Ideas and Institutions

Since Simeon's article was published, the literature on ideas and public policy has expanded dramatically, especially since the early to mid-1990s.Footnote ³ One of the central claims of this literature is that ideas are not mere epiphenomena but can exert a direct impact on policy processes and outcomes (Campbell, Reference Campbell2004). Over the years, what has become clear is that ideas matter. What recent literature on ideas does is to explore how ideas matter (for example, Jacobs, Reference Jacobs2009; Mehta, Reference Mehta, Béland and Cox2011). Based on the ideational scholarship that has emerged since Simeon's Reference Simeon1976 essay, we can say that ideas matter because they define problems that move on and off of the policy agenda and help define policy solutions to address these problems, and because they provide political actors with broad ideologies and policy paradigms that help them map the world in which they live (Freeden, 2011; Mehta, Reference Mehta, Béland and Cox2011). Such ideologies and policy paradigms also participate in the construction of the perceived interests of these actors (Campbell, Reference Campbell2004; Hay, Reference Hay, Béland and H. Cox2011). Ideas can also help actors co-ordinate their actions (Schmidt, Reference Schmidt2002, Reference Schmidt, Béland and Cox2011), build coalitions (Béland and Cox, Reference Béland and Cox2016), become framing devices used by policy entrepreneurs (Kingdon, Reference Kingdon1995) to legitimize or oppose debated policy alternatives (Campbell, Reference Campbell2004), or even simply help construct the “need to reform” in the first place (Cox, Reference Cox2001). This is only a partial list of how ideas, which we simply define as causal beliefs (Béland and Cox, Reference Béland and Cox2011; Goldstein and Keohane, Reference Goldstein, Keohane, Goldstein and O. Keohane1993), can matter in politics, but it points to the fact that ideas constitute one of the most central forms of explanation in both policy and political research (Parsons, Reference Parsons2007).

As Craig Parsons (Reference Parsons2007) claims, however, it is possible and sometimes necessary to combine explanatory factors, especially when the time comes to address empirical puzzles that make simple, univocal causal arguments impossible. More generally, a number of scholars combine different types of explanatory factors as part of broader frameworks for the political analysis of public policy. A number of policy scholars have combined ideas and institutions to formulate more effective and inclusive analytical frameworks (for example, Béland and Lecours, Reference Béland, Lecours, Benz and Broschek2013; Campbell, Reference Campbell2004; Lecours, Reference Lecours2005; Lieberman, Reference Lieberman2002; Schmidt, Reference Schmidt2002; Schmidt, Reference Schmidt, Béland and Cox2011; Walsh, Reference Walsh2000; Weir, Reference Weir1992). An early example of the type of ideational scholarship that takes institutions seriously is the work of Peter Hall (Reference Hall1986) on the relationship between economic ideas and state institutions in France and the United Kingdom. Other examples of this approach are the work of Vivien Schmidt (Reference Schmidt2002, Reference Schmidt, Béland and Cox2011) on co-ordinative and communicative discourses and the research of John L. Campbell and Ove K. Pedersen (Reference Campbell and Pedersen2014) on the construction of expertise and how national institutions shape the production of such expertise and the creation of what they call “knowledge regimes.”

These three examples illustrate how scholars have combined attention to both ideas and institutions into a coherent framework for the political analysis of policy development. Agency remains a significant component of this framework and of our analysis of the politics of territorial redistributive policies, in which both constituent unit governments (in Canada) and political parties (in Belgium) play crucial roles. Civil society actors, who have featured prominently in the public policy literature in Canada at least since the mid-1990s (Inwood et al., Reference Inwood, Johns and O'Reilly2011: 7–9), are not overly significant in our two cases of policy change. Civil society actors in Canada seldom advocate change to the equalization programme while, in Belgium, trade unions and employers' organizations have played a role in supporting continuity in the territorial administration of social security rather than seeking its decentralization.

The public policy literature also points to the added value of comparative research for the analysis of the ideas-institutions nexus on the political study of public policy. This is consistent with Simeon's emphasis (Reference Simeon1976) on the need for comparative analysis, which is especially clear when scholars deal with institutional processes that vary greatly from one country to the next, in terms of both policy legacies and formal political institutions (Campbell, Reference Campbell2004).

Comparative research is particularly important to assess how specific ideas interact with specific institutions over time (Béland and Waddan, Reference Béland and Waddan2015). Political parties and the territorial structuring of the state are particularly important in exploring how institutions may impact ideas in the policy process. Party systems are likely to influence what types of policy ideas politicians consider either relevant or taboo. Because these systems shape the ways political actors think about their strategies, and even their preferences, cross-national differences in party arrangements may explain why the ideas featured in the policy discourse and agenda can vary so greatly between countries. The territorial structuring of the state also affects the ideational landscape. A two-level political scene, such as that in federal states, is particularly favourable to the presence of ideas linked to a multiplicity of political communities, which condition the diffusion and framing of public policy.

In the remainder of this article, we explore the combined impact of ideas and institutions on the politics of redistribution, a central policy issue for Simeon (Reference Simeon1976). We study federalism, a terrain familiar to Simeon, to address key analytical points he made in a fresh way, especially the role of ideas, the impact of institutions and the issue of redistribution. By adopting an analytical and comparative approach, we are also following Simeon's advice on how to move policy research forward in political science.

Federalism and Territorial Redistribution

Simeon's extensive scholarship on federalism is useful in understanding the politics of territorial redistribution, including how it relates to the role of ideas and institutions. Sensitive to both institutional architectures and federal societies (Simeon and Robinson, Reference Simeon and Robinson1990), Simeon's work presents broad frameworks for analyzing not so much how public policies are made but rather why they are formulated in the first place, why they are debated and challenged and why they either change or keep their original form for quite a long time.

Although Simeon recognized the existence of great diversity between federal systems (Simeon and Conway, Reference Simeon, Conway, Gagnon and Tully2001), he also saw federations as a unique form of institutional arrangement, one where self-rule and shared rule are the twin guiding ideas. Simeon was always concerned with finding and striking a proper balance between the institutional autonomy of constituent units and the power of the central government. Indeed, for Simeon, “building out,” where constituent unit governments formulate and implement their own public policies, thereby acting upon their own ideational preferences and expressing their own political identity, must be accompanied by some “building in” through institutions and public policies that can support the statewide political community.

From an institutional standpoint, Canada and especially Belgium are multinational states that have relatively few common spaces where state-level redistribution and solidarity can be embedded. In Canada, equalization supports such solidarity and fleshes out an important aspect of Canadian social citizenship: the notion that all Canadians should enjoy access to public services of comparable quality at comparable levels of taxation independent of their province of residence. The 1982 Constitution Act committed the Canadian government to make equalization payments. In Belgium, social security is the country's largest redistributive programme and, although it is primarily about inter-personal redistribution, it is widely viewed as the embodiment of inter-community solidarity insofar as Flanders is a net contributor while French-speaking areas are net recipients. In a deeply dichotomized federation, social security is the most important pan-Belgian social policy and one of the only tangible signs of the Belgian political community.

Equalization in Canada and social security in Belgium are important components of the shared-rule aspect of their respective federations. In fact, although federations are characterized by a constitutional division of powers that results in political autonomy for the constituent units (building out), they also require some building in to foster the political community the central government oversees. The self-rule, building out dimension of federalism corresponds to the exercise of political autonomy by constituent units and, sometimes, by their desire to increase that autonomy. The territorial pressures on redistributive programmes in Canada and Belgium stem from this self-rule dimension. Simeon (Reference Simeon1972) showed that, for the Canadian case, the constituent units of a federation can effectively pursue their own policy ideas and perceived interests. In Canada, provinces are political communities with their own political class that can engage in the definition of provincial interests and mobilize residents to defend and promote these interests. The situation in Belgium is quite different, but the gradual empowerment of regions and communities since the beginning of the federalization process in 1970 has led to challenges to the territorial redistribution embedded in social security.

Simeon's scholarship suggests that multinational federations are especially prone to tension (Simeon and Conway, Reference Simeon, Conway, Gagnon and Tully2001) because their dynamics go beyond constituent units asserting or promoting political autonomy; they involve governments that see themselves as “national” and typically enjoy strong support from the population, who wants to see their distinctiveness reflected in policy making. Tensions over public policy making in multinational federations are also due to resistance on the part of constituent units that tend to oppose policies they view as favouring, or pandering to, their “distinct” counterpart. Strains can be particularly severe in relation to redistributive programmes. In Canada, some of the tensions around the federal equalization programme stem from the idea that Quebec is the main beneficiary of the programme and receives disproportionate benefits.Footnote ⁴ In Belgium, the most nationalist Flemish political parties argue that Francophones use social security disproportionately and that they abuse it.

Yet, from an ideational standpoint, there is at least one major difference between Belgium and Canada that requires explanation. Ideas against existing forms of horizontal territorial redistribution are present in specific regions of both countries, but only in Belgium does the notion of significantly reducing the scope of this type of redistribution profoundly and enduringly shape policy debates. In this context, how do we explain this major disparity in political discourse between Belgium and Canada? As we argue, the answer to this question requires taking cross-national institutional differences seriously, especially with regards to the territorial nature of party systems. In fact, purely regional parties like the ones found in Belgium are much more likely to spread ideas against territorial redistribution, at least when they hail from a wealthier region, than national parties that look to gain votes all across the country, as is generally the case in Canada. Our analysis of the Canadian and Belgian cases backs this claim while illustrating the broader impact of federalism and political parties on territorial redistribution.

The Politics of Equalization in Canada

Canada's equalization programme, which is managed by the federal government and funded through general tax revenues, was created in 1957 (MacNevin, Reference MacNevin2004: 188–89). The programme dictates that the federal government will make payments to provinces whose fiscal capacity falls below an “equalization standard.” For 2014–2015, the federal equalization programme made payments worth a total of $16.7 billion to six provinces: Quebec ($9.3 billion); Ontario ($2 billion); Manitoba ($1.8 billion); New Brunswick ($1.7 billion); Nova Scotia ($1.6 billion), and Prince Edward Island ($360 million).

There are episodic intergovernmental tensions around equalization (Lecours and Béland, Reference Lecours and Béland2010). The close to zero-sum nature of equalization payments,Footnote ⁵ combined with the fact that all decisions are made by the federal executive, means that there is an inherent potential for these transfers to become embroiled in the politics of the federation. Canada's multinationalism is crucial to understanding equalization, both in terms of the origins of the programme and the particular pressure on the mechanisms of horizontal fiscal redistribution resulting from Québécois nationalism (Béland and Lecours, Reference Béland and Lecours2014).

The connection between national unity and the federal equalization programme emerged in the context of the postwar debate about the tax rental system (Courchene, Reference Courchene1984: 27–35). With this tax rental system, Ottawa took over many provincial taxes in exchange for fiscal transfers to the provinces. Some provinces, especially Quebec, voiced their dissatisfaction with the tax rental system. Quebec's opposition to the tax rental system was the product of French-Canadian nationalism focused on defending the province's autonomy (Balthazar, Reference Balthazar1986; Trudeau, Reference Trudeau1968). In the late 1940s and 1950s, the Union Nationale (UN) took several steps to bolster provincial autonomy, including opting out of the tax rental system. These events directly impacted discussions over the future of federal fiscal policy (Milne, Reference Milne, Boadway and Hobson1998: 190). Quebec's autonomist politics were instrumental in the 1957 creation of equalization as a horizontal fiscal redistribution programme that allowed the central government to craft territorial redistribution in a way that tied Quebec to the rest of the country. As Bryden wrote, the allocation of “equalization payments…would be a way of ending the isolation of Quebec” (Reference Bryden, Anastakis and Bryden2009: 81).

Equalization did not stunt the growth of Québécois nationalism. Still, there are reasons to think it may have mitigated the consequences of this nationalism for Canadian federalism, at least when it comes to the secessionist option, since equalization has provided explicit fiscal incentives for Quebeckers to remain Canadian citizens (Béland and Lecours, Reference Béland and Lecours2014). Financial and economic issues have always been a weak point for the Québécois secessionist movement. In the 1980 referendum, the “no” side made various financial and economic arguments suggesting that Quebeckers would be less well off after independence. In addition to references to economic isolation, this discourse emphasized the loss of fiscal transfers, especially equalization, which would come with independence. The “no” side's arguments remained fairly similar in the 1995 referendum; Quebeckers were encouraged to vote “no” in part so that they would not lose the fiscal support federalism provided.

More generally, Quebec politicians who oppose independence have often said that Canadian federalism is a worthwhile financial proposition for the province. Long-time Quebec Premier Robert Bourassa famously spoke of le fédéralisme rentable (profitable federalism) as a way to highlight the concrete benefits of staying in Canada. Equalization payments are a big part of this fédéralisme rentable. This type of defense for Canadian federalism has only grown in importance over the years.

Sovereignist politicians typically want to avoid any discussion of equalization. When pressed to speak about Quebec's relationship with the programme, these politicians will attempt to put equalization in a broader fiscal context (interview with Bloc québécois (BQ) member of Parliament, 2011). In the overall scheme of fiscal federalism, they suggest, Quebec does not gain but probably comes out a loser, in part because the federal government has invested more heavily in other provinces than in theirs. The dominant sovereignist view on equalization is that it simply returns a part of what Quebeckers have paid the federal government in taxes and other contributions. Sometimes, sovereignists make the bolder argument that Quebec loses a lot of money as a result of equalization and would therefore be better off independent (Parti québécois, 2016). The view that Quebec is a net fiscal loser in Canada has some credibility in the province. For example, in a 2009 public opinion poll, 31 per cent of Quebeckers said that they paid more income tax to the federal government than the federal government spent in the province, compared to 23 per cent who said the opposite (the remaining 46% either felt things evened out or did not know) (L'Idée Fédérale, 2009).Footnote ⁶ Still, equalization payments represent tangible benefits that are difficult for sovereignist politicians to minimize. In sum, there are reasons to think that equalization has accommodated Quebeckers not only by providing fiscal incentives for the province to remain within Canada, but also by working to keep many Quebeckers feeling they are part of the Canadian political community.

Another key source of tension around equalization involves a reaction against nationalist politics and ideas in Quebec. In certain provinces, particularly but not exclusively in Alberta, equalization is often depicted by the media and some politicians as a programme that unfairly benefits QuebecFootnote ⁷ . This resentment is linked to five arguments about Quebec and equalization (Béland and Lecours, Reference Béland and Lecours2014). The first argument is that Quebec receives too much equalization money. Quebec has been receiving equalization payments since the inception of the programme in 1957. Provincial politicians outside Quebec have often highlighted that fact. Already in 1971, for example, British Columbia Premier W.A.C. Bennett stated, “The Government of Canada has paid out over $5,500,000,000 in equalization payments since their introduction in 1957, and they continue to increase substantially each year. One province, Quebec, received 47 per cent of this amount.” (Bennett, cited in Resnick, Reference Resnick2000: 23)Footnote ⁸ Reporting on equalization outside Quebec also tends to emphasize that Quebec receives the “the lion's share” of equalization payments (Howlett and Carmichael, Reference Howlett and Carmichael2008). Yet, provinces like New Brunswick and Prince Edward Island rely more on equalization payments as shown by per capita statistics (Perry, Reference Perry1997: 170). The media's tendency not to speak in per capita terms feeds resentment towards Quebec and the programme as a whole, especially in a wealthy province like Alberta, which has not received equalization payments since the early 1960s (Courchene, Reference Courchene1984).

The second argument is that equalization payments are a political tool to accommodate Quebec. As we have seen, there are good reasons to think this, if only because equalization provides Quebeckers with a material disincentive to leave the federation. The structural feature of the programme that leads to cynicism being generated about equalization and Quebec in the rest of the country is the management role of the federal government. Contrary to other federations, decisions on equalization payments in Canada are at the discretion of the federal executive. This allows for the politicization of equalization (on this issue, see Béland and Lecours, Reference Béland and Lecours2011; Lecours and Béland, Reference Lecours and Béland2010) or, at the very least, the perception that decisions on equalization obey some sort of political logic. Although public opinion data specifically about equalization are very limited, one could think that the notion of a programme whose primary stated objective is to help poorer provinces deliver quality public services serves the politics of nationalist accommodation could be controversial. The fact that independence remains an option in Quebec and that Quebec governments have put forward various claims for self-determination despite the perceived financial benefits of federation adds to the frustration. Former Alberta finance minister Ted Morton (Reference Morton2005: 3) once claimed that “Alberta's fate appears to be the opposite of Quebec's: the more it contributes financially, the less it receives politically.”Footnote ⁹ In other words, Quebec's perceived political clout in the context of equalization payments brings resentment towards both Quebec and the programme.

The third argument is that, through equalization, wealthier provinces such as Alberta and Ontario have been financing Quebec's progressive social policies (Holle, Reference Holle2012; Milke and McMahon, Reference Milke and McMahon2012). While there is significant policy divergence across the provinces, the cleavage between Quebec and the other provinces stands out. For example, social policy choices in Quebec have led to a publicly funded daycare system, a public drug insurance programme and the second lowest university tuition fees in the country. Perhaps even more important politically, Quebec leaders and many pundits hail these policies as better than those elsewhere in the country. They are also said to reflect a society that is more progressive than in any other province and where the values of generosity and compassion are more present than in the rest of Canada (Béland and Lecours, Reference Béland and Lecours2008). These much vaunted social policies are met with some confusion in the rest of Canada. The notion that Quebec can “afford” these policies seems impossible. In this context, the workings of fiscal federalism, and more specifically the equalization programme, are used to account for why Quebec offers several costly social programme politicians in other provinces say they cannot afford. Thus, the seemingly dominant notion in English-speaking Canada is that non-recipient provinces such as Alberta and British Columbia “help fund Quebec's lavish social programs” (Milke, Reference Milke2012). This notion is fed by newspaper commentaries and conservative think tanks (Eisen and Milke, Reference Eisen and Milke2010) and is seldom challenged. For example, in 2012, when Quebec university students began a massive mobilization to oppose a tuition hike, commentators in Alberta suggested that Quebec students needed a lesson in equalization so they understood that Albertans were paying for their modest tuition fees (Corbella, Reference Corbella2012).

The fourth argument is that while equalization unfairly favours Quebeckers, it is also inefficient because it promotes economic dependency in Quebec and other poorer provinces (Holle, Reference Holle2012). When it comes to Quebec, arguments about inefficiency and unfairness often come back to the management of its political distinctiveness within the Canadian federation. For example:

Finally, as this last sentence points to, equalization payments are often viewed as a way to “buy” federalist votes. For instance, as stated in the Toronto Star: “Especially in his first mandate, Harper seemed intent on gaining the favour of Quebecers. For instance, he recalibrated the equalization formula, giving Quebec billions of dollars more” (Chung, Reference Chung2008: A19). Sometimes, the view is that this equalization money is used to pursue the objective of independence: “Quebec uses its equalization windfall, not to improve essential services but to fund its secessionist agenda. Hence, Ontario and Alberta are being forced to subsidized [sic] Quebec's secessionist movement. Another reason to scrap the equalization program” (Sauve, Reference Sauve2004: A15). Canada's key programme of horizontal fiscal redistribution is subject to a particularly critical viewpoint resulting from nationalist ideas coming out of Quebec.

Yet these criticisms have not significantly permeated electoral politics in Canada. Whatever members of Parliament (MPs) and voters may think about equalization (and its relationship to Quebec), political parties have long believed that forming a government was not a realistic possibility without winning several seats in Quebec (something partially disproven by the last Stephen Harper Conservative government). Alienating Quebeckers by adopting a position critical of equalization and bringing up this topic during an election campaign is therefore seen as a dangerous proposition.Footnote ¹⁰ Perhaps more importantly, federal politicians making critical comments towards Quebec in relation to equalization could have been viewed as endangering national unity. The existence in Canada of national parties that not only need some electoral support in Quebec, but also seek to play a “brokerage” role with respect to the territorial and linguistic cleavage in the country, imposes restraint on the political class in their criticisms of equalization.Footnote ¹¹ In fact, this self-imposed restraint of the traditional political parties has meant there is very little discussion about territorial redistribution, let alone equalization specifically, in federal politics.Footnote ¹²

The Politics of Social Insurance in Belgium

The Flemish political class has shown no similar restraint with respect to social security, at least not since Belgian political parties split upon linguistic lines in the late 1960s. This split has its origins in the development of the Belgian state. Belgium was created in 1830 as a centralized unitary state. Although sociologically multilingual with its French-speaking and Dutch-speaking populations, the early Belgian state was dominated by Francophones and functioned almost exclusively in French. Powerful decentralizing pressures emerged over time. These pressures originated mainly in the mobilization of Flemish nationalism (Wils, Reference Wils1996). Political upheavals after the Second World War favoured a series of institutional changes, including the splitting of political parties changes. By 1993, Belgium had evolved from a centralized unitary state to a federal system.

The splitting of all political parties along linguistic lines facilitated decentralization. As a result of the disappearance of national parties, two distinct political scenes and publics developed in Belgium: the Flemish, where decentralization is viewed as positive and the idea of the nation is linked to Flanders as much, if not more, than it is to Belgium; and the French speaking, where decentralization is generally viewed with suspicion and the idea of the nation refers to Belgium. Social insurance splitting becomes a major issue in this institutional and partisan context. Most Flemish parties favour splitting at least portions of social security. They point to the unfairness of a system that transfers fiscal resources from Flanders to Wallonia, the French-speaking region that used to be the country's economic engine. Because of higher unemployment rates in Wallonia, for instance, social insurance schemes in the region collect less in social insurance contributions on average and pay more in social benefits than in Flanders. Flemish parties, especially the nationalist Nieuw-Vlaams Alliantie (N-VA), but also the Christian Democrats and the Liberals, bring up this issue during election campaigns. The aggressive position of N-VA on splitting social insurance puts great pressure on other Flemish parties to also “stand up” for Flanders, since they all court only Flemish voters.

Although the size of transfers between Flanders and Wallonia is both contested and hard to assess, the push to decentralize elements of the country's social security system is a significant aspect of Flemish nationalism's quest for further reforms to the state or, in the case of the far-right and separatist party Vlaams Belang (formerly Vlaams Blok) and even N-VA, outright independence. Most importantly, the debate about territorial social insurance transfers in Belgium is generally grounded in nationalist ideas that shape the discourse on both sides of the language divide.

On one side of this divide, Flemish nationalist writers and politicians have made numerous statements about transfers that highlight their attempt to frame the boundaries of solidarity in purely nationalistic ways, thereby symbolically excluding Francophones from their political and social policy community. From their perspective, the problem with the social security system is not so much the existence of territorial solidarity but the fact that this solidarity transcends language and national boundaries. In other words, the problem with this territorially centralized system is that it allows money from payroll contributions to flow between different regions. This is important because Flemish nationalists do not speak about internal economic and social policy disparities within Flanders; redistribution within the region is legitimate. So the problem for Flemish nationalists is only one type of territorial redistribution—where the Flemish people allegedly subsidize Walloons in key social security areas (Béland and Lecours, Reference Béland and Lecours2008).

The nationalist N-VA, the strongest party in Flanders in the 2010 and 2014 federal elections, is a key player in this discussion. In the campaign before the May 2014 elections, the N-VA clearly stated its preference to decentralize social security so that the Flemish and Francophone communities could each develop its own social policy model (for example, see the interview with Ben Weyts in Coppi, Reference Coppi and Wyets2014). Although most Flemish parties seek the decentralization of selected components of social policy, N-VA's plan involves the wholesale splitting of social insurance, including health care, pensions and unemployment insurance (Mouton, Reference Mouton2014). Its discourse in not unlike that of the more radical (but now more marginal) Vlaams Belang; that party's literature declares that Francophones and Flemings belong to two different nations (Vlaams Blok, Reference Blok2003: 1–5). As a consequence, Flemish solidarity should not extend to Francophones; instead, Flemings should treat them as they treat foreigners. There is, therefore, little solidarity between the two groups and the “rivers of money” flowing from Flanders to Wallonia, as they are described, are unfair and illegitimate, especially because Wallonia is said to be an inferior region plagued with a culture of dependence and chronic government waste (Vlaams Blok, Reference Blok2003: 7). This discourse about redistribution and territorial solidarity relies on strong cultural and national stereotypes: Flemings are described as autonomous and hardworkingFootnote ¹³ while Walloons are viewed as lazy and dependent (Béland and Lecours, Reference Béland and Lecours2008).

Although the Vlaams Belang's discourse on social security is particularly radical, attacks on the very idea of a Belgian solidarity that could transcend the language divide are also present in the N-VA (which is now a partner in the Belgian federal government), as well as in the Flemish Liberals and Christian Democrats (Poirier and Vansteenkiste, Reference Poirier and Vansteenkiste2000). The delegitimization of transfers inherent to Flemish nationalist discourse puts great pressure on social security's territorial structuring. Moreover, Flemish parties tend to justify the need to split social security based on the idea of a different “societal consensus” in Flanders (right of centre) and French-speaking Belgium (left of centre). Therefore, these parties are at once looking for a different type of social security, as well as a different territorial management of the system, all in the name of Flemish distinctiveness.

French-speaking writers and politicians, on the other hand, have formulated a counter-discourse centred on the idea of a pan-Belgian solidarity that transcends the linguistic divide (Poirier and Vansteenkiste, Reference Poirier and Vansteenkiste2000). This idea is especially powerful among French speakers since most Francophones oppose further decentralization, which they associate with the disintegration of the country. Claiming that Flemish nationalists exaggerate the scope of inter-regional transfers, some of these actors also argue that socio-economic variables like employment levels, rather than waste and cultural dependency, explain the patterns of the transfers. Simultaneously, country-wide social partners such as labour unions and the main business organizations have long opposed social security decentralization (Bouteca et al., Reference Bouteca, Devos and Mus2013).

In the field of social insurance but outside of social security, a meaningful policy change that occurred in Belgium in recent decades is the implementation of a care insurance scheme in Flanders (Jorens, Reference Jorens2006). A significant policy issue in the 1990s, care insurance was debated in federal parliament but, due to a lack of political consensus, no legislation was passed. Partly because they wanted to use the issue as a symbol for Flanders' increasing policy autonomy, Flemish politicians ended up supporting the adoption of a care insurance programme exclusively based in their region, without any equivalent in Wallonia, which could not afford to create a programme on its own (Béland and Lecours, Reference Béland and Lecours2008). Francophone experts and politicians viewed the decision to go ahead and implement a care insurance scheme in Flanders alone as an attack on Belgian solidarity. From their perspective, the creation of the Flemish programme constituted a clear step towards the end of a unified welfare state in Belgium (Jorens, Reference Jorens2006). A clear example of policy layering (Thelen, Reference Thelen2004), the addition of a Flemish disability insurance programme alongside the federal social security system was viewed in French-speaking Belgium as a nationalist move against territorial, inter-regional redistribution and solidarity in Belgium.

At the federal level, despite the existence of civil society organizations opposed to the decentralization of social security, this issue has appeared on the political and policy agenda during almost every negotiation over government formation. This was most spectacularly the case in the negotiations following the 2010 elections that lasted 541 days (Reman and Feltesse, Reference Reman, Feltesse, Germain and Robaye2011; Tissot, Reference Tissot2011). These negotiations led to the advent of a Sixth State Reform that decentralizes family allowances, among other provisions (Dumont, Reference Dumont2015; Goossens and Cannoot, Reference Goossens and Cannoot2015; Popelier and Cantillon, Reference Popelier and Cantillon2013; Reuchamps, Reference Reuchamps, Basaguren and San Epifanio2013). Because family allowances are formally part of social security, it means that the agreement among political parties over government formation created a window of opportunity for the first direct decentralization of social security in Belgium. Although the other components of social security remain centralized, the question of their possible decentralization is likely to return to the agenda as Flemish parties keep raising the issue.

Overall, this case study clearly backs our main claim that, in a multinational state like Belgium, the politics of redistribution comes under intense pressure from nationalist ideas and that these ideas prove particularly potent and politicized in the context of a party system where regionally based parties dominate. In other words, as is the case of equalization in Canada, the politics of redistribution in Belgium are shaped by a combination of nationalist mobilization and institutional forces in the context of enduring interregional fiscal and economic disparities. As our analysis suggests, unlike its Canadian counterpart, the Belgian party system allows territorial redistribution to remain a key issue on the federal political agenda, a reality that has recently facilitated significant changes to the social security system through the adoption of the Sixth State Reform, whose implementation is an ongoing process. Considering the purely regional nature of Belgian political parties, it is likely that the issue of social security decentralization will remain on the federal agenda for years to come.