Roads and Barriers towards Social Investments: Comparing Labour Market and Family Policy Reforms in Europe and East Asia

ABSTRACT

Across the OECD world, social investment policies are on the rise, which Hemerijck describes as a ‘quiet paradigm revolution’. Whilst Nordic countries are typically considered the pioneers in social investment policies, we observe that latecomer countries of not only Europe but also East Asia have made considerable efforts to catch-up with Northern European frontrunners. The rise of social investment policies, especially the expansion of family policy presents an important dimension of the recent transformation of advanced welfare capitalism, which despite the prominence of retrenchment cannot be reduced to welfare state regress. However, we observe great cross-national variation in the speed and scope of family policy expansion. Unlike family policy, labour market policy did not experience a similar social investment turn, but is instead rather characterised by retrenchment with declining efforts to improve the employability of the unemployed and labour market outsiders. In this article, we examine the ‘uneven’ social investment turn in advanced welfare capitalism and argue that family and labour market policies, and their very different outcomes, are underpinned by very different political dynamics, rather than by ‘a politics of social investment’. Not only comparing family and labour market policy but also comparing across countries within each policy domain, we analyse the roads and barriers towards greater social investments.

KEYWORDS:

• Social policy
• social investments
• labour market policy
• family policy
• Europe
• East Asia

Across the OECD world, social investment policies are on the rise, which Hemerijck (2015) describes as a ‘quiet paradigm revolution’. Whilst Nordic countries are typically considered the pioneers in social investment policies, we observe that latecomer countries of not only Europe but also East Asia have made considerable efforts to catch-up with Northern European frontrunners. The rise of social investment policies, especially the expansion of family policy presents an important dimension of the recent transformation of advanced welfare capitalism, which despite the prominence of retrenchment cannot be reduced to welfare state regress. However, we observe great cross-national variation in the speed and scope of family policy expansion. Unlike family policy, labour market policy did not experience a similar social investment turn, but it is instead rather characterised by retrenchment with declining efforts to improve the employability of the unemployed and labour market outsiders.

What are the political dynamics driving social investment policies? Not only comparing family and labour market policy but also comparing across countries within each policy domain, we find considerable variation in policy development, which allows us to analyse the roads and barriers towards greater social investments. And by including ‘deviant’ cases (i.e. cases that do not comply with the general trend in each of the policy domains), we are in a position to better capture the complexity of the political drivers in family and labour market policy reform. Rather than suggesting ‘a politics of social investment’, we argue that family and labour market policies and their very different outcomes are underpinned by very different political dynamics. In family policy, the social investment turn in Germany and South Korea will be investigated and compared against the ‘deviant’ Japanese experience, where an ‘early’ social investment turn came to a ‘standstill’ and family policy returned to more traditional family support policies (namely, the expansion of child benefits). We show the critical importance of electoral competition in Germany and Korea, where political parties on the left and right ‘modernised’ their family policy portfolios in order to attract young voters and young women voters in particular. In the German case, we also find employer support for employment-oriented family policy reforms suggesting the emergence of a new cross-class coalition, whereas business remained critical in Korea. By contrast, electoral competition did not fuel family policy modernisation in Japan. Instead, the persistence of conservative value orientation in the Japanese electorate (including young people) facilitated the returning to general family support policies by both the political right and left.

In labour market policy, we find that Sweden departed from its ‘social-democratic’ trajectory after the economic crisis in the early 1990s. Whilst retrenchment in unemployment protection and a decline in training driven by the political right might not come with much surprise, it is puzzling that social democracy pursued very similar policies after the economic crisis. Here, business breaking with the post-war, cross-class compromise appears critical, with huge impact on social-democratic policy-makers and their perception of the feasibility of the Swedish model. Likewise, neoliberal business mobilisation in Germany and the associated political confrontation are key to understanding the failed ‘social-democratic’ turn in labour market policy when social democracy re-entered government office in the late 1990s. As in Sweden, we observe the programmatic differences between the political left and right diminishing in labour market policy, which translated into far-reaching workfare activation in the 2000s. By contrast, Italy offers innovation in labour market policy. Not unlike Sweden and Germany, the country has experienced significant deregulation at the margins of the labour market, which has driven a huge increase in atypical employment, especially temporary agency work. Responding to this dualisation, organised labour has successfully campaigned for investing in the skills of temporary agency workers – i.e. Italy presents a ‘deviant’ case of investing in labour market policy. However, instead of relying on the state, unions, in the industrial relations arena, have managed the establishment of a training fund jointly run with employers, which provides temporary agency workers with exemplary access to training by European standards.

The remainder of the article is structured as follows: We start with developing the link between social investment and the activation discourse in the welfare state reform debate. It is distinguished between enabling and workfare activation, with the former corresponding with the social investment paradigm. Investigating the ‘uneven’ social investment turn in the two policy domains, we first analyse the politics of investing in families, before turning towards labour market policy and the politics of investing in labour market outsiders. With our wide selection of countries, we provide insights into social investment policies in different welfare regime contexts. Whilst Sweden is considered the archetypical case for social democratic welfare, Germany presents the prime example for conservative-corporatist regime (Esping-Andersen, 1990); and Italy, a variant of conservative welfare, represents the Southern European model of welfare, work and the family (Ferrera, 1996). Japan and South Korea are commonly considered productivist or developmental welfare states, with the former widely seen as the prototypical developmental welfare state but the latter presenting productivism in ‘purer’ form (Holliday, 2000; Ringen, Kwon, Yi, Kim, & Lee, 2011).

Social investment and activation: the ‘uneven’ social investment turn

Post-war welfare states of the Golden Age with their focus on social protection have faced increased criticism that they are ill-equipped to cope with the more complex challenges of post-industrial societies and the associated rise of new social risks (such as work-family reconciliation and low or obsolete skills). In both academic and political debates, the argument on overcoming the passivity of welfare states gained momentum from the mid-1990s. ‘Passivity’ refers to the observation that the welfare state was primarily ascribed the responsibility of providing benefits in the event of income loss (e.g. unemployment and old-age), whilst efforts and incentives to encourage participation in employment and to facilitate the reintegration into the labour market were by and large underdeveloped. This criticism of the ‘activation discourse’ was in particular targeted at labour market policy but also family policy where governments (with some notable exceptions, such as Nordic countries) made little efforts to support work/family reconciliation (Clasen, 2000; Cox, 1998; Taylor-Gooby, 2004).

Analytically, one can distinguish two ideal-typical variants of activation, based upon two different problem analyses and objectives: “activation policies may be regarded as ‘compassionate’ (tackling social exclusion) or as ‘condemning’ (tackling anti-social behaviour); as empowering or as forms of social control” (Clasen, 2000, p. 91). The objective of tackling anti-social behaviour (such as the lack of readiness to accept reasonable work) is associated with calls for negative work incentives; that is increased pressure and obligations on the unemployed (such as reduced generosity and duration of benefits as well as tightened criteria of benefit eligibility, job search and reasonable work). Non-compliance is sanctioned with the reduction, suspension or loss of benefit entitlements. Activation through workfare (also described as coercive activation) is characterised by enforced labour market integration through increased compulsion, including ‘making work pay’ measures (that is, subsidising low incomes with models combining labour income and public transfer to establish a more ‘employment-friendly’ welfare state). By contrast, the emancipatory scenario of activation can be described as an attempt of social investment. Still geared toward the recommodification of labour, social investments in the unemployed, for instance, seeks to improve the human capital and employability of unemployed persons through measures of active labour market policy, most notably training measures. Hence, the ‘umbrella term’ of activation is ideal-typically differentiated between a positive (empowering or enabling) and a negative (coercive or workfare) variant. In the domain of family policy, childcare and parental leave policies present an investment in parents (and especially mothers) by helping with the reconciliation of work and family. Instead of improving the human capital, these policies improve labour market flow; so these policies might be thought as improving skills retention rather than the promotion of their formation. However, childcare provision, as early childhood education and care policy, also promotes the development of young children, which can be viewed as an investment into their human capital (see also Dingeldey, 2007; Fleckenstein, Saunders, & Seeleib-Kaiser, 2011; Taylor-Gooby, 2004).

We see that social investment policies can be considered as being intimately linked with the enabling activation discourse, and they are typically regarded as carrying considerable societal benefits. With their firm focus on skills, social investment policies are meant to sustain the knowledge economy with its ever higher skills requirements. They help with making workforces more adaptable and able to respond to fast-changing economies, in which knowledge is widely seen as the key driver for productivity and economic growth. Thus, social policies that help with improving human capital (‘stock’) but also help with making better use of human capital (‘flow’) cannot be reduced to the consumptive function that is typically associated with social protection but present an important ‘productive factor’ in modern economies (Hemerijck, 2017; Morel, Palier, & Palme, 2012).

Looking at recent welfare reforms across the OECD world, we find that labour market policy primarily pursues a trajectory of workfare activation. Labour market reform is dominated by retrenchment in unemployment protection with reducing benefit levels, shortening benefit duration and tightening eligibility criteria. Also, when not complying with the expectations, the unemployed are typically exposed to a stricter sanction regime. At the same time, ‘costly’ up-skilling has lost importance in labour market policy, where cost containment has become a major reform rationale privileging negative incentive reinforcement in reforms across the different worlds of welfare (Bengtsson, de la Porte, & Jacobsson, 2017; Fleckenstein & Lee, 2017b). By contrast, ‘employment-oriented’ family policy reform expanding childcare provision and improving parental leave policies follows the logic of enabling activation, as reforms are thought to help working parents with reconciling work and family more easily. Also, early childhood education and care policies display a strong social investment rationale (Boling, 2015; Fleckenstein & Lee, 2014; Morgan, 2013).

But what explains the ‘uneven’ social investment turn in labour market and family policy? In light of skills shortages, one might argue that there is a very convincing case for investing in training as well, and a recent review of more than 200 labour market programme evaluations supports training as an effective social investment policy. Whilst active labour market programmes do not appear to make much immediate difference, they produce more favourable results about two to three years after completion; and this applies in particular to training programmes (Card, Kluve, & Weber, 2018).

The politics of investing in families

In the following section, we investigate the roads and barriers to social investment policies by first analysing the politics of employment-oriented family policy expansion in Germany and South Korea, followed by the exploration of barriers that made Japan return to traditional family policy after some ‘experimenting’ with progressive policy.

Germany: departing from the conservative path of family policy

Historically, complying with male breadwinner ideology, (West) German family policy was geared toward supporting the family unit with generous subsidies through the tax system (most notably, the joint taxation of married couples) and child benefits. These ‘general family support’ policies were meant to facilitate a traditional, gendered division of paid and unpaid work by ensuring enough family income, so that female employment was not needed to support the family (Korpi, 2000; Ostner & Lewis, 1995). Modest childcare expansion for the over-3s after German unification, providing legal entitlement to part-time care only modified the predominant male breadwinner model (Ostner, 1998). When a centre-left government of social democrats and green party (1998–2005) entered office in the late 1990s, perceived wisdom from the Swedish experience of employment-oriented family policy expansion under social-democratic leadership combined with feminist agency (Huber & Stephens, 2001; Mahon, 1997) would suggest policies challenging the gendered division of labour by promoting progressive parental leave and childcare policies in particular. Instead, social democracy prioritised general family support when raising child benefits, whereas childcare expansion was not even on the political agenda. The ‘Christian-democratic’, flat-rate parental leave scheme promoting traditional caregiving was made somewhat more flexible to promote part-time work but ‘Swedish-style’ parental leave was not on the agenda either. However, a right to part-time work for parents was introduced. More ambitious interventions into the workplace failed because of fierce opposition from business, which enjoyed support from the social-democratic chancellor who showed little if not no interest in the work of his family minister (Bleses & Seeleib-Kaiser, 2004; Leitner, Ostner, & Schmitt, 2008).

The status of family policy changed notably in the run-up to the 2002 election. Under considerable electoral pressure, the chancellor was persuaded that family policy not only had mobilising capacity among young voters (where social-democratic support was considered fragile) but also that family policy investments had an important economic rationale in the modernisation of Germany’s social welfare system. Despite some earlier reservations in the wider social-democratic leadership, the previously ‘sceptical’ chancellor, who also chaired the party, raised the profile of family policy, and committed the party to childcare expansion in the 2002 election manifesto. After the election, the commitment to childcare expansion was reiterated in the Agenda 2010, which was thought to shape the re-elected government’s social policy programme. Savings from the integration of unemployment and social assistance (see below) would provide the funding needed for childcare expansion. However, the government’s key agenda of ‘modernising’ unemployment protection undermined the chancellor’s authority within the party; and this made him call a snap election. Expanding the family policy modernisation agenda, the 2005 election manifesto committed the party to ‘Swedish-style’ parental leave. The political ‘gamble’ of the snap election, however, did not pay off; and the social democracy, without their chancellor, became junior partner in a grand coalition with Christian democracy (Fleckenstein, 2011b; Lee and Mohun Himmelweit Forthcoming).

In the grand coalition (2005–2009), Christian democracy secured the family ministry. Conventional wisdom would have expected family policy modernisation to come to a standstill, as Christian democracy was strongly associated with traditional family and gender values. However, almost parallel to family policy modernisation in social democracy, Christian democracy after its 1998 election defeat, driven by the then secretary general and later party leader and chancellor, modernised, outside the eyes of the public and much of the party, its approach to the family and gender with explicit consideration of the electoral fortune of the party. Based on the analysis of voting behaviour in the 1998 election, the prospect of returning into government office was considered poor without greater appeal to young women voters in particular. This agenda was further facilitated by the unexpected 2002 election defeat, which reinforced the perception of an electoral imperative to modernise the party. Building on the new Christian-democratic chancellor’s modernisation project, the new family minister not only continued with the expansion of childcare provision but also introduced a ‘Swedish-style’ parental leave scheme including ‘daddy months’, as earlier proposed by social democracy. She experienced considerable resistance among the conservative wing of Christian democracy and some parts of the Catholic church (Fleckenstein, 2011b), but skilful cross-class coalition building including the mobilisation of employers’ associations allowed the overcoming of intra-party opposition (Fleckenstein & Seeleib-Kaiser, 2011).

South Korea: departing from the Confucian path of family policy

South Korea has also been defined as strong male breadwinner country. Similar to conservative Catholic thinking in Europe, East Asian Confucian ideology not only places great weight on the family in welfare provision, but also views women as primary caregiver (Jones, 1993; Sung & Pascall, 2014). While there were no generous family/child benefits, key features in the labour market (such as the lifetime employment practice, generous enterprise welfare and seniority wages) underpinned the protection of male breadwinners (Rosenbluth & Thies, 2010). Conservative gender roles and the gendered division of paid and unpaid work went hand in hand with the developmental welfare strategy of the East Asian states. In order to maximise investment in economic development, it was critical for women to perform the role of unpaid caregivers (Lee, 2018). Accordingly, public policies promoting work/family reconciliation were rudimentary at best in the region (Sung & Pascall, 2014). In Korea, maternity leave was paid but short, and only mothers were entitled to unpaid parental leave. Childcare was regarded as a private matter, and only some very limited public childcare support was provided to low-income families in order to prevent child poverty. The first notable expansion of employment-oriented family policy, in line with power resources theory, came with the rise of the centre-left in Korean politics. Despite the weakness (though not absence) of feminists in Korea, during the ten years of centre-left governments (1998–2008), work/family reconciliation policy gained political salience, and it was particularly the second centre-left government that implemented bolder reforms. For the first time, childcare support was made available for middle-income families, in addition to ambitious pledges to increase public provision of childcare. Parental leave became much more generous in terms of both benefit level and duration, although the benefit level remained modest by international standards (An & Peng, 2016; Lee, 2018).

At the same time, the somewhat unexpected second election victory of the centre-left, which greatly relied on the support from young voters that were deliberately targeted with ‘progressive’ family policies, provided an eye-opening moment for the conservative party, which had previously been most committed to Confucian family and gender ideals. Conservatives acknowledged the critical importance of winning young voters to reverse their electoral fortune; and in the 2007 presidential election campaign, the party made a U-turn in family policy, promising policy reform as ambitious as the centre-left party’s pledges, driven by the perception that family policy was instrumental in increasing electoral appeal to young voters. Under the two conservative governments (2008–2017), family policy continued to expand, achieving free universal childcare for all, and parental leave benefit improved too. The flat-rate leave benefit turned into an earnings-related one with an income replacement rate of 40%, effectively doubling the maximum amount of benefit. In addition to progressive employment-oriented family policy measures, by which the conservatives demonstrated sustained effort to modernise the party and increase electoral appeal among young voters, they also introduced a home-care allowance, a general family support policy, to maintain appeal to their traditional support base (Lee, 2018). As the two main parties began to compete over young voters on family policy, and successfully presented family policy as pro-natal policies to address the low-fertility ‘crisis’, family policy received broad political support. Yet, organised business had initially opposed family policy reform most strongly. Employers, displaying rather liberal social policy preferences, objected to greater welfare spending, and also they saw little economic and human capital benefits in the reform. With family policy gaining political salience, however, they eventually, on purely strategic grounds, consented to childcare expansion funded through general taxation to prevent a greater direct financial burden on business (Fleckenstein & Lee, 2014, 2017a).

Japan: from pioneering employment-oriented family policy to returning to traditionalism

Japan, along with Korea and other East Asian countries, has been a strong male breadwinner country. Public measures to support mothers’ employment were meagre, as social norms expected women to quit their job in good time before childbirth. However, in the 1990s, Japan pioneered employment-oriented family policy in East Asia, when, for the first time, centre-left forces were in power. Breaking with nearly four decades of unchallenged dominance by the conservative Liberal Democratic Party (LDP), the Japanese left, prompted by the so-called ‘1.57 fertility shock’, was the key driver of the so-called Angel Plan (1994–1999), promising a substantial increase of childcare provision. This lends additional support for power resources theory in the East Asian context, although, as with the Korean case, Japanese feminists were too weak to play a meaningful role in the first expansion of work/family reconciliation policy (Schoppa, 2006).

Nevertheless, unlike the Korean case, the rise of employment-oriented policy in Japan was short-lived. The subsequent conservative coalition government of LDP and Komeito (1998–2009) continued, at first, with the childcare expansion of the centre-left by launching the so-called New Angel Plan (2000–2004). The two Angel Plans resulted in the doubling of the enrolment rate of children under the age three (from 10.1% in 1995 to 24% in 2010), but they did not overcome the chronic shortages of childcare places across different ages; and Japanese parents continued to rely greatly on unlicensed, private childcare providers (so-called ‘baby hotels’). Rather than employment-oriented policy, the LDP-Komeito coalition government prioritised general family support policy when attention shifted to the expansion of child benefits, both in terms of eligibility and generosity. The conservative public opinion in Japan, according to which childcare was still regarded as the responsibility of the family and not the state, provided the conservative forces with little incentive to modernise their stance in family policy beyond their initial concerns about low fertility. Social conservatism, not only in the wider society but also among young voters, was much more pronounced in Japan than in Korea. Despite having the same Confucian legacy in the society, the Korean public is much more supportive of state intervention in childcare than their Japanese counterparts (Fleckenstein & Lee, 2017a).

Given the strong social conservatism, it might not have come with much surprise that the centre-left Democratic Party of Japan (DPJ) continued with the expansion of child benefits rather than childcare when the left was back in power (2009–2012). The DPJ jumped on the child benefit ‘bandwagon’, a conservative initiative, to underline its support for ‘the family’. This demonstrates how the conservative appetite of the electorate shaped party competition in family policy with a strong focus on general family support rather than on employment support. Thus, despite its legacy as a pioneer of ‘progressive’ family policy in East Asia, in the absence of an electoral rationale for progressive policies, Japan shifted its focus in family policy to more ‘traditional’ measures. As with the low-fertility crisis in Korea, the ‘fertility shock’ in Japan provided an important justification for family policy expansion, at first childcare expansion and later the improvement of financial support for families. It is intriguing, however, that policy experts in the Japanese ministerial bureaucracy consider the expansion of child benefits an inadequate measures for raising ultra-low fertility in Japan (Boling, 2015). Regardless of this assessment, Japanese partisan policy-makers shifted their attention from childcare to child benefit expansion – underlining the electoral rationale in the face of strong social conservatism (Fleckenstein & Lee, 2017a).

To conclude the study of family policies, it appears that left-party rule and feminist agency that drove the expansion of progressive policies in Nordic countries (see for the power resources model including its feminist variant: Huber & Stephens, 2001; Korpi, 1983; Mahon, 1997) fails to explain the expansion of employment-oriented family policy in latecomer countries. Admittedly, left parties in all three countries brought these policies onto the political agenda, but they failed to develop political momentum. In Germany and Korea, one might want to describe the political left as ‘half-hearted’, especially if compared to the policy expansion by their right-wing successors; and in Japan, the left even turned away from childcare expansion and returned to conservative general family support policies with the expansion of child benefits. In Germany and Korea, where the political right continued with ‘social-democratic’ family policy expansion, we identified electoral competition for young and women voters as a critical rationale, which allowed the overcoming of the very traditional family and gender models that were long unchallenged on the political right. In Japan, corresponding with the electoral competition argument derived from the German and Korean cases, the persistence of traditional gender views across the Japanese society, including young and women voters, did not provide a sustainable, electoral incentive structure for the modernisation of the political right, and the left gave up on family policy modernisation. With regard to cross-class coalitions, the German case corresponds with the earlier Swedish experience, where employer support provided a broad societal base for employment-oriented family policies (Swenson, 2002). However, the Korean case, with rather hostile business, suggests that employer support is not a necessary condition for a comprehensive family policy agenda. Rather, party-political agency, as vote and office-seekers (Strøm, 1990), claims the ‘driving seat’ in family policy – its modernisation as in the German and Korean cases, but also the return to traditional policies as the Japanese case illustrates.

The politics of investing in labour market outsiders

Unlike the ‘empowering’ we can observe in family policy with the expansion of work/family reconciliation policies facilitating women’s employment participation in particular, labour market reform has been dominated by retrenchment that can be thought as an expression of ‘condemning’, with coercive activation pushing the unemployed into accepting any job rather than training efforts ‘investing’ in their employability to increase their chances for ‘good jobs’. Apparently, positive electoral competition driving up enabling activation in family policy has not been at work in labour market reform.

Sweden: departing from social-democratic activation and turning towards workfare

Sweden is not only associated with generous unemployment protection but also widely considered the ‘pioneer’ in active labour market policy, including extensive training programmes. Whilst social democracy and the labour movement are typically viewed as the key architect of Nordic welfare capitalism (Esping-Andersen, 1985; Korpi, 1983), skilled labour shortages in post-war Sweden have been argued to have provided the socio-political foundations for employers to support comprehensive social welfare provision, not only training programmes geared towards to improving employability of the unemployed but also generous unemployment protection to promote industry-specific skills formation. This suggests that Swedish labour market policy rested upon a broader, cross-class coalition (Swenson, 2002).

In the economic difficulties of the early 1990s, to cope with the massive increase in unemployment rate (from 1.5% to almost 10%), Sweden first responded in the expected manner by making great use of active labour market policy measures, including substantial training efforts. Spending on training reached 1.0% of GDP, and active labour market policy more broadly peaked at 2.8% of GDP. However, the economic crisis and the associated immense pressure on public finances paved the way for a gradual transformation of the Swedish model of social protection and employment promotion. A right-wing government (1991–1994) embarked on a comprehensive austerity path, not only reducing the generosity of unemployment benefits but also pushing back active labour market policy. Whilst this appears to comply with the power resources model, it is worth noting that social democracy largely supported retrenchment by the political right and continued with austerity when returning to power in the mid-1990s. We thus observe the programmatic differences between the political left and right declining, with far-reaching implications for the development of Swedish social and labour market policy in the 1990s and 2000s (Anderson, 2001; Sjöberg, 2011). Not only did Sweden give up its exceptionally generous unemployment protection by international standards, it also no longer provides great efforts to improve the employability of the unemployed through training measures. Whilst the Great Recession resulted in unemployment that was very similar to the economic crisis in the early 1990s, spending for training programmes dropped from 1% to 0.1% of GDP (Fleckenstein & Lee, 2017b).

Apparently, policy-makers across the political spectrum, including social democracy, have lost faith in active labour market policy, and no longer consider it a worthwhile social investment in the unemployed (Lindvall, 2010). Explaining the critical policy turn-around of social democracy, it has been highlighted that Swedish employers adopted an increasingly ‘Thatcherite’ political orientation, and they mobilised proactively for neoliberal reform in social and labour market policy. This had a huge impact on senior social-democratic policy-makers and their perception of the feasibility of the traditional Swedish model in a globalised world (Ryner, 2004). Thus, with employers taking an “aggressive neoliberal posture” (Huber & Stephens, 2001: 241; see also Kinderman, 2017), one can no longer assume the cross-class coalition that previously supported inclusive labour market and social protection policies. Instead, we find organised business at the ‘heart’ of the rolling back of both social protection and investments in the unemployed, which signals the erosion of social solidarity in ‘social-democratic’ Sweden (Fleckenstein & Lee, 2017b).

Germany: social democrats failing to make a social-democratic turn in labour market policy

Unlike the ‘historical’ Swedish approach of investing in training to improve the employability of the unemployed, traditional German labour market policy has been described of a strategy of ‘welfare without work’, which made extensive use of early retirement to cope with unemployment, in addition to temporarily ‘parking’ the unemployed in ‘cheap’ job creation schemes to keep unemployment figures down rather than expensive training programmes. Despite generous early retirement schemes, those at the margins of the labour market in particular experienced successive cuts in unemployment protection (that is, retrenchment in unemployment and social assistance), whereas unemployment benefits remained largely untouched during the centre-right governments of the 1980s and 1990s. Critically, not only trade unions but also employers are widely seen as supporting this approach of ‘welfare without work’, as publicly subsidised early retirement schemes allowed them to externalise the costs of corporate restructuring, which typically involved considerable labour productivity gains (Manow & Seils, 2000; Thelen, 2000).

As with family policy, the election of the centre-left government in 1998 provided a window of opportunity for a ‘social-democratic turn’ in labour market policy, in particular since social democracy in opposition heavily criticised workfare policies by the centre-right coalition of Christian democrats and liberals. Furthermore, not only did the Third Way ‘preach’ social investments (Giddens, 1998), but also social-democratic labour market policy-makers and ‘their’ civil servants in the ministry of labour were keen to embark upon ambitious labour market policies that broke with the workfare approach of the previous government. However, any serious attempt to strengthen training policy for ‘social-democratic’ activation was effectively blocked by the ministry of finance which, with support from the chancellery, required ‘cost-neutral’ labour market reform. This translated into an unambitious reform agenda that made some minor improvements in job placement; and it might be best described as being caught in the middle, as it avoided further workfare but failed to invest into the skills of the unemployed. This was not only the result of a ‘draw’ between social-democratic ‘traditionalists’ and ‘modernisers’ but also the consequence of social concertation through the so-called ‘Alliance for Jobs’, which failed to forge a consensus between business and labour. Negotiations in the Alliance, which was a priority of the chancellor, revealed that employers (in a similar vein as in Sweden) had broken with the previous cross-class compromise in social and labour market policy, and rather displayed firmly neoliberal policy preferences that were incompatible with large parts of social democracy and trade unions. Certainly, costly training programmes but also job-creation schemes were viewed most critically by employers, in addition to any generous unemployment protection. Importantly, increasingly aggressive neoliberal employer mobilisation not only had huge impact on Christian democracy but also, again as in the case of Sweden, on social-democratic modernisers (Fleckenstein & Lee, 2017b; Kinderman, 2017).

The balance between social-democratic traditionalists and modernisers, and trade unions and employers, experienced a massive shift when, in 2002, a scandal around manipulated placement statistics at the federal employment office discredited labour market policy as a whole. The chancellor used this window of opportunity to install an expert commission for the reform of labour market policy. The so-called ‘Hartz Commission’, with members hand-picked by the chancellery, was heavily dominated by business (with 8 out of 15 members). By contrast, the social-democratic chancellor only appointed two trade union representatives. The work of the Hartz Commission heavily fed into the chancellor’s Agenda 2010, which involved some serious retrenchment with the abolition of unemployment assistance as safety net before social assistance, cutting the maximum duration of unemployment benefit (resulting in the unemployed arriving more quickly in social assistance), and labour market deregulation (translating into greater irregular employment with poor social protection). Thus, not only failed social democracy to pursue a ‘social-democratic’ activation turn in labour market policy, it also intensified ‘neoliberal’ workfare activation started by the centre-right government. We find that employers exercised an ever greater influence on political parties that translated into diminishing programmatic differences, whereas trade unions were successively marginalised in labour market policy-making (Fleckenstein, 2011a; Fleckenstein & Lee, 2017b).

Italy: unexpected training investment in temporary agency workers

The Italian welfare state is widely thought as being geared towards labour market insiders. The by far greatest welfare efforts can be found in old-age security with not only taking a disproportionate share of welfare spending by OECD standards but also with pension spending geared towards labour market insiders. That is why the Italian welfare state has been described as a ‘pension state’ (Fargion, 2009). At the same time, labour market policy is underdeveloped, with very little spending on training measures that would improve the employability of those at the periphery of the labour market. Also, the Italian welfare state does not offer any meaningful unemployment protection for labour market outsiders that fall outside unemployment benefits, which require a level of social insurance contributions that is not typically met by labour market outsiders. Historically, the insider-oriented Italian welfare state was built by Christian democrats, but it received political support from organised labour. Trade unions put a strong focus on improving the working and pay conditions of their members in highly regulated insider employment, whilst largely ignoring public social policies as a source for material well-being – with the exception of old-age security. Admittedly, trade unions expressed the objective of standard employment for everybody – but, in practice, concentrating on workplace issues, labour market regulation and old-age security resulted in heavily insider-oriented trade union strategies, which achieved little for those at the margins of the labour market (Ferrera, 1996; Jessoula, Graziano, & Madama, 2010; Regini and Esping-Andersen 1980). Hence, despite Italian trade unions’ historical class identity (Hyman, 2001), their actual industrial and public policy strategies correspond well with insider/outsider theory (Rueda, 2005) rather than the class-analytical power resources model (Korpi, 1983; Stephens, 1979), as large parts of the working class were not effectively represented by Italian trade unions.

Labour market deregulation starting in the late 1990s further polarised Italy’s deeply dualised labour market. The rise of atypical employment, especially temporary agency work, and an associated increase in social inequality made trade unions’ objective of standard employment for all an ever more remote prospect. Unions recognised that efforts to convert irregular into regular workers were no longer a viable strategy for social inclusion. The acceptance of the changing realities in the Italian labour market translated into a revision of trade union strategy. Following the accelerating dualisation in the aftermath of legislation from 1997 (which, importantly, deregulated temporary agency work), all three confederal unions created special organisations for the representation of irregular workers. This move is seen as a genuine commitment to labour market outsiders, in addition to a recognition that traditional union structures in Italy were insufficient for meeting the specific needs of atypical workers. This organisational innovation meant that temporary agency workers in Italy are fully integrated in organised labour and enjoy all statutory union rights, which is by no means common in other European countries where workplace representation of agency workers is often compromised. In terms of policy, the inclusion of these workers by Italian trade unions translated into a major initiative for the improvement of training of these workers. Specifically, unions pushed, in collective bargaining, for the establishment of a training body jointly run by employers and unions to invest into the skills of temporary agency workers. The new organisation offers basic training focussing on general skills, professional training providing specific and technical skills, on-the-job training, and continuous training. Of the four types of training, most resources are assigned to professional training that equip temporary agency workers with skills of immediate use in the workplace. This initiative allowed remarkable access to training for atypical workers, with more than one third of temporary agency workers participating in training measures according to a European survey. This put Italy at the top of the league table, followed by the Netherlands where less than one fifth of temporary agency workers had access to training; and comparing training expenditure in Italy and the Netherlands, Italian temporary agency workers receive nearly three times the investment of their Dutch counterparts (Durazzi, 2017; Durazzi, Fleckenstein, & Lee, 2018).

The ‘deviant case’ of Italy offers some interesting insights. A laggard in training policy, as far as public policy is concerned, develops a rather unexpected innovation in the industrial relations domain with the establishment of a training fund for agency workers. Whilst organised labour in Germany and Sweden experienced increasing marginalisation, Italian unions ‘bucked the trend’ and successfully pushed employers into filling a ‘vacuum’ in the Italian welfare state (see also Johnston, Kornelakis, and d’Acri 2011). The mobilisation for training was followed by campaigning for income protection for labour market outsiders. Together with civil society organisations, unions formed the ‘Alliance against Poverty’, which campaigned, with success, for means-tested, non-contributory income protection. Parliament, in 2017, decided to introduce a ‘social inclusion income’, filling a serious gap in Italian social protection. The change in trade union strategies in the face of greater labour market dualisation and social polarisation – that is, investing into the skills of labour market outsiders, and also providing them with better social protection – shows that insider-oriented trade unions have the capacity to develop a wider notion of social solidarity, challenging insider/outsider theory (Durazzi et al., 2018).

Conclusions

Social investments in the domain of the family largely follow an activation trajectory of empowerment with supporting work/family reconciliation, which was particularly aimed at increasing women’s employment participation. This presents a remarkable social innovation in historically conservative countries, such as Germany and Korea, where a predominant male breadwinner ideology assigned home-making rather than employment to women. By contrast, one does not find a similar social investment turn in labour market policy, where investment in training could have provided similar empowerment by improving the prospect of ‘good jobs’ for those at the margins of the labour market. Instead, labour market reform was primarily characterised by retrenchment of unemployment protection and workfare activation.

In family policy, electoral competition has been identified as the ‘road’ towards greater social investment and empowerment. Responding to changes in the electorate, both the political left and right in Germany and Korea perceived the political imperative to modernise family policy, creating a broad political support for a social investment turn in family policy. In the German case, we also found organised business recognising the benefits of employment-oriented family policy reform, which allowed the forming of an even broader political coalition. We do not find similar business support in Korea, but employers at least gave up their initial opposition in the face of rising political pressure. Without these political underpinnings, as in the case of Japan, our findings suggest that ‘progressive’ family policy reform is not sustainable. Like Germany and Korea, Japan was confronted with great pressures to increase fertility. Whilst this allowed the country to ‘pioneer’ childcare among latecomer countries, the absence of electoral competition (due to Japanese people’s more conservative value and policy orientation) did not allow the consolidation of progressive reform; and instead the country returned to more traditional general family support policy with the increasing of child benefits, even though policy experts in the Japanese bureaucracy (typically perceived as being rather powerful in policy-making) doubt these will allow the reversing of Japan’s fertility decline. This comparison shows that ‘functional’ pressures, such as fertility decline and societal ageing (these are present in all three countries), are not sufficient to ‘pave the road’ towards greater social investment, but that politics remains the main ‘driver’ of welfare reform; and the more conservative orientation in Japanese welfare politics (with its foundation in a more conservative society) drove corresponding family policies (that is, child benefits rather than childcare expansion). We note though that the current Japanese prime minister, from the conservative LDP, pledged a major childcare initiative, including some free public childcare provision, the subsidisation of private childcare provision and the elimination of waiting lists by 2020, to help with women’s work/family reconciliation in the face of the country’s demographic challenges (Japanese Cabinet Office, 2017). Yet, it remains to be seen whether this childcare initiative materialises, as previous Japanese prime ministers struggled with childcare reform.

By contrast, in labour market policy, broad political support for unemployment protection and training collapsed in Sweden, with employers successfully mobilising against social-democratic labour market policy. Also, in Germany, we find business displaying increasingly liberal policy preferences in labour market reform; and corresponding political mobilisation can be observed. Whilst we might not be surprised that the political right responded to pressure from business, conventional wisdom would have expected the left to resist employer mobilisation. However, social democrats in both Sweden and Germany moved towards the political centre – not only by supporting retrenchment in unemployment protection but also turning away from training as social investment in the employability of the unemployed. The Swedish and German experiences suggest that the lack of employer support presents a ‘barrier’ to social investment in the unemployed regardless of government composition. This observation is supported by the ‘deviant’ experience of Italy, where trade unions were successful in overcoming employer resistance. Italy certainly still presents a ‘laggard’ in public training policy, but organised labour, using the industrial relations arena, partially filled this vacuum for temporary agency workers. Apparently, consensus between trade unions and employers (admittedly with Italian labour pushing ‘hard’) still presents a ‘road’ to social investments for those at the margins of the labour market. The finding that unions can still make a difference is also confirmed in recent German/Austrian comparison by Durazzi and Geyer (forthcoming), who show that Austrian unions, with high institutional resources and favourable legacy though, played a key role in creating inclusive training systems for labour market outsiders. Without effective union mobilisation, as in Sweden and Germany, employer resistance presents a very high ‘barrier’ to social investments improving the employability of labour market outsiders.

To conclude, political parties can be considered to sit in the ‘driving seat’ in family policy reform that brought considerable social investments driven by electoral imperatives; and without these imperatives ‘empowering’ family policy reform does not appear politically sustainable. Critically, the ‘modernisation’ of family policy appears possible under the leadership of both the political left and right. Conventionally, political parties of the left are associated with empowering family policy reform, whereas conservative parties are not thought to pursue reforms that invest into the employment promotion of mothers. In other words, whilst our findings support approaches emphasising party political agency in welfare reform, we question the idea of significant partisan difference in family policy reform. Not only do we find ‘unexpected’ investments in families in the conservative latecomer countries of Germany and Korea (also challenging path dependence theory; Mahoney, 2000; Pierson, 2000), but also that these investments ‘accelerated’ under conservative leaderships. The German and Korean experiences thus indicate a responsiveness of political parties to societal changes, while the Japanese case tells us, without wider societal support, transforming the policies that reinforce the gendered division of labour is incredibly difficult, despite enormous pressures from fertility decline and societal ageing. Turning to labour market reform, employers – and trade unions in the ‘deviant cases’ – appear as critical agency rather than political parties. We suggest that the traditional prominence of social partners in labour market policy-making remains largely intact, though with the power balance having shifted towards business. In this new environment, the reach of political parties seems increasingly limited, and we observe the differences between the political left and right diminishing. But unlike family policy reform where the right moves towards the left, we find the left moving towards the right in labour market reform. Apparently, left parties do no longer typically appear to make the difference they used to make in welfare state-building. Instead of providing greater opportunity, ‘condemning’ labour market reforms have contributed rising social inequality, raising the issue of political parties’ reform capacity in the age of dualisation. The differences we observe in the political dynamics behind family policy and labour market reform point to varieties of social investment politics, rather than ‘a politics of social investments’. In the social investment literature, varieties in the political drivers behind different social investment policies have not received due attention; and future research would benefit from examining additional social investment policy domains in order to systematically theorise the varieties of social investment politics.

Acknowledgments

The authors acknowledge the generous support from the National Research Foundation of Korea (2017S1A2A2041808) and the Korea Foundation.

Disclosure statement

No potential conflict of interest was reported by the authors.

Additional information

Notes on contributors

Timo Fleckenstein

Timo Fleckenstein is an Associate Professor in the Department of Social Policy at the London School of Economics and Political Science. His main research interests are comparative social policy and political economy with a focus on Western Europe and East Asia. His work has been published in World Politics, Comparative Political Studies and West European Politics, among others.

Soohyun Christine Lee

Soohyun Christine Lee is a Korea Foundation Lecturer (Assistant Professor) in Korean and East Asian political economy in the Department of European and International Studies at King’s College London (KCL). Her research focuses on the transformation of East Asian welfare states and political economies, but she also has a keen interest in comparing East Asian and European welfare states. Her work has been published in World Politics, Comparative Political Studies and Politics & Society, among others.