Shifting worlds of father politics? Comparing path-departing change in paternity and parental leave policy in Germany and the UK

ABSTRACT
 How families balance employment and the care of young children has become a focus of dynamic policy change in many high-income countries since the 1990s. While there has been a broad shift across the OECD away from male-breadwinner model work-family policy regimes, there is much variation in the extent to which policies targeted at fathers have been part of these changes. Examining this variation, this article compares two cases which both represent ‘late path shifters’ away from the male-breadwinner family model, yet whose trajectory in terms of ‘father politics’ are very different: Germany, which has introduced well-remunerated, non-transferable periods of leave for fathers, and the UK, where leave policy has remained overwhelmingly focused on mothers. This article seeks to explain these different trajectories through an analysis of the political role of ideas in the two processes of reform. Drawing on documentary analysis and interviews with policymakers, it argues that a substantial shift in ideas about the role of fathers underpinned the reforms in Germany, while no such shift took place in the UK. This difference is explained with reference to political conditions, which created similar but different windows of opportunity for change, and the impact of existing policy legacies.


Introduction
How families balance employment and the care of young children has become a focus of dynamic policy change in many high-income countries since the 1990s.Across the OECD there have been dramatic increases of public investment in early childhood education and care (ECEC) services and of policies that facilitate parents' temporary withdrawal from the labour market after childbirth, such as maternity, paternity and parental leave (Koslowski et al., 2021;OECD, 2017). 1 The focus of much comparative policy literature has been on explaining these path-shifting reforms, especially in countries previously characterized as following a 'strong male-breadwinner model', where they represent the most dramatic shift (Blome, 2016;Fleckenstein & Lee, 2014;Morgan, 2013).Yet this focus on explaining the departure from the male-breadwinner model obscures the significant variation that is evident in different countries' trajectories of reform (c.f.Ferragina & Seeleib-Kaiser, 2015).One aspect of work-family policy expansion in which cross-national variation is most evident is the extent to which leave entitlements specifically aimed at fathers have been part of the reform package.While in 2018 on average the 23 original OECD countries provided fathers with 9.5 weeks of non-transferable paid leave, this varied from more than three months in seven countries (Belgium, France, Iceland, Japan, Luxembourg, Portugal and Sweden) to no provision at all in four countries (Canada, New Zealand, Switzerland and the US) and less than one week in three others (Greece, Italy and the Netherlands) (OECD, 2021).This article examines this variation through case studies of two countries that have undertaken divergent reforms: Germany and the UK.
Several typologies have been developed for characterizing work-family policy regimes and charting the changes evident in the cross-national move away from the male-breadwinner model (e.g.Kaufmann, 2002;Korpi et al., 2013;Leitner, 2003;Lewis, 2001;Saxonberg, 2013).Yet analysis of changes in policies for fathers sits awkwardly in many of these, as the focus is often explicitly on the implications for mothers.By contrast, Michael Rush's research focuses specifically on policies that relate to fathers and posits two 'worlds of father politics' which represent opposite ideal types: Sweden and the US.In terms of leave policy, the Swedish 'world' is characterized by an 'emphasis on the improvement of child-rearing opportunities for fathers' in a context in which 'fathers and mothers are both highly de-commodified through features of parental leave insurance that provide for substantial leave entitlements' (Rush, 2011, p. 39).This focus on fathers is linked to a concern for gender equality, a long-standing feature of the Swedish welfare state which underpinned its status as an 'early bird' reformer in the departure from family policies based on the male-breadwinner family (Leira, 2006).Sweden's parental leave is well remunerated and flexible, both of which research has shown are important in ensuring fathers take-up leave (Adema et al., 2015); most importantly in this respect, fathers in Sweden have a non-transferable, individual right to 90 days of paid parental leave (Koslowski et al., 2021).Such policies aim to shift the burden of childcare away from mothers, thereby enabling higher maternal labour market participation and a dual-earner, dual-carer family model (Ellingsaeter, 2014).
By contrast, in the US 'world' fathers 'remain highly commodified as male-breadwinners with only residual recognition of their social citizenship entitlements as parents' (Rush, 2011, p. 39).In common with the liberal principles which characterize the US approach to social policy (Esping- Andersen, 1990), the state demonstrates little interest in how unpaid work is distributed within families: work-family reconciliation is conceptualized as a private matter, unsuitable for state intervention (Gornick & Meyers, 2004).The US lacks a nation-wide paid system of maternity or parental leave, let alone policies that specifically seek to shift norms about child-rearing (Koslowski et al., 2021).In Rush's 'two worlds', Sweden and the US exemplify ideal types, representing opposite ends of a spectrum along which most countries' approach to fathers can be positioned.In the context of work-family reforms widely interpreted as a shift away from male-breadwinner model family policy regimes, one might expect to see a general shift in leave policy for fathers along this spectrum away from the lack of consideration of fathers' care-giving role evident in the US 'world' towards that of the inclusive policies characterized by the Swedish 'world'.Indeed, international organizations have promoted the adoption of Scandinavian style policies, particularly the notion of individualized 'quotas' of paid leave for fathers (e.g. European Commission, 2017;OECD, 2007) which have been taken up by a number of countries (Rush, 2015).
Yet empirically this is not a general trend; some countries, such as Germany, have adopted 'Swedish-style' policies towards fathers, while others, such as the UK, while undertaking some reforms, have remained closer to Rush's US 'world'.This divergence has considerable implications for the comparative experience of fatherhood in the two countries, with the proportion of fathers taking substantial portions of leave now much higher in Germany than the UK (Koslowski et al., 2021;Twamley & Schober, 2019).The wider literature suggests that one would expect this divergence to have, in the long run, implications for fathers' participation in unpaid care work, for mothers' employment rates and for fathers' long-term involvement in the care of their children (e.g.Bünning, 2015;Duvander & Johansson, 2019;Huerta et al., 2013).However, it is important not to overstate these changes: Germany has not undergone a dramatic transformation in gender roles since the reforms.It is widely acknowledged that gendered patterns of distribution of paid and unpaid work are a function of a complex web of institutional and structural factors (Korpi et al., 2013;Lewis, 2009).While leave for fathers is certainly more generous in Germany than it was, or than it is in the UK, it nevertheless coexists with numerous institutional features which continue to incentivize male-breadwinner model families, including inter alia the family-based taxation system, the prevalence of half-day schooling and the continued entitlement of parents to long periods of unpaid leave (Henninger & von Wahl, 2019).Reflections on the outcomes of the reforms explored in this article must be made in view of this wider context, yet the reforms of leave for fathers remain significant in terms of their implications for Rush's concept of 'father politics', and worthy of comparative analysis both in their own right and in terms of their wider contribution to notions of welfare state continuity and change, particularly in relation to the ongoing departure from policies based on the male-breadwinner model.
Recent advances in the policy transfer literature highlight the importance of national contexts in shaping the way in which policies are 'translated' in different contexts (e.g.Blum, 2014;Windwehr et al., 2022), yet the comparison of reforms in Germany and the UK raises questions about why the issue of fathers' leave is central to reform packages in some contexts but not others.This article seeks to explain why work-family policy expansion in Germany included path-shifting policies for fathers, while in the UK it did not.Such an investigation can illuminate cross-national variation in reforms for fathers, but the particular comparison also poses an empirical puzzle: why did Germany make a 'Nordic turn' in its leave policies, while the UK remained on its 'maternalist' trajectory (Daly & Scheiwe, 2010;Erler, 2009)?
This is particularly puzzling because the dominant political explanations for workfamily policy change do not seem able to explain this difference.One strand of literature focuses on the role of left parties in shifts away from the male-breadwinner model (Huber & Stephens, 2001;Lambert, 2008).According to this theory, one might have expected to see change in both countries; indeed, reform came onto the agenda in both countries after centre-left parties replaced longstanding centre-right governments in 1997 in the UK and 1998 in Germany.However, neither the British Labour Party nor the German Social Democratic Party (SDP) had historically shown much interest in policies relating to work-family reconciliation, women's employment or the care of young children (Gerlach, 2010;Perrigo, 1996), prompting the question of why their positions changed.Moreover, from a comparative perspective, the political institutional structure of the two countries would seem to favour greater change in the UK: the majoritarian, two-party system and lack of veto players provide governments with considerably fewer obstacles to change than the coalition-based German political system, characterized by federalism and stronger bicameralism (Schmidt, 1996).Further, the political opportunities for path-shifting change were more promising in the UK where Labour held power for 13 years, whereas the SPD-Green coalition lost power in 2005, when a Grand Coalition headed by the centre-right Christian Democratic Union (CDU) took over government, with the SPD as the junior party.Indeed, most puzzling from this perspective is that it was under the CDU-led coalition, promoted by a CDU family minister, that the most significant German reforms to leave policy took place in 2006.
A modified version of this 'parties matter' thesis, which stresses the role of electoral competition for women's votes, has more purchase.This argument highlights that changing electoral cleavages, associated with long-term trends in voting behaviour reflecting women's increased employment (Inglehart & Norris, 2000) have prompted dramatic expansions of ECEC and parental leave by political parties that had previously shown little interest in work-family policy (Fleckenstein & Lee, 2014;Morgan, 2013).Blome (2016) and León et al. (2021) argue that whether electoral competition emerges depends on the extent to which social attitudes have shifted away from those underpinning the male-breadwinner model family.However, as Table 1 demonstrates, attitudes towards non-maternal childcare and maternal employment were more conservative in Germany than in Britain both on the eve of the reform processes in 1999 and in 2008, once the most significant reforms had taken place in both countries.Moreover, this difference holds among young women, widely considered the target group for electoral competition in this area (Fleckenstein & Lee, 2014;Morgan, 2013).Thus, while the electoral competition argument helps explain why Labour, the SPD and the CDU were in favour of expanding work-family policies, it does not seem to provide an answer to Germany's greater shift towards the 'Swedish world' of father politics.

An ideational approach: ideas, framing and institutional context
Electoral competition can therefore explain why new ideas about work-family policy reform came onto the agenda in both the UK and Germany, yet this opportunity for reform does not suffice as an explanation for divergent trajectories of reform.A key argument of the ideational literature is that without an understanding of the content of new ideas, specific policy choices cannot be understood (Béland & Hacker, 2004); as Blyth (2001, p. 29) states, 'structural explanations of institutional change are indeterminate regarding subsequent institutional form'.This is particularly true for policies relating to families, which as Clasen (2005, p. 140) notes are 'rarely based on a clear diagnosis of existing problems, needs, demands, interests, and attitudes of families or children'.
To account for the content of change, it is therefore necessary to examine the ideas involved in the processes of policy change.The ideational literature has argued that ideas, defined as 'claims about descriptions of the world, causal relationships, or the normative legitimacy of certain actions' (Parsons, 2002, p. 48), shape policy outcomes in a number of ways (Béland, 2016).In particular, the framing of new policy problems has been shown to shape future reforms, in part because how a problem is defined sets the potential range of solutions that can be considered (Mehta, 2011).Ideas also provide policymakers with 'road maps' which help them interpret political problems and chart a path towards a strategic goal (Goldstein & Keohane, 1993).Such road maps provide actors with the tools with which to set the agenda by framing ideas in ways that build political support for change by creating an 'imperative for reform' and stressing both the necessity and legitimacy of change (Cox, 2001;Schmidt, 2008).Successful framing can also foster political support through ideas acting as 'coalition magnets' which can draw together previously disparate groups of political actors and create consensus for reform proposals (Béland & Cox, 2016).Drawing on the agenda-setting literature, such an approach highlights the role of key 'policy entrepreneurs' who are able to seize 'windows of opportunity' and build coalitions of support for change (Kingdon, 2003;Stiller, 2010).
Studying ideas must therefore also involve an examination of the processes of policy change, that is, examining not only the content of new ideas but how those ideas come onto the political agenda and shape decisions.How ideas are framed is closely related to their political and institutional context: ideational scholars have highlighted that ideas and institutions, such as the existing policy context, are interconnected and should be studied together (Béland, 2016;Campbell, 2004).In particular, institutional contexts structure both the opportunities for policy entrepreneurs to institutionalize new ideas and the kind of ideas that are likely to be successful, as highlighted by the policy transfer literature (Windwehr et al., 2022).
This article therefore undertakes a comparative analysis of ideas in their respective political and institutional contexts.Its contribution is twofold.First, it demonstrates that an ideational approach can be added to existing explanations of work-family policy expansion to help explain variation in reform of leave for fathers.Second, it provides a comparison not only of ideas in the two country contexts, but of how ideas affected the respective processes of change in policies for fathers.By examining these processes in the two cases side by side, it provides an analysis of both how ideas affect politics and the conditions under which this occurs.This comparative approach builds on existing literature on ideas and family policy (e.g.Ellingsaeter, 2012;Knijn & Smit, 2009;Nyby et al., 2018;Nygård & Krüger, 2012), particularly in the German case (e.g.Blum, 2012;Bujard, 2013;Klinkhammer, 2014).This literature has, among other insights, highlighted an emphasis on economic rationales for work-family policies and a downplaying of justifications based on gender equality, both of which have helped legitimize the new policies in policy discourse (Jenson, 2009;Mohun Himmelweit & Lee, 2021;Rüling, 2010;White, 2012).Others have highlighted that work-family reforms are legitimized through multiple policy aims, which chimes with the notion from the ideational literature that polysemy, where an idea is sufficiently broad that it can be interpreted differently by different people, is a crucial feature of politically salient ideas which act as coalition magnets (Mohun Himmelweit & Lee, 2022).This polysemy has been highlighted in studies of both German (e.g.Blum, 2012;Bujard, 2013) and British (e.g.Lewis, 2006;Lloyd, 2008) reforms of work-family policy.This article builds on such literature by combining this analysis of the ideas themselves with a focus, derived from the ideational literature, on their role in the political processes of reform in comparative perspective.

Materials and methods
The UK and Germany represent appropriate choices for a comparative investigation of variation in 'father politics' because they represent divergent outcomes despite both countries undertaking work-family reforms understood as marking path-shifting change away from the male-breadwinner model.Both are formerly 'strong male-breadwinner model' countries in which policies served to reinforce traditional gender roles (Lewis & Ostner, 1994) and before the changes examined here, were reflective of the 'US world' of father politics (Rush, 2015).The UK provided no statutory provision for fathers before 1999, while in Germany, although fathers had been nominally entitled to up to three years parental leave since 1992, the design of the policy, especially its means-tested low flat-rate benefit and inflexibility, discouraged paternal take-up and led to mothers leaving the labour market (e.g.Vlasblom & Schippers, 2006).Further, in terms of the ideas that lay behind the two policy regimes, both were characterized by a reluctance on the part of the state to intervene within the family (Lewis, 2009;Ostner, 2010).
Against this background, both countries undertook substantial reforms of workfamily policy in the late-1990s and early-2000s, emblematic of 'latecomer' departers from the male breadwinner model (Morgan, 2013).Further, as mentioned above, the two countries shared many of the key political features central to comparative explanations of work-family policy reform.In both countries, a centre-right government was voted out in the late-1990s after a long period in power, during which workfamily policy had not been a political priority, and replaced by a centre-left led government.Moreover, the new governments characterized themselves has holding a shared 'Third Way' ideology (see Blair & Schröder, 1999).In both countries, the subsequent interest that governments began to show in family policy has been explained through electoral competition (Fleckenstein & Lee, 2014;Morgan, 2013).Thus both the process of change and the political contexts in the two countries were similar; in order to explain the divergent outcomes, this article focuses on the role of ideas in these processes.
As the focus is on ideas and their impact throughout the policymaking process, the chosen research method for this study is a form of historical analysis known as 'process tracing', which involves reconstructing the processes of reform and identifying the mechanisms that underpin them (c.f.Jacobs, 2015).The period of interest in the two countries begins with the election of centre-left governments in the UK in 1997 and Germany in 1998 and ends with the onset of the financial crisis in 2008, which put an end to the initial momentum for reform.The primary research conducted for this article involved document analysis and interviews with participants in the respective policymaking processes.Documents included transcripts of parliamentary debates and committee hearings, policy position papers and manifestos from political parties, statements and speeches from politicians, policy documents and consultations published by relevant government departments and consultation responses from key stakeholders (family organizations, fathers' groups, employers' associations, trade unions).All documents were publicly available on the websites of relevant organizations, government departments and parliaments.Contemporary media reports, press interviews with policymakers as well as memoirs written by key figures were also analysed.Document analysis was complemented with 42 semi-structured interviews with politicians, political advisors, civil servants and interest group representatives who were involved in the processes of reform.Interviews were conducted in 2017 and 2018, with 22 interviews in the UK and 20 in Germany; they lasted between 30 minutes and 1.5 hours and were granted on the condition of anonymity. 2Potential interviewees were identified from the secondary literature, press reports and the document analysis.
Documents were analysed thematically, to identify the ideas involved in the processes of reform.The key themes were the 'problem definitions' employed in the two cases during the reform of leave policy, that is, the ways in which the need for new or altered policies were articulated, as this level of idea sets the possibilities for the range of policy solutions that can be considered (Mehta, 2011).Particular attention was paid to the place of fathers in these problem definitions.A second stage of analysis tracked both the extent to which these ideas served as 'coalition magnets' (Béland & Cox, 2016), through examining the extent to which they were broadly shared among stakeholders and whether there is evidence that certain groups' ideas changed over time.Further, through analysing how consistently the ideas were presented over time and the ways in which 'imperatives for reform' were framed (Cox, 2001), the extent to which ideas acted as 'roadmaps' was investigated (Goldstein & Keohane, 1993).Interview data were used to triangulate findings from the document analysis and to provide 'behind-the-scenes' information to help reconstruct the processes of reform and identify the role of ideas therein.Contemporary media reports were also used to provide context for the policymaking processes and to provide details of interactions, not documented in other sources.

Results and discussion
The UK: 'maternalist' continuity Since 1987, party strategists had been concerned with Labour's male-dominated image and the gender voting gap that had favoured the Conservative Party in the 1992 election was viewed as having cost Labour the election, its fourth successive defeat (Perrigo, 1996).This had resulted in measures to increase the representation of women at all levels of the party and a search for policies that could appeal to women voters in particular, with maternity leave highlighted by women within the party as a key area (Harman, 2017;Labour Party, 1997).
In 1997, when Labour came to power, maternity leave was meagre and highly complex: all employed women qualified for 14 weeks leave, while some qualified for 29; some women qualified for 18 weeks of low, flat-rate maternity pay, while others qualified for the first six weeks being paid at 90 percent of previous salary (DTI, 1998).There was no statutory leave for fathers.Against this background, the initial priorities for the Department of Trade and Industry (DTI), the government department responsible for labour market regulation, were to improve the basic provision of maternity leave and fulfil the European Directive on Parental Leave.The problem definition focused on the incompatibility of meagre maternity leave provision with women's employment (DTI, 1998).Prompted particularly by the European Directive, 1999 saw the introduction of parental leave, the UK's first statutory leave that applied to fathers as well as mothers; however it was unpaid and policy documents made clear that while it aimed to provide 'more choice, which will help parents, women and men, to combine work with family life' (DTI, 1998 para. 5.5), the overall approach was to provide a statutory framework as a minimum on top of which employers and employees should make their own arrangements (DTI, 1998 para. 5.5).Indeed, interviews with DTI officials revealed that paid parental leave was barely considered; it was not seen as an important issue within the department and there was not believed to be sufficient demand to implement a policy that was vehemently opposed by employers. 3No problem definition specifically related to fathers' leave was articulated during this period.
In 2000 the chancellor, Gordon Brown, announced a review of 'maternity pay and parental leave' (Hansard, 2000).This was prompted on the one hand by pressure from women within the party to 'deliver what women want' before the next general election (Harman & Mattinson, 2000, p. 29), on the other, a problem definition was set out that existing leave provision was leading to many mothers leaving the labour market at childbirth, which became a concern in the tight labour market of the late-1990s (DTI, 2000).The review involved a long process of consultation, first with focus groups of parents and subsequently with interest groups (DTI, 2000).Interviews with officials highlighted that this was an unusual 'bottom-up' approach to policymaking, initiated because ministers did not have a plan for what the final policy would look like. 4Through this process, two weeks paid paternity leave was added to the policy agenda with the aim of providing greater 'support to the mother and baby (and any other children) at a critical time and more chance for fathers to bond with their babies' (DTI, 2000, p. 58).But decommodifying fathers was not part of the agenda: despite acknowledging that paid parental leave would increase its take-up, it was rejected because of employers' 'concern that any payment would increase absence from the workplace and affect business performance' (DTI, 2000, p. 31).Both Labour's priorities and the predominant problem definition pointed towards a continued focus on leave for mothers.
After the 2001 election, Patricia Hewitt's appointment as Secretary of State for Trade and Industry marked a shift in emphasis of the problem definition associated with leave policy.Hewitt was a longstanding feminist campaigner who had written about the importance for gender equality of men taking on a greater proportion of unpaid household work (e.g.Hewitt, 1993).She aimed to reconfigure the UK's leave policy, which now provided mothers with 12 months leave but fathers only two weeks, into a more genderneutral system (Hewitt, 2014).Under Hewitt, policy documents began to refer to fathers' caring role as linked to mothers' employment: 'Supporting greater participation of men in family responsibilities is important to the objective of gender equality … Mothers still bear a disproportionate burden of caring responsibilities compared to fathers, even when both are in full-time employment' (HMT & DTI, 2003, p. 14).Hewitt commissioned research which, to the surprise of officials at the DTI, demonstrated significant appetite among fathers for a greater caring role (e.g.Hatter et al., 2002). 5Despite this, the problem definition associated with fathers' leave focused on their lack of choice; arguments linking fathers' caring behaviour to mother's labour market engagement did not feature in the major policy announcements (e.g.DTI, 2005a;HMT et al, 2004).Despite Hewitt's (2004, p. 16) position that 'if parents are to have a real choice about how they balance earning a living with bringing up childrenand not merely a theoretical choice that only the rich can exercise in practicethen they need the support of government', the policies that emerged in the 2006 Work and Families Bill continued the 'maternalist' trajectory.While maternity leave payment was extended to nine months from 2007, for fathers there was just an intention to allow mothers to transfer the final six months of leave (of which three were paid).Paternity leave benefit remained paid at a low-level flat rate, and parental leave remained unpaid, despite Hewitt's stated intention to examine the case for improving both (Hewitt, 2004).Despite the claim that transferable leave was 'responding to the growing demand from fathers to stay at home and care for their child' (DTI, 2005a, p. 49;HMT et al, 2004), documents revealed that only about one percent of mothers were expected to transfer any leave (DTI, 2005b, p. 49).
Ultimately Hewitt was unable to persuade her cabinet colleagues to support more significant leave reform for fathers for several reasons.First, employers, who had been persuaded to acquiesce to the incremental increases in maternity leave, remained implacable opponents of leave for fathers. 6The problem definition that associated increasing maternity leave with reducing the number of women leaving the labour market could not be adapted to fathers.Moreover, while Hewitt had attempted to incorporate fathers' care roles into the problem definition that focused on mothers' employment, the Labour leadership were averse to promoting policies which were seen as 'social engineering' or as representative of the 'nanny state'.Interviews with politicians, advisors and civil servants confirmed the government's caution around this accusation, meaning that any policy that was seen to incentivize fathers to take leave was considered an intrusion of the state into the family. 7In this context, Hewitt was forced to rely on a problem definition that stressed fathers' demand for more 'choice', but did little to change the UK's maternalist focus of leave policy or to significantly decommodify fathers.

Germany: a 'Nordic' turn
In contrast to the UK, in Germany parental leave had been introduced in 1986, initially for 12 months, but gradually extended to 36 months by 1992.The benefit level was set at a low-level, flat-rate and by 1993 was paid for two years, although a means-test was applied after six months.While nominally gender-neutral, the long leave duration and the low levels of pay meant that the policy's main function was to permit women to exit the labour market at childbirth for a period of at least two years, conforming to a sequential model of female employment which involved withdrawal from the labour market at childbirth and subsequent, part-time return once children reached school age (Ostner, 2010).Research has found that the proportion of women returning to the labour market decreased as the period of leave was lengthened (Vlasblom & Schippers, 2006).
In 2000 the SPD-Green government, elected in 1998, introduced a number of reforms to parental leave focused on increasing the generosity of the leave payment (which had been mandated by the Constitutional Court) and on making parental leave more flexible by introducing a right to part-time work, providing a 'budget option' of a higher benefit over a shorter period of time and permitting parents to take leave simultaneously (Deutscher Bundestag, 2000a).Family minister Christine Bergmann, claiming that a priority was to encourage more fathers to take up their entitlement (Bundestag Plenarprotokoll, 2000), faced much opposition within the cabinet and from employers over the potential costs to business (e.g.Die Welt, 1999).The eventually legislated law was weaker than that initially proposed by Bergmann, and contained no direct incentives to encourage fathers to take more leave (Deutscher Bundestag, 2000b).
After the SPD-Green government's re-election in 2002 there was a marked shift in its approach under new family minister Renate Schmidt, who since 1999 had led a project in the SPD leadership to redesign the party's approach to family policy.This project was the result of analysis of the 1998 federal election results, which revealed that young people's votes were crucial to the SPD's success and those voters were particularly concerned with work-family reconciliation. 8Schmidt's working group issued a series of publications containing problem definitions that linked families' struggles with work-family reconciliation to two outcomes that threatened Germany's future prosperity: that women were leaving the labour market in large numbers after childbirth, and that couples were choosing to have fewer children than they would like (Forum Familie der SPD, 2000, 2002;Schmidt, 2002aSchmidt, , 2002b).Germany's low birth rate and tight labour market were repeatedly stressed by Schmidt in a series of speeches, publications and public appearances (Schmidt, 2002b(Schmidt, , 2003;;Schmidt & Mohn, 2004).The new approach, which became known as 'sustainable family policy', was espoused by the Family Ministry under Schmidt's leadership between 2002and 2005(e.g. Bertram et al., 2005;BMFSFJ, 2005;BMFSFJ et al., 2004;Rürup & Gruescu, 2003).Significantly, the discussion of the problem of work-family reconciliation in these terms persuaded employers, who had opposed previous reform to parental leave, to become supporters of the new approach and Schmidt made numerous public appearances alongside leading figures from employers' associations (e.g.BDA, 2006;DIHK, 2003).Notably, this shift in employers' position was not the result of changing labour market conditions; several reports in the 1990s had highlighted the dangers of a shrinking labour force if women's labour market participation did not increase (e.g.Blau et al., 1997;DIW, 1996).Rather, it was the linking of this problem definition to 'sustainable family policy' that changed employers' perspective.
In articulating the problem, 'sustainable family policy' closely linked the role of fathers to the ability of women to achieve better work-family reconciliation.For example, an economic analysis stressing the benefits of sustainable family policy highlighted that 'just as mothers must be given an opportunity for 'more work', fathers should be given the chance for 'more family'' (Rürup & Gruescu, 2003, p. 62) as 'women's employment is improved by childcare and by making full use of men's 'care-giving capacity'' (Rürup & Gruescu, 2003, p. 63).In particular, reforming parental leave benefit to make it earnings-related and to introduce two use-it-or-lose-it 'partner months', was seen as a crucial policy, in part because it would permit more fathers to take up parental leave (Bertram et al., 2005, p. 48;BMFSFJ, 2005, pp. 284-285;Rürup & Gruescu, 2003, p. 56).Unlike in the UK, the government's definition of the problem of work-family reconciliation in Germany included fathers as well as mothers.
While the SPD had adopted this redesigned parental leave benefit, Elterngeld, as party policy, they were unable to introduce it before the federal election of 2005, which the party lost and became the junior partner in a Grand Coalition.The new family minister, the CDU's Ursula von der Leyen, surprised observers when she vigorously pushed for Elterngeld to be included in the coalition agreement, which the SPD leadership had not prioritized in the negotiations. 9Like the SPD, the CDU had, under party chair Angela Merkel, also attempted to broaden their electoral appeal by modernizing their family policy since the 1998 federal election, with a focus on enabling women to reconcile employment and family life.Policy documents highlighted that many men wanted to take on more caring responsibilities (e.g.CDU, 1999, p. 3;CDU & CSU, 2002, p. 35), although unlike in SPD documents, CDU documents stressed that women's aspirations were diverse and those that wanted to be mothers and housewives should be able to.Concomitantly, policy goals for fathers were framed in terms of meeting their wishes, rather than the SPD's aim of enabling more equal distribution of unpaid labour.Indeed, unlike the clear shift for the SPD in 2002, the CDU's policies wavered between the modernizing approach of Merkel and von der Leyen and a more conservative approach that favoured male-breadwinner families (CDU, 1999;CDU & CSU, 2002, 2005).Even in this context, von der Leyen's promotion of Elterngeld was notable, as was the fact that unlike her colleagues, who had criticized the SPD's proposals in the 2005 federal election campaign, von der Leyen had expressed 'a certain sympathy' for them (quoted in Handelsblatt, 2005).
The coalition agreement in November 2005 contained a detailed proposal for Elterngeld: a 12-month benefit paid at 67 percent of previous salary, with two months reserved for each partner, on a use-it-or-lose-it basis (CDU et al., 2005).Conservatives in the CDU and its socially conservative Bavarian sister party the Christian Social Union (CSU) challenged the proposals almost as soon as the coalition agreement was signed.In particular, the partner months were seen as an infringement of the state on families' choice of how to organize employment and care responsibilities; senior CDU figures claimed the partner months served to 'tell people how they have to live and have to organise their family' (Süddeutsche Zeitung, 2005) while the head of the CSU in the Bundestag described the partner months as a 'nappy-changing apprenticeship' and demanded their removal from the reforms (Berliner Zeitung, 2006).
However these headline-grabbing statements disguised a level of consensus among German politicians in favour of reform.Von der Leyen stressed that fathers' roles were central to the broader problems related to work-family reconciliation, making a link between men and the low birth rate: What is alarming is the partial lack of esteem for child-raising, when it is perceived as an imposition that men take care of their own child for two months.More men than women do not include children in their life planning.Something has to change here.(quoted in Vowinkel & Rübel, 2005) It was particularly notable that women in the CDU and even the CSU were in favour of the partner months, citing many of the arguments that Schmidt and the Family Ministry had been making since 2002 (Bundestag Plenarprotokoll, 2006).For example, Maria Eichhorn, the CSU spokesperson on the family, argued that 'we need emancipated men' and was one of a number of prominent women in the CSU to sign a letter to party leader Edmund Stoiber calling on him to support the partner months (Geis, 2006).Eventually, a compromise was agreed with the CSU in which the partner months would be additional to a 12-month period, rather than part of it, so that the maximum a family could receive would be 14 months.This allowed the partner months to be framed as a 'bonus' rather than a penalty, but made little difference to the overall effect of the reform, which was towards the decommodification of fathers.
Ideational change: 'sustainable family policy' as a coalition magnet and strategic road map The trajectories of Germany and the UK were similar before 2002.In both, Labour and the SPD's approach to leave policy saw it as a solution to the problem of mothers' struggles to maintain labour market attachment.For fathers, providing 'choice' was the primary goal, but interventions to incentivize fathers to take more leave were not on the agenda.When fathers were mentioned, it was in reference to 'helping' mothers or to 'playing a more active role', revealing that despite the gender-neutral language used, leave policies were fundamentally seen as policies for mothers.Despite acknowledging that the design of leave policy hindered fathers' take-up of their entitlements (to unpaid parental leave in the UK and to low-paid parental leave in Germany), in neither country were there legislative attempts to provide significant non-transferable paid leave for fathers (Bundestag Plenarprotokoll, 2000;DTI, 1998DTI, , 2000)).
However, after 2002 the approach to fathers in the two countries diverged.In Germany, the appointment of Renate Schmidt as family minister saw new ideas come onto the political agenda with the introduction of 'sustainable family policy', in which the problem of the lack of work-family reconciliation was framed as fundamental to German prosperity.Sustainable family policy conceptualized fathers as a key part of the work-family reconciliation problem: in order to enable women to engage better in the labour market, men needed to take on more caring responsibilities at home.Thus, attention to the household dimension of work-family reconciliation was an important part of the new German approach (Bertram et al., 2005;Rürup & Gruescu, 2003;Schmidt, 2003;SPD, 2001;SPD & Bündnis 90/Die Grünen, 2002).This shift marked a fundamental difference to the UK, where the focus of work-family reconciliation remained fundamentally on employment and providing families with choice was as close as the government came to actively seeking to promote fathers' childcaring role (DTI, 2000(DTI, , 2005a(DTI, , 2005b;;HMT et al, 2004).A key difference was that in Germany low take-up of leave among fathers was conceptualized as a key part of the problem of work-family reconciliation, while British documents stressed the aim of providing choice, with ambivalence towards actual take-up (Bertram et al., 2005;DTI, 2005a).
That ideas shifted in Germany but not in the UK played a crucial role in comparative policy development in two ways.First, sustainable family policy acted as a 'coalition magnet' (Béland & Cox, 2016).Employers who had been opponents of the relatively minor reforms in 2000 had, by 2006 become active promoters of the concept of sustainable family policy and were enthusiastic supporters of the more substantial Elterngeld reforms.While less enthusiastic about the prospects of men taking leave, employers were persuaded by the overall effects on labour market supply. 10By contrast in the UK, employers remained opposed to providing fathers with significant portions of paid leave throughout the period of reform.Moreover, the linking of fathers' caring roles to the broader problem definition of work-family reconciliation helped foster broad support in Germany for reform, including within conservative parties; the lack of such a broad argument in the UK is evident in Hewitt's failure to persuade her colleagues to shift approach.
Second, sustainable family policy provided German policymakers with a 'road map', that is, with a strategic goal of work-family reconciliation that helped guide political and policy decisions (Goldstein & Keohane, 1993).Policy documents consistently created an 'imperative for reform' (Cox, 2001), through stressing the problem of work-family policy reconciliation as crucial to Germany's prosperity, both in terms of economic productivity and in demographic terms.This can be seen in the consistent way in which policy documents referred to fathers and the way in which both Schmidt and von der Leyen justified leave reforms.A clear example of the road map providing a course of action can be seen in the political risks that von der Leyen was prepared to take in insisting on the importance of the partner months to the overall concept of Elterngeld in the face of fierce criticism from within her own party.Here the critical importance of improved work-family reconciliation for Germany's future, and the role of fathers therein, meant that she would not compromise with conservatives in her party, even though the shift to an earnings-replacement Elterngeld would have likely prompted an improvement in fathers' leave uptake even without the partner months.The road map of sustainable family policy thus bolstered the policy entrepreneurship of von der Leyen, enabling her to take on entrenched ideas about traditional gender roles.
The contrast of this bold policy entrepreneurship with the risk aversion with which the DTI attempted to reformulate the role for fathers in Labour's work-family policy approach is striking.Hewitt's attempt to reconfigure leave policy in a gender-neutral way in the early 2000s foundered in part because she was unable to relate it to any existing broader idea about the importance of work-family reconciliation and the interdependence of mothers' and fathers' care responsibilities.With an aversion from Labour to be seen to be engaging in 'social engineering', Hewitt had to rely on arguments that fathers wanted more leave and that the government should enable that choice.Unlike von der Leyen, Hewitt was unable to draw upon already established 'road maps' which could provide a way of challenging the entrenched liberal ideas of the UK's approach to families.

Conclusions: sources of stability and change
This article has highlighted the role of ideas in the shifting worlds of father politics.It has argued that Germany made a shift from an approach to fathers characterized by Michael Rush's 'US world' towards one more akin to the 'Swedish' world.By contrast, despite undertaking much work-family policy reform, the UK has remained closer to the 'US world' with regard to father politics.During the respective periods of reform, ideas about fathers and how they are linked to mothers' work-family reconciliation changed substantially in Germany, leading to path-shifting reforms, but did not in the UK.This concluding section suggests some explanations for why ideas changed in Germany but not the UK and ends with a call for further research into political and ideational dynamics.
As stated in the introduction, given the less conservative social attitudes in the UK, one might expect stronger opportunities for a shift in ideas about fathers there than in Germany.Moreover, the ideational context of the ruling parties in 2002 was remarkably similar (e.g.Blair & Schröder, 1999).However, a closer examination of the political and institutional context of the two cases provides two interlinked reasons for the differences.
First, the 'window of opportunity' for new ideas differed in the two countries, despite opening in both cases when party leaders became persuaded of the potential electoral benefits of work-family reform.In Germany these electoral benefits were interpreted as an opportunity to create a new agenda that the SPD could 'own' in the long-term, while in the UK they were seen in a much more transactional way: to 'deliver what women want' in exchange for votes.This is evident, for example, in the way that the incremental increases in maternity leave and pay were announced shortly before the 2001 and 2005 general elections.In part this was related to the political context in the two countries.In the UK, Labour, in response to four consecutive general election defeats, had undergone a process of 'modernisation', which culminated in the election of Blair as leader in 1994 and the rebranding of the party as 'New Labour'.By 1997, the party had a very tightly controlled policy agenda, which was designed to signify that Labour had changed and that 'old' Labour was consigned to the past (Perrigo, 1996).In this context, policies associated with gender equality and the distribution of unpaid household labour were viewed by party leaders as associated with feminism and 'old' Labour, and hence off-putting to the moderate, middle-class voters the party was targeting. 11 By contrast, the SPD, despite suffering a similar run of federal election defeats in the 1980s and early 1990s, had not undergone such a comprehensive process of internal reform, in part because the German institutional structures meant that losing federal elections did not totally exclude the party from power: the SPD held a significant proportion of regional government positions throughout the 1990s for example (Busch & Manow, 2001).However, party leaders were concerned by a lack of enthusiasm for the party and a lack of dynamism in the policy proposals presented in the 1998 federal election in particular. 12Therefore unlike Labour, the SPD was actively searching for new ideas when it conducted the research that demonstrated the electoral potential of the issue of work-family reconciliation.
Second, unlike in the UK, where the lack of interest from party leaders meant that policy reforms came onto the agenda through pressure from women MPs and external campaigners, in Germany the reforms were a top-down project, emanating from the SPD party leadership and subsequently from decisive interventions of von der Leyen.This was important in terms of the extent to which fathers featured as a priority in the reforms and the way in which policy legacies impacted on the trajectory of reform.In the UK, the very meagre provision of maternity leave in 1997 meant that for women MPs and campaigners, the priority was urgent improvement in the provision for mothers.Furthermore, there was no significant voice for fathers' leave from campaigners (Hewitt, 2014).By contrast in Germany, 'sustainable family policy' highlighted that the existing system of long, low-paid parental leave was, in part, itself to blame for workfamily reconciliation problems.Thus, while in the UK, policy legacies helped shape a pattern of incremental, maternalist leave reform, in Germany policy legacies provided the opportunity for path-shifting change.
These points link the emergence of ideas and their influence on policymaking to wider political and institutional factors.Such an approach chimes with Rush's approach to the two 'worlds', which was considerably broader in scope than just examining fathers' access to leave policy.In order to analyse fully the shifting father politics of former male-breadwinner model countries and the changing ideational patterns explored in this article, further research could examine whether such changes were evident beyond fathers' entitlements to other social rights, for example in the ways in which states enforce fathers' responsibilities.

Table 1 .
Social attitudes in Germany and Great Britain, European Values Survey (EVS) 1999 and 2008.