Modernizing progress and its setbacks across Southern Europe
Despite differences in timing, the expansion of welfare provision in Italy, Greece, Portugal, and Spain between the 1950s and 1980s displayed certain similarities. In Italy, welfare state expansion took place between the 1950s and 1970s. In Greece, Portugal, and Spain social insurance was consolidated in the 1960s and early 1970s, but the expansion came only after the transition to democracy (Bermeo, 2000; Ferrera, 2010b). The maturation of social insurance co-evolved with a segmented labour market, and thus acquired its own degree of internal polarization: generous entitlements for core/regular workers, modest benefits for the peripheral/irregulars, and only meagre subsidies (if anything) for those workers unable to establish a formal contact with the regular labour market (Gallie and Paugam, 2000). As a result, Southern European social insurance schemes overprivileged the risk of old age and reserved only a marginal role for family benefits (Ferrera, 1996). Thanks to extremely generous formulas, the value of contributory pension benefits in the four southern countries was by far the highest among the EU Member States in the early 1990s, while the value of benefits for the unemployed, family dependants, and the poor in general was by far the lowest (European Commission, 1993b; OECD, 2004). This dualism set the southern European income maintenance systems apart not only from the highly homogeneous Nordic welfare regimes based on the principle of universal inclusion, but also from the Continental systems, characterized by smaller spreads between 'high' social insurance and 'low' social assistance protection.
The promotion of social assistance and the fight against poverty have been the weakest sectors of policy achievement in the southern European welfare states— at least until the late 1980s (Matsaganis etal., 2003). The so-called safety net evolved slowly, through a sequence of fragmented and mainly categorical additions (orphans, widows, disabled, elderly poor, etc.), with disparate rules, low integration between cash benefits and (generally underdeveloped) services, and with large holes. As a consequence, poverty levels have traditionally remained very high in southern Europe (Saraceno, 1994, 1997; Atkinson, 1998; Petmesi- dou and Papatheodorou, 2006). Many poor households were (and still largely are) ineligible for social assistance, because they fail to fulfil the narrow conditions stipulated by the various categorical programmes. Needless to say, those affected by this syndrome include all the largest groups of outsiders: the longterm unemployed, new entrants into the labour markets, irregular and underground workers, and, increasingly, immigrants.
As argued by Ferrera (2005b), three factors help explain the marginal role of social assistance within the southern European welfare regime: the role of the family, the incidence of the underground economy, and low administrative capacities. The extended family, comprising three or more generations (and/ or lateral kin), has historically been the cornerstone of southern European societies, functioning as an effective social shock absorber and welfare broker for its members, and responding to a wide range of risks and needs, including childcare, unemployment, care for the elderly and the disabled, and housing (Castles and Ferrera, 1996; Flaquer, 2000; Naldini, 2003; Moreno, 2004). The 'familialization' of social assistance functions has given rise to a distinct gender regime, which considers women as having primarily family roles and duties (Saraceno, 1994; Trifiletti, 1999; for a more critical reading, see Guillen, Alvarez, and Adao e Silva, 2001). Portugal's high levels of female labour market participation, however, contradicts this assertion.
From a developmental perspective, the presence of strong extended families and of a familistic culture surely delayed the arrival of public social assistance and family-oriented policies. If families would meet the care needs of all their members and guarantee basic economic security, the state could refrain from intervening on this front and concentrate instead on other priorities (among which, typically, pensions had priority). This syndrome has undoubtedly had some positive results in terms of inclusion: the poor have remained more firmly integrated in the social fabric. But it also generated low rates of female employment, especially in Spain, Italy, and Greece, and the dramatic decline of fertility among (working) women—who are sandwiched between heavy home duties and unfriendly labour markets—clearly indicates the high social strains present in the southern model of welfare (Aassve, Mazzucco, and Mencarini, 2005; Del Boca and Wetzels, 2008). The result has been the emergence of 'bad' equilibrium between low female employment (and high unemployment) and low fertility—with even more adverse fiscal consequences, especially in pensions.
Also the informal economy (estimated at between 15 and 30 per cent of total GDP in southern Europe) weakened both the functional need and the political demand for public anti-poverty and pro-family measures, by providing substantial numbers of jobs for marginal workers (e.g. seasonal workers in agriculture, the building sector, or retail trade), offering low wages with low contributions, especially to young people and women. While strong familialism and an extended underground economy operate on the demand side of social protection, low state capacity is the functional equivalent on the supply side of the southern European welfare regime (Arriba and Moreno, 2005). In this respect the role of the third sector, namely private charities (mainly linked to the Catholic church) should also be mentioned, as their prominence limited the expansion of public social policy responsibilities. This feature is particularly relevant for understanding the underdevelopment of social assistance policies in southern Europe, especially in the field of means-tested benefits and services (Cazorla, 1992; Ferrera, 1996; Sotiropoulos, 2004). As a result, the relationship between benefit administrators and beneficiaries has often come to be mediated by clientelistic 'brokerage' structures between local and central authorities, often linked to political parties and the trade unions
The combination of extended households, high rates of self-employment, large informal economies, tax evasion, and an institutionally weak administrative apparatus has created infertile grounds for the modernization of social safety nets based on criteria of (quasi-)universal coverage. In Italy and Greece particularly, fears of triggering or exacerbating 'welfare patronage' have often been evoked as political justifications for limiting the scope of targeted provisions other than those based on straightforward categorical criteria (such as old age or physical impairment) (Matsaganis et al., 2003).
Southern European welfare states entered the 1990s with severe internal imbalances. In 1993, the Maastricht Treaty came into force, envisaging a maximum 3 per cent GDP/deficit ratio by 1998. At this time, the public deficit was 7.1 per cent in Portugal, 7.3 per cent in Spain, 9.5 per cent in Italy, and 16.3 per cent in Greece as compared to the then EU average of 6 per cent. In the shadow of the EMU's marche force, and a particularly adverse demography, southern European countries were thus forced to take the politically perilous decision of severe internal restructuring: less generous benefits for insiders in order to cut down debts and deficits and—to the extent that budgetary constraints allowed—in order to finance new benefits and services for the outsiders (Guillen and Matsaganis, 2000; Guillen, Alvarez, and Adao E Silva, 2003). This set the stage for a rather ambitious welfare recalibration in the second half of the 1990s, centred on these ingredients: attenuation of generous guarantees for historically privileged occupational groups, accompanied by an improvement of minimum or 'social' benefits; introduction and consolidation of the so-called safety net, especially through means-tested minimum income schemes; the expansion and amelioration of family benefits and social services—with explicit attention to gender equality and equity issues; measures against the black economy and tax evasion; the reform of labour market legislation with a view to promoting desegmentation and modification of unemployment insurance benefits (Ferrera, Hemerijck, and Rhodes, 2000; Ferrera and Hemerijck, 2003). An added distinctive element of southern European recalibration has been politico-institutional: additional competencies have been assigned to regions and local governments and novel modes of concertation have been experimented with, promoting the involvement of social actors in the process of policy formulation and the formation of mixed partnerships in processes of policy implementation (Rhodes, 2003; Guillen and Petmesidou, 2008). Over the 1990s, especially in Italy and Spain, substantial powers were transferred from the central government to the regions. While this created numerous problems of implementation and gave rise to new inter-territorial inequities, this process of quasifederalization of important sectors of social protection system constitutes a far-reaching experiment in politico-institutional recalibration.
Italy can be regarded as an almost emblematic case of multi-dimensional recalibration in the 1990s (Ferrera and Gualmini, 2004). Functionally, the country attempted to halt the expansion of its hypertrophic pension system with a view to restoring to health its battered public budget, and making room for some upgradings in family policy and social assistance. Pensions were reformed in 1992 and then again in 1995,1997, 2004, and 2007. The so-called Dini reform of 1995 completely changed the pension formula, linking it closely to contributions in a quasi-actuarial fashion (Ferrera and Jessoula, 2007). Following the Swedish example, a public notional defined contribution system was introduced, to be implemented through a long transition process. The shift from defined benefits to defined notional contributions suspended the linkage of benefits to previous earnings. Future benefits will depend on the amount of contributions paid and the age of retirement as well as on demographic and economic trends. The Treu labour market reforms of 1997, moreover, eased the shift towards a more flexible and deregulated labour market, including the expansion of active labour market policies. Both reforms were supported by a number of 'social pacts' with the social partners. Next, family benefits were improved and a broad reform of social services and assistance was carried out in 2000. Attempts have been made to introduce a minimum income guarantee against extreme poverty, the reddito minimo d'inserimento (minimum insertion income), modelled after the French RMI analogue (Sac- chi and Bastagli, 2005). The scheme was introduced in 1998 as a pilot experiment in a number of local municipalities, to be extended throughout the national territory in the new millennium after parliamentary review. In hindsight, to be sure, the EMU deadline pressed the centre-left 'Olive-Tree' coalition governments under Ciampi, Dini, Prodi, D'Alema, and Amato to enact a wide range of comprehensive welfare state reforms in social insurance, pensions, labour markets, and social assistance, with a strong focus on cost containment, negotiated with various parties, regions, and societal interests. The EMU exam, in other words, strengthened domestic cooperation. These achievements contrast sharply with the failed attempts of the first Berlusconi government (1994-5) to recalibrate the Italian pension system further through a political style of more top-down imposition. Moreover, the reforms put the Italian welfare state, at least temporarily, on a sounder fiscal basis, while eliminating important institution distortions, including a more transparent relationship between social insurance and social assistance (see, for details, Ferrera and Gualmini, 2000).
From a distributive point of view, the reforms of the 1990s worked in the direction of levelling off social rights and obligations across various occupational groups. Within the pension system, for example, the privilege enjoyed by civil servants to retire after only twenty years of service regardless of age (which had created a mass of 'baby pensioners' since the 1960s and 1970s) was phased out; on the other hand, pension rights were accorded to atypical workers, and lower pensions were repeatedly upgraded. Outside the field of pensions, some traditional gaps in social coverage were eventually filled and new schemes were created for poor households. Little progress has been made, however, in desegmenting the Italian labour market. Moreover, despite the proliferation of new 'atypical' contracts, the insider/outsider cleavage is still very pronounced, while the reform of a faulty system of unemployment insurance is still pending.
The 1990s reforms were accompanied by the appearance of a novel discourse on the current state and future prospects of the Italian social protection system, a discourse building on various notions of 'social equity', 'inter-generational justice', 'gender equality', 'productive efficiency', 'subsidiarity', and so on. The idea that the Italian welfare state ought to give 'more to children, less to fathers' (Rossi, 1997), and that it should be aimed less at 'indemnifying' than at 'promoting' people's opportunities, became the object of an articulated public debate. The new line of argument can be seen as a clear indicator of a significant normative recalibration.
After 2000, there were important setbacks in the Italian reform momentum of welfare recalibration (Sacchi and Bastagli, 2005). The centre-right, led by Berlusconi (2001-6), took over from the centre-left cabinets of the mid to late 1990s. The Berlusconi government pushed for more labour market flexibility on the basis of the diagnosis that welfare provision undermines competitiveness and reinforces market distortions (Jessoula and Vesan, 2011). As the centre-right government privileged a welfare model based on family and community networks, it drastically cut the fund for social services and family policy and put an end to the experimentation with the minimum insertion income. Subsequently, the rather weak and short-lived centre-left Prodi cabinet (2006-8) preferred not to expand childcare, but rather regressively chose to strengthen ordinary unemployment insurance by extending duration and increasing benefits. These setbacks left Italy without provision to combat the 'new social risks' of poverty and social exclusion by the mid 2000s. After Berlusconi regained power in 2008, his cabinet reintroduced a rigid retirement age 65 for men and 60 for women, to replace the opening made through the 1995 Dini reform to allow for flexible retirement (Jessoula and Alti, 2010). It is important to reiterate how difficult it is to modernize Italian social assistance and social services as these are mainly the responsibility of fairly weak regional and local governments administering a wide range of targeted programmes with varying means tests and regulations.
The timing of the most significant modernization boost in welfare development in Spain and Portugal coincided with the period in which both countries became full members of the EU (Guillen, 2002). The influence has been twofold: first, it constituted a basis for legitimizing the new democratic regimes and, second, it served to strengthen domestic institutional capabilities, supported also financially by the European Social Fund (ESF) (Guillen, Alvarez, and Adao E Silva, 2003; Adao E Silva,2011). Particularly from the mid 1980s to the early 1990s, the Spanish welfare state underwent major transformation by adopting a national health service, universal access to education and pensions, and the introduction of minimum income schemes at the regional level (Guillen, 2010). The imperative to curtail public expenditures to qualify for the Maastricht criteria triggered a series of substantial retrenchments in the early 1990s, restricting both eligibility and generosity in the unemployment protection system. Benefits starting at 80 per cent of previous wages were cut to 70 per cent and lower, while minimum contribution periods were raised. Both Portugal and Spain engaged in restrictive pension reforms but also proceeded to improve minimum benefits in the fields of old age, family allowances, and the basic safety net. All Spanish regions introduced their own RMI schemes between 1989 and 1994 (Arriba and Moreno, 2005; Moreno and Serrano, 2011). Portugal introduced a pilot national minimum income scheme in 1996, which was adopted nationally in 1997 (Capucha et al., 2005). In Spain, the 1995 Toledo Pact, agreed among the main political parties and later supported by the social partners, is of special importance in the Spanish welfare recalibration effort (Molina, 2011). It was accepted that pensions and unemployment insurance benefits were to remain funded from social contributions, but that all other non-contributory and social assistance benefits and healthcare services would be financed by taxation. Ever since, the reliance on taxes for financing of 'outsider' social protection has grown dramatically, as in many Continental welfare regimes. Several measures of labour market reform were introduced by the Socialist government up to 1996 and by the Conservatives thereafter: the introduction of flexible forms of contract, the rationalization of unemployment benefits, new programmes and incentives to reconcile family responsibility and work (and thus gender equality), various activation measures, and a broad reform of employment services. It should be pointed out that fixed-term contracts had been introduced in 1985 and that by the late 1980s they had become one-third of all existing contracts. The reforms of the 1990s tried to counter this trend and also to foster the creation of more stable part-time jobs, but with little success. During the 1980s and 1990s, Italy and Spain witnessed a thoroughgoing process of decentralization of competencies in welfare services (healthcare, education, and social care services among others) from the central government to the regions (Ferrera and Gualmini, 2000; Cabrero, 2011). Alongside the augmented importance of social concertation, the devolution of regional social policy competencies is another important dimension of institutional recalibration.
Since the mid 2000s Spain has markedly accelerated the recalibration of its welfare system (Guillen and Leon, 2011). A new social assistance scheme was introduced in 2000, the means-tested Renta Activa de Insercion (RAI), or Active Integration Income, targeted at 45-year-olds and over who have exhausted their unemployment benefits and have family dependants, coupled to tougher activation and job offer requirements.
In 2001 and 2006 labour laws were changed, further relaxing the protection of 'core' employees and improving both the social security rights of irregular/ temporary workers and their opportunities to access the regular labour market. In 2002 unemployment insurance beneficiaries were required to sign up to an 'active agreement', compromising active jobseeking and the acceptance of 'adequate employment'. In 2003, Regional PES offices took over labour market intermediation functions from the national PES organization. In 2003, the Toledo Pact was updated. In 2006, the RAI became part of the unemployment compensation system. As a result of these changes, Spain's employment performance improved spectacularly, creating over eight million jobs in the period from 1996 to the advent of the global financial crisis. A new major reform of the labour market was agreed with the social partners in 2006. The main aim of the reform was to reduce temporality in the labour market and to gain in flexicurity. Also in 2006, another significant reform was reached on pensions, focused on an amelioration of both contributory and non-contributory benefits. When the socialist party (PSOE) gained office in 2004, reforms were passed in the area of families, improving paternity and parental leave. In 2005 a social pact with the social partners was signed on improved protection for dependent people and their care-givers under conditions of demographic ageing. A very progressive law on gender equality was passed in 2007, as well as a law to promote care for dependent people and thus facilitate the reconciliation of work and family responsibilities: Ley de Dependencia. Other progressive reforms included the legislation allowing gay marriages and adoption by gay couples, together with the liberalization of abortion.
One of the more significant policy innovations, strongly influenced by European social policy agenda-setting, was the introduction of policies for gender equality and for the reconciliation of work and family life since the early 2000s (Guillen and Leon, 2011). Over the past decade, in terms of functional recalibration, traditional job and income security for male workers has lost its control over the Spanish welfare edifice. The same applies to gender and family roles with the rapid and intense incorporation of women into the labour market. Female employment has increased rapidly, reaching 53.2 per cent in 2007 (it was 30.7 per cent in 1993)—a figure that puts Spain in southern Europe's vanguard, still behind Portugal (62 per cent), but way above Greece (47.4 per cent) and Italy (46.3 per cent) (Valiente, 2006). In the face of dramatically low fertility rates, under highly adverse demographic conditions, widely held values about traditional family roles as primary providers of social care have weakened significantly (Guillen, 2010).
The past two decades have seen important changes to the Spanish welfare state. Its Achilles heel, however, remains its high levels of unemployment, especially among youth. Some of the labour market and welfare reforms surveyed above may, in fact, have aggravated Spain's already severe labour market bifurcation (Mato, 2011). With the exception of cuts in the first half of the 1990s, the basic architecture of Spanish social insurance has remained unchanged. While additional social assistance and minimum income protection expanded, suggesting a move from Bismarckian social insurance to Beveridgean tax financing, it is important to emphasize that the two tiers of the Spanish welfare state, social insurance and social assistance, have further moved apart rather than closer together (as in France and Germany). Core workers are still the best protected, both within the labour market, by strong job protection legislation, and outside it, by generous unemployment insurance. Another third of the workforce, mainly youngsters, women, and the long-term unemployed, are on temporary contracts, undergirded by only meagre public safety nets.
Portugal's employment performance remained remarkably good throughout the 1990s. The modernization of social protection was a prime objective of the Socialist government that was in power between 1995 and 2002 and again from 2005 to 2011. From 2002 to 2005, there was a centre-right government, led by Jose Manuel Barroso (Guillen etal., 2003). Unemployment insurance was broadly reformed, occupational training and insertion programmes were expanded, and specific incentives were deployed in order to promote a 'social market for employment', based on insertion enterprises and local initiatives targeted at the most vulnerable groups of workers. In 1996, an innovative 'social pact for solidarity' was signed, with a view to mobilizing local potentials for employment creation. With their explicit reference—in official policy statements and public debates—to the principles of solidarity and social inclusion, as well as to the need to fight against fraud and abuses and to rationalize pension protection, the Portuguese reforms of the 1990s and early 2000s can be interpreted as an interesting case of normative redefinition of the tasks and priorities of this country's social policy. The 2006 pension reform particularly, which explicitly introduced a 'sustainability' factor in the pension formula, linking benefit levels to life expectancy, is surely a breakthrough.
While Italy and the Iberian countries have clearly made significant steps forward in recalibrating their welfare systems, Greece has lagged behind (Guillen and Petmesidou, 2008). Also Greece experienced an era of social modernization upon entry to the European Union in 1981; Beveridgean universal coverage was established in the health system, while social insurance was expanded to new groups, under two consecutive Pasok governments in power between 1981 and 1989. Because welfare provision was targeted at a great variety of interest groups, social policy expansion reinforced a pattern of institutional fragmentation, relative backwardness, and a limited role for the state. Moreover, the socialist government adopted a clientelist practice towards interest groups' demands, especially in the area of pensions, at the expense of investing in human capital, schools and hospitals. In addition, PASOK used EU development funds to perpetuate the inefficient structure of the economy and the welfare system in exchange for clientele political support.
The 1990s did witness some movements towards reform in many sectors and a gradual reorientation of the overall discourse on social policy, reflecting the new EU guidelines and recommendations. But the pace of institutional innovation has been very slow. The hyper-fragmented character of Greek society and its mechanisms of interest representation have posed insurmountable obstacles to the formation of adequate social and political consensus concerning reform needs (Petmesidou and Mossialos, 2005). The government did intervene more than once on the pension front, with a view to recalibrating an internally polarized and financially unsustainable system. However, none of these interventions has been able effectively to tackle the roots of the Greek pension crisis and its structural inequalities (Matsaganis, 2005). Pensions represent the highest share of Greek social expenditures, reaching 71 per cent in 1997, to the detriment of resources for other social services. Efforts to strengthen the safety net, with a view to filling in its relatively wide coverage gaps, have not resulted in any substantial reform (Matsaganis, 2005). While the EMU and the Stability and Growth Pact required fiscal discipline, reforms in the early 2000s met with severe social conflict. A new tripartite social dialogue system proved unsuccessful to help bring about greater labour market flexibility, under the new restrictive fiscal policy environment. Most of the items on the Greek-specific recalibration agenda still remain to be implemented. In 2004
a National Reform Programme (NRP) was signed between the Greek government and the European Commission, on issues of modernizing public administration, enhancing employment and education, investing in innovation and lifelong learning, which also largely remained unimplemented. Any attempt to move towards European standards met with strong popular resistance. Clientele interest group mobilization, together with extremely weak state capacities, have kept the Greek welfare state in a condition of permanent institutional fragmentation. Low interest rates coming from EMU participation reinforced Greek welfare state inertia (Spanou and Sotiropoulos, 2011).
During the second half of the twentieth century the four nations of southern Europe have gradually 'caught up' with the more advanced European countries and are now fully part of the group of rich and stable democracies. Total social expenditure as a percentage of GDP is still somewhat lower than the EU average, but the difference has gradually declined. Welfare state building has followed a distinct path, characterized by dualistic social insurance and faulty and fragmented social assistance. Under the spur of European integration, the 1990s and 2000s have witnessed substantial efforts to recalibrate and further modernize the welfare state, with a view to achieving more efficient and equitable labour markets. In terms of functional recalibration, more sustainable and internally homogeneous social insurance systems and a more effective and inclusive safety net were established in Spain and Portugal. In terms of distributive recalibration, southern European welfare states, except for Greece, became less insider-biased, although vested interests have been far better able to slow down reform processes across the region. In terms of institutional recalibration, the master trend has been regional devolution and a stronger hold of ministries of finance over social policy, supported by occasional social pacts. It has been more difficult, in coming to normative recalibration, to say farewell to deep seated 'familialism', except in Spain.
What stands out for the experience of Mediterranean social reform is the importance of the EMU entrance exam and the macroeconomic criteria of the Stability and Growth Pact on public spending and deficit financing (Feather- stone and Kazamias, 2001). These external 'productive constraints' forged the resurgence of 'social pacts' throughout the 1990s between national government and the social partners, many of which revolved around old-age pensions. In hindsight, however, the EMU proved to be something of mixed blessing. The remarkable resurgence of recalibrating social pacts across southern Europe is exemplary of the argument made by Maurizio Ferrera and Elisabetta Gualmini (2000) for Italy that EMU saved the Italian welfare state from an impending ungovernability crisis. While the entrance exam of the EMU has had evident welfare recalibration effects, this no longer applied as soon as Italy and Greece had secured their fully-fledged membership of the eurozone. Extremely low interest allowed them to abstain from further reducing public debt at extremely high levels (close to 100 per cent of GDP). In other words, participation in the EMU took the pressure off further welfare recalibration in Greece and Italy, but not in Spain. Spain kept up the recalibration effort of trading cuts for insider groups in return for enhancements in the status positions of youngsters, women, and low-income families, including significant investments in education, thus recognizing the importance of 'social investments' for successful economic performance in the EMU. Both the Spanish Partido Popular (PP) and the PS socialist party are historically far more commited to the continuous modernization of Spain's economy and society as a fully-fledged member of the EU. I will come back to the issue of stalled reforms in Southern Europe in Chapter 9 on the welfare consequences of the global financial crisis. It remains to be seen whether Italy, Greece, and Portugal, with so much weaker institutional capacities in so many social policy areas, as compared to Spain, will be able to throw off their political ambivalence in the aftermath of the eurozone sovereign debt crisis.