Implementing Social Justice within Activation Policies
The Contribution of the Capability Approach
Jean-Michel Bonvin 
Luca Perrig
Introduction
Welfare-to-work policies aim at bringing recipients of social policies back to the labor market as quickly as possible. Moving from a statutory view of rights, they follow a contractual vision of social entitlements (Handler 2003) according to which no rights should be granted without corresponding duties imposed on their beneficiaries. By the same token, they convey normative expectations about recipients that are required to comply with them in order to get entitled to benefits and services. As such, welfare-to-work policies include a normative view about how human beings are supposed to behave and what rights and duties are implied by membership in society (Gilbert 2002). They are instruments combining capacitating (improving employability, providing work experience, etc.) and constraining elements (pushing people back to work through the use of more restrictive conditionality and financial penalties or incentives). This mix of empowerment and constraint takes various shapes with respect to the countries and target groups involved: Anglo-Saxon countries tend to give more emphasis to constraining elements (Peck and Theodore 2000) while Scandinavian countries rather focus on the development of employability (Barbier 2004). Mixing restraint and empowerment raises complex issues in terms of social justice, which ought to be tackled not only at theoretical but also empirical level. This task requires extensive attention to implementation processes, i.e. how a normative prescription contained in a legislative provision or an administrative directive translates into a specific practice and, of equal significance, how it is received by the beneficiaries concerned. Three categories of actors are thus involved in such processes: those that shape or design the policy (policy-makers in parliamentary arenas, but also high civil servants in charge of designing specific directives for street-level bureaucrats), those responsible for the implementation of such policies and directives (mainly street-level bureaucrats in public administrations, but also in private and third sector providers to whom such tasks are increasingly subcontracted), and the target groups that “receive” such policies (beneficiaries, users, clients or “citizens-consumers” as they are sometimes called). Our claim is that infringements of the principles of non-interference and non-domination can happen in different ways at all three levels: at design level, where specific normative contents may be imposed by majoritarian parties or by high civil servants on the other stakeholders of the policy process; at implementation level, where managerial tools such as provision agreements or performance indicators aim at reducing or canalizing the margin of manoeuver of local agents or street-level bureaucrats, although these enjoy a significant discretionary power as emphasized by Lipsky (1980) and others; at “reception” level, where conditionalities are imposed on beneficiaries to ensure that they “deserve” benefits or services by behaving in a proper way. Dividing the implementation of welfare-to-work policies in such a way allows identifying the multiple loci and patterns of domination at work within welfare-to-work policies. We will see that welfare policies can interfere with the recipients’ projects in different ways depending on the level of observation. Our contribution investigates the mechanisms through which domination and constraint may take place on all three levels. Our contention is that Amartya Sen’s capability approach, combined with Luc Boltanski and Laurent Thévenot’s economy of conventions, is particularly relevant for completing such a task and unveiling the importance of normativity in welfare-to-work policies. The chapter is articulated as follows. Section 2 briefly presents the capability approach and how a combination with the French economy of conventions makes it particularly appropriate for the study of welfare-to-work policies. Then, we successively tackle the three levels of public policy and how they interrelate. Section 3 focuses on the design stage, which is in the hands of policymakers and high civil servants. Section 4 places the emphasis on the implementation stage, which is a task devoted to street-level bureaucrats or private providers, and how it is shaped or influenced by what happens at design level, more specifically what kind of interference or domination takes place between these two stages of public policy (and their respective actors). Section 5 pays attention to how policy is “received” by its beneficiaries, insisting in particular on how such “reception” is impacted by the two previous levels, i.e. how and to what extent the normative views of designers and implementers – which do not necessarily overlap as they may rely on diverse normative views and expectations vis-à-vis beneficiaries – are imposed on recipients. Section 6 synthesizes the main conclusions of the chapter and suggests some avenues for future reflection and research. The three empirical sections extensively draw on findings related to the Swiss unemployment insurance and to local social assistance programs, which were extensively investigated in the course of five European projects (Eurocap 2002-6, Capright, 2007-10, WorkAble 2009-12, SocIEtY 2013-5 and Re-InVEST, 2015-9).
The Capability Approach in a Nutshell
The capability approach insists that the objective of public action is to develop people’s capabilities or, according to Sen’s preferred formula, the real freedoms that people should enjoy to lead a life they have reason to value (e.g. Sen 1999a). To identify the prerequisites needed for reaching such an objective, two dimensions are to be taken into account. First, what factors are required to make people really free (and not only formally) to lead a valuable life? In this respect, Sen identifies three categories of factors, namely access to adequate resources (goods, services, incomes or transfer revenues), individual factors relating to the personal capacity to lead such a life, and social factors relating to the context or the environment in which the individual lives. He uses as an illustration the case of the bike: if a person owns such a resource, she still needs to know how to use it and to be allowed to ride it thanks to the presence of adequate infrastructure (roads or the like). The enhancement of capabilities thus requires the presence of all three elements: resources, individual and social conversion factors; the absence of one of these coincides with a limitation of real freedom. According to this perspective, the development of people’s capability for work, i.e. their real freedom to have a job that they have reason to value, requires the presence of all three parameters: first, adequate resources, which are essential to create exit options with regard to indecent jobs (as a matter of fact, unconditional access to social benefits of an appropriate level allows their beneficiaries to refuse a job that is poorly remunerated and that they have limited reason to value, since they know that a replacement income is available); second, individual conversion factors, esp. qualifications and competencies, that are necessary to find a valuable job and convince potential employers; third, social conversion factors ensuring that opportunities for having a valuable job exist in sufficient number and are open to everyone (e.g. via anti-discrimination laws, forbidding to discriminate against non-nationals, women, etc.). This first dimension thus provides a requiring yardstick to assess whether welfare-to-work policies recognize and encourage the development of capabilities. In such an approach, domination may take place if the absence of one of these three factors impedes the development of capabilities. In Sen’s view, this needs to be complemented by a second dimension, related to the expression “a life one has reason to value”, which allows placing specific emphasis on policy processes. This notion implies that aspirations and viewpoints of beneficiaries, but also those of other stakeholders, ought to be taken seriously when designing and implementing public action. However, it does not entail that such views or aspirations are to be satisfied unconditionally; as Sen repeatedly emphasizes, only those preferences that someone has reason to value should be supported via public action. Other preferences, such as aggressive preferences (e.g. a taste for torture or making others suffer) or expensive tastes (e.g. have a luxury car or spend holiday in luxury resorts), which could be considered as unreasonable, should not be necessarily supported by public action. Along the same line, it is questionable whether adaptive preferences, the level of which is adjusted to people’s actual circumstances, can be assessed as reasonable (e.g. pupils from a low socioeconomic background that have the desire to pursue tertiary education, but give up their ambitions because they seem unrealistic in their environment). In the capability perspective, it is necessary to discriminate between the aspirations that one has reason to value and those that one has no reason to value. The criterion advanced by Sen to assess the reasonableness of preferences is that of public debate: those individual aspirations that are considered as reasonable after a public discussion ought to be supported by public action; by contrast, those that are assessed as unreasonable after such a test, either because they are too aggressive, too expensive or too low (i.e. people should have higher ambitions) should not give rise to public interventions. The scope of public action, i.e. the extent to which individual preferences and aspirations should be supported by public interventions, should then be decided via a democratic discussion including all concerned stakeholders. This is the very meaning of the notion of “constructive democracy” within Sen’s capability approach (Sen 1999b), which entails that all parties have the capacity to express their views and make them count in the course of such public debates. Failing that would imply that the principle of non-domination is not respected, insofar that some actors are then able to impose their views about what is reasonable (and should be supported) to the other stakeholders of the policy process. This view has far-reaching consequences both in substantial and procedural terms. First, it entails recognizing the plurality of possible conceptions about what “a life one has reason to value” does mean. Here, the French economy of conventions (Boltanski and Thévenot 2006; Jagd 2011) is a fruitful complement to the capability approach. If we seek to uncover the mechanisms underlying policy implementation, we have to put words onto the multiple senses of justice that stakeholders mobilize in the implementation of a policy. In this respect, French conventionalists identify different “orders of worth” as they call them. They are standards of justice on which people draw to make their decisions. For instance, the value of a person or of a specific action can be decided following market criteria (e.g. with regard to productivity or the ability to generate financial returns), civic criteria (where considerations based on the common good and solidarity ought to take precedence), industrial criteria (efficiency in the use of techniques in order to increase the total output) or still other criteria (Boltanski and Thévenot 2006; Boltanski and Chiapello 2018). Hence, there is an inescapable plurality in terms of principles of justice and this plurality should be duly recognized when public action is designed and implemented. For our purpose, this emphasizes, firstly, that social policies can pursue a variety of aims such as pushing people back to work as quickly as possible, improving beneficiaries’ employability, ensuring social solidarity and social integration for all, etc. If a stakeholder is able to push one objective at the expenses of the aims pursued by other stakeholders, then there is domination of this stakeholder over all others. This goes against the principle of non-domination, all the more so if the other stakeholders are forced to comply with the dominant view because the exit costs would be prohibitive. Thus, the French economy of conventions and its focus on normative pluralism allow identifying the negative impact of normative domination (i.e. imposing one’s perspective on social justice) on real freedoms. Secondly, at procedural level, this has two main implications: on the one hand, the content of public action ought to be decided via a democratic debate giving equal weight to all viewpoints. This is what we elsewhere suggested to designate as “capability for voice” (Bonvin 2012) or the real freedom to express one’s viewpoint and make it count in the course of a public debate. On the other hand, and of equal significance, public action ought to be designed in an incomplete way so as to allow a plurality of legitimate individual preferences to be recognized and supported. More concretely, if public interventions are governed only by one order of worth and are devised in such a way that the other orders of worth are not given their due, then local agents or beneficiaries that have different views about these issues are not allowed to express themselves when implementing them. Thus, incompleteness at the designing level leaves space for the deployment of a plurality of legitimate individual preferences at the local level, where policies are implemented and “received” by beneficiaries. Accordingly, the capability perspective insists that, in the analysis of public policies, due attention should be paid to implementation processes and, more specifically, how domination may take place within such processes by impeding the expression and fulfillment of local and individual preferences inspired by an alternative “order of worth”. By using this combination of the capability approach and the French economy of conventions as a yardstick, we are able to submit welfare-to-work policies to two specific tests: first, do they adequately tackle the three factors necessary for making freedoms real and not simply formal? Second, to what extent do they recognize the capability for voice of all stakeholders? In the subsequent sections, these two questions will be successively addressed for the three levels involved in the implementation of welfare-to-work policies.
The Normative Basis of Welfare-to-Work Policies
In this section, we will unveil the underlying normativity of welfare-to-work policies in the Swiss context. We will thus emphasize the informational basis for judgment in justice (IBJJ) of these policies, i.e. the information that they consider as relevant when assessing a beneficiary or a situation (Sen 1990). For instance, when assessing the situation of a recipient, one can emphasize her needs, her entitlements or rights, her lack of competencies or employability, her lack of motivation and willingness to regain autonomy through professional integration, etc. Thus there can be a plurality of IBJJs underlying welfare-to-work policies and these may be combined in diverse ways according to the target groups. The identification of such IBJJs allows better understanding the choice of instruments used in the implementation processes, but also how these official IBJJs connect with the views of local agents and recipients downstream.
The Swiss Road to Activating Welfare
The Swiss welfare state has often been considered as a combination of the liberal and conservative models identified in Esping-Andersen’s typology (Esping-Andersen 1990). Thus, it combines a strong focus on the market as the main locus of solidarity and integration with an emphasis placed on a statutory and familialist view of social policies (Armingeon, Bertozzi, and Bonoli 2004; Bonvin, Dahmen, and Aspalter 2017). This means, first, that integration into the labor market is considered as the most appropriate solution to problems of poverty or social exclusion, and that social policies should promote such professional integration. However, benefits paid by the Swiss welfare state are rather generous and, until the 1990s at least, not primarily considered as potential sources of dependency traps. Second, it entails that the main objective of social policies is not to reduce inequalities, but to ensure that transfer revenues compensate as much as possible the lost revenue; also, Swiss social policies are based on the male breadwinner model and are not meant, at least in their initial stage, to promote female employment. These conventional features of the Swiss welfare state are challenged by the recent reforms toward reinforced welfare-to-work policies. As a matter of fact, we observe a convergence of many welfare institutions (especially the unemployment and disability insurances, as well as a great many of social assistance programs at cantonal and communal level) towards activating their recipients from the mid-nineties on, with an increasing focus on making benefits more conditional and less attractive in terms of levels and duration, together with a reinforced emphasis on the recommodification of recipients. Hence, concerning access to cash benefits, eligibility conditions have been made more restrictive in order to reduce the caseload; at the same time, assessment tools have been introduced in order to evaluate whether people behave properly while receiving benefits, with a view to avoiding paying benefits to people not fulfilling the duties and obligations imposed on them (notably in terms of active job search, attending interviews with street-level bureaucrats or participating to ad hoc programs designed to improve their employability prospects). In a nutshell, both eligibility conditions to access measures (mainly related to payments of social contributions while in work) and (mostly behavioral) conditions to maintain entitlement have been hardened for all target groups that are considered as able to work and to be activated.
Swiss welfare-to-work policies against the capability approach
When assessed against Sen’s analytical framework, this evolution can be summarized along the following lines (Bonvin and Rosenstein 2016). Redistributed resources have not significantly diminished (except for certain target groups such as young people on social assistance for whom the reduction was massive), but stricter conditionality has been progressively introduced in order to be entitled to social benefits. In terms of resources, Swiss welfare institutions tend to increasingly endorse the “dependency trap” and “moral hazard” rhetoric to justify these moves towards a (still moderate) reduction of benefits and a strongly reinforced conditionality imposed on access to benefits. One could say that “Making work pay” is a widely shared notion in the Swiss context where making access to benefits more difficult and less attractive is interpreted as a way to reduce public expenditure and balance budgets on the one hand, and create incentives towards work and financial autonomy on the other hand. It is widely accepted that too many (and too easily) redistributed resources may result into dependency traps or abusive behaviors by recipients that, accordingly, will not do their utmost to regain financial autonomy. In the capability perspective, this implies a corresponding reduction of beneficiaries’ real freedom and an increased risk to be submitted to institutional injunctions, since such compliance is required to access benefits. Indeed, if accessing or maintaining entitlement to benefits is made more conditional upon the acceptance of behavioral conditions, it implies that the possibility to refuse such conditions is also made more costly. Reinforcing conditionality can thus be interpreted as a way to constrain beneficiaries to behave in the proper way, because exit costs would be difficult to bear for them. However, in the neo-republican perspective, this interference cannot be considered arbitrary since the constraints imposed on beneficiaries are legally defined and thus common knowledge among the actors. In this sense, welfare-to-work policies per se do not go against the principle of non-domination as advocated by Pettit or Lovett. Apart from the resources provided by welfare-to-work policies, Sen’s framework also suggests taking into consideration the issue of conversion factors. Welfare-to-work tools place strong emphasis on the employability of beneficiaries and their ability to find a job. Hence, Swiss welfare institutions are increasingly turned into activation tools. As a consequence, when assessing the situation of recipients, street-level bureaucrats are called to focus on the features that may directly impact on their employability, such as their qualifications, competencies and work experience. Hence, it is not necessarily a holistic view of the person that is taken into account during the initial assessment of the person, but one that puts more weight on the features connected with employability and the likelihood to find a job. To use Sen’s words, the informational basis used to assess people’s situation focuses on specific pieces of information and discards others. Compared to the unemployment insurance, Swiss social assistance programs hold a more holistic view of the beneficiary that integrates parameters related to health, financial situation (debt), family life, etc. All the same, from the nineties on, the overall objective of social assistance programs is return to work, implying that all individual characteristics are assessed against this objective: for those who are nearest to the labor market, work is envisaged as an immediate and short-term objective while for others, e.g. suffering from addiction problems or isolation, bridging measures are privileged but with the same ultimate objective, i.e. going back to work and a situation of financial autonomy. This was a very significant transformation within Swiss social assistance policies, as they were turned from social support tools (with an almost exclusive focus on decommodification) to activation instruments insisting on the recommodification of recipients. Such an evolution also entailed a deep-seated transformation of professional practices, as will be developed in the next section. Whatever the rhythm and content of programs, the necessity to activate recipients is widely accepted, which coincides with a significant increase of so-called active expenditure in most Swiss welfare institutions. This signals that the enhancement of individual employability is considered as a social responsibility, all the more so if persons are in difficult situations calling for longer-term public interventions. To use Gazier’s terms (1998), Swiss welfare institutions tend to combine an initiative and interactive view of employability: the former insists on the individual responsibility to develop one’s employability and is inspired by the view of the individual as a self-entrepreneur in charge of maintaining her skills portfolio; the latter focuses on the social responsibility to support those who cannot carry out such a task by themselves, thus emphasizing that the development of employability is an individual and collective task. Both approaches are combined in Swiss welfare institutions along a mix that varies according to the situation of the target groups. Accordingly, the way welfare programs shape their action on individual conversion factors also impacts on the extent to which they respect the principle of non-domination: policies that insist on a quick enhancement of employability and on the individual responsibility in this respect may more easily result into practices and behaviors of domination than programs giving more time and including a social responsibility component. With regard to social conversion factors, we observe the same kind of imbalance within most Swiss welfare institutions. This is due to a specific system of rights and duties, which includes the enforcement of strong sanctions towards beneficiaries not complying with the conditionalities imposed on them (such as suspending benefit entitlement for a certain period or ending it when infringements are considered very serious), together with incentive mechanisms to promote the firms’ and employers’ willingness to collaborate, in the form of financial subsidies for those who accept to hire recipients in traineeships or subsidized jobs. While employers are considered as partners and their freedom of choice is fully respected (i.e. no sanction is planned if they do not contribute to the effort to integrate disadvantaged people), benefit recipients are required to collaborate and do their utmost to find a job, failing which they will undergo significant financial penalties. It is thus significant to note that while welfare-to-work policies do interfere in the recipients’ choices, comparatively little pressure is imposed on the employers. The power imbalance between job seekers and employers remains unchallenged. This shows that Swiss welfare institutions are envisaged as tools of adaptability, i.e. aiming at adapting individuals to the requirements of the labor market, rather than instruments of capability, where the freedom to choose of all stakeholders would be equally taken into account and respected.
Street-Level Bureaucrats as Policy-makers
This section is dedicated to how the policymakers’ vision is implemented through management tools and how street-level bureaucrats (i.e. local welfare agents in charge of implementing welfare-to-work policies) strive to reconcile these requirements with the recipients’ needs. Based on empirical research (e.g. Bonvin (2007); Bonvin and Moachon (2013); Bonvin and Rosenstein (2015); Bonvin and Rosenstein (2016)), we will emphasize how the activation objectives designed by policymakers put street-level agents in an uncomfortable position since they are not easily implemented to their target groups. In order to push street-level bureaucrats to comply with official requirements, especially in terms of activation, Swiss welfare institutions increasingly resort to managerial tools inspired by new public management (Bonvin and Varone 2004). Thus the relationship between policy-makers (and high civil servants) on the one hand, and street-level bureaucrats on the other hand, is mediated by such instruments that may have an impact in terms of interference and domination. This potential impact will be the focus of this section.
Managerialism and its ambition
Swiss welfare-to-work policies are entrapped in the dilemma of situated public action. On the one hand they require that activation measures and programs take due account of individual needs and local circumstances, i.e. that they are based on an exact assessment of the situation of the individual in her environment; this calls for leaving a large autonomy and margin of maneuver to local agents in charge of implementing such measures. On the other hand, in the line of principal-to-agent theories, there is an increasing distrust toward local agents (Bonvin and Varone 2004; Bonvin, Beuret, and Dahmen 2018) who are suspected of losing sight of official objectives in terms of activation requirements and being too empathetic vis-à-vis beneficiaries; this requires the setting up of precise targets and performance indicators, accompanied by appropriate structures of incentives or disincentives such as bonuses in cases of good performance and financial penalties for the underperformers. This also translates in the frequent use of computer systems recording data on local practices so as to ensure a rigorous control of local agents. In most Swiss welfare institutions, we observe this twofold move toward more localism (e.g. federal social insurances delegate the responsibility for implementation to cantonal agencies) and more control of local agents (e.g. implementation is increasingly monitored and piloted via managerial tools, such as computer systems, benchmarks and performance indicators), with an increased focus on placement into the labor market and collaboration with firms, and a tendency to interpret social support as a mere social buffer (Guggisberg, Egger, and Kunzi 2008). Illustrations of such a combination between managerialism and localism can be found in the recent evolutions of the Swiss unemployment insurance. The increase of local prerogatives in the field of activation was accompanied by the introduction of precise performance indicators stating the objectives to be pursued by local agents. More precisely, four indicators were selected by the federal authorities in charge of the unemployment insurance in order to assess the efficiency of the regional agencies implementing welfare-to-work policies. Three of these indicators insist on quick professional reintegration (weighting 90 The Swiss case aptly illustrates the potential consequences of an increased use of managerial tools in the field of welfare-to-work policies. In order for the designed policies to be exerted without domination as defined by the neo-republican theory, their implementation should leave no room for arbitrary decisions from local bureaucrats. To this purpose, numerous mechanisms have been designed to channel local agents’ activity. They are required to follow strict procedures and their performance is closely monitored by indicators. However, as we shall now discuss, such an endeavor leads to mitigated results when assessing welfare-to-work policies against non-domination principles.
Street-level bureaucrats’ irreducible margin of maneuver and the limitations of managerialism
Empirical research shows that, even when constrained by strict devices, the work of local agents does not boil down to the accomplishment of predefined objectives. Lipsky (1980) and others (e.g. Brodkin 2011) have abundantly documented street-level bureaucrats’ margin of maneuver and interpretation within implementation processes. Besides, local agents work in a context characterized by a chronic insufficiency of resources (i.e. the supply of public interventions cannot fully meet the demand of beneficiaries, which tends to augment at a quicker pace than the supply). If resources are insufficient, then street-level bureaucrats are well placed to select the beneficiaries on whom they will focus the existing resources, thus leading to selective or so-called “creaming” practices (Badan, Bonvin, and Moachon 2004; Rees, Whitworth, and Carter 2014). Such selectivity can be in line with managerial expectations if it focuses on those beneficiaries that allow meeting the official targets and making a good show on the indicators (Wright 2001). If by contrast, street-level bureaucrats prioritize the most vulnerable beneficiaries, then their action may go against policy-makers’ expectations in terms of activation or re-employment rates. Empirical research shows that even the most constraining managerial tools cannot fully eliminate street-level bureaucrats’ margin of maneuver in this respect. For instance, Altreiter and Leibetseder (2015) observe that even the most standardized administrations find space and margin for interpretation in order to construct their clients as more or less deserving, ultimately affecting the service provided. In the Swiss unemployment insurance (the program with the strictest use of managerial tools in the Swiss context), not all local agents complied with the expectations conveyed by managerial tools; but, due to the high costs that such non-compliance would imply for them, those who chose to go against official directives needed to “cheat” with the indicators, i.e. to record the expected behavior in the computer system while pursuing other objectives in their actual practices. Evidence however showed that such behaviors were not majoritarian among street-level bureaucrats: most of them (about 4 fifths in a sample of 30 agents investigated in an in-depth qualitative survey conducted between 2003 and 2007) preferred ritualistic behaviors, i.e. complying with the official rules and expectations even though this coincided with selective practices and low performance with the neediest recipients Brodkin (2011). The remainder (i.e. one fifth) opted for non-compliance, e.g. avoiding reporting recipients’ behaviors when it was likely to lead to a financial penalty, for instance when they did not show up to interviews or had an insufficient number of job applications (see also Remy and Lavitry 2017 for similar examples in the case of France). Such non-compliance is made possible by the weak observability of street-level bureaucrats’ practices when they are involved in face-to-face interactions with beneficiaries. The observed divide between ritualistic and resistant street-level bureaucrats seems to be more sharply cut in cases of precise and strict implementation of managerial tools. Then, street-level bureaucrats are so to say constrained to choose between loyalty (compliance) or exit (cheating with indicators is here interpreted as a kind of hidden or camouflaged exit). If implementation processes leave more room for interpretation, then there is a greater space for negotiating official targets and the way to implement them. In such cases, local agents have all three options identified by Hirschman, i.e. not only loyalty or exit, but also voice, at least to some extent (Hirschman 1970). Another important issue with regard to the limitations of managerialism relates to the compatibility of the values and objectives purported by managerial tools with those of the street-level bureaucrats’ professional environment or with their personal aspirations. Meyers, Glaser, and Donald (1998) have for instance shown that the introduction and implementation of welfare-to-work policies are faced with serious obstacles in organizations focused on administrative file-processing: the prevailing organizational culture is not easily transformed into a culture conducive to the activation of beneficiaries. In the Swiss context, this can be observed mainly in social assistance programs where the main objective has long been social support; accordingly, the ambition to turn them into activation tools meets with stronger reluctance and resistance than in the local agencies of the unemployment insurance (Bonvin, Dif-Pradalier, and Rosenstein 2013). Our empirical fieldwork also showed that managers of local agencies may resort to selective hiring practices in order to overcome such resistance, for instance by selecting new local agents from the field of human resources rather than social work (Bonvin and Moachon 2007). To sum up, empirical evidence collected in Switzerland shows that the extensive use of managerial tools may be faced with two main pitfalls (Bonvin and Rosenstein 2015). First, too precise a specification of quantitative objectives and performance indicators impedes the proper setting-up of tailor-made interventions and innovative solutions at local level. Indeed, local agents are compelled into compliance rather than encouraged into initiative. In such a context, being creative can be interpreted as a costly exit option as it does not follow official views. In the Swiss case, such tendency can be observed to the fuller extent in the unemployment insurance, whereas social assistance programs leave more space for innovative solutions. Second, managerial tools may interfere with the development of more participative ways to implement welfare-to-work policies. As a consequence, local agents tend to privilege ritualistic strategies, i.e. strict compliance with the official directives, whatever the consequences on the quality of the work performed. This may occur at the expense of program aims (Newman 2001), thereby focusing on administrative performance rather than on outcomes for beneficiaries. In such a situation, pressure is exerted on local agents, who have little or no autonomy in the choice of reintegration programs and are required to act as driving belts of the official views toward beneficiaries. However, oppositional behaviors can also be observed, whereby staff adopts strategies of resistance such as cheating with the indicators, introducing inexact information in the computerized system or postponing the reporting of the data in the system, etc. Thus, when managerialism is more strictly and precisely implemented, the space for voice and negotiation is limited. In the Swiss context, this particularly applies to the unemployment insurance, where ritualism (loyalty) and resistance (exit) are most common among local agents. By contrast, local agents in social assistance programs enjoy more margin of interpretation as well as more capabilities to express their viewpoints and make them count in the implementation processes, as well as mobilize their professional competencies (Clot 2015). This situation may, however, translate into situations of domination where the local agent abuses her discretionary power to impose her view on welfare recipients. Such undesirable situations could be remedied, e.g. via the use of external rules and procedures in order to check local agents’ use of their discretion, as Lovett (2010) advocates. This illustrates the deep-seated ambivalence of managerialism: if the objective is to carefully oversee the work of a local agent who is to be fundamentally distrusted and suspected, it leads to disqualifying her viewpoints and competencies; by contrast, if the aim is to mobilize the local agent’s knowledge and expertise and fully involve her in the implementation processes, then her voice is given fuller consideration but this may in turn translate into a misuse of her discretionary power by the street-level bureaucrat herself. Thus, we can observe that managerialism may ultimately lead to undesirable situations when assessed through the lens of the neo-republican theory. The street-level bureaucrat is facing a kind of dead end: either resisting the management devices and providing tailored assistance to the recipients, but as a consequence engaging in a form of arbitrary interference; or following the rules by the letter and be consistent with the non-arbitrariness principle. As we will see in the next section, this second option is often inefficient because of the discrepancy between the recipients’ needs and the implementation scheme provided by public management.
Welfare-to-Work Policies Faced with Plural Recipients
The last level of the policy process investigated in this chapter relates to its delivery to and reception by beneficiaries. This step is crucial, and local agents’ skills are often required to reconcile demands for efficiency and recipients’ needs.
“Ideal” vs. real individuals
Welfare-to-work policies focus on individualized and tailor-made interventions. As a consequence, there may be conflicts or dilemmas between institutional expectations about individual behaviors and the actual situations of beneficiaries. The purpose of policy-makers is to push recipients to behave in the expected way so that they reach as quickly as possible the intended outcome. In more concrete terms, this implies setting up an adequate structure of incentives and disincentives (sticks and carrots) so that people are induced to find a job on the labor market and regain financial autonomy. This entails making work pay, which in a context of increasing precariousness of work, requires revising benefit entitlements (i.e. reducing them and tightening the eligibility conditions) to guarantee that work, even though it pays less in absolute terms, still pays more than social benefits in relative terms. To this end, welfare-to-work policies have integrated the idea of contractualizing access to benefits, i.e. making it conditional upon the acceptance of certain behavioral requirements. The introduction of contractual welfare offers a ground for the imposition of sanctions in case of non-compliance (Sol and Westerveld 2007): in other words, complying individuals will get the full amount and duration of benefits, while non-compliant recipients will be inflicted sanctions and financial penalties that may be prohibitive for some of them. In such a perspective, individuals are conceived as rational beings (homo oeconomicus, one could say) that systematically choose the solution that pays most, i.e. the one with the highest (or rather the least low) financial returns. This anthropological vision of the welfare recipient lies in sharp contradiction with that of the capability approach and the French economy of conventions, insofar that both conceptions emphasize the plurality of motives that may guide and inspire human action. They insist on a plural notion of the individual, which suggests that the normative prescriptions purported by policy documents, administrative directives and managerial tools need to be confronted to real individuals endorsing a plurality of aspirations. This confrontation between institutional norms and individual realities is what Boltanski and Thévenot (2006) call a “reality test” and includes two facets in the specific case of welfare-to-work policies: on the one hand, an “ideal” individual mainly moved by self-interest is to be confronted with a real human being, pursuing a plurality of goals beyond the maximization of one’s self interest; on the other hand, the ideal of an individual supposedly able to reach autonomy and non-dependence on the welfare state (i.e. a view of the self-made man that may rely on himself for making an autonomous living) is to be confronted with real vulnerable human beings in need of long-term support and with limited prospects for full independence. Thus, two visions of the human being are at stake: one that insists on the homo oeconomicus’ concern for self-interest and capacity for autonomy, the other emphasizing a plural and vulnerable individual. The former can be read as a norm and insists on individuals as they should be, while the second pays more attention to individuals as they are. It is precisely at the crossroad between these two visions of the individual (as a norm and as a fact) that tensions may arise and require the intervention of innovative street-level bureaucrats in order to overcome them. In most Swiss welfare-to-work policies, new policy instruments have been introduced with the objective of designing more tailor-made interventions. These include assessment tools that aim at having an exact idea of individual situations in order to be able to design the most adequate program for each recipient. However, such instruments are very ambivalent as the case of the unemployment insurance illustrates. The initial assessment phase results in assigning all job-seekers to one of five categories, from the most to the least employable; these are then allotted a measure taking due account of this classification. Empirical research shows how such tools may be turned into selective mechanisms: indeed, due to budgetary limitations and pressures to enhance performance, these tools are used to select those unemployed more likely to find a job and focus available resources and time in their favor. Besides, in such a context, recipients have very little say in the choice of measures and programs, which is mostly decided by local agents; all the same, they are subjected to heavy financial penalties (up to three months of benefit suspension) in case they refuse to participate. Under these circumstances, their capability for voice is strictly limited: they are allowed to express their wishes only if these comply with the institutional expectations (Bonvin et al. 2011). As many local agents explained, they interpret their task as one of persuasion where the objective is to convince beneficiaries to adhere to the solutions designed by street-level bureaucrats. This amounts to the lowest levels of participation in Shelley Arnstein’s ladder, those of manipulation or therapy. Recipients are not partners in the elaboration of measures, but people whose ill or wrong will or opinion must be corrected or cured. In such cases, exit options (or opportunities to refuse the programs proposed) are costly and voice options are of very limited impact; by contrast, compliance and the duty to collaborate are expected to be strictly enforced. Thus, the individual as a norm tends to be imposed on real individuals and those who are unable or unwilling to abide by these standards are either to be persuaded in this direction, or to be sanctioned if they persist in their non-compliant behavior. This is, at least, how things are designed and intended by policy-makers and the representatives of the unemployment insurance. In such cases, welfare-to-work policies are conceived not in accordance, but in contradiction with the principle of non-domination. As a matter of fact, they are not designed with a view to giving voice to the powerless and vulnerable persons, but to educating them into compliance with the official view of the ideal individual. This often translates into the exercise of arbitrary power vis-à-vis beneficiaries, who are called to follow activation programs or accept jobs even though they consider them as meaningless or unsuitable.
Discrepancies between norms, practices and behaviors
Empirical research, however, does not confirm such a picture where policies and directives designed would be mechanically implemented and accepted by compliant street-level bureaucrats and beneficiaries. Discrepancies between ideal norms and actual practices can be observed along two main lines. First, when confronted with actual beneficiaries that do not match with the normative individual advocated in legislative documents and administrative directives, street-level bureaucrats are faced with two main dilemmas. On the one hand, how to push all people, independently of their degree of vulnerability, into compliance with the ideal of autonomy and non-dependence that is the objective of most welfare-to-work policies? On the other hand, how to convince recipients with a plurality of aspirations and a varying degree of endorsement of the logic of welfare-to-work policies to adopt a single format of behavior, which is characterized by compliance with duties and fulfillment of official expectations? Both dilemmas relate to human diversity in terms of competencies and desires. As a matter of fact, beneficiaries of welfare-to-work policies show up at the counter with diverse expectations and variegated situations in terms of capacities and vulnerabilities. Some have recognized qualifications and hope to find a job as quickly as possible, with varied aspirations regarding wage, status, etc. Others lack competencies or skills, and hope to be proposed activation programs allowing a professional retraining or a reinforcement of their competencies in order to go back to the labor market better equipped. Still others, more vulnerable, aim at finding appropriate resources for a better social inclusion. Besides, some are willing to collaborate and endorse official expectations; others are more reluctant and strive mainly to get welfare benefits. Faced with such plural recipients, street-level bureaucrats have difficulties to fully implement the recommendations of policy-makers or their administrative hierarchy. This is especially the case when these recommendations rely on a precise normative conception of how the ideal beneficiary should behave. Then, taking due account of the actual beneficiary’s situation and wishes may require taking distance with the existing rules and directives and, sometimes, even taking risks as actual practices may then be at odds with norms. This shows the inefficacy of building comprehensive incentive devices to channel the implementation of welfare-to-work policies. Domination in the form of arbitrary interference simply cannot be avoided by trying to anticipate the needs of a pre-defined recipient. When strict rules are in place, the recipients that do not fit the standard profile –often the most vulnerable recipients– are denied the appropriate conversion factors available to others. This leads to a degradation of real freedoms, as Amartya Sen would define it. Not only street-level bureaucrats, but also beneficiaries may impede the mechanical implementation of policies and directives. They are indeed aware that they need to present to local agents a portrait of themselves that may qualify them for the desired benefit or service. As Goffman has abundantly documented (e.g. Goffman 1959, 1961), the presentation of self in everyday life is an essential component in convincing others that we comply with their expectations and deserve their attention. Such behaviors are of course also present in welfare-to-work policies where not only street-level bureaucrats are willing to convince beneficiaries to endorse their views about the relevance of welfare-to-work policies, but also beneficiaries try to persuade local agents that they do comply with their expectations and with the official rules. Our empirical research identified numerous such processes where actual beneficiaries strove to play the role of ideal beneficiaries in order to get the benefits and services they coveted (Bonvin, Dif-Pradalier, and Rosenstein 2013). In a capability perspective, this can be interpreted as a legitimate way to avoid interference by street-level bureaucrats: in other words, when local agents try to impose too simple institutional views (or, sometimes, their own views, thus exerting undue domination) on beneficiaries, subject to sanctions in case of non-acceptance or non-compliance, this constitutes a clear infringement to the development of their real freedoms; beneficiaries are then legitimate in trying to escape the threat of sanctions by adjusting their behaviors according to local agents’ expectations. This clearly shows the difference between Sen’s view on domination and those of some neo-republican theorists: while the capability perspective assesses as illegitimate all domination practices that go against the development of capabilities (whether they are arbitrary or not), theorists like Lovett qualify domination as the presence of uncontrolled or arbitrary interference.
Conclusion
Our chapter has demonstrated the relevance of considering jointly the three levels of the policy process – namely design, implementation and “reception” – when investigating phenomena of domination within welfare-to-work policies. Our theoretical framework, based on a combination of the capability approach and the French economy of conventions, aptly shows that power and domination may be exerted at all three levels. First, at design level, the trend toward welfare-to-work policies may result in the reduction of resources distributed and the increase of conditionalities imposed on recipients, with the consequence that both access to benefits and services and availability of exit options are made more difficult. Such transformations of welfare institutions would entail in turn an increased exposure to domination power. Second, at implementation level, managerial tools may be used with a view to reducing or even cancelling the margin for maneuver and interpretation of local agents and forcing them into compliance with administrative directives and expectations. Going beyond the necessity to check street-level bureaucrats’ use of their discretionary power via appropriate rules and procedures (Lovett 2010), managerialism risks disqualifying their professional expertise and knowledge of local and individual circumstance. Thereby, it would reduce significantly their possibility to design innovative solutions, which could be better adjusted to the beneficiaries’ needs. Finally, at reception level, an overspecified conception of the ideal recipient, in terms of behavior, aspirations and capacities, may result in significant discrepancies with the actual situations of the most vulnerable beneficiaries, thus leading in most cases to inappropriate and inefficient solutions. At the beginning of our chapter, we identified two questions that would help us evaluate Swiss welfare-to-work programs along the lines inspired by the capability approach and the French economy of conventions: “Do welfare-to-work programs enhance their beneficiaries’ real freedoms?”, and “Is the capability for voice of all stakeholders promoted?”. While the actual state of welfare-to-work policies in Switzerland shows that there is a great margin for improving the real freedom of their recipients, we believe that our approach best highlighted the need for promoting the stakeholders’ capability for voice. By combining the capability and conventionalist approaches, we were able to identify the numerous points of discrepancy between the policies as they are designed and their actual implementation. Coordination between the actors at each layer of implementation is made difficult by the misalignment of orders of worth that can arise between the three levels of public policy. This often prevents the development of welfare-to-work policies that enhance the capabilities of their recipients. Our findings emphasize the relevance of a situated perspective on public action and on the issue of (non-)domination, which requires empirically investigating how prescriptions elaborated at design level interact with individual practices and behaviors of street-level-bureaucrats and recipients. On the normative side, our theoretical framework and empirical investigation of the case of Switzerland sheds light on the importance to give due attention and voice to street-level bureaucrats and to the plural recipients of welfare-to-work policies. For street-level bureaucrats, this implies recognizing that they need some discretionary power in order to adequately perform their tasks, but also that this discretion needs to be appropriately supervised and checked. In our view, managerial tools are not sufficient to this purpose, adequate rules and procedures are required that ensure that such discretion is used to enhance the recipients’ capabilities. In the perspective suggested in this chapter, then, domination is not only a matter of arbitrary or uncontrolled interference; it also takes place when the development of capabilities is impeded. We therefore call for a definition of domination that combines neo-republican and capability insights. For beneficiaries, this entails, on the one hand, unconditional access to an adequate level of resources, respectful of human dignity, be it in the form of social assistance, social insurance or even a basic income (i.e. the threat of reducing benefits under this threshold should not be used as a tool to enforce compliance with institutional norms); on the other hand, the possibility to voice one’s aspirations and have them being seriously taken into account when welfare-to-work policies are designed and implemented. Taking these two conditions into consideration would provide the beneficiaries with real exit options together with the possibility to make their voices heard in a decision-making process where the risk of domination would be effectively minimized.
References
Citation
@incollection{bonvin2020,
author = {Bonvin, Jean-Michel and Perrig, Luca},
editor = {Eleveld, Anja and Kampen, Thomas and Arts, Josien},
publisher = {Policy Press},
title = {Implementing {Social} {Justice} Within {Activation}
{Policies}},
booktitle = {Welfare to Work in Contemporary European Welfare States:
Legal, Sociological and Philosophical Perspectives on Justice and
Domination},
date = {2020-01-29},
address = {Bristol, UK},
url = {https://www.lucaperrig.ch/Research/BonvinPerrig2020.html},
isbn = {978 1 44734 001 0},
langid = {en}
}