The politics of pensions in Belgium from 2014 to 2020: the inability to implement “structural” reforms

Following modernisation and industrialisation, the first public pension programs were introduced in Europe at the end of the 19th century. By the end of the Second World War, most Western European countries had introduced some old age protection schemes with either the objective of protecting workers via social insurance schemes or talking poverty in old-age with social assistance schemes. (World Bank 2014) The post-war Golden Age is then marked by a massive expansion of pension schemes representing the “largest extension in developed capitalist democracies of politically constructed rights”. (Pierson 2001a) These systems have subsequently faced several economic, labour market and demographic pressures in ‘post-industrial’ times. Moreover, single-pillar countries relying on large public pay-as-you-go schemes were gradually approached system maturation (World Bank 2014). From the 1990s onwards, reforms started to take place in single-pillar countries. Despite the complexity of reforming such systems, we can observe deep retrenchment parametric reforms being adopted as well as structural reforms mainly by the presence of multi-pillar systems in almost all countries today. Having developed a generous public earnings-related pay-as-you-go system during the Golden Age, the Belgian system also followed this gradual path of multi-pillarization and today has three (or even four) pillars. Moreover, whereas Belgium initially focused on both fiscal consolidation strategies and pre-funding the costs of demographic ageing in order to keep the system in place, as budget surpluses were not forthcoming, it then had to turn to labour market reforms before finally considering direct reforms of the pension system in recent years. (Bayenet et al. 2019, Vandenbroucke 2020) Despite this process of reform and budgetary restrictions that began in the early 1990s, the Belgian pension system faces major financial sustainability problems and a “nuanced” social performance. (Commission de réforme des pensions 2020-2040; 2014) Pension expenses represented 10.6% of GDP in 2019. Estimates are 12%, 13.5% and 13.3% for 2025, 2040 and 2070 respectively. The demographic dependency ratio is estimated to increase by 57% and the socio-economic dependency ratio by 33% by 2070 which is considerable. (Conseil supérieur des finances 2020) In order to better represent the cost of this, an adjustment by contributions alone would lead to a 20% increase by 2040 when an adjustment by pension benefits alone would lead to a 12% decrease by 2040. (Hindriks and Baurin 2019) The financial sustainability of the system is a major issue and is linked to its social sustainability – actually two sides of the same coin. Indeed, a performant pension system must be able to protect current pensioners against the risks of old-age, but also future generations. (Commission de réforme des pensions 2020-2040; 2014) In terms of social performance, we observe that Belgium ranks in the European average when it comes to the risk of poverty among older people, while the overall generosity of the pension benefits is rather low. (Conseil supérieur des finances 2020, European Commssion 2018a) However, these general observations also hide striking inequalities among the different pension regimes that exist in Belgium for civil servants, employees and the self-employed. (Commission de réforme des pensions 2020-2040; 2014) In the face of these substantial issues, the 2020-2040 Pension Reform Commission composed of 12 independent experts from different backgrounds submitted its report in 2014 and advocated for a new “structural and reliable social contract” including the introduction of a “points-based” pension system. More than six years later, however, we observe that no “structural” reform of the Belgian pension system has been achieved during the legislature of the Michel government, despite a commitment to this end, and that the De Croo government’s agreement of October 2020 mainly contains parametric expansionary pension reforms, raising serious doubts about the willingness and ability of the current government to implement far-reaching reforms during the current legislature. Despite “absolute emergency”, it thus seems difficult to implement “structural” pension reforms in Belgium. (Devolder and Hindriks 2019) The present thesis aims to provide an answer to the following overarching research question: how to explain the inability to implement “structural” reforms in Belgium from 2014 to 2020? To this end, we adopt a political science approach by analysing the politics of pension reforms. Despite the fact that pension issues are highly technical and extensively analysed by economists, decisions taken in the field of pension reforms are always the consequence of actors’ decisions, thus responding to political dynamics and they are not the direct consequence of impersonal macro-economic forces. Hence, the contribution of political science is key if we want to understand the decisions taken in this field. In order to analyse the politics of pensions for the period 2014-2020, we use a process-tracing approach. This allows us to trace the different sequences of the Belgian political process during this period by distinguishing the stages of electoral campaigns and elections, the stages of government formation and agreement and the policy-making process during the legislature. The influence of different actors is analysed, namely government coalitions, opposition parties, social partners and experts. This qualitative study was carried out mainly through the analysis of the different electoral programs, the media interventions of different actors, the available official reports and the history of all the events related to pensions during this period on French-speaking and Dutch-speaking public news websites. This thesis is articulated around four main chapters. The first chapter introduces the broad topic by presenting pensions in Europe in a comparative perspective. After an analytical and conceptual framework focusing on the different ways of financing, managing and calculating pensions as well as on the concepts of ‘pillars’ and ‘tiers’, the chapter compares the different current pension systems that co-exist in Europe using main figures. The generosity of the different systems is outlined as well as their capacity to reduce poverty and to maintain income in old-age. We then compare the policy trajectories of pension systems throughout the twentieth century, reviewing the Bismarckian and Beveridgean original models, their massive expansion during the Golden Age followed by times of crisis and reform. The end of the first chapter is then dedicated to the political economic theories of welfare states and pension reforms in times of expansion and retrenchment. Convinced that each theoretical approach in particular can shed light on precise dynamics but cannot enable us to understand a particular situation in all its complexity, we adopt a pluralistic theoretical approach. Institutional theories (political institutions and policy legacies), partisan politics, political competition and interest groups as well as the potential impact of experts and discourses developed in this last part of our first chapter proved to be essential in the development of our different expectations with respect to our research question. The second chapter dives into our case study namely Belgium. We first describe the current Belgium pension system and architecture based on three (or even four) pillars as well as the different reforms introduced since the 1900s. We focus in particular on the first pillar and the three existing pension regimes for employees, self-employed and civil servants. The eligibility requirements, funding methods and benefit formula in the three regimes are discussed. The current architecture of the system and its recent reforms allow us to understand the difficulties of implementing structural reforms adopting a historical-institutional approach. The second part of our second chapter focuses on the Belgian consociational institutions, fragmented party system and government formation stage in the decision-making process in order to give more insight on the potential role of partisan politics, political competition, interest groups as well as political institutions may have for our case study. The third and fourth chapters represent the core of the thesis including the empirical research dedicated to the politics of pensions in Belgium from 2014 to 2020. The third chapter introduces the fourth chapter by outlining the main empirical and theoretical puzzles, related main expectations developed from the welfare state literature, as well as methodology, whereas the fourth and final chapter presents the main research results using the different sequences of the Belgian political process identified for the period 2014-2020. During the two federal election campaigns of 2014 and 2019, the electoral programs of all parties allow us to describe the proposals of the different political parties on the subject of pensions and reveal the relevance of the left-right socio-economic cleavage as well as the linguistic divide in the analysis of the positions of the different parties on this subject. Secondly, the analysis of the government formation and agreement stage, in 2014 and 2019-2020, following the elections allows us to study the impact that the composition of the governing coalition may have on the resulting agreement. The types of parties in the coalition and their number as well as their fragmentation along the left-right and linguistic cleavages seem to be relevant dimensions for the types of pension reforms proposed in the government agreement. Finally, for the 2014-2018 legislature phase, a sequential analysis of the different events relating to pensions based on agreements at the Council of ministers, parliamentary debates, social concertation moments, trade union's actions and expert's interventions was carried out and allowed us to identify four explanatory factors related to the failure of the Michel government to implement a “structural” pension reform based on a point-based pension system. These factors are related to the strategy adopted by the government, the role of the opposition parties, the influence of the trade unions and finally the role of the experts and the report of the 2020-2040 Pension Reform Commission. Discourses and communication used by these different actors also seem to be decisive in their ability to convince and mobilise.