Comparing Social Protection in France and Germany

In order to adopt the most comprehensive approach possible to the activation of social protection, it is necessary to situate the reform of unemployment compensation, minima sociaux (minimum social benefits) and employment policies in a broader context, which of course involves labour markets and employment regulation (Giraud and Lechevalier, 2010). It is not only a question of ‘activating people’ – as we often say somewhat lightly from an ethical point of view – but, rather, systems, programmes and whole sectors of social protection. The overall logic of restructuring is intended to make these systems more ‘employment-friendly’, particularly in terms of their financing. That is why, in order to analyse the reform dynamics of France and Germany, five main policies should explicitly be considered: a) unemployment insurance and the active labour market programmes it is crucially connected with; (b) social assistance, sometimes – as in France – made up of various minima sociaux, also linked to programmes intended to ‘activate’ the poor and the assisted; (c) the design and sources of social protection funding: social contributions – those of both employers and employees – were reformed because they were seen as affecting the labour market strategies of employers (hiring, in particular) and of employees (adapting their labour supply); (d) family policy because it has sometimes happened that compensation for raising children was now framed in ‘tax credit’ terms; (e) finally, in both countries, strategies were implemented in order to facilitate early exits from the labour market in the 1980s: this was of course the exact opposite of ‘activating people’.

It is especially interesting to compare the ‘activation strategies’ implemented in Germany and France, for three main reasons. Firstly, the comparison shows that when assessing the consequences of reforms of social protection one has to acknowledge a shift, especially visible in countries with extensive social insurance, from the situation that prevailed in the 1970s and into the late 1980s: boundaries that previously existed between ‘sectors’ of social protection have been increasingly blurred. Secondly, despite its undeniably innovative character at the time it was introduced, the approach in terms of “the three worlds of welfare capitalism” (Esping-Andersen, 1990) has gradually shown its limitations: the case in point here is the ascription of France and Germany to the same, improbable, category, which is an obstacle to understanding how they reform. Finally, while ‘broad view’ comparisons are very effective in stressing general trends, the loss of diversity they entail can be very detrimental to the comparison of what really happens in various countries. This can be illustrated by the explanation in terms of the ‘dualisation’ of social protection that is supposed to capture what has happened in ‘Bismarckian’ countries (Palier and Thelen, 2010).

The French and German Strategies Before the Turn of the Century

In France, two characteristics dominate the period that came to an end with the new direction taken during N. Sarkozy’s presidency, from 2007. With hindsight, and despite the fact that it may be simplifying things, this period can be said to have begun in 1988, with the advent of ‘insertion’ policies and the RMI (Revenu minimum d’insertion) – that is the first factor. This was referred to as “Republican activation” (Barbier and Théret, 2001; 2009). With undeniable originality in comparison to others, the French system did not focus its reform on compulsory labour, but rather on the goal of integrating everyone into society. During that period, a second permanent feature of the French reform emerged: the resort to a reduction in employer contributions, intended to ‘activate’ the labour market by encouraging recruitment (an action on labour demand, not supply); this characteristic was further coupled with the reform of social protection financing. There were, of course, other major reforms, such as the changes made to unemployment insurance and the public employment service. However, for the purpose of simplification, these two factors are considered to have been dominant. For a country like the United Kingdom, one might easily compare them with the general use of obligations and the strengthening of sanctions in social assistance on the one hand, and the widespread use of tax credits on the other.

In Germany, the ‘new activation’ was carried out differently. What was soon to emerge as a “paradigm shift” (Knuth, 2006; 2009) started a little later, and the reunification of Germany played a decisive role, as did federalism – two factors that were specific to that country. Firstly, unlike French ‘Republican’ activation, German reasoning was rooted in older principles: it referred to the importance of taking account of “inviolable human dignity” (die Würde des Menschen, under Article 1 of the German Constitution (Grundgesetz). The Assistance Act of 1962 (Bundessozialhilfegesetz) also contained a principle that would long remain dormant, according to which those receiving assistance had to work as long as they were able to. For the purposes of our comparison, these binding principles can be considered in terms of an approximate equivalence with the French principles of integration and insertion. Their substantial differences are clear, however, and we would not get very far if we restricted ourselves to a functional comparison: that is what happened to the authors who discussed Germany and France in the pioneering work published on so-called ‘workfare’ (Enjolras et al., 2000 ; Voges et al., 2000). Can we find an equivalent of the second element of the French strategy in the domain of funding? As we will discuss later, the issue is not entirely obvious, because the symbolic principle of equivalence of employers’ and employees’ social contributions to social insurance prevents these from being manipulated, as was the case in France. However, the share of contributions in the financing of social protection has declined in many ways less visible than in France, reaching very similar proportions today (the two great “Bismarckian” nations stand at roughly 60%). The fact is that funding issues were not raised in the same way in Germany, mainly because of the division of powers between the Bund, the Länder and the municipalities. However, as in France, financing was a key element of reform. Also as in France (where the RMI was the third pillar, alongside insurance and the allocation de solidarité spécifique (ASS), as shown by J.L. Outin, 2008), before the reform, the German system of unemployment benefits and assistance was based on three pillars (Arbeitslosengeld, Arbeitslosenhilfe and Sozialhilfe [1]). This is how the stage was set in Germany for the ‘big reform bang’, which deeply altered the structures of social protection against unemployment. The great contradiction was that what led to an unexpectedly radical reform with unintended outcomes was the very cost-containing policy of the federal government – the gradual closure of Arbeitslosenhilfe for direct access by those who had failed to earn insurance entitlements, protracted over a long period from 1976 to 2000 – and thus the refusal to adapt Bismarckian social insurance to the contingencies of modern labour markets and employment careers.

On these bases, summarised schematically here, it is interesting to study the way in which reforms ‘activating’ social protection have evolved in the two countries. Although they have displayed common characteristics, they have by no means resulted in identical systems.

From a comparative point of view, the ‘insertion’ aspect of the RMI was ‘activation avant la lettre’, with a rationale very different from ‘workfare’ (Barbier, 2002). It however represented a major turning point in the French system. By comparison, changes to the strategies introduced in the period after 2000 were a great deal more superficial than those implemented by the Hartz German reforms (Knuth, 2006; 2009). In 2007, however, in a new reversal, the French reform seemed to draw closer to that undertaken in Germany. In all these developments – which arose not from a grand design based on reason, but rather from trial and error, and conflicts between actors – national patterns were quite different.

Hartz Reforms: Public Employment Service and Benefits

When, in 2002, the Schröder government reacted to a scandal about statistics of the Public employment service (PES) by setting up a commission headed by Peter Hartz, the intention was to modernise the Bundesanstalt für Arbeit, but not to fundamentally change the institutional set-up of the PES or break it up into several strands of service provision. Nevertheless, this was precisely the unintended result of a reform project taken up in the name of ‘activation’.

Even before the Hartz Commission was established in the mid-1990s, and culminating around 2000 in the discourses transported through the Bündnis für Arbeit, the idea had been propagated that the problems of the German labour market and labour market policies were related to the institutional incongruence between Arbeitslosenhilfe and Sozialhilfe and that these two benefits somehow had to be harmonised or even merged. In August 2002, the Hartz Commission therefore stated that in order to have ‘Modern Labour Market Services’ (Hartz Commission, 2002), i.e. “one-stop” jobcentres in the fashion of the day, Arbeitslosenhilfe and Sozialhilfe would have to be merged into a new, streamlined benefit – known as Arbeitslosengeld II (unemployment benefit II) – for all jobseekers who could not claim insurance benefits. When draft legislation came out in 2003 as the fourth stage in the implementation of the Hartz Commission’s proposals – therefore called “Hartz IV” – the following features became apparent.

Arbeitslosengeld II was to be a flat-rate benefit only slightly higher than Sozialhilfe. Disregard of assets and other sources of income was greater than it had been with Sozialhilfe but more restrictive than with Arbeitslosenhilfe. On average, former recipients of Arbeitslosenhilfe lost out (even to the extent of not qualifying at all) whereas former recipients of Sozialhilfe experienced slight gains. Despite borrowing heavily from the Sozialhilfe regime, explicit reference to “human dignity” (Würde des Menschen) was not carried over to the new legislation [2]. By contrast, the Hartz IV legislation stated its primary objectives to be strengthening people’s sense of individual responsibility and helping them to break free of benefits by taking up employment. The regime of Arbeitslosengeld II is very comprehensive in that it covers all persons of working age and not permanently incapacitated, whose means do not suffice to support themselves and who have no other sources to turn to, namely claims to any insurance-based social protection.

Whereas Arbeitslosenhilfe was a wage replacement for the former breadwinner leaving ‘inactive’ spouses out of the administrative process, Arbeitslosengeld II construes the formerly inactive spouse as a needy person potentially to be activated. Consequently, a reform justified on the grounds that it would reduce unemployment initially boosted the national unemployment count considerably, to just over 5 million when implemented in 2005. In a situation which showed many similarities, the French response was very different, and the fear of seeing an increase in unemployment figures no doubt explains the reason why the law on the Revenu de solidarité active (RSA – see below) did not go as far as the Hartz reform.

Criteria of job “acceptability” (Zumutbarkeit) were carried over from the legacy of Sozialhilfe, not from the insurance regime. This meant that any job was acceptable as long as it did not interfere with childcare or place demands beyond personal capacity, and provided it was not considered immoral on account of the nature of the work, the working environment or extremely low pay. In the absence of a legal minimum wage, pay is considered immoral by the courts if it is more than 30% lower than the ‘going rate’. Thus there is no safeguard against a downward spiral of the ‘going rate’ itself. Whereas sanctions in Sozialhilfe were at the discretion of frontline workers, there is now a meticulously prescribed regime of legal sanctions by which defined percentages of the benefit are supposed to be withdrawn in response to certain offences. In case of repeated offences, the benefit may be withheld completely or substituted with vouchers. Public employment programmes (namely the Arbeitsgelegenheiten) are now implemented on much larger scales. These are dominated by the ‘working-for-the-benefit-plus’ type (therefore called ‘1-Euro-jobs’ because the allowance for extra expenses is between 1 and 2 Euros per hour), which means that they do not require a job contract. Incentives to take up work even if it does not yield a living wage were increased by constructing a sliding scale of disregard for earnings. Where low earners receive tax credits in countries like the UK (and, marginally, in France), German low earners not supported by partners or other means will receive complementary Arbeitslosengeld II. There are currently about 1.3 million of them, which implies hidden subsidies for low-paid jobs of 11 billion Euros annually.

Since Arbeitslosengeld II would now be clearly distinct from insurance principles, the social partners (as representatives of the contribution payers) were stripped of their statutory influence as far as this benefit was concerned. So all in all, the regime under which Arbeitslosengeld II falls can be characterised by an unequivocal ‘work first’ orientation; strong financial incentives to take up work (though hampered by the fact that wages for unskilled work have a tendency to fall below the poverty line); disposing of status maintenance and protection as well as considerations of the Arbeitsmarktordnung (‘good order on the labour market’); a vast re-commodification of circumstances that were previously considered ‘out of the labour market’; and, consequently, an administrative and potentially activating impact on the whole household with the intention of optimising all family members’ attitudes towards becoming independent of the benefit.

Against the fundamental paradigm shift implicit in the merging of Arbeitslosenhilfe and Sozialhilfe into Arbeitslosengeld II, the long-term effects of the other parts of the Hartz reforms (I-III) seem limited. The maximum duration of Arbeitslosengeld proper for older workers was lowered from 32 months to 24.

Exemption from job search for older workers was closed to new entrants from 2009. Subsidies for ‘gradual retirement’ (Altersteilzeit – old-age part-time working - in practice usually involved taking early retirement by working full-time in the first half and taking time off in the second) were phased out for new entrants from 2010. Following earlier legislation that has gradually come into force, all premature old-age pensions before the age of 63 are being phased out and the statutory pension age at which one can draw a ‘full’ pension without penalty for premature take-up is scheduled to rise stepwise from 65 to 67 between 2012 and 2027.

Changes in Labour Market Performance

The effects of this new regime on the labour market are ambivalent. During the 2005-2007 upturn, the ratio between employment growth and unemployment reduction was more favourable than in the previous upturn. Transitions from unemployment into employment were considerably more enhanced than in the two preceding business cycles (Knuth, 2010). On the other hand, fear of job loss has increased among the employed, while labour turnover rates have declined and attachment to current jobs has stabilised. While the employment rate increased by 4.4 percentage points between 2003 and 2007, the full-time equivalent employment rate

grew only by 2.1 percentage points during this period, indicating the expansion of part-time work. Employment growth generated by the last economic upturn has been carved up into more and smaller jobs. Increasing wage dispersion, a growing low-wage sector, growing shares of fixed-term contracts and increasing market shares of temping agencies complete the picture of a downward spiral in working conditions and pay levels. Nevertheless, since the crisis, developments in the respective labour markets in France and Germany have been quite different: while the high unemployment rates in France have jeopardised the new ‘activation strategies’ planned since 2007, this is not the case now in Germany, where a 5% decrease in GDP in 2009 barely affected employment figures. The average annual unemployment rate (national count) was lower in 2010 (7.7%) than in 2008 (7.8%) and only slightly higher in 2009 (8.2%).

Comparing the German Activation Strategy

A comparison of German and French reform strategies can also be carried out by determining whether one country has an ‘equivalent’ of the other’s strategies. While reforms of minima sociaux and the various employment programmes show frequent similarities, there are also a great many differences. In addition, two aspects of the German strategy contrast with that of the French: the challenges of financing social protection, considered from the point of view of the relationship between “levels of government” on the one hand and the role of social partners on the other; also, the role of what has been known as “marginal work” in Germany (geringfügige Beschäftigung).On the issue of funding, in Germany we encounter two very different points in the reform. Firstly, the separation between the federal government and the states (Länder) – and, even more so, the municipalities – is very different and plays a decisive role. Secondly, both the institutional and cultural processes used in the management of social protection provide social partners in the two countries with very different roles. Overall, the issue of what in France is known as “fiscalisation” (transfer of the funding from social contributions (social security funds) to taxes (the budget)) is very different in the two countries and leads to different constraints. In Germany, beyond the main rationale for the merger between Arbeitslosenhilfe and Sozialhilfe, lay the idea of introducing a “single reception point” (einheitliche Anlaufstelle) for all unemployed people (which, in France, was referred to as “single-counter” (guichet unique), but Pôle emploi has never had this “single” role if one considers the overall numbers of unemployed and poor).

In the end, that streamlined service did not emerge in either Germany or France, although for different reasons. In Germany, the Hartz reforms were also intended to resolve the urgent budgetary problems faced by the municipalities, which had long found themselves having to bear increasing costs. Consequently, the Bundesanstalt, meanwhile renamed Bundesagentur in order to express its organisational modernisation, was foreseen to take charge of the recipients of the two remaining benefits, Arbeitslosengeld proper and Arbeitslosengeld II, in a uniform and streamlined process, whose intake process would be designed by a ‘customer centre’. However, this plan was fiercely opposed by the Christian Democrats. The political-ideological justification was that the municipalities were allegedly better prepared than the centralistic and bureaucratic Bundesagentur to address the multi-faceted social problems of the long-term unemployed and the former recipients of Sozialhilfe and to administer a strict activation regime. The political interests behind this were those of county administrations (with the Christian Democrats in the majority in county councils) who feared losing their administrative function because Sozialhilfe was being taken away from them (Knuth, 2006).

The final outcome of the compromise (Knuth, 2009) has been a mixed implementation and governance structure quite different from a ‘single gateway’ for all workless people: 69 municipalities (out of roughly 440) are fully in charge of Arbeitslosengeld II and thus, in their territories, running a completely separate and second ‘public employment service’ for the non-insured workless population in parallel with the local branches of the Bundesagentur für Arbeit. However, in the majority of local areas, the respective municipality must form a joint venture (presently: Arbeitsgemeinschaft; in the future: gemeinsame Einrichtung) with the local branch of the Bundesagentur. As from 2012, there will be around 40 more municipalities fully in charge alone, following a compromise struck to resolve constitutional conflict about the ‘mixed administration’ between a federal and a municipal unit for which there was no justification in the Grundgesetz (the federal constitution). So the irony of German activation may be summarised as follows: instead of adapting the “Bismarckian” system of unemployment insurance to the contingencies of modern labour markets and life courses, it was actually increasingly closed off in a process lasting more than two decades, thus shifting part of the burden of underemployment to the municipalities. Even so, the former Bundesanstalt für Arbeit was unable to “activate” the long-term unemployed it was left with. The merged benefit Arbeitslosengeld II was recast in the mould of Sozialhilfe in order to bring dormant principles of activation inherent in Sozialhilfe to the fore. In addition, due to the intricacies of German federalism, an implementation structure involving both federal and municipal actors cannot have straightforward governance, which further impedes “activation”. Instead of creating “single gateways” for all jobseekers, Germany now has a two-tier PES – like before the reform, if one considers the former municipal social assistance bureaucracy as a kind of PES, a situation comparable to the French one.

The ratio between employers’ and employees’ contributions is another important different feature in both countries. Unlike France, in Germany, the equal sharing of contributions – the Beitragsparität – is a key principle of “Bismarckianism” as far as Germany is concerned, deeply enshrined in the German perception of social justice and in the idea of Soziale Marktwirtschaft, and the corporatist governance of the social insurance bodies (equal votes and equal contributions). The trade unions fiercely resist a situation in which there would be no lower limit to employers’ contributions. In reality, however, the parity principle has been undermined in many ways, and great pains have been taken in order to uphold its façade, despite pension and health care insurance reforms. Besides, something like the French contribution sociale généralisée (CSG), a hybrid between contributions and taxes, would be impossible to introduce in Germany: contributions are dedicated to a particular branch of social insurance and must only be used for purposes related to the specific social risk covered by it (Versicherungsprinzip); taxes may not be strictly earmarked (Non-Affektationsprinzip). However, like in France, and to a degree rarely appreciated in public debates, social insurance contributions in Germany are complemented by federal revenue. The Hartz IV reform may be seen as an example of shifting certain risks, clients, provisions or operations from the responsibility of an insurance scheme into a tax-funded scheme.

Another difference, one that would be overlooked in a purely functional comparison, characterizes the concept of so-called “marginal” employment (geringfügige Beschäftigung), sometimes wrongly understood in France as a functional equivalent of the development of “precarious” employment as a way of activating social protection. In the Hartz reform, the traditional tax and contribution privilege for “marginal employment” was amended once again, having been modified many times before, under the political label (not introduced in the legal texts) of “mini-jobs”. Though not traditionally considered an instrument of “active” labour market policies for the unemployed, the Hartz reform context caused the mini-job reform to be seen as a means of “activation” by some observers. An employee with a mini-job as his or her only job, now defined as a job paid at a maximum of 400 Euros per month, is exempt from employees’ social insurance contributions and taxes for that job. In turn, such a job does not earn social insurance entitlements except a very small increase in pension entitlements stemming from the employer’s contributions. The employer now (2011) pays 31.08 per cent in contributions and as a non-progressive income tax. In order to overcome apparent lock-in effects, a sliding scale for employee’s contributions was introduced, setting in gradually between 401 and 800 Euros, but with little effect in terms of take-up of these so-called “midi-jobs”. Far from being a means of “activation”, like the reduction of labour costs in France, this type of employment means that, for the employer, the “non-wage labour costs” are even higher than normal, while the employee is largely excluded from social insurance despite the contributions the employer is paying as a sort of penalty for creating non-standard employment. This apparent riddle can be solved by looking at the legacy of “marginal employment”: it was invented during the labour shortage of the 1960s as a means of promoting female labour without changing gender arrangements (the male breadwinner model) which had just been restored through the 1950s social protection reforms. By limiting the number of hours worked to 15 (at that time  now the threshold is expressed in terms of earnings), the domestic role of women would not be interfered with, additionally leaving the tax privilege for married couples untouched.

France’s New Strategy Caught in the 2008 Economic Crisis

With hindsight, certain characteristics of the ‘French activation strategy’ (Barbier, 2008b) led to a relatively stable combination of measures. The start of the new year in 2008 did not signal the end of that combination, but it radically changed certain aspects of it: the general reduction in employers’ social contributions, still dominant in terms of financing, remained intact; so did family policies, which were not affected by “activation”. However, changes to the working age emerged, and others were implemented in the area of minima sociaux. In 2007, the new French president announced a considerable number of reforms. The list of Acts passed in 2008 is impressive: reform of the Public employment service, merging of minima sociaux (a clear instance of similarity with Germany) simplification of subsidized contracts, reform of “social partnership” (i.e. the rules for unions to be formally accepted as “representative”), reform of employment contracts in the labour market. On all these issues, it was widely considered by French actors at the time that sharp policy turns were involved. It was believed that the new government intended to reform the French arrangements in a systematic and systemic way, in the direction of more flexibility and more explicit “activation strategies”. However, bar the reform of social partnership, all implementation processes directly clashed with the 2008 crisis

. It is thus impossible to tell whether the post-2007 decisions would have eventually resulted in a really new (and not only theoretical) “activation strategy”, in the absence of the crisis. It is certain that new conceptions were developed that could, just like the Hartz IV reforms in Germany, amount to a “paradigm shift” in France in these policy fields. Nevertheless the manifestations of it are yet to materialise in actual terms.

A New Public Employment Service

The first reform to be implemented was the merging of the unemployment insurance funds (ASSEDICs) and the PES (Agence nationale pour l’emploi). From January 2009, when the merge was started (which, in any case, was to take a long time), it was intended that placement and income replacement activities would be combined under the responsibility of a single institution. UNEDIC only kept its overall decision-making role for the social partners’ jurisdiction, while ASSEDICs and local ANPE branches were merged under the new name, Pôle Emploi. Social partners retained their competence in the management of unemployment insurance (establishing social contributions and the balancing of budgets for the fund, in particular), while the unemployed were now the ‘clients’ of a single entity, which not only paid their benefits (PARE and ASS) but also delivered them the whole range of services and ALMPs. The reform of the PES and its merging with the unemployment insurance local funds had been on the agenda for many years. It was presented by the new government, in 2007, as a key factor for enhancing the effectiveness of labour market policies and institutions. However, this has not yet materialised, because of the sudden increase of unemployment (from 7% in 2007 to about 10.5% in 2010), which produced grave and persisting internal difficulties in the new institution Pôle Emploi.

Another “activating” measure might eventually well be in jeopardy in the near future: the DRE (dispense de recherche d’emploi) is supposed to be completely cancelled by the year 2012, forcing many – if not all – recipients concerned to look for work. However, in early 2010, around 320,000 still enjoyed the “dispense”. It is difficult to predict today what the situation of the labour market will be like in two years’ time. Finally, from mid-2008, a new Act established a new definition of “offre raisonnable d’emploi” (normal or acceptable job offer), which restricted the range of reasons for which the unemployed could refuse offers. It was clearly an “activation” mechanism. Inasmuch as something is known about the present difficult situation of the PES (especially in terms of training and human resources management), it is generally considered that the actual implementation of new sanctions has been, de facto, very limited [3].

RSA, the End of a Legal, Unconditional Right to RMI in Turbulent Times

Certainly, however, the reform that the government wanted most to change the French “activation mix” was the introduction of Revenu de solidarité active. As in Hartz Germany, this reform dealt with the lowest social safety net. While the reform was being prepared, some actors wanted all working-age minima sociaux to be merged into a single one: ASS (for long term unemployed), API (for lone parents), RMI (the mainstream minimum income benefit) and AAH (for disabled persons). At the same time, the idea was circulated according to which all recipients would be legally obliged to register at the new PES (Pôle Emploi) and to look for work – an obligation that had never been imposed nor implemented formally before. Thirdly, another simplification was announced: instead of the four types of subsidised contracts existing in 2007, a single contract would replace them, streamlining employment conditions in the private and non-profit sectors. Finally, RSA was also to be merged with the existing Prime pour l’emploi (a supposedly “activating” and incentivising tax credit introduced in 2000, which all existing evaluation studies have considered ineffective in this respect).

This comprehensive programme of reforms was hailed as consistent, winning support in many sections of the public and across party lines. At the end of the day, the system was to be based on a new “activation rationale”, that of “working more to earn more” (“travailler plus pour gagner plus” was the presidential motto): as with Arbeitslosengeld II, any working-age minima social recipients would be “incentivised” to take a job, and at the same time would keep his/her benefit to a certain limit, as long as his/her new job did not pay more than the limit. It was a clear shift, paradigmatic certainly, when compared to “republican insertion”.

However, things did not go as predicted. The RSA Act – symbolically voted on 1^st December 2008, twenty years after the 1^st December 1988 (RMI Act) – was to be implemented from 1^st June 2009, and so it was. However, AAH and ASS benefits remained outside the scope of the merge. Only API and RMI were merged. Moreover, the Prime pour l’emploi was merged on the same day for recipients with an income lower than the minimum wage (‘SMIC’) [4], but, as statistics showed in 2010, non-take-up of the new merged benefit by former PPE recipients was high. All in all, the immense majority of former RMI and API recipients make for the bulk of the 2011 caseload and, on two counts that can indicate a degree of “activation”, present figures are not very favourable: (a) the proportion of RSA recipients registered with Pôle Emploi is at about the same level as it was before the reform [5]; (b) the proportion of RSA recipients who combine their benefit with a supplement linked to their employment – although tricky to compare with former proportions – has not altered significantly (this is another clear difference with the German situation in which the number of recipients of Arbeitslosengeld II rose sharply). In a way, these still inconclusive and partial statistics lead to the assumption of a very limited impact in terms of “activation”, while, on the other hand, one of the main provisions of the RMI Act was cancelled, that is, the unconditional nature of the right to RMI: the right to RSA is now conditional to looking for work, whether by registering at Pôle Emploi or by organising similar contacts with the département’s local administration (in theory, this applied – as it did in Germany – to all household members who were RSA recipients).

Finally, as of January 2010, a new Contrat unique d’insertion (CUI = single subsidized contract for insertion) replaced, as was expected, the various contracts that existed before: conditions of wage-based employment under the contract are now homogenized, and apply to the private and non-profit sectors. These subsidised contracts, part of the “activation supply” for the long-term unemployed and target groups selected by the employment policy, are not very different from the previous mainstream Contrat emploi-solidarité analysed above (a part-time contract of at least 20 hour or a full-time contract for 6 to 24 months). When the crisis began in 2008, the government announced the funding of more of these contracts for the year 2010. Subsequently, the early 2010 stock of holders of such contracts again started to rise with unemployment, and pre-2000 numbers of such contracts could compare with what happened at the time. Such a dispensation was however threatened by fresh measures introduced to curb the budget deficit in autumn 2010, but the situation is uncertain in this respect for 2011.

Did the crisis erase shifts in activation strategies? All in all it is difficult to analyse the state of new “activation strategies” in France after the sharp turn introduced by the new policy that was established after the presidential election. Yet, with the massive impact of the 2008 crisis on the margin of manoeuvre for implementing the new measures, in 2010, the main elements of the 2006 “activation strategy” were still in place: the structural reduction of employers’ social contributions, a large stock of employment programmes, a diversified minima sociaux system with associated subsidised contracts and other actions, and finally, a limited impact of “activation” both in the family policy area, and in the domain of “activating the older workers”. However, for obvious reasons, taking stock of the set of reforms passed in 2007-2008 remains transitory (not to mention the unexpected pension reform being passed 2010).

The Diverse Dynamics within “Bismarckian” Traditions

Comparing France and Germany is not only interesting for learning about the striking similarities that exist between the two countries: the common elements present in their long history of social insurance, those they borrowed from one another, and, on the other hand, their persistent differences. It is also interesting because, in the European Union, these two large countries play a considerable role while being generally considered as part of a common, vaguely defined “model”, the so-called “Conservative-Corporatist” type of welfare regime. The detailed empirical inquiry that we have presented in the previous pages allows drawing three conclusions to be set in a wider perspective.

First of all, if it allowed understanding that “welfare systems” could not be seen as mere variations over an universalistic design, the traditional classification of welfare regimes was always in doubt for its ability to explain and foresee the dynamics of social protection systems (Barbier and Théret, 2001; 2009). Moreover, it has also been especially difficult to match regimes with corresponding realities across all sectors of social protection. “Activation strategies” is certainly a case in point that serves to illustrate this difficulty. The stylised institutional setting of both countries only explains some characteristics of the story. It is true, as Clasen & Clegg (2006) have shown, that social insurance and the implication of social partners in its governance and steering are factors that have impeded an “activation rationale” to smoothly settle in Germany and France, as they had in the UK and Denmark in the mid-1990s. That common heritage of the big “Bismarckians” is without question, as is their long tradition of accommodating various groups of interests and statuses even among insurance beneficiaries, with different benefits. However, many other factors can be observed in the functioning of present systems. Political cultures active in social protection are one example: symbolic conceptions of justice are present in both countries, which are nevertheless strikingly different. Although the manipulation of levels of social contributions eventually became one key factor of the French strategy, it was less readily available in Germany, because of the traditional importance of an equal sharing of contributions in that country (Beitragsparität). Elements pertaining to these political cultures can be traced far back in history: this is the case of the religious roots discussed by S. Kahl (2005), but also the case exemplified by the amazingly lasting impact of the French Revolution on the invention of Republican insertion, which was certainly a precursor of ‘activation’ avant la lettre. Although very different in their moral and political justification, the German principle of die Würde des Menschen as an obligation for the state to protect, and the French principle of the state’s debt to its citizens (Barbier and Théret, 2001) make the German and French systems closer to one another than they are to the “Beveridgeans”. However, many crucial differences still remain.

Another case in point is the role played by the minimum income benefits we have exemplified in both countries: in a way, the “activation rationale” – at least the part of it concerning the unemployed and the poor – came from this area, but, even so, the pace and outcomes of reforms in both countries is now, in 2011, very different. If we see “activating” principles once dormant in a residual system interacting with the insurance system and having an influence over a country’s “employment model” as a whole, the exact nature and tradition of a country’s residual system of minima sociaux becomes a relevant question. In this paper, we have highlighted differences between RMI and Sozialhilfe in their cognitive and ideological frameworks, but we still stand short of an adequately qualifying term. Yes, if we superficially apply conventional typologies, the German system of social protection against unemployment has become more “liberal” because “residual” provisions have displaced wage-replacing provisions to a considerable extent (Mohr, 2008). But is the tradition of Sozialhilfe itself a “liberal” one? Does it make a difference whether a national “employment model” and welfare regime are “activated” by building on principles of German Sozialhilfe, French RMI, or Danish sociale hjælp? In order to answer such questions, one has to go back to the religious roots of poor relief and follow up on their secularisation and institutionalisation, which could only partly be done in this paper. It is worth mentioning in passing that – to a certain extent but in a different way – “residualism” is also present in the Scandinavian welfare regime, and that the role of social assistance (or its equivalents) within national welfare regimes has probably been under-researched in an international perspective. So, by saying that France and Germany are “hybrids” because, in both countries, Bismarckianism is combined with residualism, we are not really saying anything new. This has long been said of France (Barbier and Théret 2001; 2009).

Another aspect of the differences concerns the collective reference to the wage-earner status. We refer here to an analysis inspired by the “Régulation” school of economists in France, who identify the rapport salarial (Lohnverhältnis) as a key element of the economic and political ‘nexus’ (or, the social compact). They have always considered Germany and France to be differing types of a wage-earner society, adapting differently to crises and pressures (Boyer, 1986). What we learn from comparing both countries is that the relationship of social assistance to the mainstream situation of a wage-earner has remained different, despite the predominantly common economic features, when considered abstractly. Whereas the French state was always expected to act as an employer of last resort (although it performed this role in a very petty manner when compared with the generous Scandinavian tradition), the German state was never expected to do so as a fundamental obligation. While, from the 1980s, the French state systematically resorted to large-scale ‘employment programmes’ (of which we analysed the typical ‘contrat aidé’) this was not the case in Germany. This difference also has to do with political and economic choices made after the war, and the special understanding of the roles of the state and the market.

Accordingly, we saw that many among the ‘activation strategies’ that directly targeted the unemployed and assisted persons have tended to originate in the assistance sectors. This dynamic would not exist, of course, if France or Germany had only contribution-based insurance-type or only tax-funded minima sociaux. However, a national welfare regime exclusively based on entitlements earned through employment-related contributions is incomplete because there will always be circumstances that prevent people from earning an insurance-based claim. Esping-Andersen’s (1990) contempt for “residualism” being a measure reciprocal to “de-commodification”) corresponds to his failure to establish a difference between the Mediterranean and the Conservative type of welfare regimes: worse than having a large share of “residual” provisions is the lack of them in Southern countries. Therefore, developed “conservative” or “Bismarckian” welfare states will always have an undercurrent of “residualism”, which may well tend to increase as life courses and employment careers become more contingent, with the exact proportions between the two depending on whether insurance systems open up or close themselves off against the volatilities of the employment regime. It seems nevertheless that in both countries, dynamism for far-reaching “activating” reforms did not emerge out of the unemployment insurance regime but more from the safety net provisions for income support, although this had potential repercussions on the insurance regime. Hence, the blurring between traditional “sectors” of social protection that we were alluding to in the introduction has become increasingly evident.

Both French RMI and German Sozialhilfe comprised “activating” principles, however dormant they may have been for decades. Dormant activation principles in Sozialhilfe were more narrowly work-oriented; they lacked economic justifications for not being able to work, and were directly in conflict with competition-dampening labour market regulation and status-preserving rules for the acceptability of job offers. This may explain why these rules were kept dormant for so long, but also how they are now beginning to show such radical potential after having been transferred from the field of poverty relief to mainstream labour market or “employment” policies. In the French RMI context, by contrast, one only has to discover that the best way to achieve social inclusion (insertion sociale) is through inclusion in employment (insertion professionnelle) in order to place greater emphasis on work and employment for recipients (more so, obviously with the introduction of RSA), without, however, the deregulatory implications we observe in Germany, at least so far. It remains to be seen which dynamic the RSA reform will develop in the next upturn, if it is to survive until then, and whether this will have repercussions on the employment options of the insured unemployed. The survival of ASS as a buffer, if only of a psychological nature, will make a difference. With the Hartz reform, Germany went much further than France, which seems to be hesitating in the midst of the present economic crisis.

Secondly, differences between sectors of social protection and policy areas are significant. This cannot easily be captured by the welfare regime approach. For instance, it is well known that family policies in France have remained specific to the country, with labour market impacts, especially where young people are concerned. However, ‘sectors’ are not administered in the same way in France and Germany. We have documented the importance of the various “levels of government” that differ in the two countries. Their differences – which, as is well known, lead to the existence of various ‘veto points’ – are illustrated here in the case of the Public Employment Service and the role of municipalities in Germany (which cannot play the role of a functional equivalent of the départements in France). Obviously, the two political systems react differently to external shocks. The centralised French system tends to react more flexibly, not to say in opportunistic ways, thus possibly going back on its policies after reforms have already got underway, or letting them falter in the vagueness of implementation. Often characterised as immobile because of its numerous checks and balances, the German political system simply cannot go back on itself once a reform process has been set in motion, for the simple reason that there is no centre of power that could take responsibility and reap a short-time political dividend for doing so. So, instead of restoring dwindling popularity by calling off a reform, as French presidents would do, the German Social Democrats went through the ordeal of reforms they had initiated at the expense of a prematurely terminated chancellorship, the breaking away of a new left-wing party, and subsequent tremendous electoral losses.Clear differences also separate Germany and France in terms of their reform of the funding mechanisms of social security. The symbolic political importance in Germany has probably helped a great deal in maintaining equal sharing of social contributions between employers and employees, at least on the surface; in contrast, this principle was not applied in France, leaving the door open for employers’ social contributions to be manipulated on a massive scale. At the end of the day, however, both countries fund a similar share of their social protection expenditure through taxation. Clear stakes linked to institutions par excellence, e.g. the institutions of federalism/centralism and decentralisation have also had clear, important consequences on the eventual strategies chosen, especially in the 2000s.

In both countries, we find comparable benefit reforms, namely a merging of non-insurance benefits in Germany and a partially-failed attempt to do the same in France, and an organisational reform of what is commonly called the Public Employment Service. However, this is only a temporal coincidence in France: the merging of unemployment insurance funds with the PES (ANPE) was not directly related to the fact that the two benefits, administrated at the regional level, became more work-oriented. In Germany, by contrast, the unintended differentiation within the PES (to the extent of its actual duplication through municipalisation in one quarter of the territorial units) and political and constitutional struggles around this issue have been at the forefront of the reform process, to the point where they have obscured and hampered “activation” issues. Nevertheless, as with a Hawthorne experiment in which turning the light up or down has the same “activating” effect, the strain of reorganisation and the inherent competition between different organisational forms has had an “activating” effect in itself. This mechanism seems to be lacking in France, and it remains to be seen whether the merging of the insurance and the job placement function will have any ‘activating’ effect.

Thirdly, despite numerous similarities, one very different element is the timing of reforms. This timing is linked to internal national decisions and to power struggles taking place among relevant actors – whatever the dissemination of common, “neoliberal” ideas, as they are sometimes known. Corrections of reforms, when they happen – as for instance the failed reform in France in 2003-2004, and the changes made to the situation of the older unemployed in Germany under the Grosse Koalition – require compromises to be made and internal decisions to be achieved within each polity. We think that this explains a great deal about what happened in the two major “Bismarckian” countries.

Whereas France, compared to Germany, was a latecomer in introducing both a national scheme of unemployment insurance (1958 as opposed to 1927) and a modern equivalent of social assistance (RMI, 1988, as opposed to 1962), its republican-based discourse about insertion always carried an “activating” undertone, although not in the narrow sense of ‘work first’. French attempts to adopt the philosophy of “activation” in the insurance regime came earlier than in Germany (AFR 1988, AUD 1992, PARE 2001, attempt for ASS reform 2003). German “lateness” can be explained by the problems of unification which forced Germany to resort to the widespread use of traditional “active” policies of providing paid alternatives to unemployment and thus reducing the excess of labour supply. When Germany finally came around to making “activating reforms”, these were all the more radical, and happened to be debated and implemented during a recession, which therefore meant they were already in place in time to be effective, in certain ways, when economic recovery came in 2006 and 2007. France, by contrast, debated its latest reforms during that recovery, so that the start of their implementation currently seems to be thwarted by the shock of the crisis.

A more comprehensive review of the differences between France and Germany would also need to explore the differing impact that “activation strategies”, particularly those in the 2000s, have had on different sections of the population: the young, the elderly, women, migrants and their families, notably. Initial insights, especially with regard to the gender dimension, have shown that women have, again, been disadvantaged (Giraud and Lechevalier, 2010; IAQ, 2009), whereas results with regard to clients with migrant backgrounds are mixed (Knuth 2010).

Finally, in France and in Germany, as in other countries, the “promises of activation” were essentially not met (Barbier, 2009). Colleagues, however, have recently discussed a process that they see as one of ‘dualisation’ especially in “Bismarckian” countries (Palier and Thelen, 2010). In “regulationist analysis” terms, the prevalence of the wage-earner relationship is linked in both countries with what is called Normalarbeitsverhältnis in German, and to CDI – contrat à durée indéterminée in France). Because of the development of atypical employment, and the emergence of Prekarität in Germany (later than in France), it is undeniable that traditional full-time employment has tended to lose its quasi-monopoly. Is this, however, a consequence of “dualisation”? We would tend to think that there is much more to the present evolution than a simple opposition between the classic “insiders” and “outsiders” economists like to stylise in their analysis. Certainly, labour market specialists identify such “outsiders” and “insiders” in firms. Yet this divide tends to be non specific and universal; it has also existed in other countries that are not Bismarckian (such as the UK, Laparra et al., 2004), and it is doubtful whether the highly stylised opposition between “insiders” and “outsiders” can provide much useful information about actual actors – individual or collective. The recent “dualisation” analysis, in the case of Germany and France, probably fails to capture most of what is happening, however. The social insurance principle was always ‘status-saving’ in both countries; the idea that both their systems had to be “encompassing” and “universal” was always to be taken with a pinch of salt, contrary to what some seem to assume (Palier and Thelen, 2010, p. 139). Rather than ‘dualisation’, what we observe in France and Germany are the growing consequences of increasing and manifold inequalities that lead to a patchwork of situations, and, as far as social protection programmes are concerned, the fragmentation of types of rights and entitlements that vary according to traditional sectors of social protection (education, housing, pensions, healthcare coverage, unemployment insurance and assistance). But what we see in both societies is further fragmentation into various sub-groups, with “activation strategies” contributing to this fragmentation (women, the young, sectoral differences for pensions and unemployment, working poor people and “marginal” employment, migrants, etc.).