Scholarly interest has grown in understanding the changing relationship between party politics and the welfare state. Traditional ‘parties matter’ theories established a durable programmatic link between party families and welfare reform direction. However, clear left–right differences in welfare state politics during the post-war decades have blurred alongside the transition to a post-industrial society. In particular, centre-right parties have shifted from promoting male-breadwinner family models to supporting mothers' employment and incentivizing men to participate in domestic care work (Morgan Reference Morgan2013). Parental leave reform has garnered significant attention for its direct impact on gendered work–family arrangements (Blum Reference Blum2010).
New theoretical perspectives have emerged to reassess the impact of parties in government on policy output. In the case of family policy, research has focused on the motivations of centre-right parties to deviate from their historical commitments. Namely, studies have examined the political and economic factors that may influence the party in charge of the ministry developing the policy in question to purposefully introduce the reform. For instance, it has become commonplace that the German Christian Democratic Union (CDU) consciously altered its family policy repertoire as a strategy to attract female and urban voters (Fleckenstein Reference Fleckenstein2011). This focus is understandable, as parties are held accountable for the policies implemented under their administration and they claim credit for them if they succeed. Nevertheless, the approach risks disregarding the complex interactions and the various stakeholders involved in policymaking.
By updating insights from the institutionalist literature, this article complements existing research by reassessing the agency of the party controlling the responsible ministry and reviewing the influence of partisan veto players. We focus on contexts of coalition and minority governments, where ‘veto points’ at the executive and legislative levels turn policymaking into a complex process of negotiations between diverse parties (Immergut Reference Immergut1990). However, classic institutionalist accounts consider that, in such contexts, ministers hold agenda-setting power while coalition partners or other legislators can merely veto (Tsebelis Reference Tsebelis2002). We take this idea further by arguing that parties other than that of the responsible minister may also set the agenda for a reform, design it and exert pressure to implement it.
We theorize three mechanisms through which parties can introduce reforms in ministries led by other parties. Relying on the literature on minority and coalition governments, we argue that partisan veto players may impose their policy priorities in exchange for enabling government formation and ensuring government survival (Strøm Reference Strøm1990). Further, we argue that parties may establish ministerial resources to influence policies of future cabinets, something we conceptualize as ‘bureaucratic continuity’. By tracing the different strategies of partisan veto players to influence policy, we explore the reasons why the party of the minister in charge may not be the ‘protagonist’ of the reform, fully responsible for its design and intended purpose, but rather a ‘consenter’ bending to external pressure (Korpi Reference Korpi2006).
Empirically, we apply deductive process tracing to test whether the theorized mechanisms can explain the adoption of two progressive parental leave reforms approved by centre-right parties in familialistic countries. Parental leave policy regulates the entitlement, access and payment for each parent to be away from the labour market for childrearing, directly influencing the gender distribution of paid and unpaid work and care activities. Hence, while other family policy instruments such as childcare expansion open opportunities for dual-earner households without negatively affecting single-earner families, changes in parental leave may favour some family models over others (Leitner Reference Leitner2003). The reforms analysed in this article benefit dual-earner households at the cost of traditional male-breadwinner families. In Germany, the CDU adopted a path-breaking parental allowance reform in 2006. It transformed a long and low-paid leave mainly taken by women into a shorter leave, serving as an income substitute with a quota for each partner. In Spain, the Popular Party (PP) expanded paternity leave from 13 days to five weeks between 2017 and 2018, rather than expanding maternity or transferable parental leave. While these reforms have garnered much attention from scholars, previous studies concentrated primarily on the potential motivations of the CDU and PP in approving the reforms, as part of broader processes of ideational or strategic change. Complementing existing research, this article inquires if the CDU and PP were protagonists or rather consenters of the respective 2006 and 2017 parental leave reforms. We provide novel empirical data which suggest that coalition and support parties designed these reforms and used various mechanisms to influence centre-right administrations to approve them.
By underscoring the relevance of interparty dynamics and bureaucratic continuity, these findings pave the way for further research on why policy output may deviate from the historical trajectory of the party formally responsible for a policy. More broadly, the correct attribution of reforms has important implications for our understanding of the evolution of the welfare state, the interactions between different parties across diverse institutional settings, and the validity of theories on the partisan politics of the welfare state.
Existing explanations for the centre-right as a protagonist of progressive leave reform
In line with their traditional ideology, Christian-democratic and conservative parties have historically shaped ‘familialistic’ welfare states, where a lack of childcare services, generous child benefits and long parental leaves support single-earner families and hamper work–life reconciliation for working mothers (Lewis Reference Lewis1992; van Kersbergen Reference van Kersbergen1995). Amid widespread adoption of ‘defamilializing’ policies providing care services in support of working mothers, conservative governments have sought to defend policy options for families to ‘choose’ traditional models (Morel Reference Morel2007). This has led to what is known as an ‘optional’ version of familialism (Leitner Reference Leitner2003).
In contrast, centre-right parties have recently performed a radical transformation by reducing options for male-breadwinner families and promoting female employment and male caregiving roles. While expanding childcare infrastructure supports working mothers without causing direct negative consequences to single-earner households, the case of parental leave reform stands out for its potential to redefine gender roles by constraining families' choices concerning work and care, either by curtailing mothers' time away from the labour market or by reserving leave time for use by the father exclusively. Therefore, progressive leave reforms by centre-right governments pose a challenge to ‘parties matter’ theories that predict consistent differences in policy output across party families, sparking a diverse academic discussion (Häusermann et al. Reference Häusermann, Picot and Geering2013). These can be classified into two theoretical perspectives.
On the one hand, some scholars have focused on the socioeconomic changes associated with post-industrialization. It has been argued that the growing share of female human capital in the private workforce, together with the increasing relevance for governments of raising employment levels, would make employment-oriented family policy attractive to governments and employers (Bonoli Reference Bonoli2005; Fleckenstein et al. Reference Fleckenstein, Saunders and Seeleib-Kaiser2011). International organizations and the EU may also recommend these policies to governments as part of a social investment recipe for welfare state recalibration (Hemerijck Reference Hemerijck2015). For instance, Martin Seeleib-Kaiser (Reference Seeleib-Kaiser2016: 234) argued that socioeconomic transformations in Germany led employers to develop new ideas about work and welfare which encouraged modernizers in the CDU to push for the 2006 parental allowance reform.
On the other hand, a burgeoning strand of literature has argued that these policies are being instrumentalized for party competition. Given socioeconomic and cultural transformations such as growing female employment, changing gender norms and secularization, a growing share of the electorate may favour progressive family policy (Emmenegger and Manow Reference Emmenegger and Manow2014; Schwander Reference Schwander, Manow, Palier and Schwander2018). Several studies have argued that the pursuit of votes in a more progressive, feminized and young electorate explains the German 2006 parental allowance reform (Fleckenstein and Lee Reference Fleckenstein and Lee2014; Morgan Reference Morgan2013) and the Spanish 2017 paternity leave expansion (León et al. Reference León, Pavolini, Miró and Sorrenti2019). Other case studies have also addressed intra-party conflict, tracing how modernizers within the CDU gradually changed the conceptualization of the family inside the party until progressive family policies were introduced during the first Angela Merkel government (Blum Reference Blum2010; Fleckenstein Reference Fleckenstein2011; Morgan Reference Morgan2013; Seeleib-Kaiser Reference Seeleib-Kaiser2010). Moreover, case studies have also recorded how coalition partners developed policy concepts later adopted purposefully by conservative administrations (Blum Reference Blum2010; Leitner Reference Leitner2010; Seeleib-Kaiser Reference Seeleib-Kaiser2010).
Despite resting on varied approaches, existing research focuses on the potential motivations of centre-right parties to introduce progressive leave reforms which they had previously rejected. However, the strong focus on change within the party responsible for the relevant ministry risks overlooking important drivers of reform (Alvariño Reference Alvariño2024). In particular, when policymaking takes place within minority and coalition governments, the relationship between parties and their policy output is mediated by interparty negotiations and coalition dynamics. This article offers a complementary explanation for these unexpected reforms by systematically examining the influence of the relatively more progressive coalition and support partners of the centre-right CDU and PP in two relevant parental leave reforms. This extends the validity of ‘parties matter’ ideas to institutional contexts where political power is less concentrated.
From antagonists to consenters: Explaining unexpected reforms through institutional mechanisms of influence
By focusing on the reasons leading centre-right parties to adopt progressive family policies in their ministries, existing research has conceived these parties as ‘protagonists’ of the reforms. By contrast, in this study we explore the mechanisms through which progressive parties may have influenced the centre-right to ‘consent’ to these policies. According to Walter Korpi's conceptual distinction (Reference Korpi2006: 181–182), reform protagonists are those who, following their preferences, initiate, design and exert pressure for the adoption of a reform. Meanwhile, antagonists are those who oppose them, and consenters are those who concede to reforms proposed by protagonists.
Korpi (Reference Korpi2006) used these concepts to argue that employers may switch from antagonists to consenters of social policies due to the agenda-setting power of progressive actors, and not because they had developed a first-order preference for them. In a similar vein, we argue that centre-right parties may become consenters to progressive family policies proposed by centre-left parties through diverse mechanisms of institutional influence. The distinction between protagonists and consenters is crucial, as it clarifies actors' preferences and their role in the process of welfare state reform, altering our understanding of the causal mechanisms at work and the validity of theories about welfare state change. Although parties, our unit of analysis, are complex organizations with a certain level of ideological heterogeneity, they are characterized by formal processes that determine joint positions later expressed by documents (e.g. manifestos) and representative agents (such as ministers or spokespersons).
To study party interaction in the executive and legislative arenas, we update insights from institutionalist literature. An essential idea from this research strand is that veto points enable actors to prevent policies proposed by the agenda-setter, creating the need to negotiate (Immergut Reference Immergut1990). Although institutional fragmentation can refer to diverse layers of governance, we apply this idea to minority and coalition governments, where the party in formal control of a ministry may be influenced by coalition partners or external support parties, here referred to as ‘partisan veto players’ (Strøm Reference Strøm1990; Tsebelis Reference Tsebelis2002). While classic accounts have concentrated on how veto players may block or allow policies proposed by the agenda-setter, usually a minister, we argue that partisan veto players may also act as protagonists, initiating change as agenda-setters designing a reform, and exerting pressure for its approval. We theorize three institutional mechanisms through which partisan veto players may influence a government's reform agenda.
First, building on Ellen Immergut (Reference Immergut1990), we highlight government formation as a veto point for both coalition and minority governments. Parties that want to lead either type of government need the support of other parties. In the case of coalition governments, parties reach coalition agreements on several issues, constraining the later activity of cabinet ministries (Indridason and Kristinsson Reference Indridason and Kristinsson2013). Therefore, a minister may be required to implement policies against his or her ideology (Moury Reference Moury2011). Moreover, research shows that coalition agreements focus on issues where the partners have ideological differences, consequently becoming a tool for partners to constrain the activity of ministers from other parties (Krauss and Kluever Reference Krauss and Kluever2022). On the other hand, minority governments may acquire external support by reaching an informal compromise. However, they usually rely on a formal ‘support agreement’, which Kaare Strøm (Reference Strøm1984) defines as being explicit, comprehensive, long term and negotiated before government formation (Bale and Bergman Reference Bale and Bergman2006).
Second, after coalition or minority governments have been established, coalition and support parties have the power to determine government survival, which gives them leverage to influence policies. Minority governments must constantly seek the support of the parliament, facing the threat of a motion of censure (Müller Reference Müller2022). Support parties can use these opportunities to influence policies in exchange for sustaining the government. Coalition governments not only confront parliamentary pressure, but they can additionally be challenged by parties within the government. Coalition partners can threaten to exit the coalition, which gives them great bargaining leverage for policy change. For instance, Mark Kayser et al. (Reference Kayser, Orlowski and Rehmert2022) find evidence that coalition parties' credible threats to exit are the best predictor of policy change. In addition, external support and coalition partners can use parliamentary questions and committees to oversee the application of the agreements reached (Sozzi Reference Sozzi2024).
Third, beyond direct interaction, parties can also rely on ‘bureaucratic continuity’ to influence the policies formulated in ministries headed by other parties. By bureaucratic continuity we refer to the feedback effects between consecutive governments that are created through the establishment or transformation of ministerial resources such as knowledge, personnel and ideas. When parties take over a ministry, they recruit political bureaucrats to ensure that civil servants work towards the newly defined political goals (Peters and Pierre Reference Peters and Pierre2004). However, the replacement of previously recruited ministerial bureaucrats is usually a gradual process and, to avoid loss of administrative capacity, not all politically recruited personnel is replaced. Even top civil servants may continue to lead policy development for ministers with different partisan affiliations (Fleischer Reference Fleischer2016). Furthermore, non-political, permanent public servants continue their work on the same topic in the same position after the change of the political figurehead of a ministry, creating legacy effects for policies that have not been approved by parliament (Fleischer Reference Fleischer2016). These bureaucrats serve as a channel of influence between governments led by different parties. In addition, policy changes may introduce new routines and knowledge into the administration, influencing the activity of civil servants in the future (Heclo Reference Heclo1974; Moynihan and Soss Reference Moynihan and Soss2014).
Tracing progressive parental leave reform under centre-right parties
We have designed and implemented a deductive, theory-testing process tracing to assess if centre-right parties were the protagonists or merely the consenters of the reforms selected (Beach and Pedersen Reference Beach and Pedersen2019). The deductive approach allows us to assess the empirical validity of the mechanisms we theorize above. We formulate a series of necessary and sufficient conditions that serve as tests to gauge the value of two rival hypotheses. A hypothesis can be rejected if it does not fulfil any of the necessary conditions, measured through ‘hoop tests’. Achieving a sufficient condition, tested through ‘smoking-gun tests’, implies robust evidence in favour. To understand how partisan veto players influence policies that are formally under the control of another party, we identify which observable implication can be considered a sufficient condition for each mechanism of influence.
The null hypothesis is that the party that appoints the minister in charge of the reform is its protagonist. The alternative hypothesis is that the party appointing the minister in charge is a consenter, conceding to pass a reform supported by a different party through any or several of the mechanisms outlined.
The process-tracing tests analyse different moments of the policymaking process: the electoral campaign preceding the reform, government formation, government survival and policy adoption. These are summarized in Table 1. Hoop tests help to decide whether to discard the alternative hypothesis or not. If the centre-right party consented to a reform proposed by others, then it could not have supported the reform in its electoral manifesto or the debates preceding the election (Test 1). However, it should feature in the manifesto of the partisan veto player (Test 2) and then appear in the coalition or support agreement (Test 3). If these conditions are not met, we must reject the alternative hypothesis developed above. We investigate these conditions by studying manifestos, agreement documents and election debates.
Process-Tracing Design
Following Korpi (Reference Korpi2006), we argue that a party can be considered a protagonist if it initiates and designs the reform and then leads its adoption. Hence, the first smoking-gun test is whether the inclusion of the reform in the coalition or support agreement can be attributed to the party of the responsible minister, or rather to the coalition or support party (Test 4). The second is the identification of clear exit threats by the coalition or support parties to force the centre-right party to comply with the coalition or support agreement (Test 5). The last test evaluates if the reform was designed before the centre-right party was in government and not changed significantly thereafter (Test 6). This is a sufficient condition to determine that a centre-right party has consented to a reform proposed by another actor.
We assess the smoking-gun tests through interviews and document analysis. We have conducted semi-structured interviews with key stakeholders: high-level politicians, policy advisers, experts and members of civil society. A total of eight interviews for each case were conducted between January and October 2023 (full list in Supplementary Material A.1). Moreover, we triangulated these interviews through the analysis of media reporting, official documents and parliamentary debates, following Rebecca Natow (Reference Natow2020). We refer to the interviews by using the party acronym followed by a number, such as SPD_1. Alternatively, we specify the country (ES or DE) followed by a code to refer to an expert (Exp), Catholic organization (Church) or feminist organization representative (Fem), such as ES_Exp1.
In addition to the formalized hoop and smoking-gun tests, our in-depth case analysis has uncovered further evidence. This information can be seen as what Derek Beach and Rasmus Pedersen call ‘straw in the wind tests’ (Reference Beach and Pedersen2019). While they are not independently a necessary or sufficient condition, they contribute to bolstering confidence in one hypothesis over another. We have therefore chosen not to formalize these additional tests. Some of these elements include the role of other parties in agenda-setting, the involvement of bureaucratic personnel, the significance of specific parliamentary institutions, the activities of civil society organizations or distinctions between party factions.
We apply this process-tracing design to two cases: the parental allowance reform of 2006, adopted by the German CDU, and the increase in paternity leave by the Spanish PP in 2017 and 2018. In the German case, the reform reduced the total available leave time and reserved two months for fathers, prompting an earlier return to work for mothers and care incentives for fathers. While it is less problematic for the CDU to expand childcare services, the leave reform effectively reduced options for male-breadwinner family models. Meanwhile, the PP expanded paternity leave in a context of fiscal constraints and a short maternity leave, and it was the only instance of progressive family policy in the PP government, as already signalled by Reto Bürgisser (Reference Bürgisser, Garritzmann, Häusermann and Palier2022). In both cases, these reforms contrast with the historical political positions of both parties. Expert surveys in the supplementary material (A.2) show that the PP and the CDU have held conservative positions towards family relationships and gender equality. Moreover, the CDU was even qualified as slightly more conservative before the reform, while the assessment of the PP was generally stable.
The two reforms are selected as most-similar and most-likely cases. The cases are similar because of their akin institutional contexts, via their parliamentary systems, their proportional representation rules and because of their common familialistic welfare legacy (Lewis Reference Lewis1992). However, they are also sufficiently different to allow us to test the theorized mechanisms in two scenarios of institutional fragmentation, namely a coalition and a minority government. They are also most-likely cases because neither features an absolute majority government, where the governing party is expected to be the protagonist of the reform. We argue that it is justified to select most-likely cases given the novelty of the theoretical argument for a political event which has attracted vast scholarly attention. However, our research may also provide explanations for partisan behaviour in other cases of unexpected reforms in contexts of minority and coalition governments.
A junior partner in the driver's seat and a paradigm shift in German leave policy
The political dominance of Germany's centre-right parties, the CDU and Christian Social Union (CSU), in the post-World War II era created a legacy of family policies that explicitly favoured single-earner households (van Kersbergen Reference van Kersbergen1995). In 1986, Helmut Kohl's government introduced three years of parental leave with a two-year low-paid benefit which was almost invariably taken by women (Morel Reference Morel2007). In contrast, in 2006, a CDU-led government transformed parental allowance to 12 months, transferable, and with an extra two months reserved for the partner, effectively benefiting dual-earner households. This radical change in leave policy has attracted an impressive amount of scholarly attention.
Ideational origins and policy legacies: Towards a progressive parental leave policy
To explain the puzzling support of the CDU for a progressive parental allowance scheme, existing studies have traced the gradual ideological change steered by the ‘modernizers’ of the CDU (Blum Reference Blum2010; Seeleib-Kaiser Reference Seeleib-Kaiser2010). Timo Fleckenstein (Reference Fleckenstein2011) explains how CDU intra-party commissions in 1999 and 2004 switched a strict male-breadwinner definition of the family to a more diverse concept, including dual-earners and single parents. However, the new CDU position on family policy revolved around a narrative of protecting families' ‘freedom of choice’, as shown by Timo Fleckenstein and Soohyun Lee (Reference Fleckenstein and Lee2014) and Nathalie Morel (Reference Morel2007). The new principle of ‘choice’ – resonating with the concept of optional familialism (Leitner Reference Leitner2003) – allowed the CDU to support policies such as the expansion of public childcare services, enabling both parents to participate in the labour market without imposing costs on single-earner families. Meanwhile, the CDU opposed policies with a value judgement on whether parents should ‘share their responsibilities equally or not’ (CDU_1; see also Deutscher Bundestag 2005a: 16115) or ‘dictate to families that father and mother must work’ (Deutscher Bundestag 2006: 5365). Contrastingly, shortening the available parental leave time for mothers while introducing a paternal quota contradicts the choice principle, as it benefits dual-earner households at the expense of single-earners.
Within the social democratic SPD, a more fundamental and consequential shift in leave policy ideas took place. When Gerhard Schröder became chancellor, there was not much interest in family policy. In fact, Schröder introduced his minister for families, senior citizens, women and youth, Renate Bergmann, as minister for women and stuff (‘Gedöns’). Yet, soon after, technocrats within the SPD secretariat and soon-to-be family minister Renate Schmidt ‘moved family policy from the periphery into the centre’ (Ristau Reference Ristau2015: 6) of social democratic social policy.Footnote 1 Based on new research, they argued that the familialistic allowance system of Erziehungsgeld was highly inefficient and that the same budget distributed through a progressive parental allowance scheme would result in increased female employment and higher birth rates (Ristau Reference Ristau and Bergmann2022). Armed with this evidence base, Renate Schmidt and social democratic technocrats around Malte Ristau and Petra Mackroth, policy officers in the SPD secretariat and central figures in the reform process (SPD_1; DE_Exp; DE_Church), convinced Schröder that the introduction of a progressive parental allowance scheme would be instrumental in winning over the new centre of the electorate and hence in securing his re-election (Bujard Reference Bujard2014; Ristau Reference Ristau2015).
After the 2002 election, Renate Schmidt became the new minister of family affairs. During her time in office, she and her ministry developed a parental leave and allowance concept, Elterngeld, which was modelled after the Swedish parental benefit (CDU_1, SPD_1, SPD_2, DL_1, DE_Exp, DE_Church). An SPD family policy adviser at the time reveals that Renate Schmidt ‘had two close confidants. One was Malte Ristau and the other was Petra Mackroth. And together those two developed the concept of Elterngeld, pushed it forward, made it popular in the SPD and then wrote it into the coalition agreement and implemented it’ (SPD_1). Further, Schmidt relied on scientific evidence and the creation of large societal alliances to produce favourable media reporting and public opinion (Bujard Reference Bujard2014). The ministry commissioned academic publications by leading scholars such as the Seventh Family Report or the Rürup opinion (DE_Exp). Moreover, the ministry also launched the ‘Alliance for Families’ and the ‘Local Alliance for Families’ to connect labour unions and business actors in support of progressive family policies (Blum Reference Blum2010; Bujard Reference Bujard2014; Leitner Reference Leitner2010). Academics, as well as the political stakeholders who were involved, emphasize the importance of Schmidt's strategic approach of partnering with academia and societal actors to initiate an overall societal discourse on family policy for the successful adoption of the reform (DL_1; DE_Exp, Blum Reference Blum2010; Ristau Reference Ristau and Bergmann2022).
Elections and government formation: Turning antagonists into consenters
In parliament, Renate Schmidt's proposal for an Elterngeld reform by 2008 was heavily opposed by the CDU and CSU (Deutscher Bundestag 2005a, 2005b). Ingrid Fischbach, a member of parliament belonging to what social scientists and journalists have called ‘modernizers’ in the CDU (Seeleib-Kaiser Reference Seeleib-Kaiser2010) strongly rejected Schmidt's Elterngeld shortly before the federal election: ‘You want to link the payment of parental allowance to the obligation of both parents to take at least partial parental leave. This clearly interferes with the parents’ freedom of choice, and we don't want that to happen’ (Deutscher Bundestag 2005a: 16115).
During the subsequent elections, Elterngeld became a contested topic with fundamentally different positions taken by the CDU, CSU and SPD. In their campaign and manifesto, the SPD promised to introduce Elterngeld as an income substitute and to expand childcare, while also emphasizing the goal of promoting female employment and equal partnership between parents (SPD 2005). In contrast, the CDU and CSU proposed a monthly benefit of EUR 50 per child and a flat income tax deduction to ensure the ‘special protection of marriage and family’ (CDU/CSU 2005a: 24) and to enable families to decide how to distribute care work (CDU/CSU 2005b). Even modernizers such as the future family minster Ursula von der Leyen demanded per-child transfers in line with the single-earner household model (Schmergal Reference Schmergal2005).
The elections resulted in a coalition of the CDU/CSU and SPD. The negotiation of a coalition treaty provided the junior partner SPD with the opportunity to bring its leave policy concept onto the governmental agenda. An SPD policy officer responsible for family policy during the coalition negotiations remembers that the CDU and CSU were ‘not even capable of negotiating the details because getting involved in detailed negotiations would have meant getting involved in the concept itself. Elterngeld just was not their concept. And they confined themselves to basic criticism, especially that it's all too expensive’ (SPD_2) – a criticism hard to maintain as the SPD was able to claim the Ministry of Finance (SPD_2). Further, the CDU and CSU reportedly ‘struggled with the family image behind it. That's why there wasn't much reason to talk about the details of such a concept, because they thought it was strange, unnecessary and harmful anyway’ (SPD_2, see also Deutscher Bundestag 2006). In fact, a participant in the negotiations who would become the family policy spokesperson of the CDU/CSU parliamentary group confirms that the centre-right parties had not yet become fully ‘operational’ (CSU_1), as roles within the parliamentary group were not allocated. These perspectives from both camps are confirmed by a lobbyist involved in CDU family policy development during the Elterngeld reform, who argues that the SPD dominated the negotiations (DE_Church), while the CDU and CSU were ‘argumentatively unprepared’ (DE_Church).
The CDU and CSU, who entered the negotiations demanding a flat income tax deduction per child (FAZ 2005b), managed to secure support for a multigenerational housing policy. An SPD policy officer responsible for the 2005 coalition negotiations reports that ‘there were no negotiations on individual issues. Instead, it was clear that the SPD would introduce the concept of Elterngeld and then Ms von der Leyen would come up with her multigenerational homes’ (SPD_2). The coalition treaty strongly reflects the SPD manifesto and fully keeps the SPD promise of introducing Elterngeld as developed by minister Schmidt prior to the election (SPD_1; SPD_2; Blum Reference Blum2010; Schmidt Reference Schmidt2005), including the allocated budget, level of income compensation, overall duration, as well as the maximum level of Elterngeld.
Last but not least, during the portfolio allocation, the SPD was able to choose the ministries it wanted in return for the CDU taking the chancellery with Angela Merkel as chancellor (FAZ 2005a). Giving the CDU the family ministry was therefore a consequence of SPD priorities rather than a deliberate choice by the centre-right party (SPD_2).
Coalitional dynamics and the legislative process: Consenting without conviction
When the CDU took over the ministry, ‘the policy was already written’ (DE_Church) and agreed in detail in the coalition treaty. As previous empirical studies show, coalition agreements in Germany strongly restrict the ability of ministers to diverge from the initial agreements made by parties, especially when they outline policies in detail (Klüver and Bäck Reference Klüver and Bäck2019). The CDU/CSU family policy spokesperson at the time puts it as follows: ‘We found a compromise, as is customary in politics, and that is also crucial for democratically responsible action’ (CSU_1). The treaty was a decisive reason for minister Von der Leyen to follow through on implementing Elterngeld as emphasized by involved stakeholders from the conservative, liberal and social democratic camps (FDP_1, DE_Church, SPD_1, SPD_2).
Loyalty to the original Elterngeld concept was further reinforced as the personnel brought in by Renate Schmidt continued to play a vital role. Interviewees highlight the significance of key individuals such as Malte Ristau, head of the family policy department, who served as the ‘mastermind behind Renate Schmidt's Elterngeld’ (DE_Church) and remained a prominent figure throughout the whole legislative process (SPD_1). The decision not to replace Ristau is in line with the role of the CDU to consent to the previously developed policy instead of adjusting the law to own priorities. The political weight of bureaucratic continuity in the Elterngeld reform is strong as the minister does not show any intention to amend the reform. Instead, she is perceived as disinterested in the reform in the family policy committee of the Bundestag: ‘She had no interest in discussing the details of the reform’ (FDP_1). Her successor as family minister emphasizes that it is ‘a fair assessment’ that ‘it was an SPD project’ and that ‘the idea came from Renate Schmidt’ (CDU_1). Ursula von der Leyen, on the other hand, had ‘the courage and influence in the party to make sure that the CDU agrees’ (CDU_1). In other words, minister Von der Leyen's role was to make sure her party consented to a policy developed by its coalition partner.
While the legislative process of adopting the Elterngeld was quick, there were some obstacles along the way. The CDU/CSU parliamentary group did not support Elterngeld and only ‘grudgingly let the project pass without inner conviction’ (SPD_1). Ingrid Fischbach, leader of the group of female CDU MPs and the party member responsible for the reform in parliament, confirmed during the parliamentary debate on its adoption that her party was against using the parental allowance reform to promote equal caregiving and female employment. She states, ‘The coalition partners have very different ideas when it comes to family policy. We do not want to dictate to families how they should live together’ and ‘we do not want to dictate to families that father and mother must work’ (Deutscher Bundestag 2006: 5365). Further, voices from within her parliamentary group publicly attacked the minister, who had to implement the reform following the coalition agreement and strong pressure from her ministry and the SPD parliamentary group (SPD_1, DE_Church).
The adopted version of Elterngeld, approved on 29 September 2006, exhibits two significant differences compared to the original concept: the mandated distribution of Elterngeld between parents and its social policy dimension. The concept developed by minister Schmidt proposed a model of ten months plus two additional months if both partners took at least two months of Elterngeld, to incentivize fathers to share the burden of domestic work. This was strongly opposed by the CSU, but also by family policy experts within the CDU/CSU parliamentary group (DE_Church; Deutscher Bundestag 2006; FAZ Reference FAZ2006). As a compromise, Elterngeld was adopted with a duration of 12 months to be taken by one or both partners and a bonus of two additional ‘partner months’ if both parents participate (CSU_1). It is widely observed that increases in the length of parental leave are associated with a familialistic logic, as reintegration into the labour market becomes more difficult (Bürgisser Reference Bürgisser, Garritzmann, Häusermann and Palier2022; Leitner Reference Leitner2003). In the case of the Elterngeld reform, the CSU, the Bavarian sister party of the CDU, managed to reduce the level of defamilialization introduced through the leave reform. Furthermore, giving in to pressure exerted by the SPD leadership, the coalition agreed to make the minimum Elterngeld of EUR 300 per month a bonus that would be paid on top of unemployment benefits (Welt 2006).
Overall, the introduction of the progressive parental allowance scheme Elterngeld cannot be attributed to the protagonism of the CDU, but to its coalition partner the SPD. The concept was first developed and subsequently negotiated into the coalition treaty by the centre-left party. In fact, the law was drafted under the leadership of a previous minister, adopted as part of the electoral manifesto of the SPD and only modified through pressure exerted by the SPD and CSU. Starting as antagonists of the progressive parental leave scheme, the CDU became a consenter after signing the coalition treaty. Yet, as best illustrated through the final parliamentary debate before its adoption, instead of defending the reform with conviction, the CDU rapporteur for the reform continued to emphasize that her party had a different position on leave policy than the SPD and that they ‘do not want to dictate to families how they should live together’ (Deutscher Bundestag 2006: 5365).
Navigating external agreements in a Spanish minority government
Between 2011 and 2015, the PP implemented harsh austerity cuts in childcare, elderly care and gender-equality promotion programmes (León and Pavolini Reference León and Pavolini2014). Surprisingly, by 2017 the same party doubled the duration of fully paid and non-transferable parental leave from two to four weeks, and within a year, extended it to five weeks. How can we explain this change in the direction of welfare reform?
We argue that the policy was not motivated by an electoral or economic strategy. Instead, it can be attributed to three factors exogenous to the PP. First, the extension had originally been developed by a previous socialist government, providing an administrative blueprint. Second, the 2016 elections weakened the parliamentary position of the PP, leading it to accept the reform in an agreement for government formation with the Citizens (Cs) party. Third, leveraging their parliamentary influence, Cs effectively pushed for approval of the extension.
Ideational origins and policy legacies: Persistent partisan differences and bureaucratic roots
Spain was traditionally described as an example of ‘unsupported familialism’, where women in families played a central role in welfare provision. However, after the transition to democracy in 1975, Spain witnessed significant changes in gender roles, female employment and state welfare capacity (Guillén and León Reference Guillén and León2011). In this context of modernization, the PP, originally founded by former officials of the Francoist dictatorship, moderated its authoritarian image under the new leadership of José María Aznar in 1990. Aznar championed women's careers within his organization and introduced ideas favouring gender equality and female employment, becoming relatively more similar to the Socialist Party (PSOE) (PP_1).
Yet, cross-partisan differences about family ideals and the role of the state in society persisted. During Aznar's tenure (1996–2004), family policy reforms promoted female employment but did not challenge traditional gender roles in caregiving. The government introduced the option of transferring part of the maternity leave to the father and reducing working hours with a proportional pay penalty to working families.Footnote 2 The result was that many women turned to part-time employment after maternity leave, while the proportion of men taking transferable shares of the leave was negligible (Meil and Escobedo Reference Meil and Escobedo2018). This diverges from the future expansion of paternity leave by the same party, which created explicit incentives for men to engage in care activities. Overall, despite modernization, the PP continued to defend traditional values and large families while it opposed gay marriage and abortion, to preserve a religious constituency and strong ties with the Catholic Church (León et al. Reference León, Alvariño and Soler-Buades2022).
In contrast, José Luis Zapatero's socialist government (2004–2008) challenged the distribution of care responsibilities by expanding public childcare, introducing long-term care programmes and passing gender-equality laws (León and Pavolini Reference León and Pavolini2014). A relevant impact came from the development of the ‘Law for the Effective Equality of Men and Women’, passed in 2007, which was intended to advance comprehensive gender-equality measures for private companies and public bodies. Negotiations with social partners spanned two years. Given employers' resistance to a non-transferable, fully paid, four-week paternity leave, the newly established general secretariat of equality reached a compromise by approving two weeks of paternity leave and a mandated future increase to four weeks by 2013. Soledad Murillo, the general secretary at that time, declared that ‘what we did was to freeze its value, so to speak’, maintaining the social dialogue while strategically postponing implementation to influence future administrations (PSOE_1).
The law generated strong feedback effects at different levels, positively affecting the prevalence of a gender-equality lens in future policy development. It brought innovative practices to public administration, introducing new routines and knowledge which influenced the policy practices of future cabinets, particularly concerning gender equality (ES_Exp2, PSOE_1). For instance, it introduced ‘equality units’ which made specific civil servants responsible for introducing a gender perspective or gender programmes into policymaking (ES_Exp2). In addition, new gender quotas increased the share of women within parties and company boards.
The law also stimulated civil society mobilization. The ‘platform for non-transferable paternity leave’ evolved into the ‘platform for egalitarian and non-transferable birth and adoption leave’ (PPiiNA), advocating for 16 weeks of paternity leave. After criticizing the government for not endorsing the four-week extension, activists close to the PSOE abandoned the organization (ES_Fem). The platform then started to lobby different parties by presenting expert reports and influencing public opinion, as reported by several respondents (ES_Exp1, PP_1).
In 2009, influenced by PPiiNA, the Catalan Convergence and Union Party approved the government's general state budget in exchange for a law which set the four-week paternity leave extension for 2011 (ES_Fem; Europa Press 2011). However, rising debt and pressure from the European Union in 2010 led the government to introduce a one-year delay in the budget for 2011, as part of broader austerity measures (León and Pavolini Reference León and Pavolini2014). The extension of paternity leave to four weeks became a ‘pending reform’ awaiting the necessary political will.
The 2011 elections resulted in a substantial electoral setback for the PSOE, granting Mariano Rajoy of the PP an absolute majority until 2015. During Rajoy's initial cabinet, the extension of paternity leave was deferred every year, concurrent with austerity measures impacting a wide array of social programmes. However, antagonism was rooted in budgetary concerns rather than ideological differences. When questioned about the rationale behind suspending the increase in paternity leave, the then minister of employment argued that the escalating costs of pensions and unemployment benefits, together with the mandate from European institutions to curtail budgetary expenditures, left no fiscal leeway for the enhancement of work–life balance policies (PP_2).
Moreover, Carmen Quintanilla, the chair of the Equality Commission, also a member of the PP, argued that the Ministry of Finance possessed the authority to quash any gender-equality proposals if they carried financial implications (PP_1). Regarding the issue of increased paternity leave, she lamented, ‘I, as chairwoman of the Equality Commission, tried my best to get our parliamentary group to approve it, but in vain, as it was a question of budget’ (PP_1). Meanwhile, the Ministry of Finance defended its stance by asserting that rigorous budgetary measures would bolster market confidence, reduce interest on debt and promote the sustainability of welfare (Congreso de los Diputados 2012).
Elections and government formation: A new party system and minority compromises
In the elections of 2015 and 2016, two new parties made a strong entrance to parliament, capitalizing on popular resentment of the austerity policies implemented. The new left-wing party, We Can (Ps), embraced the proposal put forth by PPiiNA, advocating for equalizing paternity and maternity leaves at 16 weeks. Meanwhile, the liberal Cs placed significant emphasis on the needs of working families, proposing an extension of paternity leave to eight weeks and the introduction of an additional 10 weeks, transferable between both parents. PSOE pledged an immediate increase to four weeks of paternity leave, with a commitment to reach 16 weeks in the future. In contrast, the PP remained silent on the paternity leave issue and maintained that its budgetary control strategy would achieve economic recovery and welfare sustainability.Footnote 3
Party fragmentation greatly increased following the results of the 2015 and 2016 elections. The PP started to negotiate a minority government supported by Cs. However, this would also require the abstention of other parties, given that their seats fell short of securing a congressional majority. Fátima Báñez (minister of employment from PP) and Toni Roldán (head of Cs' economic office and programme coordinator) declared that negotiations for government formation were marked by cooperation among politicians and experts on social policy and economics, who engaged in cordial discussions (Cs_1, PP_2). At the technical level, Cs negotiators strategically used scientific evidence to convince PP members of their proposals, gradually overcoming resistance rooted in budgetary concerns (Cs_2). This indicated a degree of pragmatism and a willingness to collaborate on the design of a shared policy programme.
However, Cs deliberately maintained a clear distance from the PP. It opted not to participate in the cabinet, instead positioning itself as an external support party with policy influence (Cs_1). This approach allowed the Cs' most public-facing figures to preserve their identity as members of a new and distinct political party, untainted by corruption and separate from established political elites (Cs_1). In fact, President Rajoy noted in his memoirs that ‘they seemed to be much more comfortable acting as the opposition, setting conditions and marking their own territory, than sharing management’ (Rajoy Brei Reference Rajoy Brei2019).
Negotiations culminated with a written government formation agreement, published in August 2016. The Cs team expressed satisfaction with the outcome, as the final document incorporated numerous provisions outlined in its manifesto, including the phased increase of paternity leave to four weeks by 2017 and eight weeks by 2018 (Cs_2). Despite economic recovery and the gradual relaxation of budgetary pressures from European institutions, the PP preserved its aversion to increasing public expenditure (PP_1). As a result, each measure in the agreement was accompanied by a cost analysis and a proposal to offset these costs through an increase in fiscal revenues (Ciudadanos and Partido Popular 2017).
Coalitional dynamics and the legislative process: Parliamentary pressure and final approval
After Mariano Rajoy secured office again, Cs developed three mechanisms to ensure compliance with the agreement and guarantee their influence throughout the government's tenure. First, a panel to monitor the government agreement secured continuous communication between both parties. Second, Cs and the PP shared control of the Bureau of the Congress, through which ‘the control of time and pace was maintained by the government's initiative’ (Cs_2). Crucially, to hold a majority the PP needed the votes of Cs, who could therefore threaten to vote against the government and side with the opposition instead (Cs_2). Third, parliamentary voting was the main venue of strategic negotiations between the PP and Cs. Due to its position in parliament, Cs was able to either endorse or oppose government legislative proposals, which gave it leverage to facilitate approval for its own policies. It also frequently influenced the government's output by introducing legislative amendments. Through these mechanisms, Cs adopted an ambivalent strategy: while it generally supported the PP, ensuring governance by voting in alignment within the Bureau and in the Congress, it also possessed leverage to vote against the PP if policy agreements were not met.
The budget negotiations for 2017 and 2018 were vital for the minority government. Starting after the government was formed in October 2017, Cs insisted on securing funding for its reforms in education, labour market and family policy with its counterparts in the ministries of labour and finance. It was during these negotiations that the extension of paternity leave to four weeks was approved for the next year.
Interviews with decision-makers from the PP and Cs, as well as the congress diary, provide clear evidence that the increase was driven by Cs, and that the PP ultimately agreed to it. Fátima Báñez (PP minister of employment) declared in parliament, ‘In accordance with what we have agreed with Citizens, we are ready for the 2017 budget to extend paternity leave to four weeks’ (Congreso de los Diputados 2016). In fact, Toni Roldán, who became Cs' representative responsible for social and economic policy, stated, ‘If you ask me now in retrospect what was the greatest achievement of Cs in policy terms, it was the extension of paternity leave that we negotiated with Fátima Báñez and [her secretary general]’ (Cs_1). Meanwhile, the president of the equality commission, a PP member, recognized that Cs achieved something she had been fighting for unsuccessfully inside her party:
In 2015–2019 or 2018, we were able to move forward many policy positions of Toni Roldán … because we had to negotiate other issues or other legislative initiatives where we needed the support of Citizens. Therefore, at times there is some ‘give and take’ at the legislative level. If you want to know why the change happened, that's why. (PP_1)
Nevertheless, the events of 2017 diminished Cs' policy influence. The party's technical faction tried to persuade the government to fulfil the remaining parts of the agreement, including extending the leave to eight weeks (CS_1). However, the intense political competition that followed the Catalan referendum in October 2017 hindered policy negotiations between the two parties. The referendum and the subsequent imprisonment or escape of pro-independence leaders shifted the political focus towards nationalism and territorial conflicts. In this scenario, Cs became the most voted-for party in the Catalan elections, with national polls even suggesting it might surpass the PP (Nieto Reference Nieto2018). This sparked ‘a battle for power and domination of the right-wing axis after the referendum … [where] the priorities of the party's leadership begin to change, and policy issues become totally secondary’ (Cs_1).
As Cs' political wing intensified its criticism of the government, the PP decided to back-pedal on its concessions to proposals by Cs. During the 2018 budget negotiations, the PP used economic arguments to limit the extension of leave to five weeks, falling short of the promised eight weeks (PP_2). Eventually, a few months later, before the budget could receive parliamentary approval, significant corruption scandals involving the PP leadership precipitated the premature end of Rajoy's government through a successful motion of censure.
Discussion and conclusion
This article re-examines the attribution of two reforms that defy traditional ‘parties matter’ expectations. It explores the German 2006 parental allowance reform and the Spanish 2017 paternity leave extension. Parental leave is particularly interesting for analysing the family policy positions of centre-right parties, as they directly influence the gender distribution of work and care activities. Parental leave reforms benefiting dual-earner households at the cost of single-earners are particularly troublesome for centre-right governments, as they cannot be reconciled with optional familialism (Leitner Reference Leitner2003). Existing literature has primarily focused on explaining why the German CDU and the Spanish PP may have willingly pursued reforms misaligned with their traditional ideologies. We complement this research by exploring the role of partisan veto players in designing and pushing for the implementation of these reforms.
We have theorized three mechanisms through which partisan veto players may accomplish the adoption of a reform: their support for the formation of a government may be granted on certain terms; they may use their leverage in the legislative process as a condition for government survival; or they can rely on bureaucratic continuity to influence future governments. To investigate these mechanisms, we designed a formal and deductive process-tracing approach. Table 2 summarizes our results. Our analysis of manifestos, government formation documents and electoral debates shows that indeed neither the PP nor the CDU featured the reform in their manifestos nor advocated for it before the election, while the PSOE, Cs and SPD did (Tests 1 and 2). Furthermore, in both cases the reform appeared in the coalition or support agreement (Test 3). These hoop tests provide initial evidence supporting the hypothesis that alternative actors designed and influenced approval of the reform.
Process Tracing Test Results
We found compelling evidence that refutes the attribution of the reforms to the PP and CDU. First, both reforms had already been drafted by previous cabinets: the family ministry controlled by the SPD and PSOE's general secretariat of equality (Test 6). In Spain, approval was subsequently delayed, favouring its eventual enactment. Second, our evidence shows that the veto players were responsible for introducing the reform into the coalition or support agreement (Test 4). In Germany the SPD successfully incorporated their Elterngeld concept in the coalition treaty, while in exchange the CDU secured funding for a multigenerational housing initiative. Meanwhile, in Spain the liberal Cs introduced the increase of paternity leave in the support agreement for a PP minority government, while the PP required a careful budget assessment of the reform. Third, in Spain, Cs employed credible threats to withdraw support from the government during the budget negotiations of 2017 and 2018 to ensure the enactment of the reform (Test 5). In contrast, this was not necessary in the German case, where loyalty to detailed coalition treaties is the norm, and actors ranging from social democratic technocrats to the SPD parliamentary group ensured commitment to the agreement. Hence, the findings consistently show that neither the CDU nor the PP were the protagonists of the parental leave reforms under study. Instead, the CDU consented to the reform initiative and influence of the SPD, and the PP to the pressure of Cs to approve and expand a measure drafted by the PSOE.
These findings have significant theoretical implications for research on the partisan politics of welfare state reform. Existing research often concentrates on the factors that may motivate the party in charge of the relevant ministry to design and approve welfare reforms and focus on processes of ideational change, electoral strategies or intra-party conflict. Instead, we update earlier institutionalist literature to argue that institutional fragmentation requires researchers to view partisanship as a more complex variable. Coalition or minority governments require agreements and compromises, which potentially bestow partisan veto players with substantial policy influence. Therefore, other parties may uphold the intent of a specific reform, drafting and introducing it for reasons more aligned with the traditional ‘parties matter’ thesis. The distinction between antagonists and consenters is relevant to enhance our understanding of the evolution of party preferences, the effect of party interaction on policy output, and the current validity of different theories about the party politics of welfare state reform.
Moreover, the findings from these reform cases point to a need to re-evaluate the political attribution of other welfare reforms. We do not contest the importance of electoral and socioeconomic considerations in bringing about ideational change. Reforms themselves may generate policy feedback effects such as public opinion support or policy learning, contributing to parties' effective change of political position. In fact, centre-right parties such as the CDU have developed a track record of more progressive family policy ideas over time. However, in contexts of institutional fragmentation, the need to reach agreements with other parties and the influence of bureaucratic politics should be considered as potential explanations for unexpected reforms. This is especially pertinent given that coalition and minority governments are the norm in many advanced welfare states. In such scenarios, shifting the focus to the dynamics of interparty cooperation and bureaucratic influence becomes essential to understand better the partisanship of policy change.
Supplementary material
The supplementary material for this article can be found at https://doi.org/10.1017/gov.2024.25.
Acknowledgements
We thank Anton Hemerijck, Ellen Immergut, Margarita León and two anonymous reviewers for their invaluable feedback, as well as the participants of the panel on ‘The Causes and Consequences of Social Investment Welfare States’ at the CES 2023 conference for their insightful comments and suggestions. We are also grateful to all of our interview respondents. The authors also thank Eve Leckey for providing editorial support.
Financial support
This work has received funding from the European Research Council (ERC) under the European Union's Horizon 2020 research and innovation programme (grant agreement No. 882276).
Ethics statement
The research involving human subjects has been approved by the ERC under the Grant Agreement 882276 in the Ethics Monitoring Note EMN-882276-001, and also by the Ethics Committee at the European University Institute under Project code ‘20210921_WellSIRe’.
Open access
