A priori, Italy and France would appear as unfit for activation policies. Yet, when activation is seen as a diversified strategy of restructuring welfare states across Europe, and not only the ‘activation’ of individuals from assistance, then the empirical picture over the last 20 years is much more complex. One of the main differences between both countries is with regard to the role of the state, whilst at the same time they share many common features. However, French achievements in matters of activation seem to be important, whereas in Italy institutional features have consistently impaired the development of comprehensive strategies.

Building on the current literature, France and Italy appear as unfit for ‘activation’. Irrespective of their typological classification, both countries seem unable to ‘activate’ working age population, and reform their social protection system, with its contributory funding and high labour costs (Esping Andersen 1999). However, by considering the reforms passed over the last 20 years, the following pages will present a significantly different picture.

Although France and Italy differ under many respects, it is interesting to compare these two countries, because they also have quite a lot in common. The comparative exercise helps understanding more precisely the differences in the rationale of current policy changes, but also the reasons why so far reform has been more successful in France than in Italy. Both countries adopted the mainstream activation political discourse, along the lines of the European Employment Strategy (EES). Both introduced important innovations, but with very different timing indeed, and probably with a crucial differing factor, namely the role of the state, which Ferrera (1996: 30) considers a main element distinguishing the conservative from the Southern welfare regime. The comparison between France and Italy does not allow us to add an additional model of activation to the two which are discussed in the introduction to this special issue. Indeed, both countries partly draw on some features of either model, and sometimes – as the Italian case shows best – seem to oscillate between the two. The evidence presented here suggests that neither country followed the American ‘workfarised’ route to a residual social citizenship. Although the general conception of what is acceptable for citizens has certainly changed in the wake of the activation reforms, a universalistic interpretation1 of this change is inadequate both in France and in Italy.

Before reviewing the main domains of ‘activation’, we need to recall the main institutional features of the French and the Italian system before the 1990s new turn. These features pertain to labour market but also demographic and social protection major developments.

‘Traditional activation’ has certainly existed in all countries to a certain extent (Barbier and Ludwig-Mayerhofer this issue). Indeed, where ‘activation’ has been difficult to identify as a clear new strategy, ‘old features’ are all the more at work, as in Italy and France. The contextual differences of the Italian and French systems largely pertain to the distinction between the ‘continental’ and ‘Southern’ clusters, but extend beyond.

2.1 Labour markets: mostly similarities

The French and Italian labour markets have both been characterised by high unemployment from the late 1970s to the late 1990s (even into the 1997–2001 labour market boom). In France, the unemployment rate has remained high: doubled from 1975 to 1983 to 8.0 per cent, it grew to more than 12 per cent in 1994; subsequently, because of the labour market boom, it decreased to less than 8 per cent in 2001, but it has increased again since 2003. Italy has been characterised by slightly higher levels of unemployment during the same period; however the trend was reversed between 1998 and 2002, and unemployment dropped from 11.5 to 8.5 per cent – a quite remarkable record, although in the South the relevant figure was still 18 per cent.2

A second common feature can be characterised here as dualism or rather, segmentation of the labour markets. Very often in the comparative international literature, France and Italy feature as having ‘rigid’ markets (Esping Andersen 1999), but, actually, this presumed rigidity only applies to segments of these markets. An increasing part of the labour force in both countries has been affected by the flexibilisation of employment relationships, which has happened ‘by stealth’, through the increasing creation of jobs excluded from standard protection3. In the French case, this is illustrated by the development of ‘formes particulières d'emploi – FPE’ and in Italy, by the rapid increase of new contracts, particularly the so-called ‘co-co-co’ (‘collaborazioni coordinate e continuative4) but also fixed-term and part-time contracts. The proportion of such ‘flexible’ employment relationships approximately now amounts in both countries to about 10 per cent of the workforce in the formal sector (Barbier 2004a). The Italian labour market displays additional features: a North–South cleavage, but also an important role played by small enterprises and the informal economy. In the mid-1990s Italian employees effectively covered by the standard status – that is working in private firms with more than fifteen staff or in the public sector – accounted for only about 40 per cent of total employment. Moreover, the proportion of small firms and self-employment remains much higher than in France, while illegal work accounts for as much as 22 per cent of total employment.

A third similar feature pertains to labour market participation patterns. Italy and France are characterised by comparatively low participation rates at both ends of the age spectrum (the young and the elderly). Even more visible in Italy, low participation rates are further explained by a lower female rate of employment. Despite these differences, Table 1 demonstrates that the French and Italian patterns are closer to each other than to the British ones.

Table 1. 
Employment rates according to age and gender (2001)
%ItalyFranceUK
The young (15–24) 26.3 29.5 56.9 
The older (55–64) 28.0 31.0 52.3 
Men: all 68.5 70.3 78.3 
Women: all 41.1 56.1 65.1 
Women (25–54) 52.8 71.2 73.5 
Women (15–24) 22.1 25.7 54.3 
%ItalyFranceUK
The young (15–24) 26.3 29.5 56.9 
The older (55–64) 28.0 31.0 52.3 
Men: all 68.5 70.3 78.3 
Women: all 41.1 56.1 65.1 
Women (25–54) 52.8 71.2 73.5 
Women (15–24) 22.1 25.7 54.3 

Source: Employment in Europe, 2002 (European Commission).

In both countries, low employment rates of the young are explained by a mix of high rates of participation to education and high rates of unemployment; low participation rates for the older, for their part, are explained by frequent early retirement.

2.2 Contrasted demographic situations

If the labour markets are overall similar, Italian and French demographics differ significantly. Like the Scandinavian countries, Britain and Ireland, France enjoys a comparatively favourable situation. While in 2001 the French fertility rate was about 1.89, Italy's had decreased to 1.23, among the lowest in the EU. The Italian population is also ageing more rapidly than the French. Notwithstanding a possible influence of future immigration, both these characteristics aggravate the internal pressures for activating the system of social protection in Italy.

2.3 Inherited systems of social protection

Although the differences between the ‘Southern’ and the ‘conservative-corporatist’ model are well known, it is useful to specify basic features of social protection systems which directly bear on the present comparison. The proportion of GDP spent on social protection certainly represents an important starting point. Table 2 shows in the French case the relevant share of GDP is 4 percentage points higher. Yet the more striking difference lies in the distribution of outlays.

Table 2. 
Proportion of GDP for expenditures according to functions (2000)
% GDPHealthPensionsFamily and housingEarly retirement, unemployment and povertyTotal
France 9.7 12.4 3.8 2.3 28.3 
Italy 6.1 15.4 2.8*  24.3 
% GDPHealthPensionsFamily and housingEarly retirement, unemployment and povertyTotal
France 9.7 12.4 3.8 2.3 28.3 
Italy 6.1 15.4 2.8*  24.3 

Source: French and Italian sources, own calculations.

*Includes public assistance disability benefits while early retirement is included in pensions.

More than half the Italian social protection bill is devoted to pensions, with the health risk accounting for the essential rest. Hence funding for all other risks is comparatively marginal. The distribution between risks is more balanced in the French case, pensions and health expenditure accounting for 78 per cent of the total bill, as against 88 per cent in Italy. In France, the combined weight of family, housing, early retirement, unemployment and assistance benefits amounts to 6 points of GDP, an area of social protection which provides a good number of programmes susceptible to be affected by the drive to ‘activation’; in Italy, the situation is very different, because both the unemployment sector and the family and assistance benefits remained marginal throughout the 1980s and 1990s. Structurally then, the potential scope for activation in this area is limited. Additionally, the pension sector would a priori appear as a better domain for implementing activation strategies in both countries.

2.4 Different timing

Important changes to the French social protection system were introduced rather early, in the late 1970s, but not under an ‘activation’ label. With hindsight, these innovations nevertheless fit very well into the ‘activation’ scope. In this respect, whatever its very recent importance, the European Employment Strategy decided in Lisbon has played only a minor role. The main external influence having affected the transformation of the French system in the years after 1993 has been the European Monetary Union (EMU) economic policy (Barbier and Théret 2003). On the contrary, in Italy, fresh activation initiatives and discourses appeared much later, with the 1996 centre-left government. These developments were also essentially boosted by the subsequent Italian participation in the common currency (Ferrera and Gualmini 1999). Very different political agendas contrast the Italian and the French situations notwithstanding the fact that both countries have obviously been subjected to common pressures and influences (the economic crisis, the transformation of economic paradigms from Keynesianism to a more ‘monetarist’ agenda (Hall 1993), the common opening of their economies to international competition, the rush to the Euro). In both cases nevertheless, the emergence of the very term activation in the discourse is not so dissimilar; and the distance between a comprehensive and consistent implementation of activation reforms and their policy discourse is certainly also high in both countries.

These general features can be better illustrated by investigating social policy developments in the areas appearing as primary candidates to activation strategies: unemployment insurance, active labour market policies (ALMP), assistance and family benefits; tax and benefit systems; early retirement and pensions.

3.1 Unemployment insurance and assistance in France and Italy

With hindsight, French programmes introduced from 1975 under the banner of ‘insertion’ could be seen as precursors to activation reforms later implemented elsewhere (Barbier 2004b). Yet, when they were introduced for the disabled and the young unskilled, their justification was to ‘activate’ these groups in a very specific sense: exactly at the opposite of a punishment or as the necessary way out from ‘dependency’, work was here promoted as a positive channel for integration into society and accessing political citizenship. This doctrine originated from militant civil society initiatives and only at a second stage the state took over with insertion policies. In the initial solidaristic insertion philosophy, ‘social integration’ was neither meant primarily nor solely in terms of constraining people to take jobs on the market. In fact, due to the particularly low rate of job creation in France at that period, many programmes entailed the opposite function of keeping people in ‘welfare’ rather than transferring them over to work.

Throughout the 1980s and into the 2000s programmes involving both minimum income benefits and job creation schemes have played a key role. Although a mean quantity of 400 to 500,000 places have been funded yearly,5 this provision was never able to accommodate all potential candidates, thus ruining the programme's claim to ‘universal solidarity’. Moreover, in a system of social protection built upon the hegemony of social insurance principles, minimum income benefits did not significantly feature before the second half of the 1970s. Unlike general assistance benefits, these were introduced for special cases of need in ‘three generations’ (Barbier and Théret 2003). They have included benefits for lone parents (API, allocation de parent isolé, from 1976); benefits for the disabled (AAH, Allocation d'adulte handicapé, from 1975) and Allocation spéciale de solidarité (ASS) (unemployment assistance, from 1982). RMI (revenu minimum d'insertion) appeared in 1988 as a totally new benefit (a universal minimum income for all those not entitled to any of the others). Its main innovation lay in the introduction of a ‘contrat d'insertion’, defining a ‘projet d'insertion’ (plan) which stressed the scheduling of actions the benefit's recipient was supposed to undertake. Activities were in the areas of health, housing, various forms of counselling and employment and training. However, even today when a fresh reform is tentatively implemented,6 conditions for eligibility have never included the obligation to actively seek work: indeed, successive 1988 and 1992 RMI Acts established RMI as an unconditional citizenship right.7

Yet political and economic pressures have fostered reforms aiming at eliminating possible disincentive effects. This rationale has presided over the reform of minimum income benefits in the late 1990s, in line with the overall influence of the ‘activation’ political discourse, and notably its justification among orthodox economists, but also with the political discourse of the EU co-ordinations (Barbier 2004b). Yet AAH has never entailed problems of inactivity comparable to for instance the disability benefits in the UK or The Netherlands. As for API, contrary to British benefits for lone parents, it was ‘activated’ from the start, because served only for the period when the children were under three: presently, 40 per cent of API recipients receive it for less than 1 year, and 40 per cent of them are also active. On the contrary, for ASS, although rules similar to the unemployment insurance have applied, the majority of its older recipients have tended to be exempted from active job search, thus resembling early retirees (see further).

In unemployment insurance, activation has been present from the start (the ‘old activation’ obligation to seek work). Contrary to the Italian case, the ‘actively seeking’ clause has never been rhetoric for mainstream insurance recipients. However, in a comparative perspective, sanctions were always limited and they only recently started to increase sharply. Moreover, in the mid-1980s, the French insurance fund innovated with the provision of more active support to the unemployed: AFR (allocation formation reclassement) allowed for an extension of the compensation period and additional support for training periods. Since then, the philosophy has been both to activate unemployment compensation by giving the unemployed better opportunities to improve their employability and, more marginally, to influence employers’ decisions to hire long-term unemployed people.

Activation has also featured high in the long and conflictual process of reforming unemployment insurance in 2000, when PARE (plan d'aide au retour à l'emploi – back-to-work support plan) was introduced. In October 2000, employers’ organisations and only three of the five French representative trade unions agreed on the reform. But it was eventually only implemented from July 2001, after more than a year of conflict among unions and between unions and employers’ associations and the government.

The mainstream benefit, which used to decrease with time, was upgraded and remained unchanged over the total duration of entitlement. This reform, which also entailed an increase of the stress put on a ‘back-to-work’ logic, certainly look more comparable to Scandinavian activation than to the liberal ‘welfare-to-work’. However, means have never been available in France at levels comparable to Scandinavia: the French system remains well short of a comprehensive logic of universal provision. During the same period, management reforms occurred in the Public Employment Service, which, to this day, are difficult to assess. The latest period of the implementation of the PARE, now the mainstream benefit and form of ‘activation’, has also seen a very significant increase of sanctions for the unemployed,8 while a new decrease of the benefit, decided in December 2002 and to be implemented fully in 2004, was eventually cancelled after the conservative government lost the 2004 regional elections.

Given the persisting absence of social assistance comparable to the French and the wide limitations of unemployment insurance, the Italian situation appears very different.9 Only the core industrial workforce was entitled to the most generous unemployment benefit granted by CIG (Cassa Integrazione Guadagni)10 which was initially designed as ‘short-term’. During the 1980s, the programme became very palatable for both unions and employers because it granted firms the opportunity to externalise the social costs of industrial restructuring whilst freezing firings. The measure offered generous income compensation at 80 per cent of previous earnings. But the programme soon ceased to provide ‘temporary’ benefits: renewals were granted for up to as many as 10 years, especially in large industrial plants, irrespective of whether or not the restructuring process would eventually create new working opportunities within the firm.

However, none of this applied to the large majority of the unemployed, which could only claim the ordinary unemployment benefit, provided the 2 years contribution requirement was met. Moreover, this benefit was payable for a maximum of 180 days and (until 1988) at a very low flat rate. From 1988 to 1994 it was upgraded first at 28 per cent and then at 30 per cent of the previous wage. Last but not least, no public assistance programme existed for employees whose ordinary unemployment benefit expired or for those who did not even qualify for such ‘second-class’ measures. From the 1970s onwards, depending on their budget certain municipalities in the Centre-north provided means-tested income support; but the Southern part of the country in which the largest share of the poor is concentrated did not follow the same path.

Given these premises, one can better understand why Italy could hardly develop a coherent ‘activation’ policy aimed at moving citizens from ‘welfare dependency’ to work. At best, this kind of approach could apply to the relatively small ‘citadel’ of workers eligible to CIG compensation. This is precisely what happened with the 1991 Act no. 223. Its explicit aim was to end the freezing of employment status for redundant workers who had no real possibility of being reintegrated in firms. In order to bring CIG back to its original temporary purpose, strict limitations were set on both duration and eligibility for benefits. Indeed in just 5 years the number of workers receiving CIG benefits was halved.

Act 223's major innovation however consisted in the establishment of the ‘mobility benefit’ (mobilità) for workers in firms with more than 15 staff. While severing every link with previous employers, it provided a comparatively generous 12 months allowance. This benefit was progressively scaled down to speed-up job search and beneficiaries were required to attend training programs or to engage in ‘socially useful jobs’.11 However, working requirements were soon abandoned because it was feared that mounting pressure by the programme participants would result in turning these temporary jobs into permanent public sector positions. In short, the 1991 Act opened the way to the introduction of activation principles within Italian social legislation but was in fact unable to trigger a general reorientation of unemployment policies (Fargion 2003a). Quite clearly, in order to extend this type of approach, one had to first reform existing ammortizzatori sociali12 by providing adequate unemployment compensation. So far, this remains an unfulfilled task. The Prodi Cabinet started moving in this direction but subsequent centre-left governments proved unable to overcome opposition within their own majority and from the trade unions. In 1999, in a context of continuing fiscal austerity, Parliament formally delegated the executive to design a new framework according to the three-tier system originally suggested in 1997 by the Special committee headed by the economist Onofri. The system was to provide: (a) insurance-based ‘short-term’ compensation extending coverage to previously excluded categories; (b) considerably upgraded ordinary benefits; and (c) a public assistance safety net. However, very little has actually come through. Thirty-nine municipalities were selected to carry out an experimental minimum income programme for a 2-year period ending in December 2000. The scheme – largely inspired by the French RMI – was supposed to be extended to the whole country and funded by central government, a commitment never fulfilled. The centre-right majority decided to end the programme altogether in 2003. Hence, the reform of ‘social shock absorbers’ is still on the waiting list and the only real improvement was the upgrading (in 2000) of the ordinary unemployment benefit from 30 to 40 per cent of previous wages (to be compared with a compensation rate between 75 and 57 per cent in the mainstream insurance provision in France). All in all, the regulations introduced over the last decade, while gradually shifting resources among different schemes, did not propel them in a consistent ‘activation’ direction.

The common adoption in Italy and France of an ‘activation’ discourse has not meant that the systems have come nearer to each other. In Italy, an incipient dynamics, started in the 1990s, eventually stumbled upon the institutional capacity of the system to reform. In France, the insertion dynamics, coupled with the introduction of diversified minimum income benefits has provided for an increasingly complex benefit mix, along with the traditional unemployment insurance. The French picture of activation which emerged during this period is of a mix of state-led initiatives to introduce more universal coverage, under the justification of Republican solidarity and, on the other hand, an early adjustment to increasingly active compensation of the unemployed in the insurance sector, which culminated in 2001 with the introduction of PARE's ‘new activation’ as the standard provision for all insured unemployed.

3.2 Employment programmes and the public employment services in Italy and France

There is a similar structural opposition in the field of employment programmes. What is obviously common to both countries over the period is that they failed to achieve full employment and saw their unemployed population grow. Hence, governments were confronted with the ‘employer of last resort’ question, the state being expected to provide temporary (or ‘secondary market’) jobs when the market failed to deliver them. Yet, quantitatively, France and Italy have had very different expenditure patterns in matters of ‘active labour market programmes’: the proportion has been from 3 to 4 per cent of GDP in France (Barbier and Gautié 1998) to no more than 1 per cent in Italy (Gualmini 1998). However, a qualitative analysis is much more telling.

Actually, in France, from the 1980s, politiques publiques de l'emploi have gradually emerged as a new, significant and consistent policy area for social protection (Barbier and Théret 2003). In a first period, from the late 1980s, extensive programmes were introduced.13 The corresponding expenditure went from 0.9 to more than 4.0 per cent of the GDP from 1973 to 1995, a considerable expansion, which has only slightly been slowed recently. Programmes have encompassed: (i) training schemes for the unemployed; (ii) temporary subsidised employment in the public and non profit sectors; and (iii) subsidised contracts in the market sector for certain hard-to-place groups. Except for training programmes, almost all participants enjoyed an employee status (statut de salarié) and, consequently, were entitled to standard social protection rights (nevertheless, there has been a clear relationship between these schemes and the emergence of a ‘working poor’ stratum in France). The number of participants in the various employment programmes has increased to a little over 10 per cent of the active population (2.8 million in 2000). This figure includes a stock of about 300–500,000 places for the temporary subsidised jobs in the public and non-profit sector, of whom the CES (Contrats emploi solidarité) have been the mainstream contract. This means that during the period, all governments – despite obvious reluctance from the more liberal ones – have stuck to the logic of the state as an employer of last resort to a certain degree, for fear of being confronted with even higher unemployment figures and with recurrent social demonstrations, as those which occurred in 1995 and 1997.

Italy has only developed marginal employment programmes. Nevertheless, it was also an early starter, launching a job-creation program for unemployed youth in 1979. However, this programme's final outcome largely discouraged subsequent governments from following that path and certainly contributed to de-legitimising this policy option in the public opinion, because under trade unions’ pressure, the originally temporary jobs were in fact turned into permanent public jobs. As mentioned above, in the context of the embryonic activation endorsed in the early 1990s, ‘short-term’ and mobility allowance recipients were required to participate in ‘socially useful jobs’ (LSU), primarily to avoid irregular work. But LSU received a boost only in 1995 as the first wave of 300,000 mobility benefits expired with almost no recipients re-integrated into the workforce. Considering the vast majority would remain without any kind of support, expanding LSU was the only option to avoid social unrest, especially in the South where unemployed workers were highly mobilised. As a result, in the public opinion LSU were increasingly perceived as the equivalent of a passive measure to the benefit of a comparatively advantaged group of unemployed workers (the bulk were prime age male blue collar workers). Not surprisingly, the workers involved in the programme pressed to stabilise their positions while the governments in office endeavoured to ‘emptying the reservoir of socially useful jobs’ by resorting to every imaginable incentive. The policy reversal was initiated under centre-left rule and continued with greater vigour under the centre-right majority: between 1998 and 2002 the relevant spending was more than halved.

The LSU trajectory epitomises three distinct but interrelated factors which shed light on the Italian case: first, the entrenched institutional weakness undermining the implementation of legislative goals and principles; second, the labour market's structural difficulty in creating new jobs in the South; third, the current social protection system's structural unfitness for consistent activation. Yet, to fully appreciate how Italy approached the activation issue one also needs to consider the reform of public employment services. In 1997, the Prodi Cabinet initiated a radical restructuring of the inefficient public placement offices (collocamento) and passed the relevant functions from the central to the regional level, a devolution process which was completed by 2000. The rationale underpinning the reform was that the public administration had to do its job properly before introducing behavioural requirements on individuals. The new rulings updated the Italian conception of ‘availability’ along ILO standards.14 Additionally, local employment services were required to interview young unemployed people within 6 months from registration, and to provide individually tailored labour market ‘re-insertion’ or training programmes to various categories. The new functions were supposed to be phased in by 2003; however, considerable differences in regional performance remain; further, as the reform of ‘social shock absorbers’ kept being postponed, ‘active’ and ‘passive’ programmes remained largely de-coupled. Considering the current government gives priority to opening up job placement to the private sector, future prospects appear uncertain – particularly as to whether the Public employment service (PES) will mostly be left to deal with the hard-to-place.

In the area of employment programmes, France and Italy score differently. Globally, over the period, both have failed to actually provide hard-to-place people (and the unemployed more generally) with effective transitions from employment programmes to conventional market jobs. Only a minority of CES (contrats emploi solidarité) participants succeeded in gaining such access. However, when compared with the LSU, this programme appears as a relative success. Other forms of temporary subsidised jobs, like the emplois jeunes, have also brought positive outcomes for participants. Accordingly, subsidies targeted for contracts for the long-term unemployed or RMI beneficiaries in the private sector have proved effective in France. While more similar to the Scandinavian ‘employer of last resort’ rationale than to the liberal one, the French record appears as implemented only half way, because of limited funding. In a way, Italy exemplifies the limits encountered in France, but as a caricature. Hence, in France – and in Italy all the more – a significant proportion of employment programmes cannot be considered as really effective paths to activation but, like in many other countries, have certainly acted as ways of decreasing ‘open unemployment’.15

3.3 Family benefits: unlikely candidates to activation in either country

In the liberal mode of activation, tax credits, and particularly family-based tax credits have now gained a central role (see Wright et al. in this issue). None of this exists either in the universalistic model or in France and Italy. Both countries have however had a very different record for ‘familialisation’, a feature supposed to be adverse indeed to ‘activation’. The legacy in France has been of a large scale family policy: it has encompassed a wide array of benefits, the most prominent of which have been benefits for families with two children or more, and housing benefits which have acted as effective mechanisms to alleviate poverty. In certain cases, and very recently again, familialistic and even natalistic features of the old times have resurfaced, clearly contradicting the mainstream policy of enhancing child care and the reconciliation of work and family.16 Yet, despite very important persisting gender differences, the French system has long been ‘de-familialised’ although obviously not to a similar extent as in the Nordic countries. Contrary to Esping Andersen's interpretation (1999), this feature clearly distinguishes France from Italy. However, when considering the young, French family benefits de facto allow many young people to remain dominantly inactive and France is among the number of EU countries with the lowest participation and employment rates among the 15–24 age group (Table 1). Universal housing benefits are also at a significant level for students not living with their parents and they certainly do not foster job search or labour market participation. Hence, the combined effect of high participation in education and family-linked benefits can be seen as directly contradicting or hindering the extension of the activation dynamics for the young.

Italy, for its part, is still marked by a familialistic bias. Family benefits, traditionally limited to dependent workers, leave wide sectors of the population ineligible and, due to their structure, tend to discourage wives and mothers’ participation in the labour force. Following their electoral victory in 1996, Centre-left governments actually introduced a variety of innovations: on the one hand, they steadily increased tax allowances for children (a trend continued by the centre-right majority), and established a means-tested benefit for large families; on the other hand, they improved parental care leave and part-time arrangements along with providing project-funding for child-care facilities. However, the overall results remain rather limited and far from reflecting a coherent approach. Indeed, with the centre-right government, the balance tipped once again in favour of a more traditional vision of the family and its caring functions – as one can easily see by taking a look at the ‘White Book on Welfare’ published by the Minister of Social Affairs, Maroni. Not surprisingly, the family maintains its traditional ‘broker’ function within the Italian welfare state and young persons continue to live at their parents’ home much longer than in any other European country.

3.4 Social contributions and tax credits: activating the funding or the people?

Another policy area for activation stands between social policy, tax policy and indirect labour costs. The Italian and French social protection systems are still hugely funded by social contributions to the level of about 66.5 per cent of the total bill in 2003. However, in the late 1990s, indirect labour costs, it was agreed, were completely employment unfriendly, and the solution lay in the decreasing of social contributions paid by employers, in order to foster job creation.

In France, this form of activation has entailed a fundamental overhaul of the funding of social protection. The creation of a special contribution (Contribution sociale genéralisée, CSG) has represented a major ‘Beveridgean’ innovation (Barbier and Théret 2003). In contrast to the British tax credits, this shifting of previous employers’ social contributions over to the budget is demand-oriented, aiming at fostering job creation rather than at ‘incentivising’ people to take jobs. Yet the French government has also introduced a new tax credit, the prime pour l'emploi, which, albeit still marginal, has represented an additional shift in labour market thinking. Decisive successive reforms were taken from 1993 and from 1998 (the reduction of the working time, RWT). With the return of a conservative government from May 2002, the previous RWT logic has been reversed, but the reduction of employers’ social contributions has been even extended. Altogether, state budget expenditure for compensating reduced contributions soared from 1993: while the aggregated outlays for unemployment insurance and traditional ‘active’ programmes has remained roughly unchanged, the amount spent on this compensation has multiplied by almost six over 7 years. The present proportion of the workforce affected by the reduction (around 60 per cent) is bound to increase further in the coming years, which means that this most important feature in the French activation strategy has become permanent.

In this respect Italy appears more similar to Spain than to France: instead of introducing systematic social security contribution cuts, over the 1990s, subsequent governments introduced an impressive variety of incentives for employers, thereby creating a sort of jungle in which each firm was left to figure out the best combination to lower its labour costs. The flexibilisation of the labour market and the legitimisation of atypical work contracts – more than thirty different types – went hand in hand with the introduction of over ninety different measures aimed at bolstering job creation by reducing the ‘social wedge’. The 1998 Social Pact actually envisaged the elimination of social security contributions for family and maternity benefits, partly shifting the burden to general taxation but the relevant figures were rather small. The changes introduced in health funding appear far more important, particularly the reforms falling under fiscal federalism which are now being phased in. Yet, compared to France, these reforms have so far only moderately reduced non-wage labour costs, despite the fact that during the mid and late 1990s incentives for employers – which were mostly funnelled to the South – consistently represented over 60 per cent of total spending for active labour market policies. All in all, while activation in the strict sense remained secondary, starting with the 1997 ‘Treu package’, the main emphasis was on the flexibilisation of the labour market – a trend strongly reinforced by Act no. 30 which the centre-right government passed in 2003.

In the context of similar structural problems, policy responses have been diverse in France and Italy. From the early 1990s, France has embarked on a structural transformation of its social protection both on the side of funding (the significant increase of the funding by tax and special new contributions to the state budget) and on the side of its labour market impact (reductions for employers). This has resulted in a rather clear strategy to ‘liberalise’ the system, mainly stressing the demand side of the labour market, and only marginally the supply side and individual incentives. The overall impact, in terms of employment creation actually appears controversial in current evaluation studies. But gradually, unions, parties and the employers have been at least partly convinced of the strategy of funding reforms, as linked to the reduction of labour costs, and coupled with the ‘solidaristic’ policy of the employment schemes. On the other hand, a ‘de-coupled’ strategy has prevailed in Italy. Here, a growing flexibilisation of the labour market has been accepted by the state and the social partners, without actual reform of the assistance and pension systems and, inevitably, without ‘consistent’ activation. Despite the argument by late 1990s governments, that the introduction of flexibility would open up a whole new range of opportunities when employment started to grow, especially after 1999, almost all new jobs were atypical with limited social protection attached.

3.5 Exit or activation for early retirees and disabled people?

A last area of social policy susceptible to be affected by activation strategies encompasses early retirement programmes and disability benefits. In some countries, and this is definitely the case in France and Italy, governments have resorted extensively, from the late 1970s and into the 1990s, to early retirement.

In France, early retirement was particularly used to remove older and redundant workers from the labour market, their early retirement benefits being funded by the state budget. This mechanism became very palatable both to employers and to unions, because it allowed for a ‘socialisation’ of individual consequences. Recourse to such arrangements was at its highest in heavy industries and large firms (in 1982–88). It then started to decrease significantly; however, new, more targeted schemes were introduced, of which only some were actually ‘activated’ in the sense that the older left the labour market in exchange for the recruitment of young people. In 1994, the statistics showed a stock figure of participants in various early retirement schemes at about 210,000 people. At the same time, 284,000 older assisted unemployed were allowed not to seek jobs. In 2001, the corresponding figures were at 204,000 and 365,000. These figures point to a persistent inability to turn away from inherited practice. Fresh initiatives were heralded in the wake of the 2003 pension reform (which increased the number of years of contributions to 40 in the public sector), but all odds are that France will prove unable to reach the targets for employment rates which were fixed at the EU Barcelona summit. For all the talk of reform in both countries, Italy has experienced very similar problems and will probably not achieve the demanding targets either.

As in France, piecemeal reforms were attempted but they were even more difficult to implement because of the central role of pensions within the Italian social protection system and its substitutive function. Legislative restrictions were introduced in the mid 1980s to curb the disproportionate growth of disability benefits (Fargion 2003b). Moreover, three major pensions reforms were passed over the period surveyed here. First, explicit early retirement procedures were abandoned, albeit only gradually, in favour of other types of measures; second, retirement age was gradually increased from 60 to 65 years for men (from 55 to 60 for women); third, the public sector privileges (for pensione d'anzianità) were progressively scaled down. Last but not least, after a temporary blockage of new uptakes for pensione d'anzianità, the rules were repeatedly tightened by moving up the age threshold (from 52 to 57 between 1996 and 1997 and nowadays) for a full pension with 35 years contribution. These restrictions appear to have had an inconclusive impact on employment rates.

Despite rhetorical acceptation of the current mainstream activation recommendations, early retirement and equivalent benefits for the old have clearly illustrated a common French and Italian inability to actual effective reform.

From the present survey, the picture actually rather shows different combinations or hesitations between orientations picked from the liberal or the universalistic types of activation. Yet, despite their common Bismarckian legacy and important similarities, France and Italy stand in distinct positions. From France tentative features have emerged but they certainly are too weakly institutionalised to really model for a consistent activation ideal-type (Barbier 2004c). In the case of Italy, on the other hand, despite efforts to activate the system, a sustainable mismatch emerged between social needs and institutional capabilities.

The elements of the French activation reforms seem to be picked either from the liberal and from the universalistic ideal-types: (1) a sector of employment programmes, well established, mostly wage-based, certainly amounts to significant activation, although the ‘hidden unemployment’ question also apply in France as in many other countries. (2) Social services, family and housing benefits, have so far been spared any linkage to any particular work incentivising. In this respect, contrary to the liberal type of activation and closer to the universalistic model, there seems to remain a significant – although decreasing – room for manoeuvre for ‘traditional’ social policy. (3) While activation was introduced from the late 1980s, unemployment insurance and ‘assistance’ have recently been at the forefront of the public debate; reforms have tended to target remaining regulations leading to possible ‘inactivity traps’, especially for minimum income benefits’ recipients. (4) Finally, the gradual decrease of employers’ social contributions’ has taken the leading role in the activation dynamics, and was recently ‘embedded’ into the working time reduction process, along with emerging tax credits.

Rather than specific measures, it is perhaps the underlying assumptions and concerns which bring France and Italy close to one another: both countries have left stigmatising and punitive provisions largely in the background. Indeed elements of the ‘workfare’ ideology did filter through to Italian policy makers at the beginning of the 1990s but they were progressively discarded in practice. With the Prodi Cabinet, labour market policy (politica del lavoro) was put on the front burner, but most of the attention was devoted to eliminating structural barriers and bottlenecks hampering wider access to the regular labour market (Fargion 2003a). In line with this approach, financial incentives aimed at bolstering new hirings became the primary tool along with the piecemeal loosening of labour market regulations governing the standard contract. The results appear rather mixed. Contrary to the original expectations, these policies have ended up adding a further generational cleavage to the already existing ones. Actually, as a result of flexibilisation and in the absence of a concurrent reform of the social guarantees attached to atypical work, Italy now witnesses a stronger than ever generational split: it is increasingly emerging between the adult core labour force enjoying standard permanent job contracts and growing numbers of young workers – especially women – on part-time, fixed-term, temporary jobs. Today Italy has no significant assistance programme; its unemployment insurance is still very fragmented and unequally distributed; despite the reforms of the public employment service, the system is weakly activated. The pension system – in many cases – still provides a crucial source of income for family members not adequately covered by the social protection system. Given the current distribution of Italian social expenditure and the practical impossibility to further exacerbating fiscal and contributory pressure, the pension sector appears as the first candidate for any redistributive effort. However, this avenue is practically barred by the trade unions which stand adamantly united on this front.

The contrasted reform paths in both countries are certainly linked to the legacies of the Southern and the continental welfare regimes. Well established in the Italian case, veto points have played and continue to play a key role, while the French state proved able to redirect significant parts of resources, and undertake a complete overhaul of funding mechanisms (Barbier and Théret 2003). From this different background, important similarities nevertheless also emerge: a similar structural situation in terms of labour market participation at both ends of age groups (the young and the older) and a clear inability so far to reforming early exit from the labour market. Moreover, whatever the reforms, they have occurred on the background of low employment creation and of an increased segmentation of the labour market, with certain categories in the workforce particularly affected by insecurity and instability, low quality of jobs, while the majority has retained significantly better security and social protection rights. In this respect, ‘activation’ reforms of the 1980s and the 1990s have obviously not delivered their promise of integrating all in the labour market. The French system has been able to only limit some of the negative consequences, with ambiguous outcomes (see the insertion case). In the case of Italy, centre-left governments had an interpretation of activation similar to the French but suffered from the veto powers of the unions when it came to really redefining employment policies. On the other hand, the present centre-right government shares an approach to activation closer to the Anglo-Saxon model with conditional and supply-side punitive elements; however, as the preceding majority, it ‘suffers’ from the same vetoes: while pushing the flexibilisation of the labour market even further, it is unable to reform the ammortizzatori sociali. These elements paint the general context for understanding the recent evolution of social citizenship in Italy and France.

The countries appear similar with regard to gender-bias. Although labour market statuses of women and men have been more similar to each other in France than they have in Italy (Table 1), caught in their ‘Republican’ (and presumably ‘gender-blind’) tradition, French policy makers have only begun to seriously address the gender question in the second half of the 1990s. Although the French system has been much more ‘de-familialised’ than Italy's, as individuals, women in France overall enjoy lower quality social protection and lower quality provision of services related to employment. Another common feature concerns the way social policies treat young people. In France, despite significant exceptions (the emplois-jeunes was one in 1997–2002), in the last 20 years, the young have been eligible to lower quality social benefits and programmes. This is of course linked to the previously ‘familialistic’ legacy. Many entitlements and benefits the young under the age of 20–25 are eligible to still depend on the family policy system and are not individualised. Moreover, the young experience worse conditions than adults and older employees in terms of labour market participation and precariousness of employment, especially the less qualified and young women. The situation in Italy is very similar, but increasingly worsening in the recent years, with the prospect of freezing the family's traditional role as the welfare state ‘clearing house’.

An ‘inability to universalism’ (particularly to the ‘generous’ universalism typical of the Scandinavian countries) provides another important feature of social citizenship in both countries. As is typically illustrated in the case of the minimum income benefits, and especially RMI, whereas entitlements and rights are theoretically designed as universal, de facto eligibility is selective. This situation can be explained by many reasons: overload of services, scarcity of resources, actual targeting mechanisms. A similar situation applies to the unequal access to places in employment or vocational training programmes. It leads to some polarisation (the opposition between better quality mainstream insurance-linked provision and lower quality assistance-linked provision, but also the increasing emergence of a working poor stratum in French society). From what we have seen, Italy witnesses an even more fragmented and insufficient provision than France.

Moreover – and this is probably one of the main realities ‘universalistic’ analyses fail to grasp – another common feature in both countries emerges: whatever the activation discourses and the effective implementation of reforms, compared with the strict American or British welfare-to-work (or even ‘workfare’) rationales (Torfing 1999), France and Italy are bound to experience limited pressure for job search, and the absence of a consistent punitive orientation. In the French case, this is not only explained by scarce market opportunities and lack of resources for provision of opportunities, but also by the persisting Republican citizenship reference. In the case of Italy, lack of resources is probably prominent, but the 2002 social movements about employment protection (Article 18 of the workers’ statute) and the embedded difficulty to reform pensions also point to a deeply established implicit contract about the central role of pensions.

Considering the ‘participatory’ or ‘political participation’ aspect of social citizenship in the context of activation, the French RMI can be used as typical. Its core political justifications entail the extension of the effective exercise of political participation to the polity. This ‘Rousseauist’ feature sharply differs from the Anglo-Saxon conception of ‘rights and responsibilities’ (Barbier and Théret 2003). Insertion professionnelle retains a positive function of political integration. At the same time, ‘collective participation’ to the definition of rights is marked by the domination of central state and a relatively secondary role ascribed to social partners. Phases when unions resort to confrontation and street demonstrations are also characteristic of the French polity. In the Italian case, experimental RMI programmes (reddito minimo) most clearly endorsed the commitment to enhance participation by the poor and the socially excluded in the country's social, cultural and political life – not simply to provide them with a job. Actually, between 1998 and 2001 the centre-left majority passed a number of other measures aimed at expanding the substantive content of social citizenship. The 2000 Acts, respectively, on equal opportunities and social care services, are a case in point (Fargion 2003b). Both measures reflect the incorporation of post-materialist values as self-fulfilment with a new emphasis on combining personal, family and professional life. Precisely this type of legislation brings to the fore the coexistence within the previous centre-left coalition of two different perspectives on social policies and the role of the welfare state: a more traditional one deeply entrenched in the old left-wing parties’ membership and the trade unions, and a more innovative one promoted by a minority group of reformists and policy experts. One needs to bear in mind that during the 1990s turbulent reshaping of the Italian party system, the trade unions remained and still are the actors with the strongest mobilising capacity.17 Under centre-left governments, albeit having to accept compromises, the unions were able to limit the redistribution of resources from ‘over-protected’ to ‘under-protected’ risks, largely on the basis of elites negotiations. Quite differently, in the face of the confrontational style of the centre-right majority, unions have been able to present themselves as the staunch defenders of the welfare state per se – bringing millions of people on to the streets to protect the existing social protection system, irrespective of its growing mismatch with the reality of social risks and needs. Needless to say, conflating principles with reality – as both sides on the political spectrum are inclined to do – will certainly not help to disentangle the inequities of the Southern model.

Considered in the context of the current Europeanisation process and, especially of the gradual extension of the Open Method of Coordination, and with the hindsight of welfare states reforms, the assumption would appear credible as to the persistence through time of various social citizenship models across the European Union and not only a model tending to emulate the British, or, even less, the American.

1.

We term ‘universalistic’ the epistemological posture which contends that social phenomena are to be studied in the context of universal laws, and which assumes that, in any country, ‘functional equivalents’ can be found of phenomena observed in one country. The ‘universalistic’ posture has polar counterparts, in the ‘relativistic’ or ‘culturalist’ ones. In our introductory paper (Barbier and Ludwig Mayerhofer, this issue), we have used the term ‘universalistic’ with another, different meaning, referring to the universal provision of services and benefits, which characterise the social-democratic, Scandinavian welfare model.

2.

The calculation of unemployment has long been impacted in Italy by the limited capacity of its statistics to take into account whether the unemployed were or were not available; a reform was introduced in 2000, but not yet fully implemented, see next section.

3.

Standard employee status in France is contrat à durée indéterminée; in Italy the mainstream employment relationship is regulated by reference to the 1970 Statuto dei Lavoratori.

4.

The majority of their holders are low qualified employees or workers; for instance call-centre operators are in this way ‘disguised’ as self-employed to save on non-wage labour costs. In just a few years the number of co-co-co has increased to about 2,500,000.

5.

The active population in France has been in the range of 24–25.5 million in the 1990s (between 24.5 and 25.3 in Italy).

6.

Recipients after a certain period of eligibility will be transferred to a new benefit (revenu minimum d'activité, RMA), which is deemed to entail enhanced employment obligations in 2004.

7.

Article 2 of the RMI Act reads as follows: ‘Every person residing in France whose income (…) does not reach the amount of the minimum income (…) and who is at least 25 or is in charge of one or several child (ren) (…) and who accepts participation in the activities, determined with him/her, that are necessary for his/her social or labour market integration, has a right to RMI” (1988, 1992 amended, Acts).

8.

The number of sanctions has sharply increased with the introduction of the PARE: in 2002/2003, there were about 35,000 monthly sanctions (radiations) for a total of 340,000 people leaving the PES register (in 1991, when the labour market was also very depressed and the number of people leaving the register was about the same, the number of sanctions was about 50,000 annually.)

9.

Although both populations are similar: from 1990 to 2001, the French population grew from 54.6 to 57.7 million; the Italian population from 56.2 to 57.2 million.

10.

There were 80,000 CIG beneficiaries in 2000, out of 1.4 million of unemployment benefit recipients.

11.

Lavori socialmente utili, See further.

12.

Ammortizzatori sociali, in Italian, literally ‘social shock absorbers’, refer to unemployment insurance and assistance programmes for the unemployed.

13.

In a second stage (see next section), the very notion of what was considered ‘active labour market policy’ in France was to be transformed through the introduction of the systematic decreasing of employers’ social contributions.

14.

Because the ‘availability’ criterion was not applied, the number of unemployed officially registered with the job placement offices was unduly boosted. With the introduction of a new registering procedure in August 2003, availability is supposed to be checked, but this is just starting to be implemented.

15.

It must be stressed that similar discussions developed even in the most successful countries, including Denmark (Jørgensen 2002).

16.

This was for instance the case with the introduction of the allocation parentale d’éducation (APE).

17.

One should recall during the transition period, given the parties’ weakness, the executive resorted to the social partners and particularly the trade unions to legitimise crucial policy decisions.

Barbier
,
J-C
, (
2004a
). “
‘A comparative analysis of ‘employment precariousness’ in Europe’
”. In:
Learning from Employment and Welfare Policies in Europe.
(
2004a
). pp.
7
18
, in M.T. Letablier (ed.), European Xnat Cross-national research papers, http://www.xnat.org.uk/.
Barbier
,
J-C
, (
2004b
). “‘The European Employment Strategy, a channel for activating social protection?’”. In:
The Open Method of Coordination in Action: The European Employment and Social Inclusion Strategies.
Brussels
. (
2004b
), in J. Zeitlin, P. Pochet and L. Magnusson (eds.), (forthcoming).
Barbier
,
J-C
, (
2004c
). “‘National systems of social protection in Europe: two contrasted paths to activation, and maybe a third’”. In:
Labour and Employment Regulation in Europe.
Brussels
. (
2004c
). pp.
233
54
, in J. Lind, H. Knudsen and H. Jorgensen (eds.).
Barbier
,
JC
, and
Gautié
,
J
, (
1998
).
Les politiques de l'emploi en Europe et aux Etats-Unis.
Paris
. (
1998
).
Barbier
,
J-C
, and
Théret
,
B
, (
2003
). “‘The French social protection system: path dependencies and societal coherence’”. In:
Changing Patterns of Social Protection.
New Brunswick
. (
2003
). pp.
119
68
, in N. Gilbert and R. Van Voorhis (eds.).
Esping Andersen
,
G
, (
1999
).
Social Foundations of Post-industrial Economies.
Oxford
. (
1999
).
Fargion
,
V
, (
2003a
). “
The European employment strategy and the Italian case
”. (
2003a
), paper presented at the conference‘Changing European Societies – The Role for Social Policy’, Danish National Institute for Social Research, Copenhagen, http://www.sfi.dk/graphics/ESPAnet/papers/Vfargion.prn.pdf.
Fargion
,
V
, (
2003b
). “‘Half way to the Ford: the Italian Welfare State at the start of the new century’”. In:
Changing Patterns of Social Protection.
New Brunswick
. (
2003b
). pp.
309
38
, in N. Gilbert and R. Van Voorhis (eds.).
Ferrera
,
M
, (
1996
). ‘
The “Southern Model” of Welfare in Social Europe
’,
Journal of European Social Policy
6
((1)) ((
1996
)), pp.
17
37
.
Ferrera
,
M
, and
Gualmini
,
E
, (
1999
).
Salvati dall'Europa? Welfare e lavoro in Italia.
Bologna
. (
1999
).
Gualmini
,
E
, (
1998
).
La politica del lavoro.
Bologna
. (
1998
).
Ichino
,
P
, (
1996
).
Il lavoro e il mercato.
Milano
. (
1996
).
(
2003
). “
I servizi per l'impiego. Monitoraggio 2002
”.
Roma
. (
2003
).
Jørgensen
,
H
, (
2002
).
Consensus, Cooperation and Conflict, the Policy Making Process in Denmark.
Cheltenham
. (
2002
), (ed.).
(
2003
). “
Rapporto di monitoraggio sulle politiche occupazionali e del lavoro
”. (
2003
), http://www.welfare.gov.it.
Sestito
,
P
, (
2002
).
Il mercato del lavoro in Italia. Com’è, come sta cambiando.
Bari
. (
2002
).
Torfing
,
J
, (
1999
). ‘
Workfare with welfare: recent reforms of the Danish welfare state
’,
Journal of European Social Policy
9
((1)) ((
1999
)), pp.
5
28
.
Treu
,
T
, (
2001
).
Politiche del lavoro. Insegnamenti di un decennio.
Bologna
. (
2001
).

Jean-Claude Barbier is Directeur de Recherche (CNRS, Sociology) at the Centre d'Etudes de l'Emploi (CEE, Noisy le Grand, France). He works on the international comparison of social policies, social protection systems and, particularly employment policies. He also currently studies the European Employment strategy, from the point of view of dissemination of ideas. He also does research on the epistemological and methodological stakes of the inter-national comparative posture.

Valeria Fargion is associate professor of Political Science at the University of Florence; She teaches Public Policy and is also Jean Monnect Professor of European Social Policy. She has been the Italian representative in the management committee for COST A13 (‘Changing labour markets, welfare policies and citizenship’) and is currently a member of the ESPANET Board and of the ISA RC 19 Board. Her main research interests include comparative social policy and the Italian welfare state with particular reference to decentralisation and centre-funds on Italian Regional Governments.

This is an open-access article distributed under the terms of the Creative Commons Attribution-NonCommercial 4.0 International (CC BY-NC 4.0) license, which permits unrestricted use, distribution, and reproduction in any medium, provided the use is non-commercial and the original work is properly cited. For a full description of the license, please visit https://creativecommons.org/licenses/by-nc/4.0/legalcode.