Does the New Managerialism Stabilise Gender Asymmetries in Street-Level Interactions? The case of Germany after ‘Hartz IV’
Karen Jaehrling, Institute for Work, Skills and Training (IAQ), University of Duisburg-Essen
The transition within the German social security system to policies of activation was accompanied by a ‘recategorisation of risks’ (Clasen und Clegg 2011), i.e. a change in the administrative categorisation of certain groups. Thus the new ‘basic allowance for jobseekers’ (known as ‘Hartz IV’), introduced in Germany from 2005 onwards, changed the status with regard to means-tested benefits of many non-working women with children, who had previously not been registered as ‘unemployed’. After the reform, they too were regarded as unemployed and were now explicitly included in the target group for the labour market authorities’ ‘activating’ measures. Similar regulations now apply in Germany and other countries to other groups of benefit recipients, such as those with health problems or older workers.
At the same time that the activation paradigm was introduced into Germany's labour market policy, gender equality was more strongly enshrined in legislation (cf. Jaehrling, 2009). This formal entrenchment of the goal of equality and the elimination of the special statuses associated with gender or gender roles are in keeping with the global trend towards the ‘de-institutionalization of gender differences’ (Heintz/Nadai 1998; Heintz 2008). Thus, regulations that directly or indirectly infringe on the equality imperative are increasingly regarded in functionally differentiated societies as illegitimate. As a result, gender differences and asymmetries are losing their direct institutional foundation and hence their stability.
At the same time, however, the nature and objectives of employment promotion have changed. As in other countries that have introduced ‘activating’ labour market reforms, the emphasis now is on returning individuals as quickly as possible to the labour market while at the same time imposing stricter requirements on jobseekers. Because of the associated risk of a mass influx of women into low-paid jobs that might also be difficult to combine with care responsibilities, the activating labour market reforms have from the outset been viewed critically by gender researchers (cf., among others, Lewis 1998; Ostner 2004; Letablier et al. 2011). As long ago as the early 1990s, scepticism was growing among feminist researchers on the welfare state about the disproportionate emphasis in social policy on the adult worker model which, in their view, did not take sufficient account of care responsibilities, and particularly those of women with children (among others, Fraser 1994). In the context of activating labour market policy, therefore, does the de-institutionalisation of gender differences in fact produce the opposite of gender equity? Does equality of treatment actually lead to levelling down, resulting in precarious labour market biographies and ‘marginalized care work’ (Knijn 2000)?
These questions lie at the heart of the following analysis. The aim of the analysis is to investigate the implications for gender equality when gender differences in treatment are no longer closely prescribed by policy directives. The study is based on the assumption that the actions of so-called ‘street-level bureaucrats’ and the relevant framework conditions, such as the introduction of managerialist principles into the employment service, play a decisive role in this regard. The analysis will therefore focus on the interactions between street-level workers and their clients and aims to ascertain the extent to which institutional incentives or, conversely, caseworkers' gender beliefs serve to stabilise gender asymmetries in these interactions. The first section (2) is given over to a discussion of studies in the sociologies of gender and public administration which lend support to the assumption that street-level bureaucracy is playing an increasingly important role in reinforcing gender differences. The next section (3) turns to the example of Germany and begins by analysing how legal and administrative regulations on parents' obligations to seek work have changed in the context of ‘activating’ policy reforms. The practices and routines that have developed in the interactions between job centre caseworkers and benefit recipients with dependent children will be investigated in the subsequent section (4). The analysis is based on organisational case studies in 10 job centres carried out in the first half of 2008. The findings are then discussed with reference to the question of the importance of gender beliefs and institutional incentives in the interactions between street-level workers and their clients (Section 5). The analysis confirms that, despite the formal de-institutionalisation of gender differences, in many cases it is actually only the reproduction mechanisms that have changed. The main factors producing gender differences in the interactions between job centre staff and benefit recipients are not primarily gender stereotypes among the former but rather time-honoured routines, standardised actions enshrined in law and conflicting legal goals that structure the interactions in a way inimical to equality of treatment. In contrast to the US welfare-to-work programmes, however, the practices adopted towards women with children tend not to be repressive but are intended rather to be protective. The final section (6) is given over to the requirements of equality policy under these changed conditions.
2 Street-level bureaucrats, the new managerialism and the impact on gender inequalities
If we accept the argument that gender differences have been de-institutionalised, then the actions of street-level bureaucrats become the decisive factor in the reproduction of gender asymmetries. This is because a de-institutionalisation of gender differences does not automatically lead to their disappearance, according to Heintz/Nadai (1998:79): In place of the routine application of rules that differentiate directly between the sexes, the preservation of gender asymmetries depends increasingly on the deliberate and targeted actions of the actors concerned, as well as on specific framework conditions that provide a basis for practices which create gender asymmetries. The assumption that a considerable degree of explanatory power can be attributed to the interactive production of gender differences and asymmetries also lies at the heart of constructionist gender theory (West/Fenstermaker 1995; Gildemeister/Wetterer 1992). Accordingly, in the sphere of activating labour market policy, it might reasonably be assumed that it is the interactions between the unemployed and caseworkers in local job centres that largely determine if and to what extent formal equality of treatment paradoxically strengthens gender inequalities or, conversely, contributes to their elimination. Various assumptions have been formulated in the sociology of gender literature as to which framework conditions foster practices conducive to gender asymmetries. According to Ridgeway's studies, for example, this depends essentially on how strongly entrenched ‘hegemonic gender beliefs’, which are ‘hierarchical presumptions about men’s greater status and competence’, are among the actors concerned (Ridgeway/Correll 2004). The gender composition of the groups interacting with each other is an important factor here. In a mixed group, there is a high probability that any hegemonic gender beliefs will become ‘salient’ and influence the behaviour of the interacting parties. However, it is generally assumed in this approach that hegemonic gender beliefs predominate in the vast majority of contexts and situations. This approach can be described as a culturalist explanation of gender inequalities, since it attributes considerable explanatory power to “hegemonic gender beliefs” and conceives them as relatively self-contained influencing factors operating independently of the societal structures involved in the distribution of power and resources (cf. Ridgeway 2014: 12f). Heintz (2008), in contrast, places greater emphasis on the institutional context. Drawing on theories developed in organisational sociology, she takes as her starting point the hypothesis that in highly formalised, bureaucratic organisational structures, in which recruitment and performance assessment are governed by explicit rules, the agents involved in the interactions will tend to prescind from ascriptive characteristics. As a consequence, criteria leading to gender differentiation will play a lesser role.
As far as our present subject is concerned, this brief survey of the sociology of gender raises the general heuristic question of how far gender stereotypes and/or institutional framework conditions contribute to the stabilisation or elimination of gender asymmetries in the employment service. However, it is far too general for us to be able to infer any more specific hypotheses about the possible effects of changes in the institutional context, in this case the labour market reforms. We need additionally to take into account studies in the sociology of public administration that focus on the particular type of structurally asymmetrical interactions between public service workers and citizens or ‘clients’. The ‘street-level bureaucracy’ theory (cf. Lipsky 1980; Maynard-Moody/Portillo 2010) has focused attention on the discrepancies between policy objectives – such as regulations on equality of treatment – and their day-to-day implementation in interactions between clients and representatives of officialdom. This may be because of conflicting goals, vague legal concepts, limited resources or even because the rules intentionally give front-line workers some leeway in their decision-making. In these interactions a balance is by no means always struck in individual cases between clients' needs and the provisions of the legislation, as many studies have shown. Instead, street-level workers develop routines in which stereotyping classifications (by gender, class and race) may find a place. Such attempts to simplify matters by classifying clients on the basis of subjective moral judgements and beliefs are one of the coping mechanisms used by street-level workers to deal with the number and diversity of cases they have to process (Wright 2003).
This raises the question of how street-level decision-making is affected by changes in the organisational framework, and hence in the context of interactions between public service workers and the unemployed. One fundamental element of these organisational changes that have accompanied labour market reforms is a reduction in hierarchical control. In accordance with the new public management model, target-based management systems have gained in importance, with regional and local agencies themselves being left to decide how they intend to achieve the targets that have been set. This is also consistent with the claim that services can be more closely tailored to regional and individual needs (van Berkel/Valkenburg 2007). And even in the interactions between officials and the unemployed, unilateral administrative decisions have been replaced, at least in formal terms, by a contract in which the entitlements and obligations of both parties are standardised (cf. Sols/Westerveld 2007; Weinbach 2012). At the same time, caseworkers have been given greater discretionary authority, since clients’ right to participate in vocational training measures and other active labour market programmes has been partly abolished and financial benefits have been more strongly conditionalised by linking them to possible sanctions, to be applied if certain conditions are not met.
However, many gaps have been identified in these organisational reforms, as well as hybrid forms combining the ‘old’ and ‘new’ modes of administration (Van Berkel/Borghi 2008: 335). Thus, various studies are unanimous in finding that the increase in formal terms of the discretionary authority granted to local agencies and individual caseworkers in their decision-making is being reduced again. This is happening, on the one hand, as a result of direct instructions, guidelines or ‘guidance material’ on how this increased discretion is to be used (Evans/Harris 2004; Rice 2012). Even more significant, however, are the indirect restrictions on discretionary authority imposed by a highly detailed set of new public management (NPM) instruments used to operationalise and monitor the targets that are set (target agreements, performance indicator systems, controlling instruments etc.). These instruments create de facto incentives for local agencies and caseworkers to adopt certain advice and placement strategies, contradicting the claim that more account is being taken of individual needs (Bonvin/ Maochon 2007; Brodkin 1997; 2011, Soss et al. 2011, Steinke et al. 2012). Caseworkers are using their discretionary authority primarily to develop creative solutions that enable them to meet their performance targets. As a result, interactions become highly standardised; computerised classification tools inhibit the proper assessment of clients’ needs and resources (Caswell et al. 2010); and greater use is generally made of punitive measures (Soss et al. 2011). It is true that the discretion caseworkers enjoy also enables them to adopt strategies that take greater account of clients' individual situations or even of caseworkers' subjective preferences. However, such strategies are more costly for caseworkers, in the sense that they require more time and may also cause them to miss their performance targets, and require “attitudes of resistance and non-compliance with administrative requirements” (Bonvin/Maochon 2007: 411). Overall, the analyses identify the institutional framework conditions as the main explanatory factor and attribute a lesser role to caseworkers' subjective preferences: “personal preference is not irrelevant, but it is mediated by organizational conditions” (Brodkin 2011: 260; cf. also Soss et al. 2011: 255).
Thus, various hypotheses as to the effects of organisational reforms can be inferred from studies in the sociologies of gender and public administration. On the one hand, the formal increase in the discretionary authority caseworkers are allowed could mean that their advice and placement work is also more strongly influenced than previously by their own personal value judgements. This would probably lead to considerable variation in street-level performance, with gender asymmetries being strengthened or reduced depending on the caseworkers' gender awareness and norms. On the other hand, the available empirical studies on discretionary authority among caseworkers in the employment service would lead us to expect that gender stereotypes are also losing their significance in the interactions between caseworkers and the unemployed or are being modified by other, institutionally entrenched incentives that shape the interactions to a large extent. At the same time, this would not necessarily lead to a reduction in gender asymmetries, since such work practices obviously do not lead to individualised case processing (which would also be suited to overcoming gender discrimination) but rather to the adoption of certain coping strategies, with street-level workers putting increasing pressure on the unemployed to take ‘any’ job on offer, which in turn may stabilise gender asymmetries overall, as feminist research on the welfare state suggests.
To date, however, there has been little empirical research that has explicitly examined the interactions between street-level workers and their clients in the changed framework conditions from a gender perspective (but cf. Knijn/van Wel 2001; Karl 2012). This is the starting point for the following analysis. Taking Germany as an example, it draws on qualitative case studies and additional quantitative material to investigate what practices and routines have emerged in German job centres as they deal with women with dependent children claiming the new basic allowance for job seekers.
3 Discretionary authority on work obligations for parents after ‘Hartz IV’
When the so-called Hartz IV reforms were introduced in Germany, the previous unemployment assistance and social assistance benefits were merged from 2005 onwards to form the new basic allowance for jobseekers (also known as Arbeitslosengeld II or unemployment benefit II, hereinafter UB II). The reforms increased the obligation on UB II beneficiaries to seek and accept ‘reasonable work’ – which, except for very few limitations, includes any paid job regardless of pay and skill levels or necessary commuting time. The reforms simultaneously increased caseworkers' discretionary authority in respect of parents' employment obligations. This is despite the fact that the statutory regulations on exemption for reason of childcare responsibilities laid down in the law governing social assistance (Article 18, sub-section 3 Bundessozialhilfegesetz) were largely reproduced in the ‘reasonableness’ regulations governing UB II. The regulation stipulates that a claimant who is fit for work can reasonably be expected to accept any work if the youngest child has reached the age of three. However, the regulations leave considerable room for interpretation and, following the reform, this interpretative space is now occupied by administrative regulations differing from previous rules and court rulings. Under the previous regulations governing social assistance, the duration and scheduling of the working time deemed reasonable were restricted by case law and administrative regulations. Decisions were strongly influenced by civil case law on child support and the widely used ‘age-phased model’, also known as the ‘0-8-15 rule’ (cf. Wapler 2010). According to this rule, the parent chiefly responsible for childcare could reasonably be expected to take a part-time job once the youngest child reached the age of 8 and to work full-time once he or she was 15. It was not only court decisions that were guided by this rule; in many Länder, internal guidelines on the implementation of social security regulations also contained similar provisions. The new internal directives contain no such upper limits on what is regarded as reasonable. It is true that Article 1, sub-section 1, no. 4 of Social Security Code II (SGB II) stipulates that, in implementing the regulations governing payment of the basic allowance for jobseekers, account should be taken of ‘the family circumstances’ of claimants looking after children or family members in need of care. However, the corresponding internal directive on reasonableness issued by the Federal Employment Agency provides no further clarification but merely stipulates that children up to the age of 15 have ‘different needs for care depending on their age’ (cf. Bundesagentur für Arbeit 2014). The directive does not contain any more detailed provisions as to what might constitute a reasonable working time or work schedule or who determines the age-related need for care.
The vagueness of the legal terms (‘in consideration of the family’s circumstances’, ‘variable need for care’) means that the actions of the agencies responsible for implementing the legislation are not subject to any clearly defined limits. At the same time, changes in the legislation on post-marital spousal and child support and a strong increase in public childcare provisions, supporting the emerging norm of the adult worker model, may serve as a source on which caseworkers draw when assessing parents' obligation to work. In principle, the increased possibilities for cutting benefits give caseworkers the grounds to expect women with dependent children aged 3 or older to participate in the labour market to a greater extent than the part-time employment that is the norm among large sections of the population, and hence to make households in receipt of benefits a sort of ‘avant-garde by law’ against their will. It is true that, in the view of legal commentators, restrictions can be inferred from parents' right of care and custody as enshrined in the German Basic Law (cf. Rixen 2008). However, this view has to date had little effect on the case law on UB II, since the social security courts have dealt with virtually no cases in which this question was the matter at issue (cf. Wapler 2010).
The absence of legal proceedings on this question is in itself revealing and reinforces the impression conveyed by examination of the empirical findings on parents' labour market participation, namely that in practice only limited use has been made of the increased discretionary authority in order to push through a change in the family division of labour.
4 Routines and practices in job centres
At first glance, the quantitatively measurable indicators on the results of the activation process seem to be relatively unambiguous: at least in the years immediately following the introduction of UB II, women with children requiring care were less likely to take part in employment promotion measures, while the presence of children in some cases had the opposite effect on men (cf. IAQ/FIA/GendA 2009, 145f.). Overall, women are less likely to take paid work, and when they do, the vast majority – a far greater proportion than among men – find employment in so-called ‘mini-jobs’ (until the end of 2013: up to 400 euros per month; since 2014: up to 450 euros per month). Gaps in the provision of childcare, particularly in the afternoons, are undoubtedly one important factor standing in the way of women seeking to work longer hours, particularly in Western Germany. However, this is obviously not in itself sufficient explanation for the gender differences that can be observed in the activation process since, as further analysis shows, women with dependent children are significantly less likely to take part in employment promotion measures even when, according to their own statements, the child care situation would in principle allow them to do so (cf. IAQ/FIA/GendA 2009, 148).
Thus, the available quantitative findings do not really suggest that women with children are involved to an excessive extent in activating measures. On the contrary, it is clear that the restrictions on their availability lead to them being underrepresented in employment promotion programmes. Nevertheless, little can as yet be inferred from these figures, because the reform has also brought about a more frequent use of measures that are partly, or indeed primarily, designed to verify claimants' readiness to work. It is precisely because of this melange of support and coercion, of ‘carrot and stick’, that more frequent participation in employment promotion measures is even less likely than previously to indicate that the individuals in question are being supported in accordance with their needs. Rather, it may very well also indicate that they are being more closely monitored because the job centre has serious doubts about their readiness to work. A positive interpretation of the lower level of female participation in ‘activating’ measures would therefore be, conversely, that it merely indicates they are receiving support more closely tailored to their needs. To that extent, it can be said that, in their case, ‘less is more’. However, it is not only the poor outcomes in terms of women's labour market integration (see above) that gainsay such an interpretation. The following qualitative analysis of day-to-day practices in job centres also provides little evidence to support an inference of that kind.
4.1 How many hours can mothers be expected to work? Interpretations of the ‘reasonableness’ rules
As noted above, the discretionary authority accorded to job centre staff means that they have to assess in each individual case what hours (in terms of both duration and scheduling) parents can reasonably be expected to work, and which parent is allowed to restrict the time spent on their job search activities and their obligation to cooperate with the employment service in favour of looking after children in need of care. As far as our question is concerned, it would be interesting to ascertain what reference points caseworkers employ to guide how they use their discretionary authority and to what extent their personal values, institutional incentives or administrative routines have an influence.
It is evident from the individual interviews and group discussions that caseworkers use a number of different reference points, each one to varying extents. They include the availability of external childcare services, the employment patterns or attitudes to employment prevalent in the region, individuals' own employment biographies or the situation on the demand side of the labour market. Not only do these influencing factors vary between Eastern and Western Germany but they can also differ considerably within individual job centres between the various generations and on the basis of other characteristics. This diversity of practices is also linked to the fact that there seldom seems to be any explicit organisation-wide agreement on how to deal with restrictions on individuals' time. Only in one of the job centres investigated was a verbal agreement arising out of team discussions reported that could be regarded as an organisation-wide rule – in this case on the approach to be taken in dealing with couple households. Caseworkers in several West German job centres stated in interviews that, in the absence of any such agreement, they managed by resorting to the practices and regulations that had prevailed prior to the reforms. Thus caseworkers in two job centres declared, with reference to the age limit up to which it was reasonable for one parent to work part-time only, that they were continuing to apply the rules that applied in the old social welfare system, which varied from municipality to municipality. In one local authority, the upper limit was 15, in another 11.
With regard to the number of hours constituting a part-time job, it was evident from several interviews with caseworkers and managers in the West German job centres that the statutory requirement on parents with care responsibilities to undertake paid work is construed in such a way that mothers (but not fathers!) are generally expected to work half-time in the mornings or even to work ‘for an hour or two’ in off-peak periods:
“In this regard, we expect what the legislature has prescribed, namely that individuals should make themselves available to work for an hour or two [orig: “stundenweise”] or half-time, that's basically what we expect. Because the legislature has declared that ‘everybody must make their contribution to preventing the need to claim benefits”. (job centre team leader)
At the same time, however, the caseworkers often stressed that there are virtually no jobs left in the labour market in which working time is confined to the mornings. In some cases, these severe time constraints were cited as justification for concentrating the placement efforts on the man (see next section). From a gender equality perspective, therefore, this practice has to be regarded somewhat ambivalently. On the one hand, it is in line with the still limited opportunities for afternoon childcare. On the other hand, it creates what is presumably supposed to be a protected space, but one that seems virtually impossible to bring into step with the realities of the labour market.
However, this tolerance of jobs involving very short hours is by no means motivated solely by caseworkers' desire to protect women with children. It also seems to be connected to the fact that placing people in mini-jobs allows them to achieve their performance targets more quickly than trying to place clients in jobs offering (somewhat) longer hours, which can be both more time-consuming and cost-intensive. As part of the system of management by objectives, each job centre concludes a target agreement each year with its superordinate authority, in which targets are set for certain indicators. Besides the number of clients placed in insurable employment, the German system of management by objectives, unlike that in other European countries, until recently attached the same importance to the objective of reducing benefits as to that of increasing the number of individuals placed in employment (cf. for more details Kaltenborn et al. 2010). Since earned income is, except for some small allowances, set off against benefits, even a ‘mini-job’ helps a job centre to reach its target for the ‘total passive benefits’ indicator. The gender officer in one job centre suggested that economic considerations of this kind were the reason why many women are not even offered training courses but are placed immediately in mini-jobs:
This is confirmed, at least indirectly, by other caseworkers, as the following statement by a case manager shows:
“This also means that we can achieve the goal of ‘reducing passive benefits’ only by pushing clients to take mini-jobs. That's not our original objective, it's true, but in any event it's better than nothing at all.” (caseworker)
In the case of households that, because of earned or other income, only receive benefits of 400 euros per month at most, benefit payments can be stopped completely by a mini-job paying the corresponding amount. According to a department manager in one job centre, such households were being deliberately identified and given preference when mini-jobs were being offered, with a view to ending benefit payments altogether:
“One successful measure is the (...) practice of deliberately identifying households receiving less than 400 euros in benefits and placing them in mini-jobs.” (caseworker)
In Eastern Germany, on the other hand, the working time requirements are considerably greater than ‘an hour or two’, particularly because nurseries and schools have longer regular opening hours. It is striking that, in the interviews conducted there, none of the interviewees made any reference to rules that restrict the working time regarded as reasonable until the youngest child has reached a certain age. On the contrary: if UB II recipients placed temporal restrictions on their readiness to work, citing a lack of childcare as the reason, then this gave some caseworkers cause to doubt their basic willingness to work. A common argument in these circumstances was that caseworkers themselves or their close family members were willing to make considerable concessions in order to stay in work, including travelling for several hours to their place of work. This is an indirect legacy of the considerable importance attached to paid work in the former GDR (cf. IAQ/FIA/GendA 2009, 237).
4.2 ‘Priority for sole breadwinners’ – the sequential integration strategy
Up to this point, we have been examining the extent to which a parent with children requiring care is allowed to reduce his or her availability for work because of those responsibilities. In both the Western and Eastern German case studies, such concessions were granted only to the women, even in couple households. Besides the question of the extent to which such an unequivocal arrangement was decided by the job centres themselves, it is also of interest what consequences this practice has for the activation process.
As is evident from the interviews with caseworkers, restrictions on one carer's availability do not necessarily – as might be supposed – lead to a strategy of getting that partner into part-time employment. In some cases at least, the granting of the concessions is used to justify postponing implementation of any integration measures at all for the caregiving partner. Such a course of action, described here as a ‘sequential integration strategy’, is sometimes more the rule than the exception, as is clear from the following statement made by a team manager:
“We discuss with both partners who of the two is going to make himself available for work – actually generally speaking both of them are available for work – but on which of the two the help is to be focused. Usually that's enough – depends on the job someone gets – but under some circumstances it's sufficient if one of them can get placed in work. (…) It's not laid down in writing, it's just developed like that. There are team manager meetings every two weeks at which such decisions are made.” (Team manager)
It is also consistent with such an approach, which completely excludes the person with caring responsibilities, that in two of the few cases in which caseworkers reported of attempts to work towards a role reversal, this did not happen until it became evident after some time that attempts to get the male partner into the labour market were not bearing fruit.
“The man was initially in our woodworking shop [a vocational guidance measure, K.J.]. Then he stopped going and said to me quite clearly: ‘I don't want to do that anymore. I'm not going any more’. Then I invited them both in and in the course of the interview it emerged that the woman was really better qualified and was also more motivated, more committed”. (Caseworker)
Here too, as with the practice of determining what constitutes a reasonable working time, the extended discretionary authority is based on a set of customs and practices that largely corresponds to that which prevailed in the former social assistance system. This is sustained and reinforced to some extent by the gender role models of individual caseworkers, as is evident from further statements made by the case manager quoted above:
“From my point of view, there are groups of people who actually shouldn't be drawing UB II. For example, the standard family with children, where one partner is in insurable employment, while the other looks after the children.” As he had declared earlier, in such a case he would not “take any further measures to help [the woman] find paid work”. (Caseworker)
However, the strategy of giving ‘priority to the sole breadwinner’ may also arise out of a conscious decision by caseworkers to respect people's ‘private sphere’ and to refrain from disrupting the established family division of labour. This attitude was frequently expressed in the interviews, even by caseworkers whose personal role models were more egalitarian (cf. in greater detail Rudolph 2010).
Ultimately, the structural framework conditions also stabilise a sequential integration strategy of this kind, as experiences in the East German job centres suggest. The social assistance tradition has less effect here, not least because social assistance concerned far fewer people in East Germany than unemployment benefit or unemployment assistance. Caseworkers also tend to have more egalitarian attitudes with regard to gender. However, several caseworkers and benefit recipients stated in the interviews that employers have reservations about employing women with children requiring care, even when the women have all-day childcare available, because there is always a possibility that the child could fall ill and that in that case mothers – but not fathers – would be expected to take time off work. This statistical discrimination by employers is obviously used by some caseworkers as a reason to adopt a sequential integration strategy, which can even run counter to the claimant's explicit wishes, as is clear from the following statement by a mother:
“I was certainly interested in getting a place [on a programme], but they said I was the children's mother. I'm responsible for childcare”. (Female UB II recipient)
Furthermore, because of the way the demand side of the labour market is structured, supporting mothers in a search for suitable part-time or full-time jobs – which is actually more time-consuming and laborious – comes up against the limits of caseworkers' restricted time resources. To some extent it also conflicts with the business targets that were criticised by several caseworkers in the East German job centres because of the pressure they exert on staff, whose success is measured by performance indicators. Thus, the organisation-wide targets on frequency of contact – i.e. the obligation to summon clients to the job centre at intervals of so many months – were bemoaned because they force caseworkers to work to extremely tight schedules that do not allow time for longer interviews and more exhaustive searches for job offers with individual claimants. There was also criticism of the fact that, in any case, the time spent on job searches etc. was not rewarded, but merely the successful placements. Thus this pressure to achieve successes is likely to be another reason for the focus on the male partner, who is – supposedly or in fact – easier to place in work.
The qualitative findings presented here are supported in part by a quantitative investigation using administrative data to examine the participation of women in couple households in active labour market programmes (ALMPs) (cf. Kopf/Zabel 2014). This study included information on the family division of labour prior to benefit receipt (measured by the cumulative earnings of each of the two partners over the previous ten years) as an explanatory factor. The quantitative results show that, firstly, ALMP entry rates are generally lower for women with a partner than for single women, regardless of which division of labour couples previously had and, secondly, that women in former male breadwinner households have lower ALMP entry rates than women in households without a clear former division of labour. However, this second finding applies only to Western Germany; in Eastern Germany, women in couple households with a traditional division of labour take part in ALMPs more frequently than other women in couple households. The authors attribute this to differences in gender role attitudes among case managers in Eastern Germany who – in line with the more egalitarian family policies in the former GDR – do not seem to accept a former homemaker role as a reason to be exempted from activation programmes (Kopf/Zabel 2014: 30). This confirms our qualitative findings on the attitudes of East German street-level workers (cf. Section 4.1). Thus, the fact that, on average, women in couple households in Eastern Germany are nevertheless also less likely than single women to take part in ALMPs is obviously due less to respect for the ‘private’ division of labour within such households and more to do with some of the other factors mentioned above. On the other hand, one of this study's findings seems to contradict the qualitative findings presented above, at least at first sight: women with a partner who is himself taking part in an ALMP or is in employment (but whose earnings are insufficient to cover the household's needs) are more likely to take part in ALMPs than other women in couple households. This applies to both Eastern and Western Germany. Thus the sequential integration logic, i.e. the concentration of placement efforts on one of the two partners, seems to be overridden by countervailing logics, at least in a considerable number of cases. At the same time, this finding could be regarded as an indication that caseworkers' practice is being guided more strongly by the performance measures. The authors thus suggest that caseworkers may be “especially motivated to increase women's chances of employment if their partner is already employed, but not earning quite enough for the household to be independent of UB II” (Kopf/Zabel 2014: 26), so that even a small additional income will suffice for the household to pass the income threshold and no longer be eligible for UB II/ALG II. This is consistent with the results of the qualitative interviews with regard to caseworkers' motives for placing women in mini-jobs (cf. Section 4.1).
5 Discussion: administrative stubbornness or policy indecisiveness? On the stabilisation of gender asymmetries in German job centres
Gender asymmetries, it seems, do indeed survive their direct institutional foundations. By replacing across-the-board sub-statutory regulations on parents' obligation to secure income by decisions made on a case-by-case basis (cf. Section 3), the direct institutional forms of support for practices that differentiate claimants by gender have been eliminated. However, as the above analysis has shown, the increased discretionary authority granted to caseworkers is frequently not used to generate tailor-made solutions and support; instead, caseworkers have recourse to blanket rules similar to the routines used in the now-defunct social assistance system. Nevertheless, this tells us nothing about how these rules come about or about the mechanisms by which the gender-differentiating practices are reproduced. Is it ‘traditionalist’ gender stereotypes and ‘old’ administrative routines or rather new institutional incentives that help to stabilise gender asymmetries? Are we dealing with routines to which caseworkers have become habituated and which now live on without any statutory basis, backed up by their own gender stereotypes? There are in fact some striking parallels between the implementation of Social Security Code II and that of a similar reform in the Netherlands, where social security office staff, acting on their own normative convictions that ran counter to the intentions of the reform, established a sort of ‘protective buffer’ against the implementation of the newly introduced obligation for single parents to seek paid work (Knijn/van Wel 2001).
According to the hypothesis of the de-institutionalisation of gender differences, a shift of this kind in the reproduction mechanisms, away from action guidelines enshrined in statute to more or less consciously gendered interactions, is a classic example of the stabilisation of gender asymmetries, which comes into effect particularly in contexts in which the scope for freedom of action and interpretation is relatively unstructured. As far as the present object of investigation is concerned, it can be argued, as outlined in Section 3, that the context is less structured than previously because of the increased discretionary authority granted to caseworkers dealing with women with children aged 3 and over. To that extent, there is some evidence to support the notion that the altered framework has set free hegemonic gender beliefs and that wayward behaviour on the part of caseworkers is subverting the legislature's intentions on gender equality.
At the same time, the comparison between Eastern and Western German practice in particular shows that even different gender beliefs on the part of caseworkers can lead to similar results, namely less coercion and less support for women to take up paid work. This tends to contradict the assumption that the main obstacle to implementation of the legislature's (good) intentions is inertia on the part of the caseworkers responsible for putting the law into practice. One possible alternative explanation would seem to lie in the indecisiveness of the legislature and the competing principles underpinning the legislation itself. The resultant uncertainties give caseworkers a justification for regarding their use of established routines from the previous social assistance system as legitimate. Given that the statutory regulations on parents' obligation to obtain an income were taken virtually word for word from the previous system; and given that the sub-statutory directive, which speaks of ‘different needs for care depending on age’, is virtually an open invitation to treat caregivers differently depending on the age of their youngest child; and given that it was not in any case the primary purpose of the reform to modernise gender roles but rather to harmonise organisational structures and performance levels, then caseworkers would have to be utterly dedicated to the cause of gender equality to see that the perpetuation of past practice contradicts the legislature's intentions.
One important principle that conflicts with gender equality goals is the priority given to savings targets. The objective of reducing the need for state benefits, which is given considerable prominence in both the legislation and the system of management by objectives, creates incentives to place partnered women into ‘mini-jobs’ or even to concentrate placement efforts in couple households solely on the person closest to the labour market – all the more so when staffing levels in the job centres are too tight to enable caseworkers to engage in the more time-consuming task of supporting unemployed individuals who are more remote from the labour market. In this case, therefore, sub-statutory guidelines have a structuring effect on the interactions. Here too, caseworkers would need to be stubborn in setting their own priorities in order to circumvent the implicit gender bias in the guidelines. Activation policy under these specific institutional conditions and in a gender-segregated labour market encourages precarious forms of employment among women.
Thus, though the routines may often have remained the same as in the previous social assistance system, the legal and administrative bases for them have diversified to a certain extent. Despite caseworkers' greater discretion, gender stereotypes have not become more significant as a mechanism through which gender differences are reproduced. Instead, caseworkers' inclination to tolerate or even support traditional gender roles among UB II recipients may also stem from a professional ethos of non-involvement in the private sphere or from more instrumental considerations relating to performance measures. In both cases, caseworkers consciously seek to strike a balance between organisational objectives and client preferences rather than merely enacting their personal ideal. At the operational level, at any rate, the interactions are also structured by the principles implicitly or explicitly enshrined in the legislation. All things considered, the German reform can therefore be characterised as an incomplete de-institutionalisation of gender, since caseworkers feel incited and justified to follow rules that indirectly preserve gender asymmetries.
However, the findings also make it clear that the mere absence of rules that indirectly preserve gender asymmetries would not be sufficient to overcome the difficulties that arise in practice when efforts are made to promote greater gender equality in the labour market. These difficulties include a gender-segregated labour market and statistical discrimination against women. As the statements by both caseworkers and benefit claimants make clear, an oversupply of labour means that employers are able to discriminate statistically, particularly in low-paid segments of the labour market – independently from any behavioural pattern of caseworkers. Moreover, dealing with long-established life plans (including traditional models of labour division) is a challenge, to which there is not simply one ‘right’ or ‘best’ answer from an equality perspective (cf. also Callerstig 2012). Both legislative and organisational changes more systematically and dedicatedly devoted to the gender equality objective are therefore required.
6 Conclusion: gender equality in the activation regime – a relationship in need of reconfiguration
As currently configured, in the German case at least, gender equality as a labour market policy objective relies to a large extent on the stubbornness of caseworkers willing to make the best of the situation, in defiance of incentives that indirectly maintain gender differentiation. Since the barriers in the labour market and among jobseekers are high, this is not sufficient and also places excessive demands on the attitudes and behaviour of care workers. Equality policy, and indeed policy management in general, cannot rely on implementation practices that are subversive; rather, it must be organised so as to create structures that prompt even caseworkers with conservative beliefs and perceptions about gender to act in ways that promote equality.
Decentralised countermeasures implemented by means of the standard set of equality policy instruments (gender quotas) are not sufficient here. Rather, a comprehensive set of gender mainstreaming measures must be incorporated into the very fabric of the legislative framework. For one thing, the omnipresent threat of sanctions and the intensive monitoring of clients' own efforts have a disheartening effect that render constructive collaboration between state ‘service providers’ and their clients difficult and therefore need to be revised. Promoting gender equality, however, requires more than the absence of pressure: in particular, it calls for advice and support adapted to individual needs but at the same time reflecting on structural causes for gendered work trajectories. In this regard, the management instruments that have mainly been deployed to date are not only inadequate but to some extent also counterproductive. What is required is, on the one hand, a different type of management by objectives. The decision of the Federal Employment Agency to refrain from setting target values for the savings goal and instead merely ‘monitor’ this goal (Bundesagentur für Arbeit 2013) might be a first, albeit timid step in this direction.
Even more important, however, is a shift from management by objectives to what Brodkin has referred to as “management by enabling”, breaking with “NPM's essential indifference to how policy work is done” (Brodkin, 2011: 273f.). Caseworkers' professional skills certainly constitute an important feature here. It is not simply a question of whether they have undergone a specific type of training – this is generally the case in Germany, in contrast to the US and UK, where studies have frequently pointed to the low levels of qualification among street-level workers (cf., among others, Wright 2003; Soss et al. 2011). What is at least as critical is the content of initial and further training programmes. In the 1990s, in accordance with the principles of the new public management, elements of economics were introduced into the courses provided by the Federal Employment Agency's training academy, replacing some of the public administration content (cf. Osiander/Steinke 2011). Recently, however, further training courses following a new ‘counselling concept’ (‘Beratungskonzeption’, cf. Rübner 2010) have been introduced for caseworkers in Germany in an attempt to bring counselling practices in job centres and employment agencies up to the standard of professional case management. However, gender equality issues hardly seem to play a role either in the design of these courses or their implementation (cf. ISG/IAQ/tifs/genderbüro Berlin 2013: 145f.). This neglect is partly backed up by widely shared perceptions among both caseworkers and job centre managers that individualised service provision automatically goes hand in hand with support strategies which, in a “gender-neutral” way, are more closely adapted to both women's and men's needs (ibid.). However, the first empirical results on the implementation of the reform provide rather limited evidence in support of a generalised trust in such “automatic” implications for gender equality (ibid: 108 and 135ff.). This is not to argue against the benefits of professionalising and individualising service provision; rather, it reconfirms the point that the reliance on individual street-level bureaucrats' norms (regarding gender or their profession) is not sufficient in itself. What is required, instead, are policies and management strategies that marshal discretion on behalf of greater gender equality, by establishing organisational practices that make gender equality a collective concern and that help street level workers to continuously and explicitly reflect on how to deal with the irreducible tensions and dilemmas arising from the goal of increasing gender equality in the context of strongly gendered life courses and labour markets.
Berkel, R. van, & Valkenburg, B. (2007). Making it personal: Individualising activation services in the EU. Bristol: Policy.
Berkel, R. van., & Borghi, V. (2008). Introduction: The Governance of Activation. Soc Pol Soc 7 (3): 331–40. doi: 10.1017/S1474746408004302.
Bonvin, J. M., & Moachon, E. (2007). The impact of contractualism in social policies: The case of active labour market policies in Switzerland. International Journal of Sociology and Social Policy 27 (9/10): 401–12.
Brodkin, E. Z. (1997). Inside the Welfare Contract: Discretion and Accountability in State Welfare Administration. The Social Service Review 71 (1): 1–33.
Brodkin, E. Z. (2011). Policy Work: Street-Level Organizations Under New Managerialism. Journal of Public Administration Research and Theory 21 (Supplement 2): i253.
Brodkin, E. Z., & Marston, G. (Eds.) (2013). Work and the Welfare State: Street-Level Organizations and Workfare Politics. Public Management and Change series. Washington: Georgetown University Press.
Bundesagentur für Arbeit (2014). Fachlicher Hinweis zu § 10 (Zumutbarkeit) vom 20.6.2014. Nürnberg.
Bundesagentur für Arbeit (2013). Jahresbericht 2012 zur Zielerreichung im Bereich Grundsicherung. Nürnberg
Callerstig, A. C. (2012). Gender equality and public services. In K. Lindholm (Ed.), Gender mainstreaming in public sector organisations: Policy implications and practical applications (pp. 93-115). Lund: Studentliteratur.
Caswell, D., Marston, G., & Larsen, J. E. (2010). “Unemployed citizen or 'at risk' client? Classification systems and employment services in Denmark and Australia.” Critical social policy 30 (3): 384–404.
Clasen, J., & Clegg, D. (Eds.) (2011). Unemployment protection and labour market change: Towards a ‘triple integration’? In J. Clasen & D. Clegg (Eds.), Regulating the risk of unemployment. National adaptations to post-industrial labour markets in Europe (pp.1-12). Oxford: Oxford University Press.
Evans, T., & Harris, J. (2004). Street-Level Bureaucracy, Social Work and the (Exaggerated) Death of Discretion. British Journal of Social Work 34 (6): 871–95.
Fraser, N. (1994). After the Family Wage: Gender Equity and the Welfare State. Political Theory 22 (4): 591-618.
Gildemeister, R., & Wetterer, A. (1992). Wie Geschlechter gemacht werden. Die soziale Konstruktion der Zwei-Geschlechtlichkeit und ihre Reifizierung in der Frauenforschung. In G. A. Knapp (Ed.), Traditionen Brüche: Entwicklungen feministischer Theorie (pp. 201-54). Freiburg / Breisgau: Kore Verlag.
Heintz, B. (2008). Ohne Ansehen der Person? De-Institutionalisierungsprozesse und geschlechtliche Differenzierung. In S. M. Wilz (Ed.), Geschlechterdifferenzen - Geschlechterdifferenzierungen: Ein Überblick über gesellschaftliche Entwicklungen und theoretische Positionen (pp. 231–351). Wiesbaden: VS Verlag.
Heintz, B., & Nadai, E. (1998). Geschlecht und Kontext. De-Institutionalisierungsprozesse und geschlechtliche Differenzierung. Zeitschrift für Soziologie 27: 75-93.
IAQ, FIA, & GendA (2009). Bewertung der SGB II-Umsetzung aus gleichstellungspolitischer Sicht. Forschungsbericht des Bundesministeriums für Arbeit und Soziales No. 396. Berlin, Download unter: http://www.bmas.de/DE/Service/Publikationen/Forschungsberichte/Forschungsberichte-Arbeitsrecht/forschungsbericht-f396.html
ISG, IAW, tifs, & genderbüro Berlin (2013). Gleichstellungspolitischer Auftrag der Bundesagentur für Arbeit im SGB III. Forschungsbericht des Bundesministeriums für Arbeit und Soziales No. 443. Berlin, Download unter:http://www.bmas.de/DE/Service/Publikationen/Forschungsberichte/Forschungsberichte-Arbeitsmarkt/fb443-gleichstellung.html
Jaehrling, K. (2009). Gleichstellung und Aktivierung – Wahlverwandtschaft oder Stiefschwestern? In S. Bothfeld, W. Sesselmeier & C. Bogedan (Eds.), Arbeitsmarktpolitik in der sozialen Marktwirtschaft – vom Arbeitsförderungsgesetz zu Sozialgesetzbuch II und III (pp. 117-30). Wiesbaden.
Jaehrling, K., & Rudolph, C. (Eds.) (2010). Grundsicherung und Geschlecht. Gleichstellungspolitische Befunde zu den Wirkungen von ‚Hartz IV’. Münster: Verl. Westf. Dampfboot.
Kaltenborn, B., Wielage, N., von Bothmer, A., Henkel, A. I., Weishaupt, T., Buchholz, H. L. & Oschmiansky, F. (2010). Zielsteuerung in der Arbeitsverwaltung – ein europäischer Vergleich. Forschungsbericht des Bundesministeriums für Arbeit und Soziales FB 409. Berlin, Download unter: http://www.bmas.de/DE/Service/Publikationen/Forschungsberichte/Forschungsberichte-Arbeitsmarkt/fb-fb409-arbeitssteuerung.html
Karl, U. (2012). Vergeschlechtlichte Kategorisierungen und ihre praktischen Zwecke. In D. Böhringer, U. Karl, H. Müller, W. Schröer, & S. Wolff (Eds.), Den Fall bearbeitbar halten: Gespräche in Jobcentern mit jungen Menschen (pp. 97–137). Opladen, Berlin, Toronto: Budrich.
Knijn, T. (2000). The Rationalized Marginalization of Care: Time is Money, Isn’t it? In B. Hobson (Ed.), Gender and Citizenship in Transition (pp. 201-19). New York.
Knijn, T & van Wel, F. (2001). Does it work? Employment policies for lone parents in the Netherlands. In J. Millar & K. Rowlingson (Eds.), Lone Parents, Employment and Social Policy: Cross-national Comparisons (pp. 107-28). Bristol: Policy Press.
Kopf, E., & Zabel, C. (2014). Active labour market programmes for women with a partner - challenge or replication of traditional gender roles. IAB-Discussion Paper 06/2014.
Larsen, F., & van Berkel, R. (Eds.) (2009). The new governance and implementation of labour market policies. 1st ed. Copenhagen, Portland, OR: DJØF Pub.
Letablier, M. T., Eydoux, A., & Betzelt, S. (2011). Social Citizenship and Activation in Europe: A Gendered Perspective. In S. Betzelt, & S. Bothfeld (Eds.), Activation and Labour Market Reforms in Europe. Challenges to Social Citizenship (pp. 79-101). Basingstoke: Palgrave Macmillan.
Lewis, J. (1998). 'Work', 'Welfare' and Lone Mothers. Political Quarterly 69 (1): 4–13.
Lipsky , M. (1980). Street-Level Bureaucracy: Dilemmas of the Individual in PubIic Services. New York: Russell Sage Foundation.
MaynardMoody, S., & Portillo, S. (2010). Street-level Bureaucracy Theory. In R. F. Durant (Ed.), The Oxford handbook of American bureaucracy (pp. 252–77). Oxford: Oxford University Press.
Ostner, I. (2004). Review-Essay: Aus Anlass eines Geburtstags: „Gender and Welfare Revisited“. In S. Leitner, I. Ostner & M. Schratzenstaller (Eds.), Wohlfahrtsstaat und Geschlechterverhältnis im Umbruch. Was kommt nach dem Ernährermodell? (pp. 44-62). Wiesbaden: Springer VS.
Rice, D. (2013). Street-Level Bureaucrats and the Welfare State: Toward a Micro-Institutionalist Theory of Policy Implementation. Administration & Society 45 (9): 1038–62.
Ridgeway, C. L. (2014). Why Status Matters for Inequality. American Sociological Review 79 (1): 1–16.
Ridgeway, C. L., & Correll, S. J. (2004). Unpacking the Gender System: A Theoretical Perspective on Gender Beliefs and Social Relations. Gender & Society 18: 510-31.
Rixen, S. (2008). SGB II § 10 RdNr 3. In W. Eicher & W. Spellbrink (Eds.), SGB II: Grundsicherung für Arbeitsuchende; Kommentar. München: Beck.
Rudolph, C. (2010). Die Rolle des Staates bei der Regulierung von Privatheit. In K. Jaehrling, & C. Rudolph (Eds.), Grundsicherung und Geschlecht. Gleichstellungspolitische Befunde zu den Wirkungen von Hartz IV (pp. 214-28). Münster: Westfälisches Dampfboot.
Sol, E., & Westerveld, M. (2007). The individual job seeker in the sphere of contractualism. International Journal of Sociology and Social Policy 27 (7/8): 301–10.
Soss, J., Fording, R., & Schram, S. F. (2011). The Organization of Discipline: From Performance Management to Perversity and Punishment. Journal of Public Administration Research and Theory 21 (Supplement 2): i203.
Steinke, J., Koch, S., Kupka, P., Osiander, C., Dony, E., Güttler, D., Hesse, C., & Knapp, B. (2012). Neuorientierung der Arbeitsmarktpolitik. Die Neuausrichtung der arbeitsmarktpolitischen Instrumente aus dem Jahr 2009 im Blickpunkt: Mehr Flexibilität und größere Handlungsspielräume für die Vermittler? IAB-Forschungsbericht 2/2012, Nürnberg: IAB.
Wapler, F. (2010). Was kommt nach dem Altersphasenmodell? Die Erwerbsverpflichtung des alleinerziehenden Elternteils im Unterhaltsrecht und im Recht der Grundsicherung (SGB II). In K. Scheiwe, & M. Wersig (Eds.), Einer zahlt und eine betreut? Kindesunterhaltsrecht im Wandel (pp. 251-72). Baden-Baden: Nomos.
Weinbach, C. (2012). Extra-vertragliche Zumutungen im New Public Contractualism: Die doppelte Logik der Eingliederungsvereinbarung und die Rechtsstellung des Klienten im Sozialgesetzbuch II. Der moderne Staat 5: 377-99.
West, C., & Fenstermaker, S. (1995). Doing Difference. Gender & Society 9: 8–37.
Wright, S. (2003). The street level implementation of unemployment policy. In J. Millar (Ed.), Understanding social security: Issues for policy and practice (pp. 235–53). Bristol: Policy.
Dr. Karen Jaehrling
Institute for Work, Skills and Training (IAQ)
University of Duisburg-Essen
 The author would like to thank two anonymous reviewers for their very valuable comments.
 While this is described in the empirical studies carried out in the US as being very much the exception, the conclusions elsewhere are more nuanced (cf. the contributions in Larsen/Berkel 2009 and Brodkin/Marston 2013) – which could be interpreted as confirmation that the design and effects of new public management (NPM) systems vary depending on the political and administrative context (including administrative traditions and socio-political principles, among other factors).
 ‘Mini-jobs’ were introduced in Germany as early as the 1960s, as a way of encouraging part-time employment for married women, but the number of mini-jobs increased considerably in the 1990s and 2000s. ‘Mini-jobs' are jobs paying €450 or less (until 2013: €400 or less); they are exempt from income taxes and employee social security contributions. Hence, they do not provide a living wage and nor are they covered by health, unemployment or old-age insurance (although, in the case of old-age insurance, employees can waive their exemption from insurance contributions on a voluntary basis). Mini-jobs are therefore financially attractive primarily for those employees who have additional income sources (spouse's income, student grants, pensions etc.).