18 The German Welfare State

This chapter asks what characterizes the German welfare state today and to what extent it has been fundamentally transformed during the last thirty years since reunification. Looking further back, we can see that the history of the German welfare state has always been one of striking institutional continuity in the face of massive political and economic turbulence (Kaufmann, 2013). The pioneering Bismarckian reforms introduced social insurance for industrial workers in case of sickness (1883), work accidents (1884), and old age and invalidity (1891). The subsequent welfare state expansion over the course of almost a century tended to follow a path of increasing generosity and incremental extension of that initial institutional template to new groups and new kinds of risk (at least in West Germany). All major reforms before 1990—apart from universal child benefits and social assistance—took place within the framework of male wage earner-oriented and mostly status-preserving social insurance. Bismarck’s imprint survived two World Wars, three regime changes, and several massive economic shocks. Even the retrenchment after the oil price shocks remained—compared to what happened in Britain under Thatcher, for example—gradual: a ‘smooth consolidation’ (Offe, 1991) within the confines of the dominant social insurance path.¹

At that time, (West) Germany continued to be the prototype of Gøsta Esping-Andersen’s ‘corporatist’ or ‘conservative’ welfare regime, which is to be distinguished from the egalitarian ‘social democratic’ regime of Nordic countries and the more market-oriented and residual ‘liberal’ regime mostly of English-speaking countries (Esping-Andersen, 1990). Social policy in the conservative model is not aimed at reducing but preserving traditional status differences in society via social rights linked to previous earnings and occupational status (and status as male breadwinner). Scholars of political economy research, moreover, have highlighted the complementarities of status-preserving arrangements with the ‘coordinated’ nature of post-war German capitalism (Hall and Soskice, 2001; Manow, 2020). Redistribution in this regime is therefore limited and ‘stratification’ by the welfare state is high, notably according to occupation. Moreover, the conservative model does not assign welfare production solely to the state, but to a significant degree also to the family (via unpaid female care work) and to non-profit, including faith-based organizations. Other conservative welfare states are France, Belgium, and Italy.

However, the literature is divided on what happened after 1990. Some scholars emphasize continuity (Anderson, 2015; Clasen and Goerne, 2011; Leisering, 2016), others transformation and paradigm change (Bleses and Seeleib-Kaiser, 2004; Bosch, 2015; Hinrichs, 2010). I will argue that policy changes since 1990 combine continuity and change in the specific sense that, while the inherited ‘shell’ of a Bismarckian order still seems intact, it no longer delivers on its traditional goals, or has been charged with new tasks. Part of this is the product of ‘policy drift’ (Hacker, 2004), that is, insufficient updating to harsher social and economic realities. Deliberate reforms since 1990, as will be shown, have sometimes reinforced, sometimes alleviated such outcomes.

The chapter is structured as follows. To set the scene, I will start with locating the German welfare state within the group of Organisation of Economic Cooperation and Development (OECD) democracies, using the latest available data. I will then describe the most important welfare state changes since 1990, tracing a sequence of problem-solving attempts. By no means does this imply that problems are beyond contestation and somehow objectively given, nor that the solutions offered were without alternative or always effective. On the contrary, political actors likely perceive problems and choose solutions based on their ideological leanings and cognitive biases and their problem-solving is part of an interaction with others within a specific institutional setup, a dominant discourse and shifting public opinion. Starting with German reunification, the narrative is loosely chronological, dealing subsequently with welfare state reforms that responded to mass unemployment, demographic change, female labour market participation, and a succession of crises. I will then briefly address the question of whether the German welfare state delivers on key social and economic outcomes before concluding.

The German welfare state at first sight looks strikingly unremarkable when compared to its peers (for similar assessments, see Alber, 1998; Obinger, 2014). In terms of social expenditure, it is clearly above average, but not among the very top spenders in the group of rich OECD countries (see Figure 18.1). With 25.9 per cent of GDP, current German social spending is slightly lower than at its peak in 2009 (26.8 per cent).

The composition of public social spending (see Table 18.1) is again close to the average OECD country, except for survivors’ benefits and the health care branch which, relative to the size of the economy, is among the most expensive public health systems in the OECD. Perhaps surprisingly, given Germany’s demographics and its ‘conservative’ welfare state legacy, social spending is not particularly biased towards the elderly (Vanhuysse, 2013) and the relative spending share going towards old age is below average.

How generous are individual benefits? The Comparative Welfare Entitlements Dataset reports comparative net replacement rates, that is, the average share of previous income typically replaced when unemployed, sick, or retired. Most of the data extend no further than the year 2010, but the OECD reports unemployment benefit replacement rates up until 2019. Note that these data must be interpreted with caution, because they are based on model households and do not reflect the actual distribution of generosity across all households. Table 18.2 demonstrates that Germany occupies a somewhat incoherent position with respect to benefit generosity. Clearly in the top group for sickness insurance and still quite high up in standard pensions for singles, Germany is average at best when it comes to providing for the unemployed and distinctly ungenerous in pensions, apart from singles (OECD, 2019a). This suggests a welfare state characterized by serious horizontal inequities in terms of risks and household situations, a characteristic I will come back to below.

Over the last thirty years, the German welfare state has dealt with a number of distinct challenges, that is, problem complexes that were dominant in the public discourse for several years. In what follows I will discuss these dominant problems and reforms in roughly chronological order.

The single most important event for the German welfare state during the last decades was the reunification of 1990. Its significance remains somewhat underappreciated by the comparative literature. Prior to reunification, social policy in East and West had evolved in different directions for forty years. Compared to West Germany, the socialist GDR provided ‘basic security at an austere level’ (Manfred G. Schmidt, 2005, p. 132), accompanied by a multitude of special schemes, privileging certain occupations and politically important groups. In addition, female employment was actively supported through comprehensive child care. Low public spending masks a great deal of ‘welfare production’ provided by firms and via subsidies and price controls. A de facto job guarantee coupled with rising minimum wages also lowered the need for working age benefits—and explains why the GDR did not have an unemployment insurance scheme.

Reunification of the two states was executed through the wholesale accession of the GDR to the Federal Republic of Germany (FRG) which required the mammoth task of transferring the entire body of law to the East. In the case of social legislation, this largely happened already in the summer of 1990. Moreover, to prevent mass migration from East to West, the new rules were rolled out with only few transitional provisions.

An important strategic decision of the Christian democratic-liberal coalition of Chancellor Helmut Kohl (in office: 1982–98) was to finance reunification to a large extent through the social insurance system. This happened partly because the new citizens in the East were immediately entitled to benefits equivalent or near-equivalent to those available in the Western part.² In addition, special labour market schemes were set up to deal with the virtual collapse of the no longer competitive East German industry, with the consequence that social expenditure in Eastern Länder peaked at a staggering 68 per cent of (East German) GDP in 1992 (Manfred G. Schmidt, 2005, p.147). Active labour market policy spending alone reached 18 per cent of GDP in the East (Ritter, 2013, p. 233).³ Although the effectiveness of many retraining and job creation schemes is doubtful, to say the least, the welfare state became the key shock-absorbing institution to compensate for the tremendously disruptive process of extending the West German economic and monetary system eastward within such a short period of time.

As if reunification alone was not a herculean task, several external shocks and domestic debates converged in the early 1990s to seriously challenge the status quo. In 1992–93, the initial unification boom turned into a recession. Deepened European integration increasingly revealed the missing ‘social dimension’ of that process. The 1990s also saw the beginnings of the globalization debate (see Seeleib-Kaiser, 2001 on the origins), discussed in Germany under the headline ‘Standort Deutschland’ (‘Germany as an investment location’). High non-wage labour costs were soon singled out as one of the key competitive disadvantages. And as wages or productivity are difficult to influence directly, high social contributions became the prime indicator of a need for radical reform.

Figure 18.2 shows the development of total social contributions—jointly paid by employers and employees—since 1970.⁴ While rising social contributions were nothing new (Manow, 2020), the level quickly increased in the 1990s. Reunification alone probably added 3 percentage points (Bönker and Wollmann, 2000, p. 517) and the addition of Long-term Care Insurance in 1994 another 1.7 points (see the section on demographic change below). In the context of the ‘Standortdebatte’ about globalization and European integration arguments about excessively high contributions proved particularly powerful (Scharpf, 2000; Streeck and Trampusch, 2005).

At the same time, political problem-solving capacity was at a low point. The traditional ‘welfare state consensus’ between the main political parties, employers, and trade unions fractured, as evidenced by the Kohl government’s failure to negotiate a quasi-corporatist social pact in 1996 (the first ‘Alliance for Jobs’). What is more, the relationship between government and opposition became increasingly adversarial, which led to frequent stalemate due to the many veto points of the German political system.

Still, change did happen. And while the Centre-Left opposition fiercely fought most reforms, especially the 1997 pension reform, once in office they nonetheless largely continued on the path of ‘revenue-oriented fiscal policy’ initiated by Kohl. Figure 18.2shows that, from the late 1990s onwards, benefit cutbacks, increased tax subsidies, and a more favourable economic situation eventually stabilized and then lowered contributions.

As social contributions rose, so did unemployment (see Figure 18.3). Starting at a level of just over 5 per cent, the harmonized unemployment rate peaked at 11.3 per cent in the mid-2000s (and almost twice that in the East). At the same time, the US and Denmark—two key reference countries—were experiencing labour market ‘miracles’ with rates around only 5 per cent. Accordingly, unemployment was clearly the single most salient political issue in opinion surveys from 1993 up to 2010 (Forschungsgruppe Wahlen, 2017). The picture after the Global Financial Crisis (see the section on crisis management below) was starkly different, even though the turnaround already started in 2005.

Traditionally, Germany had a two-tier unemployment benefit system: the first tier (Arbeitslosengeld) was a contribution-funded, earnings-related benefit at 63–68 per cent of previous income. The second tier (Arbeitslosenhilfe) was aimed at the long-term unemployed no longer eligible for Arbeitslosengeld. As an earnings-related, but tax-funded and means-tested benefit (replacing 53–58 per cent of income) it was a strange hybrid between an insurance and assistance benefit. It nevertheless embodied the conservative notion of status preservation—linked also to Germany’s skills system (Estevez-Abe, Iversen, and Soskice, 2001)—because long-term unemployed could receive earnings-related benefits indefinitely. This made it a route to de facto early retirement so that, in the 1990s, the number of beneficiaries in the second tier grew massively. Below the second tier social assistance (Sozialhilfe) existed as a benefit of last resort.

In 1996, the Kohl government proposed significant structural reforms limiting access to Arbeitslosenhilfe, redefining acceptable employment, decentralizing employment services, and rolling back some of the public employment measures used in Eastern Germany. The opposition tried to block these measures in the Bundesrat, the second chamber of parliament, with only partial success (Hassel and Schiller, 2010, pp. 100–6). However, partisan conflict (and to some extent conflict between the federal level and the Länder) still impeded more far-reaching change.

During its first years in office, the government of Gerhard Schröder (1998–2005) initially reversed cuts in sick pay and labour market deregulation. Yet these early expansionary decisions masked a deep conflict within the Social Democratic Party (SPD). When the Left’s figurehead Minister of Finance Oskar Lafontaine stepped down in 1999 (and founded the party The Left), things started to turn. After a second tripartite social pact (also called ‘Alliance for Jobs’) failed, Schröder seemed to have come to the conclusion that unilateral action would be preferable to consensual decision-making (read: trade union support). A public scandal in 2001 that the Federal Employment Service’s placement statistics had been grossly misleading created a rare opportunity for such action.

The Hartz reforms of 2003–05 comprised four laws with focus areas from governance of the public employment service, through the deregulation of so-called ‘mini-jobs’ to a restructuring of unemployment and social assistance benefits. Sidelining traditional concertation with social partners, the reform was prepared in a series of expert commissions (Dyson, 2005), the most important of which was chaired by Volkswagen top manager Peter Hartz (for details, see Hassel and Schiller, 2010). The aim was to activate the long-term unemployed through the ‘carrot and stick’ of individual support and obligations—or Fördern und Fordern, as the slogan went at the time. The duration of the first-tier insurance benefit was reduced from a maximum of thirty-two to eighteen months (later extended to twenty-four months). The second-tier Arbeitslosenhilfe was abolished as it was merged with social assistance. The consequences were wide-ranging. Benefit rates decreased especially for long-term unemployed with relatively high previous earnings, but they increased for others, especially families with children and low-income earners. Ironically, the number of registered unemployed rose at first (Bosch, 2015, p. 188) because social assistance beneficiaries who were able to work were being ‘activated’ and because take-up was higher than for the older social assistance. The definition of an acceptable job offer was widened and long-term unemployed received extra support for starting a business. Along with reforms on the benefit side came massive organizational restructuring and the attempt to outsource parts of employment services to private providers (reversed in 2008). Interestingly, the reforms also led to a reduction in active labour market spending per unemployed (OECD, 2016b) which shows that for the most part ‘stick’ trumped ‘carrot’.

In addition to activation and some benefit retrenchment, the German labour market was deregulated in the 1990s, especially at the margins, that is, for temporary agency work and marginal part-time jobs (so-called ‘mini jobs’ which fall below legally defined earnings thresholds and are not subject to social contributions). The OECD’s 5-point ‘employment protection legislation’ index for Germany shows almost perfect stability at a moderately high level for regular contracts (moving only from 2.58 to 2.68 on a five-point scale between 1990 and 2013) but a massive drop in the regulation of temporary contracts (from 3.25 to 1.13 during the same period). This deregulation came in several incremental steps, under both Kohl and Schröder (Eichhorst and Marx, 2011). The result was a considerable ‘dualization’ of the German labour market, further fuelled by the deepening segmentation of German industrial relations into a well-covered manufacturing core and a weakly organized service sector.

To what extent the Hartz reforms were successful is still debated. Many observers emphasize low unemployment and one of the highest employment rates in the OECD (currently at 76 per cent), others criticize rising inequality, a low-wage sector of Anglo-Saxon dimensions, and the ineffectiveness of certain parts of the Hartz package. Moreover, the ‘job miracle’ that followed the Hartz reforms happened in the context of an exceptionally favourable world economy (for Germany at least) and a decade of wage moderation. It is important to note that in response to some of these negative effects and political discontent there have been numerous part-reversals after 2005. Benefit duration was extended, private placement services largely discontinued, and start-up grants made more stringent. Since 2015, there is even a statutory minimum wage, previously anathema in German labour market policy (Marx and Starke, 2017).

With an elderly population share of 21 per cent, Germany is now among the ‘oldest’ countries in Europe (OECD, 2018a), with the number of people needing care projected to rise from 2.6 million currently to 3.5 million by 2030. Not surprisingly, such outlooks—combined with the immediate concern about rising contributions—dominated the German pensions debate of the 1990s and early 2000s. The extent of policy change in pensions during that period is perhaps underrated by the wider public, especially compared to the Hartz reforms. A first major structural reform limiting future spending increases was, ironically, passed by the Bundestag on 9 November 1989, the day of the fall of the Berlin Wall. Soon, it was considered insufficient and more radical ideas were put on the agenda. The politics of old age care became more conflictual, and the old pension consensus came to an end. The next reform step was the 1997 pension reform, which introduced a ‘demographic factor’, again to dampen future pension increases. The opposition of Social Democrats and Greens criticized and then repealed the provision in 1998, only to reintroduce it under a different name shortly afterwards. But not only that. A series of major reforms between 2001 and 2004 changed the pension formula in a dramatic, but highly technical way.

As a result of these reforms, pension increases no longer follow wages so that the standard pension level⁵ decreased from about 53 per cent of previous gross earnings⁶ to 48.2 per cent currently and without further policy changes is expected to hit 44.6 per cent by 2031 (Bundesministerium für Arbeit und Soziales, 2017). This is widely seen as a break with the status preservation goal, the ‘primary legitimation basis’ of the German old age pension system (Leisering, 2016: 227). And yet, the government’s reasoning was that much, if not all, of that drop could be offset with additional private provision. Retrenchment in the public pension pillar was thus accompanied by subsidized, yet voluntary private pensions in the form of the so-called ‘Riester-Rente’ (unofficially named thus after the Minister of Social Affairs). Expectations were not met, however—far from it. Over 16 million eligible pension contracts are currently in place, but it is also estimated that perhaps as many as a fifth of these are currently inactive due to lack of ongoing contributions.

These were just the more prominent reforms of that period, together with the rise in the statutory pension age to 67 years (phased in until 2029)—clearly one of the least popular decisions of Angela Merkel’s first coalition. Numerous further changes were introduced, from organizational changes to the closing of early retirement opportunities. Germany, one-time champion of early retirement now has one of the highest older employment rates in Europe. A key result of all these reforms is that public old age spending against global trends has been decreasing for a number of years—from 11.3 per cent of GDP in 2003 to 10.2 per cent in 2017—which, given continued demographic ageing is remarkable (Manfred G. Schmidt, 2012). Yet sharp cutbacks have, predictably, led to gaps in the adequacy of old age benefits especially for women, some self-employed, and marginal workers. While the current situation looks manageable at first sight with old age poverty still below average—although significantly ‘gendered’ (OECD, 2017)—the number of older social assistance recipients is rising and prospects for many, especially those at the margins of the labour market, look decidedly bleak.

Political parties have begun to address these worries incrementally. The first break with the purely contribution-oriented strategy was the decision to temporarily loosen scheduled pension level ‘brakes’ (which prevent benefits to rise in line with wages) in 2008 and 2009 and introduce a protection against nominal pension cuts. In 2014, even early retirement at age 63 (under certain conditions) was reintroduced and women with employment gaps due to child-rearing received additional pension top-ups. After protracted negotiations, a minimum pension was legislated in 2020 which tops up low pension benefits, but eligibility is tied to having a relatively long, but ultimately insufficient contribution record in the public pension insurance.

There is no space to delve into the highly complex changes in health care and long-term care. A few key trends and events must suffice. Traditionally, German health insurance was the epitome of a highly stratified ‘conservative regime’ with more than 1,000 occupationally-based funds with different contribution rates by the early 1990s. Only white-collar workers could choose between funds. An additional stratifying feature of German health insurance was—and still is—that high-earning employees have an exit option out of statutory insurance and into a separate private insurance system. The corporatist, non-interventionist character of the German system lies also in its degree of self-regulation by insurance funds and provider representatives. Since the 1990s, this system has come under threat due to a simultaneous increase of competition and hierarchical regulation (Rothgang, Schmid, and Wendt, 2010). The market mechanism entered with the Health Care Structure Act of 1993 which broke with occupational stratification and introduced free choice of insurance funds (with an accompanying system of risk redistribution between insurers). The number of insurance funds consequently fell to just under 100 (mostly through mergers). Reforms in the 2000s brought in new forms of cost control for hospital care (e.g. via diagnosis-related groups) and experiments with differentiated service packages, but also benefit cuts and new co-payments. Public health spending had massively increased after 1990, from 6.1 per cent of GDP to 8.1 per cent by 1996 (OECD, 2016b). And again, rising contribution rates became a focal point of reform. In 2005, employer contributions were de-coupled from increases to control non-wage labour costs.⁷

In 1994, after complex negotiations, the Kohl government introduced the Long-term Care (LTC) Insurance. As with health reforms, demographic changes (as well as rising female labour market participation) loom large in explaining that reform, but the key proximate cause was the overburdening of municipal and state budgets with costs for eldercare (Götting, Haug, and Hinrichs, 1994). Social assistance was increasingly used to cover the cost of long-term care of the elderly and municipalities and state governments lobbied the federal government for a bail-out. The solution, a mandatory social insurance scheme funded from social contributions and with free insurance of family members, at first sight confirms the strong normative power of ‘Bismarckian’ ideas (Evers, 1998). However, underneath the hood, LTC insurance is much more of a hybrid. In contrast to health insurance, it covers needs only partially, through capped benefits. Recipients have the right to choose at two levels. First, benefits can either be paid out in cash to informal caregivers (often family members) or provided in kind by residential or home-care providers. Second, those providers—which are overwhelmingly for-profit and non-profit private organizations—compete for care contracts with consumers. They are reimbursed by the LTC insurance funds, but significant private co-payments remain, forcing many to resort to social assistance. To sum up, in the spirit of 1990s social policy, the universal LTC insurance was designed as a partial insurance embedded within a market. Its flat-rate benefits are a far cry from the status-preserving ideal of the ‘conservative welfare regime’. Still, LTC insurance was an important innovation and, at the time, ‘virtually unparalleled in Western welfare states after 1975’ (Götting et al., 1994, p. 288). Since 1994, spending and contribution rates have increased and issues like cost and quality of eldercare have moved up the political agenda.

The conservative welfare regime traditionally depends on informal and unpaid female care work and the narrow role of the father as the principal wage earner. The state has a hands-off role and support for families comes through cash transfers rather than services. In Germany, gender roles were remarkably resilient to cultural change and rising female education levels for a long time. Early reforms in the 1980s introduced a parental leave scheme with low flat-rate benefits, but the changes were still in line with the male breadwinner model. By the 1980s female participation rates were seriously lagging the trend in other OECD countries at least in West Germany—East Germany looks very different in that respect. The child care services gap was particularly serious for under-3-year olds, whereas provision for children from about age 4 had been expanded significantly already from 1970s onwards. Attendance rates for that age group were high even by international standards. A right to child care from age 3 until 6 was legislated in 1992 in parallel to a reform of abortion rights and also influenced by much higher acceptance of public child care in East (Meyer, 1996). However, kindergartens often had short and inflexible opening hours, which—combined with one of the largest gender wage gaps—prevented mothers from working full-time.

Things began to move only in the 2000s and, to the surprise of many, it was a Christian democrat-led government that implemented major reforms. Angela Merkel’s grand coalition government designed the new parental leave benefit (Elterngeld) as an earnings-related benefit, replacing 67 per cent of former income up to a limit of €1,800, for twelve months. It was deliberately intended to be attractive for well-educated middle-class families. An extra two months are granted if care is shared between both parents. To achieve lasting change, public child care, especially the gap in provision below age 3 in West Germany, needed to be addressed, too. The constitution prevents direct provision by the federal level so that special grant schemes to the states and municipalities were used, from 2013 onwards also complemented by a right to child care for small children. As a result, enrolment rates of under-3 year olds have soared from 13.6 per cent in 2006 to 32.9 per cent in 2015 (Bundesministerium für Familie, 2016a).

What explains the puzzle that Christian Democrats, the traditional champions of the male breadwinner model, in concert with Social Democrats led family policy towards the ‘adult worker’ model? The literature highlights the role of party competition for the female, urban vote in the 2005 election as the trigger of this change of heart (Fleckenstein and Lee, 2014; Morgan, 2013). Moreover, an element of policy learning from Nordic countries which combined high employment rates with high fertility is evident, as the Elterngeld is closely modelled on the Swedish policy of the time. Much of the oft-cited cultural shift in favour of working mothers, however, has to do with reunification rather than changing individual attitudes, as socialization in East Germany still has a large effect on gender-egalitarianism and support for public child care (Goerres and Tepe, 2012).

Total spending on families has not really changed since 1990, fluctuating around 2 per cent of GDP (OECD, 2016b). Yet the share of service spending has increased, from about a quarter to a half, reflecting a shift in the role of the state from passive income support to provision of education and care. As a result, we can see movement towards the dual earner/dual carer family model in Germany, but within limits. While maternal employment in 2014 stood at 69 per cent, only 30 per cent of German mothers with children below age 14 worked full-time. For Denmark, the rates were 82 and 72 per cent, respectively (OECD, 2016a). And while fathers’ leave-taking before the introduction of the new parental leave law was below 5 per cent, a third of fathers now take at least some of their entitlement. Yet 79 per cent of fathers who take leave go back to work after two months or less, which indicates that the new norm is to only make use of the two bonus months, not more (Bundesministerium für Familie, 2016b).

Welfare state development in recent years was shaped by the succession of transnational crises, beginning with the Financial Crisis and the Great Recession, through the Eurocrisis and the 2015 Refugee Crisis to the Covid-19 pandemic. During each of these shocks, the German welfare state has played a major direct or indirect role. Especially in the Great Recession and the Covid-19 crisis, that role has been one of a stabilizing crisis management tool (Blum and Kuhlmann, 2016; Starke, 2014). The built-in automatic stabilizers of the social insurance system were on both occasions complemented with discretionary stimulus packages and improved access to social benefits. One policy tool, in particular, needs to be highlighted as it has become the epitome of German crisis management: short-time work (STW) (Kurzarbeit). STW has become the vehicle of choice, also because it fits well with the German ‘coordinated’ production model. It allows for sweeping working time reductions at the company level and partly compensates employees for the ensuing loss of income. This buys employers time to adjust production while avoiding mass layoffs—especially of workers with sector or firm-specific skills. In 2008 and early 2009 trade unions and employers in the manufacturing sector successfully lobbied the government to expand the scheme, so that take-up soon peaked at 1.5 million or 5.1 per cent of the labour force (Brenke, Rinne, and Zimmermann, 2013). In 2020, the use of STW even surpassed these levels, also thanks to relaxed eligibility requirements (Eichhorst, Marx, and Rinne, 2020). Part of the effect was that the rise in unemployment (both in 2009 and 2020) was small compared to the US, for example. The decisive welfare state-based crisis response limited the immediate social fallout from the crisis so far: Incomes have been stabilized to a remarkable extent, counteracting a rise in inequality (Bruckmeier et al., 2021).

While it rarely competes with the Nordic countries for the top spots in international rankings, the German welfare state achieves decent outcomes on several dimensions. The latest Social Justice Index Report by the Bertelsmann Foundation ranks Germany in tenth place in the EU and OECD, based on a wide range of indicators (Hellmann, Schmidt, and Heller, 2019). But upon closer inspection the picture is quite mixed, also when looking at changes over time. The most remarkable positive development of the last decade, without doubt, concerns labour market performance. The times of ‘welfare without work’ (Esping-Andersen, 1996) are over. Yet there are visible cracks in social cohesion. Relative poverty rates have increased from 11.6 to 15.8 per cent of the population between 1995 and 2014 (Bundesministerium für Arbeit und Soziales, 2015, p. 551) with a high concentration among, for example, single parents. Moreover, twenty-eight years after reunification, many indicators of socio-economic deprivation look significantly worse in East Germany. Income inequality is also on the rise, which is mostly due to sharp rises in market inequality (now reaching US levels!), not welfare state cutbacks. If anything, the German welfare state has become more, not less redistributive over time, measured in terms of ‘Gini-reduction’ (Caminada, Goudswaard, and Wang, 2017), thus defying its ‘conservative’, status-preserving heritage. Against the background of the extremely expensive German health care system, broad health outcomes in the population are fortunately comparatively good (Eurostat, 2017). The resulting picture is one of a highly interventionist and equalizing welfare state within the context of an increasingly unequal political economy.

When it comes to legitimacy of the German welfare state, surveys show that the German public overwhelmingly (i.e. 80 per cent and higher) supports state intervention in economy and society for social policy goals (Andreß and Heien, 2001; Ullrich, 2008). Citizens in the Eastern Länder are particularly interventionist. At the same time, perceived flaws and injustices of the existing welfare state are seen critically so that the support of the status quo is remarkably low (Ullrich, 2008). Little evidence, however, can be found of the often evoked ‘generational conflict’ in attitudes towards pensions or family policy (Goerres, 2008).

Is Germany still a conservative welfare state? It is fair to say that Germany remains a ‘social insurance state’. The five big insurance schemes—pensions, health, work accident, unemployment, and long-term care—are still dominant in terms of coverage and fiscal weight and social contributions provide most of the funding. And yet, while the ‘shell’ of the social insurance state is intact, policy drift as well as deliberate policy innovation have changed its nature and impact—arguably much more so than in the other ‘conservative’ welfare states such as Austria, Belgium, and France.

New goals and instruments have either replaced or supplemented the traditional aims of the conservative welfare state. First, the status preservation of the male breadwinner is in demise. The Hartz reforms in unemployment benefits, pension reductions, and flat-rate benefits in eldercare are cases in point. Second, while stratification based on occupational categories has disappeared in health care and pensions, new horizontal inequities have emerged, based on varying generosity of different benefits and the differential treatment of the marginal (but numerically quite large) group of labour market outsiders. Third, traditional governance based on hierarchical and corporatist mechanisms has been supplemented with internal markets and/or regulated private provision, notably in health and eldercare as well as in pensions. Fourth, the welfare state has shed its conservative non-interventionism in family matters and shifted from a purely transfer-based to a more service-based model with incentives for female employment. Finally, higher redistributive effort may also be counted as a sign of new priorities and of a move away from ‘security’ as the overarching goal of the welfare state and towards minimum protection and ‘social investment’.

The wave of market-liberal ideas in social policy—across all major political parties—clearly left its mark on reforms up until the mid-2000s. Not only labour market policy, but also a good deal of family policy and even pensions are increasingly redesigned to raise the employment rate. Together with policy drift stemming from a liberalized economy and labour market, and an increasingly segmented wage-setting system, we are dealing with what has been termed an ‘exclusive Bismarckian model’ (Bosch, 2015), which allows for a much higher degree of inequality in the labour market and, despite considerable redistribution, tends to reproduce divisions between labour market insiders and outsiders.

However, there is no reason to think that this model is stable, either. Since the late 2000s, the neo-liberal wave has rolled back, prompted by some of the negative consequences of earlier reforms combined with a stronger economy The introduction of a statutory minimum wage—also a decidedly pro-insider policy—in 2015 was an important example of self-correcting dynamics (Marx and Starke, 2017) as was the basic pension of 2020. The German welfare state is no longer politically under attack as it was in the 1990s and is seen more as a part of the solution than a part of the problem. And yet, this seeming stability should not be taken for granted. More than any other German institution, the welfare state has been handed the enormous task of holding an increasingly fragmented and unequal society together.