a “comparable” workplace-based plan; it would go into effect starting in 2017 in the absence of federal action (Christensen, Reference Christensen2020). With the Trudeau government in office, all of the CPP-participating provinces agreed to a gradual expansion from 25 to 33 per cent of earnings (up to average earnings) financed by increased payroll contributions, a move that the Couillard government in Quebec eventually agreed to match (Béland and Weaver, Reference Béland and Kent Weaver2019).
While debate on CPP expansion took place during the Harper years, an attempt at retrenchment, rather than expansion, targeted Old Age Security and Guaranteed Income Supplement (OAS/GIS)—two programs that, in contrast with CPP, are financed through general revenues of the federal government rather than through dedicated payroll contributions. In 2012, shortly after winning a parliamentary majority, Prime Minister Harper announced the federal government would increase the age of eligibility for OAS/GIS from 65 to 67, with the shift occurring between 2023 and 2029. This measure generated swift and intense opposition from both the Liberals and the NDP, who later pledged to cancel the measure. In a context of strong electoral competition and the more progressive social policy preferences of these two centre-left parties, the pledge of reversing the Harper government's decision on eligibility age for OAS and GIS made its way to both the Liberal and the NDP platforms for the 2015 federal elections. Soon after forming government in late 2015, the Liberals cancelled the anticipated increase in eligibility age for OAS and GIS. Although it is clear that fiscal issues, electoral competition and distinct partisan preferences shaped the OAS/GIS eligibility age episode, the impact of federalism is much less obvious here, in part because the provinces have no veto point over the reform of these two programs, in contrast to the situation prevailing in the earnings-related CPP, which Ottawa cannot reform without first securing provincial approval.
Employment Insurance and Parental Leave
EI is another federal cash benefit program that, like CPP, witnessed both retrenchment and expansion in recent decades. While the traditional cash benefits for the unemployed are less generous now than they were in the 1970s, “special benefits” for other categories were added over the years and financed through EI contributions, which do more than simply finance unemployment benefits (Campeau, Reference Campeau2005). The fact that it is entirely within federal jurisdiction makes it a favoured vehicle for action by Ottawa: most notably, the program has since 1971 offered paid maternity (later parental) leave, and those benefits have been expanded in length and access over time (Canada Employment Insurance Commission, 2019; Mahon and Brennan, Reference Mahon and Brennan2013).
Starting in the mid to late 1970s, fiscal pressures, stemming from higher unemployment rates and a major benefit expansion that took place in 1971, led to efforts to control spending in unemployment benefits (Lin, Reference Lin1998). Activation became a major priority of both the Conservatives and the Liberals in the 1980s and 1990s, and successive governments made it harder for workers to qualify for unemployment benefits. In 1993–1994, income replacement rates were also lowered from 60 to 55 per cent (Campeau, Reference Campeau2005). Federal general funding for unemployment benefits that had existed alongside payroll contribution financing since the creation of the program in 1941 (Pal, Reference Pal1988) was phased out in the name of fiscal responsibility, a policy preference that both Conservatives and Liberals embraced over time. Thus, the program came to rely entirely on payroll contributions to finance both traditional unemployment benefits and the “special benefits” discussed below.
Finally, in 1996, the adoption of the Employment Insurance Act led to the penalization of repeat users, lower average replacement rates, tightened eligibility criteria and a reduction of the maximum period of eligibility (van den Berg et al., Reference van den Berg, Parent and Masi2004: 2). At the same time, this legislation implemented new “activation” policies financed through existing EI contributions (Campeau, Reference Campeau2005; Hale, Reference Hale1998). Although some of the retrenchment provisions enacted in 1996 were reversed, basic EI unemployment benefits have not witnessed any major expansion since the late 1990s, in sharp contrast with what has been witnessed in CPP and child benefit reform. This situation reflects converging partisan preferences over the need to promote activation in unemployment benefits (Béland and Myles, Reference Béland and Myles2012).
While federal unemployment benefits faced retrenchment during the 1980s and 1990s, the pattern for EI “special benefit” programs since the early 2000s has proved quite different. Parental benefits, for example, were substantially expanded under the Chrétien government in 2001 (Evans, Reference Evans2007). During both the Harper era and Justin Trudeau years to date, numerous changes were made to these programs, all in the direction of expansion (Canada Employment Insurance Commission, 2019). This expansion is the direct product of exacerbated partisan competition and a convergence of partisan policy preferences toward a pro-family agenda meant to address new social risks. Moreover, a much more gender equality-friendly version of parental leave in Quebec served as an example of policy diffusion encouraging advocates in the rest of Canada to push for change to the federal program (van den Berg et al., Reference van den Berg, Plante, Raiq, Proulx and Faustmann2017).
Expansion of EI special benefits has taken three forms. First, Conservative and Liberal federal governments each introduced a new EI special benefit: the benefit for parents of critically ill children in 2013 by Harper; the family caregiver benefit for adults in 2017 by Trudeau. Both initiatives indicate a common policy interest between the major federal political parties in one of the new social risks facing many voters by offering better income protection to working families while balancing care responsibilities.
Second, Conservative and Liberal governments both extended the duration of existing EI special benefits. Under Harper, in their final budget just prior to the 2015 general election, the compassionate care benefit was extended from six weeks to six months effective for 2016, heeding the recommendation by an all-party House of Commons committee and reflecting a distinctive form of partisan cooperation. Under Trudeau, in 2019 the parental sharing benefit was extended by five weeks or eight weeks depending on the length of time the benefit was being drawn, a fiscal commitment projected at $1.2 billion over five years (Canada, 2019).
Third, throughout the Harper and the Trudeau years, the main form of expansion to EI special benefits has concerned enhancing access to specific benefits by broadening eligibility criteria. These tend to be targeted and modest expansions in the name of improving the flexibility and responsiveness of EI to families. The most significant in the Harper period was making special benefits available to self-employed workers on a voluntary opt-in basis. Conservative preferences were also apparent in improving access to special benefits for military families, families caring for critically ill children, and families affected by a crime. Preferences of the Trudeau Liberals are apparent in expansively broadening the definition of family members eligible for certain special benefits and in framing certain reforms, such as the parental sharing benefit, as supporting the economic empowerment of women and “greater gender equity in home and in the workplace” (Morneau, Reference Morneau2018: 46). Overall, the case of EI special benefits suggests a mixture of distinctive and convergent policy preferences between the Conservative and Liberal governments, with fiscal resources generally available through the EI fund to enable a suite of targeted expansions.
Regarding institutional factors, federalism played no central role in EI reform except classical federalism in the sense of a clear exercise of exclusive jurisdiction (Banting, Reference Banting, Obinger, Leibfried and Castles2005a). Ottawa can unilaterally reform EI, which facilitates both retrenchment and expansion, depending on the partisan policy preferences and the levels of electoral competition at hand (Béland and Myles, Reference Béland and Myles2012).
Discussion
Canada's mixed experience with welfare state expansion in the first two decades of the twenty-first century suggests that core arguments in Pierson's analysis of the “new politics of the welfare state” continue to provide a powerful analytical framework for analyzing postwar welfare state development. Retrenchment pressures have remained intense but have been felt unevenly. Most programs have been resilient, but pensions much more so than Employment Insurance, reflecting differences in the strength and attentiveness of program clientele. Immature programs like the ELCC agreements are most vulnerable to major cutbacks, or even outright abolition. Indeed, the cases suggest that “ratchet effects” may be asymmetrical and conditional: changes toward more generous benefits are likely to require substantial effects to dislodge (1) if they are visible, which those embedded in the tax code are not to the same degree, (2) once people have developed substantial expectations based on them, (3) especially if blame cannot be passed onto to someone else—for example, provincial governments. Cutbacks, on the other hand, are likely to be continuously contested if they have organized clienteles whose grievances are primed by politicians.
The Canadian case also suggests that some qualifications and modifications are needed both in Pierson's framework and in research by later scholars that engages that framework. Most notably, selective welfare state expansion in Canada has taken all three forms noted in the first section: while some expansions (for example, parental leave and child benefits) are consistent with a coalitional and policy recalibration around new social risks, others have involved benefit expansion to existing clienteles (the 2016 CPP expansion) or rollbacks of previously enacted reforms (reversal of the OAS eligibility age increase). These patterns are consistent with Noël's (Reference Noël2020: 876) conclusion that income protection and social investment policies are not in a trade-off relationship with each other and “are explained by similar institutional, political and electoral dynamics.”
Overall, the relationships outlined by Pierson are heavily mediated by five factors—fiscal conditions, increased recognition of new social risks, partisan competition, electoral leverage for parties of the left, and political institutions—that vary across time, place and individual programs. Understanding these mediated relationships can deepen the capacity of the framework to explain variation in policy change across countries and sectors and over time.
Fiscal conditions are an especially important mediating factor. In the absence of favourable fiscal conditions, social policy expansion is unlikely to move onto the agenda except as relatively modest compensatory elements in a reform package that is primarily concerned with expenditure reduction. Fiscal conditions, however, vary across policy sectors, depending on the financing arrangements and institutional legacies—notably whether there is a dedicated social insurance fund.
Canadian evidence suggests that while “permanent austerity” remains a constant pressure in an era of slower economic growth, there is also a “politics of cyclical austerity,” in which governments that face deep fiscal pressures, such as Chrétien's government in the mid-1990s, are likely to push for deep budget cuts regardless of their partisan/ideological preferences. The later shift in Canada from what had seemed to be “permanent austerity” to selective expansion is not unique to Canada: as the Great Recession beginning in 2008 receded into memory, pension retrenchment enactments became less frequent in OECD countries, a number of countries backpedalled on previously enacted reforms, and selective enrichments occurred (OECD, 2019: 35–37).
The fiscal situation interacts with four other key factors in shaping both retrenchment and expansion: the recognition of new social risks; the level of electoral competition; the nature of partisan policy preferences; and the characteristics of specific political institutions, such as federalism and the electoral system. Conservative governments have been reluctant to expand a state role regardless of fiscal conditions, typified by the 2010 decision of the Harper government to not pursue CPP expansion, despite the pressures stemming from its minority situation and the intense electoral competition prevailing at the time.
With respect to new social risks and accompanying opportunities for coalitional change and policy responses bases on social investment strategies, Canada's record is, like its welfare state record generally, quite mixed (White, Reference White2012). The impact of new social risks was not independent and transformative of policy across all policy sectors; instead it was felt more in some areas than others: it was strong in EI 1990s cutbacks and EI special benefit expansion, contested along ideological lines in child benefits and childcare, and more of a background factor (through decline in workplace pensions) but still contested along ideological lines in pension cutbacks under Harper and CPP/QPP expansion under Justin Trudeau. Indeed, social investment is an aspect of policy and political strategy for Trudeau's Liberals, alongside compensatory policies. Canada under Trudeau shows that both social investment and compensatory policies can be pursued simultaneously as part of a broad yet moderate social policy agenda. Both types of policies can serve as political tools in terms of credit-claiming, and that is a critical consideration for most politicians.
The impact of new social risks was mediated by other factors, moreover. Federalism has played an especially important constraining role in some policy sectors. A shift toward more active labour market policies has been inhibited by struggles between Ottawa and the provinces over which level of government should control them. Expansion of support for childcare and early childhood education has been inhibited by the need to negotiate agreements with the provinces that delayed their entrenchment under the Martin government, making it more vulnerable to reversal after a change in government (White, Reference White2017: 215). Social conservative pressures under Conservative governments not to increase incentives for mothers of young children to work outside the home also increased barriers to initiatives in this sector. The fact that parental leave had to be financed through the awkward mechanism of EI to remain under federal jurisdiction reflects the enduring impact of Canadian jurisdictional arrangements on Canadian welfare state development, and in particular the difficulties of a robust and uniform national approach in social investment policies, most of which fall under provincial jurisdiction. These differences are exacerbated both by provincial state-building aspirations in the case of Quebec and by wide variations in the ideological preferences and fiscal capacities of provincial governments (Jacques, Reference Jacques2020).
Partisan competition has also played an important role. Two features stand out. One is the importance of campaign dynamics themselves. Almost all welfare state research on effects of political parties on welfare state policy making focuses on what happens after elections—for example, the share of cabinet seats held by parties of the left, centre and right (see, for example, Huber and Stephens, Reference Huber, Stephens, Baramendi, Häusserman, Kitschelt and Kreisi2015). Not enough attention is paid to what happens during election campaigns—for example, promises made, and whether competition has been muted through informal signalling and agreements not to compete on issues like rolling back retrenchment (Hering, Reference Hering2008). Second, the details of intra-party coalitions matter. In particular, the fact that social conservatives, who exhibit an enduring lack of enthusiasm for formal childcare (see Noël, Reference Noël2020: 877), have been a key constituency of the Conservative party in Canada in recent years has been an important cause of the shift in childcare and child benefit policies over time.
Other political institutions also played an important mediating role. Canada's single-member plurality parliamentary system meant that there were few veto points in purely federal programs, increasing opportunities for sharp policy shifts and even reversals. This occurred both with the OAS age of eligibility increase and the Conservatives’ UCCB.
Beyond the Canadian case, our analysis offers important lessons to students of comparative social policy about the conditions under which expansion remains possible in the context of the austerity politics of welfare states. Variations in the fiscal situation, the level of electoral competition, the degree and nature of partisan polarization, and the structure of political institutions and existing policy legacies, along with general trends such as increased awareness of new social risks, shaped the likelihood and form of selective welfare state expansion at specific points in time.
Future comparative research on contemporary welfare state expansion could explore the interaction among these factors to explain how and when expansion is likely to take place. This research is particularly essential in the aftermath of the ongoing COVID-19 crisis, which witnesses the adoption of large-scale yet temporary social programs as well as sudden variations in our five factors, especially the fiscal situation. This suggests that the aftermath of this unprecedented crisis is likely to offer many opportunities to assess the importance of, and interaction among, these factors in the ever-changing politics of social policy. Retrenchment is far from the only policy strategy in contemporary welfare states.
Acknowledgments
The authors thank Karen Anderson, Michel Fournier-Simard and the anonymous reviewers for their useful comments and suggestions.
