Introduction
Over the past decades, Western European welfare states have introduced social investment reforms. Social investment aims at addressing structural changes associated with globalization, technological changes, and shifts in family relationships (Morel et al., Reference Morel, Palier and Palme2012; Hemerijck, Reference Hemerijck2013; Garritzman et al., Reference Garritzman, Häusermann and Palier2022). Although it takes different forms across countries, this approach generally features policies that aim to enhance individuals’ capacities to fare well in post-industrial societies, provide them with more opportunities to tackle related (new) social risks, and prevent the clustering of disadvantages. As such, social investment introduces a radical shift in the normative underpinnings of European welfare states with an emphasis on capability-building based on ex ante interventions, in contrast to ex-post compensations (Andersen and Larsen, Reference Andersen and Larsen2024; Bagadirov et al., Reference Bagadirov, Hemerijck and Puertas Roig2025). Crucially, however, the social investment welfare state coexists with the ex-post compensatory and workfarist welfare states (Wagner et al., Reference Wagner, Pöyliö and Hemerijck2025). This is particularly the case of active labour market policies (ALMPs) (Hemerijck and Bokhorst, Reference Hemerijck and Bokhorst2025). Universal, so-called “passive” policies have been partly replaced by activation policies that make entitlements conditional on behaviours deemed to facilitate a “return to employment” – this is their workfarist component – and that provide lifelong learning opportunities – this is their social investment or capacitating component (Schulze and Jørgensen, Reference Schulze, Jørgensen, Van Gerven, Rothmayr Allison and Schubert2023; Bagadirov et al., Reference Bagadirov, Hemerijck and Puertas Roig2025). This observation that European welfare states in general, and ALMPs in particular, display hybrid normative underpinnings raises the question of the extent to which citizens endorse social investment and how it has reshaped their understanding of social solidarity. In this article, shifting the focus away from the politics and policies of social investment (Hemerijck, Reference Hemerijck2013; Garritzman et al., Reference Garritzman, Häusermann and Palier2022), we question the normative outcomes of social investment in the context of ALMPs from a citizen perspective. We ask: how do hybrid ALMPs – combining capacitation with conditionality – contribute to shaping citizen-level norms of social solidarity?
Thus far, the literature has studied citizens’ preferences for social investment in comparison to compensatory policies, and has explored variation in preferences associated with employment risks linked to technological change and the generosity of existing welfare states, among other factors (e.g. Busemeyer and Tober, Reference Busemeyer and Tober2023; Eick, Reference Eick and Greve2023). In this article, we take a further step by studying the normative feedback of social investment. Grounded in feedback studies on mass publics (Bussi et al., Reference Bussi, Dupuy and Van Ingelgom2022), the concept of normative feedback rests on Schneider and Ingram‘s argument that policies are not only instruments of resource allocation, but that they also convey normative messages about what is fair, legitimate, and politically acceptable (1993). Normative feedback, thereby, comes from individuals’ policy experiences and representations that co-construct their normative expectations (Svallfors, Reference Svallfors2010). In this article, we are specifically interested in one normative feedback of social investment: citizen-level norms of social solidarity. How do they reflect the normative logics of social investment, coined as stepping-stone solidarity (Bagadirov et al., Reference Bagadirov, Hemerijck and Puertas Roig2025; Wagner et al., Reference Wagner, Pöyliö and Hemerijck2025)? Citizen-level norms describe shared understandings, expectations, and points of reference among group members (Weber, Reference Weber1971). A norm emerges when a given understanding is not contentious and is held across relevant divides in a society (Vila-Henninger, Reference Vila-Henninger and Vila-Henninger2020).
To study citizen-level norms of social solidarity in the context of hybrid ALMPs, we conduct a comparative qualitative secondary analysis of two large-scale focus-group datasets collected in Francophone Belgium and France in 2006 and 2019. The cross-country and over-time comparison is intended to strengthen the robustness of our interpretive claims (Gerring, Reference Gerring2012) by assessing whether key patterns in citizens’ discussions recur across two highly comparable hybrid-activation settings and across two points in time that feature an incremental layering of enabling instruments alongside persistent conditionality and sanctioning.
We focus on participants’ unprompted discussions of ALMPs, which is instrumental in describing how they understand social solidarity in their own terms, while not being prompted either to discuss active labour market policy or their understanding of social solidarity. As such, our data covers active labour market policies as they are perceived by participants – thereby shifting attention away from policy design per se. While participant perceptions may not be fully accurate or attuned to the complexities of each policy instrument, they nonetheless are true to what participants think of ALMPs – hence making them crucial to the study of citizen-level norms. Last, our operationalization of citizen-level norms is two-fold: first, we searched for diagnostic linguistic cues associated with each of the policy norms to map which ones are present in participant discussions of ALMPs; second, we inductively analysed frames participants rely on to discuss social solidarity to study how they combine different policy norms and what citizen-level norms may emerge.
The article proceeds as follows. First, we discuss the notion of welfare state norms and citizen-level normative feedback. Second, we present our comparative qualitative secondary design and explain our two-step operationalization of citizen-level norms. Third, our empirical section presents our analysis of citizen-level norms in the context of discussions of ALMPs. Last, we discuss our results and conclude.
Theory: Hybrid ALMPs, norms of social solidarity, and citizen-level normative feedback
Welfare states and their respective norms of social solidarity
Since the 1990s, based on the comparative study of welfare state reforms, comparative social policy research has converged on a threefold distinction between (i) a passive or compensatory welfare state, which provides ex-post income protection; (ii) a workfarist welfare state, which relies on behavioural conditionality and sanctions to enforce labour-market participation; and (iii) a social investment welfare state, which emphasises ex-ante capacitation through education, training and work-life policies, to secure high-quality employment and resilient life-course transitions (Esping-Andersen, Reference Esping-Andersen1990; Morel et al., Reference Morel, Palier and Palme2012; Hemerijck, Reference Hemerijck2013). Whereas compensatory measures ensure income after shocks, and workfarist measures are meant to accelerate re-entry into employment, social investment develops capabilities through early-childhood education and care, schooling and vocational training, guidance and counselling, active labour-market programmes with a developmental orientation, work–life reconciliation policies, and measures that improve job quality and employment security. Social investment is not always treated as a coherent or uniformly “capacitating” paradigm, but may be articulated alongside stronger expectations of individual responsibility and disciplinary activation, and may generate distributive trade-offs that require explicit compensatory complementsFootnote 1 (Cantillon and Lancker, Reference Cantillon and Lancker2013; Garritzmann et al., Reference Garritzmann, Häusermann and Pinggera2025).
Social investment policies not only rely on different policy instruments, but they are also underpinned by a normative conceptualization of social solidarity and social rights that is distinct from that of compensatory and workfarist policies. Historically, the compensatory welfare state has been grounded in a Rawlsian understanding of social justice as redistribution, centred on the notion of fair equality of opportunity (Rawls, Reference Rawls1972). Social solidarity warrants fair access to opportunities and compensates for the adverse conditions that individuals may face over the course of their lives. The workfarist welfare state rests on another conceptualization of justice, one that emphasises individual choice and responsibility. Social solidarity should not compensate individuals for circumstances that result from their own choice-making. Individuals should be compensated only when they are not responsible for these circumstances. As a consequence, welfare dependency is often portrayed as a moral failing (Turgeon et al., Reference Turgeon, Taylor and Niehaus2014), and there are frequent assumptions that unemployed individuals are idle or lack the willingness to work (Krinsky, Reference Krinsky2008). These assumptions are reflected in the representations of target groups in the media and public discourse (Dencker-Larsen and Lundberg, Reference Dencker-Larsen and Lundberg2016; Boulus-Rødje, Reference Boulus-Rødje2019) and in the very principle of workfarist policies that make benefit receipt subject to work-related obligations.
In contrast, a prominent capacitation-centred articulation of the social investment welfare state is what Bagadirov, Hemerijck, and Puertas Roig (Bagadirov et al., Reference Bagadirov, Hemerijck and Puertas Roig2025) term “stepping-stone solidarity”: a third type of social solidarity complementing redistribution (“Robin Hood”) and insurance (“piggy bank”). It aims to de-cluster disadvantages across the life course by capacitating individuals to navigate transitions from early childhood to late career, thereby securing both objective and subjective well-being, such as security, capability, and recognition. In this regard, effective safety nets remain an essential complement to social investment. Yet, its distinctive and unique contribution lies in capability-enhancing interventions and their socio-political payoffs (see also Morel et al., Reference Morel, Palier and Palme2012; Hemerijck and Bokhorst, Reference Hemerijck and Bokhorst2025).
Normative feedback of social investment
Because of its radically distinct normative foundations, the social investment welfare state raises the question of whether citizens endorse the notion of stepping-stone solidarity and alter their normative understandings of social solidarity. From a policy perspective, social investment has introduced a clear shift, but does it translate to a shift in citizen-level norms of social solidarity? Research has shown that, over time, recipients may adjust their beliefs and attitudes in response to policy-based expectations, occasionally adopting the underlying normative assumptions of the policy (Camus and Berjot, Reference Camus and Berjot2015; Demazière and Zune, Reference Demazière and Zune2016). However, the extent to which ALMPs contribute to entrenching new norms of social solidarity in broader public perceptions, beyond policy recipients, remains unclear.
Related work on social investment has studied its macro-level performance and the formation of support coalitions among the public. Conflicting preferences between beneficiaries of monetary transfers (e.g. the unemployed) and beneficiaries of services (e.g. parents of young children), in particular, have been documented, as have the budgetary trade-offs that render the expansion of social investment politically contentious (Morel et al., Reference Morel, Palier and Palme2012; Hemerijck, Reference Hemerijck2013; Busemeyer and Neimanns, Reference Busemeyer and Neimanns2017). In this article, we bridge both strands of scholarship and focus on the normative feedback of social investment: we examine how citizens understand social solidarity, whether they have a direct or indirect (through family or friends) experience of ALMPs, or are exposed to public discussions about these policies.
Feedback studies on mass publics have conceptualized mechanisms to account for how policy experiences and representations impact policy attitudes and behaviours (Bussi et al., Reference Bussi, Dupuy and Van Ingelgom2022). The resource-based mechanism centres on the resources and incentives a policy allocates, while the interpretive mechanism focuses on the messages policies convey to their recipients. More recently, several works have suggested a third feedback mechanism: the normative feedback, which echoes Ingram and Schneider’s earlier analysis (1993) that public policies convey normative understandings of their target populations, constructing some of them as deserving and others as undeserving. A normative feedback “is present where public policies provide citizens with a sense, not only of what their material interests are and who is responsible for different political decisions but also of the desirable state of affairs Footnote 2” (Svallfors, Reference Svallfors2010, p. 120). Based on the adaptation of preferences to existing policies, the concept of normative feedback thereby suggests that policy exposure may result in specific citizen-level norms, like norms of redistribution (Svallfors, Reference Svallfors2010) or norms pertaining to the desirable territorial structuring of the state (Verhaegen et al., Reference Verhaegen, Versailles, Van Ingelgom and Laloux2022; Kopec, Reference Kopec2025). Following a Weberian definition, norms are understood as collective reference points that structure what citizens deem acceptable or legitimate (Weber, Reference Weber1971). Crucially, recent research has reported that policies may exert feedback effects beyond direct beneficiaries, including on those citizens who are exposed to public discussions about them in the media or everyday political discussions (Bakkær Simonsen, Reference Bakkær Simonsen2021; Rosenthal, Reference Rosenthal2021).
The hybridity of ALMPs and our expectations
ALMPs provide a strong case to study citizen-level normative feedback because they include the three main welfare logics within a single domain and are associated both with frequent and salient encounters (requirements, appointments, letters) and also with visible and recurrent public debates which centre on policy norms. Existing scholarship suggests that normative feedback effects are not straightforward, but hinge on the consistency of policy norms. A crucial feature of active labour market policies is their hybridity (Wagner et al., Reference Wagner, Pöyliö and Hemerijck2025). They are layered and include both capacitating and workfarist components that are associated to diverging norms of social solidarity (for variation across welfare regimes, see Taylor-Gooby et al., Reference Taylor-Gooby, Hvinden, Mau, Leruth, Schoyen and Gyory2019; Taylor-Gooby and Leruth, Reference Taylor-Gooby and Leruth2018). Because of ALMPs’ hybridity, it is crucial to study how citizens navigate diverging policy norms and how citizen-level norms of social solidarity are shaped. To do so, we adopt a two-fold approach where, first, we study whether the multiple policy norms layered in ALMPs are present in citizen-level norms of social solidarity; and second, how citizens interpret them and what the associated citizen-level norms of social solidarity are. The expectations below reflect this approach.
First, we expect that compensatory, workfarist, and capacitating components will coexist in citizen-level norms of social solidarity. Because ALMPs combine compensatory, workfarist, and capacitating instruments, they expose citizens to a plurality of policy norms through both direct encounters (e.g. requirements, appointments, letters) and indirect exposure via public debate. In line with the conceptualization of normative feedback that stresses the role of policy messages in shaping citizens’ conceptions of what is fair and legitimate (Schneider and Ingram, Reference Schneider and Ingram1993; Svallfors, Reference Svallfors2010), we therefore expect citizen discussions about ALMPs to include mentions consistent with these different welfare approaches. At the same time, we anticipate that the mentions to each of them will be uneven. Specifically, we expect workfarist cues to be particularly prominent given the salience of conditionality and sanctioning both at the point of delivery and in public debates (Brodkin and Larsen, Reference Brodkin and Larsen2013; Barrett et al., Reference Barrett, Gray, Farrall, Barrett, Gray and Farrall2023; Pennetreau and Laloux, Reference Pennetreau and Laloux2025); while compensatory mentions will persist as a protective baseline and capacitating mentions may appear, but are less likely to be spontaneously elaborated on in lay discussions (Bürgisser et al., Reference Bürgisser, Häusermann, Kurer and de Pinho Tavares2025; Hemerijck and Bokhorst, Reference Hemerijck and Bokhorst2025). The underlying rationale is that protective entitlements (income maintenance, minimum protection) are more visible. By contrast, capacitating instruments are often more targeted, technocratic, and mediated through specific delivery encounters, making them less visible and less frequently used in citizens’ discussions (Mettler, Reference Mettler2011).
Second, we expect that the hybridity of ALMPs and the conflicting policy norms they convey will generate a normative feedback. More specifically, we expect this feedback to be characterized by normative tensions where citizens combine solidarity as reciprocity and individual responsibility with a stepping-stone (capacitating) understanding of social solidarity. As ALMPs convey partly divergent policy norms, citizen-level norms of social solidarity are unlikely to mirror a single, internally coherent policy norm. Our expectation of normative tensions in citizen-level norms of social solidarity follows from the broader insight that activation reforms frequently layer enabling and workfarist elements, thereby exposing citizens to competing policy norms (Taylor-Gooby and Leruth, Reference Taylor-Gooby and Leruth2018; Taylor-Gooby et al., Reference Taylor-Gooby, Hvinden, Mau, Leruth, Schoyen and Gyory2019; Wagner et al., Reference Wagner, Pöyliö and Hemerijck2025).
Methods
Comparative design: ALMPs in France and Belgium in 2006 and 2019
We conduct a comparative qualitative secondary analysis of citizen discussions of active labour-market policy in France and Francophone Belgium at two points in time (2006 and 2019). This design has three main justifications.
First, ALMPs are typical cases of hybrid policies that layer compensatory, workfarist, and social investment instruments, making them well-suited for studying how citizens navigate conflicting policy norms. Attitudes toward activation are “two-faced,” with citizens distinguishing between “demanding” and “enabling” dimensions – a duality that reflects the hybridity of ALMPs (Meuleman et al., Reference Meuleman, Van Hootegem, Rossetti and Abts2025). Second, policy changes between our two data points feature the gradual introduction and implementation of additional social investment instruments, which allow for the analysis of the normative feedback of social investment over time. Between 2006 and 2019, Belgium expanded counselling and training, while intensifying sanctions; whereas France reorganized delivery (Pôle emploi), which has reinforced control and monitoring on both unemployed and job-centre workers, and introduced subsidised contracts and mentoring (PEC) (Bürgisser et al., Reference Bürgisser, Häusermann, Kurer and de Pinho Tavares2025; Meuleman et al., Reference Meuleman, Van Hootegem, Rossetti and Abts2025). These policy changes impact on what citizens are likely to experience and recall through both “direct policy experiences and indirect exposure via public debate” (Rosenthal, Reference Rosenthal2021, p. 1100). Third, while ALMPs in France and Belgium are typical of hybrid policies that include social investment, the comparative design introduces variation in policy design, which strengthens the generalizability of our findings beyond context-specific observations (Gerring, Reference Gerring2012). Specifically, in Belgium, job counselling and training have been strengthened (Plan d’accompagnement des chômeurs, Plan Activa) alongside monitoring. In France, reforms have focused on public employment programmes and individualized support (Pôle emploi, PEC).
Table 1 summarizes the main citizen-facing components of ALMPs for each country at each point in time. The aim is not to provide an exhaustive presentation of policy changes, but to outline the policy instruments and related policy norms that citizens have been exposed to in everyday experiences and public debates.
Table 1. Outline of the main policy instruments of ALMPs in France and Belgium
Source: Clegg, (Reference Clegg, Clasen and Clegg2011), Laenen (Reference Laenen2020), Avlijaš et al. (Reference Avlijaš, Hassel, Palier, Anke and Bruno2021), Schulze and Jørgensen (Reference Schulze, Jørgensen, Van Gerven, Rothmayr Allison and Schubert2023), Bürgisser et al. (Reference Bürgisser, Häusermann, Kurer and de Pinho Tavares2025), Hemerijck and Bokhorst (Reference Hemerijck and Bokhorst2025), Meuleman et al. (Reference Meuleman, Van Hootegem, Rossetti and Abts2025), Pennetreau (Reference Pennetreau2025, Reference Pennetreau2024).
Data: Secondary analysis of focus groups
We conducted a comparative qualitative secondary analysis. Secondary analysis refers to “a research strategy which makes use of pre-existing quantitative or pre-existing qualitative research data to investigate new questions or verify previous studies ”(Heaton, Reference Heaton2004, p. 16). In qualitative research, this often involves using existing data – such as interviews (individual or focus groups), fieldnotes, and observational records (Irwin, Reference Irwin2013).
Our corpus consists of two main datasets shared with us by the primary researchers (see Table 2 for an overview and Appendices 1 and 2 for more details). It comprises 30 focus groups, including a total of 159 participants. Our corpus is composed of datasets that were collected at two different points in time in France and Belgium, thereby offering variation in time and policy contexts (see above). Each dataset also offers socio-economic and political variation among participants, as primary researchers sampled them based on both criteria (in addition to gender and age).
Table 2. Datasets included in the corpus
The shared focus on a comparable and consistent research question – how citizens perceive and understand European integration – and the use of similar methods in both datasets – cross-national qualitative research based on focus group data – provide a solid basis for cross-study comparability (Hughes et al., Reference Hughes, Frank, Herold and Houborg2023). The remaining heterogeneity in data-collection methods and original research questions, rather than being a limitation, enhances the robustness of our findings.
Interviews were non-directive, allowing participants to speak freely on any topic. Crucially, policy discussions were not prompted by primary researchers. None of the discussions of ALMPs were prompted. The presence of active labour-market policy discussions thus reflects participants’ own ways of engaging with focus group discussions rather than any influence of research design. However, it may be that the EU focus of the primary studies may have heightened the salience of cross-border competition in participants’ accounts. Overall, these features of data collection ensure that participants discuss their understandings of social solidarity in their own terms, uninfluenced by focus group scenarios or follow-up questions. Thus, we avoid the pitfall of designs where participants are prompted to discuss specific topics and, thereby, discuss them even when they are neither salient nor relevant to them (see Bourdieu on topic imposition, 1984). Also, the unprompted nature of discussions of ALMPs is a consistent feature across our dataset.
Operationalization of citizen-level norms
In the initial phase, we applied a deductively developed team codebook to the primary datasets (Vila-Henninger et al., Reference Vila-Henninger, Dupuy, Van Ingelgom, Caprioli, Teuber, Pennetreau, Bussi and Le Gall2024). Drawing upon the categorization of policy domains outlined by the Comparative Agendas Project, we coded the data accordingly (see Dupuy et al., Reference Dupuy, Teuber and Van Ingelgom2022). For this article, we selected the policy codes pertaining to employment policy. A broad definition of employment was adopted, reflecting our focus on lay citizens’ discussions rather than expert accounts, particularly in research contexts where participants were not prompted to discuss specific policy areas (for further details regarding the coding of policy discussions, see Author 1, 2024 – see also Appendix 3).
Our operationalization of citizen-level norms of social solidarity is two-fold. First, we operationalized each policy norm based on diagnostic linguistic cues and searched for them in participant discussions of ALMPs. By diagnostic linguistic cues, we mean recurring lexical markers and expressions that, in lay talk, reliably indicate a topic and allow segments to be assigned to analytic categories (Schreier, Reference Schreier2012). Each cue is reflective of common phrasings that encapsulate defining features of each policy norm. We used allocations/indemnization for the compensatory approach; contrôle/sanction/offre raisonnable/il faut/on doit for the workfarist approach; and qualité de l’emploi/reconversion/formation for the social investment approach. In addition to training, reconversion, job-quality cues, we also coded references to recognition through interpersonal interaction, agency/autonomy through active choice, optimism regarding future opportunities, and cumulative life-course security in the context of experiences or perceptions of ALMPs under social investment cues (Wagner et al., Reference Wagner, Pöyliö and Hemerijck2025). When no cues were identifiable, we coded as “other” to avoid over-inference. This first operationalization captures whether, and with what relative salience, participants’ unprompted discussions refer to diagnostic cues consistent with the three policy norms embedded in hybrid ALMPs. However, cue identification alone does not allow to uncover how citizens make sense and interpret these different policy-level norms, how they combine them and the type of normative feedback that may result from ALMPs.
As a consequence, we use a second operationalization of citizen-level norms of social solidarity that consists of the frames participants rely on in their discussions of social solidarity in the context of ALMPs (Van Gorp, Reference Van Gorp, D’Angelo and Kuypers2010). A frame includes a causal narrative that identifies the causes and consequences of a given situation or policy problem, and characters – such as targets and beneficiaries (Schneider and Ingram, Reference Schneider and Ingram1993; Stone, Reference Stone2012). In line with this, we coded for participants’ references to both the causes and consequences of ALMPs problems, and their identification and description of target populations and beneficiaries. Our codebook thus includes each of these categories, which is instrumental to systematic coding and, relatedly, to minimizing the interpretive bias in frame identification – one of the central challenges in qualitative frame analysis (Matthes and Kohring, Reference Matthes and Kohring2008). Frames are identified inductively based on the repetition of patterns associating specific causes, consequences and target populations. They are instrumental in studying in detail how citizens understand social solidarity in the context of ALMPs through their focus on the problems and issues ALMPs are to address, the identification of the policy target population as well as the solutions the policy should implement. Crucially, this second operationalization of social solidarity accounts for how participants articulate conflicting norms of social solidarity and the tensions that may arise.
This distinction between diagnostic cues and frames allows us to stay close to what participants themselves foreground as they discuss social solidarity in the context of ALMPs (frames) while also identifying when and how distinct policy norms are mentioned (cues). It is important because ALMPs, being hybrid policies, do combine different, if not conflicting policy norms (Taylor-Gooby and Leruth, Reference Taylor-Gooby and Leruth2018; Taylor-Gooby et al., Reference Taylor-Gooby, Hvinden, Mau, Leruth, Schoyen and Gyory2019; Meuleman et al., Reference Meuleman, Van Hootegem, Rossetti and Abts2025). Thus, citizen-level norms of social solidarity must be operationalized in a way that is open to the combination of divergent policy norms. In fact, in our data, participants sometimes discuss the importance of social rights, while simultaneously conditioning them on behavioural compliance; they also sometimes describe insecurity as a general feature of the labour market condition, or as the direct result of ALMPs; or they express solidaristic claims in protective (compensatory) or in disciplinary (workfarist) terms. We therefore consider that these combinations are substantive evidence of how participants understand social solidarity in ALMPs and the normative tensions that may arise.
Viewing ALMPs like a citizen
In this empirical section, we study the relative salience of compensatory, workfarist, and social investment policy norms in participant discussions of ALMPs based on the analysis of diagnostic linguistic cues. We then report on the frames participants rely on to discuss social solidarity in the context of ALMPs and show how these frames combine different policy norms. Finally, building on these analyses, we study the normative feedback of social investment under hybrid activation.
Tracing policy norms with linguistic cues: the sparse presence of capacitation
Across cases and points in time, participants predominantly refer to the compensatory and workfarist policy norms, while social investment cues are comparatively rarer. As Table 3 reports, most discussions of ALMPs cluster around benefit protection and entitlements (the compensatory approach), and conditionality, monitoring and sanctions (the workfarist approach), while social-investment’s defining features (training, guidance, work–life reconciliation, job quality) are present, but are comparatively less frequent.
Table 3. Policy norms in participant discussions
Compensatory cues appear most clearly when participants discuss basic unemployment-related entitlements (e.g. income maintenance, access to support) as matters of minimum protection. For instance, Michèle, a 26-year-old Belgian employee, contrasts unemployment insurance and minimum-income protection with contexts where such guarantees are absent: “In Belgium… if you lose your job, you get unemployment benefit. And if you have nothing left, you go to the CPAS (municipal assistance officeFootnote 3) and you get the minimum income. But in other countries, you wouldn’t get that – you’d end up on the street” (Brussels, 2006). The compensatory rationale is used as the background against which the merits of welfare are evaluated. Participants consider compensatory protection as the taken-for-granted baseline – against which they judge whether ALMPs are fair, too strict, or insufficient in both France and Belgium.
Workfarist cues, by contrast, are frequently expressed through reciprocity language and behavioural expectations that qualify solidarity as conditional. Yasmina, a 35-year-old French worker, puts this in explicitly exchange-based terms: “Even if a guy is on minimum income, even if he is paid to be in training or whatever… if you give him money, he has to give something in return…” (Grenoble, 2019). In this type of discussion, participants assert solidarity while also framing it as contingent on effort and compliance with requirements. Most of the time, these statements serve to exemplify a more general and moral point raised by participants as to how people should behave. Participants, thus, rely on workfarist cues mainly to state a general moral principle about deservingness and proper conduct (“people should give something back”), rather than to discuss specific ALMP rules or instruments.
Capacitating cues do appear, but they are typically voiced as pragmatic, individual-level tools for improving employability (e.g. “training helps me keep up”), rather than as a shared, general justification of what solidarity should require. For example, Jérôme, a 23-year-old Belgian worker without diploma, frames training as a future-oriented requirement of employability: “In my job you always need to stay up to date… maybe take additional training… for the future” (Louvain-la-Neuve, 2019). In a small number of instances, participants explicitly praise individualized employment guidance as useful in practice, as argued by Guy, a 59-year-old French activist, discussing the Blair reforms in the UK: “He’s done some interesting things with how unemployment’s handled through the job-centre. He pushed through reforms we French just can’t manage – actually doing proper one-to-one employment support.” (Paris, 2006). However, this type of mention is isolated.
While overall capacitating cues are rare in participant discussions, they are slightly more frequent in the Belgian discussions than in the French ones (12 vs 9 occurrences), while France contains a higher number of segments coded as “other” (8 vs 2) (see Table 3). In addition, participants occasionally anchor their evaluations in nationally specific frontline agencies (e.g. references to the CPAS in Belgium or to job-centre/employment-agency interactions in France). Importantly, however, these small differences do not translate into a divergent pattern: across both countries and both time points, compensatory and workfarist cues are more prevalent than capacitating ones. This is analytically consequential, because if social investment had become part of participants’ understanding of social solidarity, we would observe a clearer increase over time and/or stronger cross-case differentiation in their unprompted discussions of social solidarity, which we do not observe.
While this first step of our analysis maps the co-presence of mostly compensatory and workfarist cues in participant discussions of ALMPs, with rarer instances of social investment cues, it overlooks how participants articulate these different policy norms and what citizen-level norms of social solidarity may emerge from these combinations. This is the next step of the analysis.
Citizens’ frames of ALMPs: Hybridity, insecurity and tensions
Based on our analysis, five main frames are present in the discussions, regardless of participants’ socio-economic backgroundFootnote 4. Importantly, first, participants routinely combine competing policy norms when discussing social solidarity in a single frame. Second, the frames participants rely on to discuss ALMPs are of unequal importance and three of them dominate across datasets (see Table 4). Third, there is no noticeable evolution between 2006 and 2019. We now examine each frame in turn.
Table 4. How participants discuss ALMPs: five frames
The “insecurity” frame emphasizes that labour-market policies are experienced in a way that results in insecurity for people in general, and not only for workers. Career trajectories, employment more generally and the entire life course are deemed insecure. Participants debate social solidarity by mentioning competing diagnoses of why insecurity arises and who should bear the costs of protection and adjustment. On the one hand, insecurity is discussed in ways that support a compensatory approach (e.g. demands for stability, rights, or safeguards). On the other hand, insecurity is also sometimes discussed in line with workfarist instruments and emphasis on control, obligations, or compliance. As a result, the “insecurity” frame foregrounds the tension between protection and reciprocity as some participants call for collective protection, while others normalize insecurity as something individuals must manage.
Extract 1: Brussels, White-collars, 2006 [Insecurity – Workfarist]
Tina: Well, what bothers me is mostly things like contracts: we make contracts to encourage employment, but on the other hand, yeah, it only encourages employment for a year.
Michèle: Yes but that, that benefits the employer, not the employee, that’s another debate, but still, that’s…
Tina: but I think that for that, for example, they should do something, like say, look, allow companies to hire people, but at some point they should be required to hire them on permanent contracts.
The “social-rights” frame highlights how labour market policies either grant or erode enforceable entitlements – enhancing protection or, conversely, depriving workers of rights and, thereby, generating labour-market precarity. Participants primarily consider social solidarity as the protection of rights that should be enforceable and provide security. The frame emphasizes most often a compensatory approach, which is reflected in the use of compensatory cues like entitlements, protection and enforceable rights. However, the same rights-based frame is also sometimes voiced in more conditional, workfarist terms, when rights are discussed alongside behavioural obligations or monitoring. Solidarity is affirmed as a matter of protection, but it can also be simultaneously qualified as a protection that is contingent on compliance – shedding light on a recurring tension between entitlement and reciprocity. Philippe, a 32-year-old recruitment officer, discusses the paradox of activation policies that put protection under pressure.
Extract 2: Grenoble, White-collars, 2019 [Social-rights – Workfarist]
Philippe: The market certainly does take care of everything, but there’s something contradictory, er… One thing and its opposite, in fact. We’re effectively asking people in work to remain in employment for longer, to retire later. Meanwhile, we’re also asking young people to enter the labour market, and they find themselves sidelined in terms of employment. We want fewer taxes but we want, we want public services and… There are contradictions there as well, there are paradoxes."
The “individual-responsibility” frame holds that employment outcomes primarily reflect individuals’ choices and behaviours. Workers are held responsible for having, or not having a job, and ALMPs should focus on incentivizing them to adopt the right behaviours, particularly through requirements, monitoring and sanctions. Participants think of social solidarity as a reciprocal obligation whereby support is warranted insofar as individuals are deemed to behave appropriately and make the “right” efforts. Thereby, the frame is mainly expressed in workfarist terms, especially when participants emphasize conditionality, monitoring, and sanctions. At the same time, these terms coexist with affirmations of the principle of solidarity; while redefining it based on deservingness and compliance with expected behaviours. Extract 3 below, from a focus group with white-collars participants held in Paris in 2006, illustrates this frame.
Extract 3: Paris, White-collars, 2006 [Individual-responsibility – Compensatory]
Laetitia: I see people who take advantage while unemployment benefit is, well, is more profitable, and then after a certain time – I don’t know exactly how it works but it drops. And at that point they think, actually it would be better to work. But in the meantime they stay unemployed.
Hadia: Yes, yes, but in any case, if it’s not them who are unemployed, it’ll be someone else, because there aren’t enough jobs for everyone, so…
Clélia: There are jobs available, but they don’t want them…
Laetitia: I think there are tonnes of jobs. Jobs, if you want one, you can find one.
Clélia: They don’t want to. At the job centre or there are plenty of agencies, except that…
The “market” frame is used by participants to foreground social solidarity less as a question of enforceable rights (“social-rights” frame) or moral principle (“solidarity” frame, see below), than as a question of how the economy works: ALMPs are evaluated through their impacts on job creation, labour costs and incentives, and social solidarity becomes implicitly conditional on “not distorting” the labour market. This frame is therefore more often expressed with workfarist and economic ("other") cues, rather than with compensatory or capacitating ones. However, the use of this frame also features normative tensions, that are apparent in the way the meaning of “good” ALMPs shifts away from solidarity as protection, thereby leaving open to debate the question of what principle should replace it. For instance, in a Belgian workers’ group, tensions emerge between Farouk, a 28-year-old security agent, who defends a more generous welfare provision on compensatory grounds and insists that employer tax cuts come at the expense of workers and solidarity, and Dona, a 56-year-old domestic helper, who argues on economic and workfarist grounds that ultimately it is employers who provide workers with jobs, also adding a moral component to her position: “so everyone can make an honest living without staying on unemployment, the CPAS [municipal assistance office] and so on.”
Extract 4: Brussels, Workers, 2006 [Market – Workfarist]
Dona: There are plenty [of workers] who want to work, that’s why we should reduce employer taxes: if only they were taxed less, they’d hire more easily.
Farouk: What I was getting at is that there are more closures than job vacancies, companies shutting down – I’m talking about Belgium. With Europe, it’s about getting more gains, more profits, let’s say, that’s what they’re calculating – with less manpower. That’s how employers think.
[Later during the discussion, coming back on the topic]
Dona: Well, it still puts people to work though.
Farouk: Yeah but who takes the benefits? In reality, it’s the taxes the employer has to pay for the worker – so it’s actually the worker’s own money being taken from the employer.
Dona: Yes but the employer has to invest, so he puts in his own money first of all for equipment, for the factory itself; the worker isn’t going to provide that. He’s the one who has to invest, who has to put up capital beforehand, so he’s committing himself. So when he hires staff, the staff also benefit despite everything, instead of remaining unemployed. You get hired, with a reduction thrown in, of course. Nobody does anything for nothing, but meanwhile you’re working and you’re benefiting, you’ve got earnings too. Of course you’ll pay taxes, just like the employer, but the worker and employee will still make a living at the same time. And for that, there should be even more aid for the self-employed, to be able to hire more staff, so everyone can make an honest living without staying on unemployment, the CPAS and so on.
Finally, in the “solidarity” frame, social solidarity itself is the explicit object of evaluation. The norm of solidarity remains uncontested in participant discourse: across datasets, not a single participant opposes it when it is mentioned, and solidarity is explicitly endorsed as a key foundation of the political community. In that respect, solidarity is understood as a principle that extends well beyond the realm of the welfare state: it is perceived as a fundamental ethical value, which makes it difficult to contest even when there is an argumentative tension with divergent or even opposite principles. Yet, whenever it is discussed, the “solidarity” frame is almost invariably associated with conflicting considerations. In some discussions, participants discuss solidarity in relation to the issue of the costs that solidarity entails. Gaston, a 60-year-old Belgian manager questions whether the (compensatory) welfare state is financially sustainable (CITAE focus group, Brussels, managers, 2006). He is faced with conflicting values and principles (solidarity and public spending considerations) that he endorses, but that stand in a seemingly unsolvable tension.
Extract 5: Brussels, Manager, 2006 [Solidarity – Compensatory]
Gaston: Do you find it normal that one pays between 45 and 60% of his income, his taxes, to people one never even sees?
[…]
Bruno: It must annoy you to have unemployed people around who take your money, but it’s, it’s –
Gaston: No no no there I – because you’re attacking me there, no no no no because, no –
Bruno: Because that’s your philosophy, that’s how it is.
Gaston: I said that I’m in solidarity, I said it. What I said is I cannot accept 65% of my taxes, that’s what I said.
Bruno: How can you be in solidarity if you’re only giving 20%?
In other discussions, participants are critical of the workfarist approach. They stress that, in the absence of stable employment, it is not only the job that becomes precarious, but more profoundly, individuals themselves. However, at the same time, participants are quick to suspect, or even assign blame to, the individual responsibility of unemployed people and, to a lesser degree, that of workers or employers. In general, participants consider that being proactive in job search is enough to secure employment. The suspicion of a lack of willingness to work is particularly strong when the joblessness of young people and the unemployed is discussed. Strikingly, unemployed participants also express such views. Taken together, these recurrent moves – from endorsing solidarity to qualify it through responsibility and compliance – reflect what we conceptualize as an ideological dilemma.
Compensation or reciprocity: ALMPs feed back into an ideological dilemma
In their unprompted discussions of ALMPs, participants repeatedly draw from evaluative criteria that extend beyond instrumental judgements about effectiveness. Consistent with Svallfors’ argument that a normative feedback is present when policies provide citizens with “the desirable state of affairs” (Svallfors, Reference Svallfors2010, p. 120), the discussions we analysed are structured by claims about what social solidarity ought to entail. Across countries, time periods, and social groups, participants discuss and assess principles such as entitlement and protection, reciprocity and deservingness, and, more sporadically, capacitation. These are not merely policy opinions; they operate as normative reference points through which participants discuss ALMPs and articulate expectations about legitimate social relations between the state, workers, and benefit recipients.
Crucially, the normative feedback we observe is shaped by the hybridity of ALMPs. Because ALMPs layer enabling instruments onto persistent compensatory instruments and conditionality, monitoring, and sanctioning, they expose citizens to competing policy norms. In our data, this hybridity is reflected in the fact that the same interpretive frame can be expressed based on distinct welfare-approach cues: rights are affirmed, yet qualified by behavioural requirements; insecurity is narrated either as a structural condition warranting protection or as a context that justifies intensified conditionality; and discussions on responsibility coexist with explicit endorsements of compensatory solidarity. In other words, policies do not simply convey a single norm; they generate a setting in which citizens’ sense-making of social solidarity is repeatedly organized around conflicting evaluative criteria.
For this reason, the normative feedback of ALMPs we report takes the form of a dilemma between compensation and reciprocity. We build on Billig’s conceptualization of ideological dilemmas, which refer to socially shared repertoires of “common-sense” reasoning that are structured by opposing, but widely available, normative imperatives, here solidarity as protection versus solidarity as conditional reciprocity (Billig, Reference Billig1992). Rather than signalling confusion or inconsistency, these dilemmas are analytically consequential (Andreouli et al., Reference Andreouli, Greenland and Figgou2020): they show how citizens can endorse solidarity as an uncontested principle, while simultaneously re-specifying it through stringent conditions and suspicion towards claimants. In our case, the recurring references to a “generous yet stringent” social solidarity capture precisely this form of feedback: hybrid activation fosters a durable tension in which solidarity as compensation is both stated and bounded by reciprocity obligations, resulting in a dilemmatic – yet stable – normative orientation towards ALMPs.
Discussion and conclusion
Acknowledging the normative change introduced by social investment policies in European welfare states, this article examined citizen-level normative feedback of ALMPs, describing how participants interpret and endorse policy norms in collective discussions.
To address this question, we designed a comparative qualitative secondary analysis of two datasets in which French and Francophone Belgians from various social backgrounds and political leanings discuss employment policy, in their own terms and without any policy prompts at two points in time (the early 2000s and the late 2010s). This design is instrumental in examining how unprompted participants, whether they are recipients of ALMPs or not, discuss norms of social solidarity in the context of ALMPs and what norms exactly. Based on the analysis of diagnostic linguistic cues reflecting compensatory, workfarist and capacitating policy norms and inductively built frames of social solidarity, the article investigates the existence of normative feedback of ALMPs in the form of citizen-level norms of social solidarity shared across social groups, points in time and country cases.
Our first expectation recognizes that real-world ALMPs are hybrid policies that include not only capacitating instruments, but also compensatory and workfarist measures. We expected that these policy norms would co-exist in citizen-level norms of social solidarity associated with ALMPs. Our empirical analysis partly confirms this expectation. Across datasets, we report that when discussing ALMPs, participants draw from the workfarist approach of the welfare state, which emphasizes individual responsibility. We also report that the compensatory approach lingers in citizen-level norms as we evidence its presence across social groups, points in time and country cases. Importantly, however, our results deviate from our expectation because while our data includes diagnostic cues associated with the social investment approach, only a few participants use them in their discussion of ALMPs.
Our second expectation zeroes in on how participants articulate the conflicting policy norms of ALMPs. We expected that they would not resolve the normative tensions resulting from the hybridity of ALMPs. Our empirical analysis confirms this expectation. We analysed 5 frames of social solidarity that participants rely on to discuss ALMPs across socio-economic background, time, and country cases. None of these frames reflects a single policy norm. In contrast, while providing different understandings of what social solidarity is in the context of ALMPs, each of them features normative tensions. The most recurring tension is that between social solidarity as compensation and as reciprocity. This dilemma between compensation and reciprocity is the normative feedback of ALMPs. While strongly supporting the principle of solidarity, participants also express significant suspicion regarding the extent to which individuals are responsible for their own circumstances. Paradoxically, even participants who themselves face precarious employment or are unemployed endorse such views. This tension leads participants to accept, and at times even support, ALMPs that contribute to the deterioration of working conditions, an outcome that they themselves regret or actively criticize.
Besides, another result concerns the role of time in citizen-level normative change. In line with other works (Author 1, 2024; Svallfors, Reference Svallfors2010; Vila-Henninger, Reference Vila-Henninger and Vila-Henninger2020), we observe that citizen-level norms change at a slower pace than policy norms. In our data, we witness few traces of the normative underpinnings of the social investment welfare state. It may be that social investment is simply too recent and too weakly present in hybrid ALMPs to result in a clear capacitating normative shift at the level of citizen-level norms. In addition to being design-contingent, it is likely that the normative change related to social investment is only temporally lagged.
Alternative mechanisms may also contribute to the patterns we observe. Some frames – especially insecurity and responsibility claims – can plausibly be explained by broader labour-market transformations (Jørgensen and Schulze, Reference Jørgensen and Schulze2024), media debates (Pennetreau and Laloux, Reference Pennetreau and Laloux2025), and national political discourse (Bakkær Simonsen, Reference Bakkær Simonsen2021), rather than by ALMPs narrowly construed. As regards media debates and political discourses in France and Belgium, they both are depoliticized as they mainly frame ALMPs through a “market” lens and, to a lesser extent, through an “individual-responsibility” lens (Pennetreau and Laloux, Reference Pennetreau and Laloux2025). This discursive context is not specific to France and Belgium but is also observable more generally in Europe and the US (Dencker-Larsen and Lundberg, Reference Dencker-Larsen and Lundberg2016; Jørgensen and Schulze, Reference Jørgensen and Schulze2024). Moreover, the relative scarcity of capacitation-oriented public debates may reflect not only policy design (Mettler, Reference Mettler2011), but also the fact that enabling instruments may often be experienced based on individualized interactions and may, therefore, be harder to generalize into shared public justifications in unprompted discussions (Soss, Reference Soss1999; Brodkin and Larsen, Reference Brodkin and Larsen2013).
In conclusion, we report that active labour market policies in France and Francophone Belgium have a clear normative feedback effect, one that foregrounds the ambivalent, and unresolved, coexistence of the conflicting norms of compensation and individual responsibility, and that social investment policy norms have not (yet) reshaped citizen-level norms of social solidarity.
Funding
This research was funded by the project ERC Starting Grant Qualidem - Eroding Democracies. A qualitative (re-)appraisal of how policies shape democratic linkages in Western Democracies. The Qualidem project is supported by the European Research Council (ERC) under the European Union’s Horizon 2020 research and innovation programme (grant agreement 716208). This research also benefited from the support of an FNRS Fresh Grant awarded to Damien Pennetreau.
Open access
