Between structure and agency: the use of conditional welfare through the lens of social capital Open Access

This article assumes that the theory of social capital can contribute to explaining some of the processes behind decision-making independence and the informal aspects of relationship in the implementation of social policies. To support this theoretical proposition by means of case-based empirical evidence, the article focuses on the implementation of the principle of welfare conditionality, which links the access to income support benefits on the acceptance of job search activities and training. Three comparative cases show that its implementation varies significantly in the different contexts: conditionality embodies a very weak principle in Barcelona, which is not applied in practice; in Lyon it has a symbolic function representing a formality that provides access to an unconditional right; in Gothenburg it is implemented as a stringent and effective proof of activation. Among the many institutional and structural factors that help explain these differences, the article shows that there are also specific forms of trust, group identity and mutual recognition that motivate civil servants’ agency and discretionary choices.

Over the last decades, comparative political economy and welfare state research have devoted particular attention to the patterns of change of welfare regimes and social policy (Esping-Andersen 1990; Hemerijck 2017; Hassel and Palier 2021). Especially in the European context, where the reallocation of authority upwards and downwards has drawn attention from a growing number of scholars, specific research has been dedicated to both exploring social policy harmonisation and understanding the maintenance of national and regional differences. While a tendency towards convergence is represented by the sharing of policy ideas and paradigms, such as social investment or activation, that have widely found place in the EU and national policy agendas, when looking at policy implementation, divergence between European countries seems to persist (Hassel and Palier 2021; Burroni et al. 2021; Scalise 2020). Two complementary approaches help to explain such divergence. On the one hand, macro and institutionalist analysis demonstrate that structural factors and the historical legacies of European capitalisms cause countries to continue following a certain trajectory that remains within the boundaries drawn by their own model of political economy (Esping-Andersen 1990; Amable 2003). On the other hand, meso and micro perspectives reveal that social policy-making, implementation and delivery are also influenced by factors related to social practices, relations, networks, and the agency of actors embedded in specific contexts (Andreotti et al. 2012; Johansson and Panican 2016). Especially the street-level bureaucracy (SLB) literature highlights the responsibility of local managers and civil servants in social service organisations for policy outcomes and acknowledges the formal and informal practices that lead to diverse implementations of common regulations (Lipsky 1980; Brodkin 2011). Although front-line workers’ action represents the ‘last link’ in the policy-making chain, it is crucial because only in the interaction between them and welfare recipients ‘formal social policy comes to life’ (Rice 2013). This perspective, which is often overlooked in social policy studies, needs to be taken into consideration for a better understanding of policy impact, especially today that we are observing a shift towards service-oriented welfare states (Hemerijck 2017).

This article builds on the SLB literature on discretion in social policy implementation and contributes to this debate by including an additional theoretical perspective that focuses on the informal aspects of relationship structures, namely that of social capital. Despite its widespread use in social sciences, the SLB literature has barely referred to the notion of social capital, which remains under-researched in the SLB realm (Rothstein and Stolle 2001). We bring the social capital theory and the SLB approach into dialogue and assume that the notion of social capital can contribute to explaining some of the processes behind decision-making independence and discretion rooted in social policy delivery. To support this theoretical proposition by means of case-based empirical evidence, we focus on the implementation of the principle of welfare conditionality, which links the access to income support benefits on the acceptance of job search activities and training.

The article is organised as follows. The next two sections discuss the SLB literature and the theory of social capital. Section Four explains the research design and the three comparative case studies performed to assess the implementation of income benefits conditionality in the cities of Barcelona, Lyon and Gothenburg. Section Five outlines the results obtained from the analysis, showing that despite the fact that the legal requirement for jobseekers’ behaviour monitoring has been introduced in all three countries, its implementation varies significantly in the different contexts. In Barcelona, conditionality embodies a very weak principle which is not applied in practice; in Lyon it has a symbolic function representing a formality that provides access to an unconditional right; and in Gothenburg, it is implemented as a stringent and effective proof of activation.

As discusses in the final section, institutional and structural factors weight in explaining these differences. However, civil servants’ discretionary choices are also motivated by different forms of trust, mutual recognition and sense of belonging which, especially in the Catalan case study, can be associated with some operating principles of social capital. Indeed, in Barcelona a source of social capital based on solidarity seems supported by a strong, cohesive sense of local community and belonging, and guides civil servants’ actions according to the principle of group solidarity. Instead, forms of reciprocity and redistribution can be seen in Lyon, underpinned by a universalistic recognition emerging from this society's normative structure. Finally, the Gothenburg case study revealed a generalised trust based on confidence in public institutions and in the returns in welfare that these institutions guarantee.

The differences that persist between European countries, regions and cities in social welfare policy and delivery are a matter of both theoretical and political relevance, as shown by the numerous studies which discuss developments of convergence and divergence between welfare regimes and social policy implementation. Despite a process of policy hybridisation is taking place – encouraged by EU incentives, exchange of policy practices and learning processes – institutional theories explain that the lengthy process of recalibration that welfare states are undergoing over the last decades are not undermining distinct regimes: path dependency and institutional diversity still shape the different social protection systems, and the forms of access to, and the provision of, welfare services (Esping-Andersen 1990, Hassel and Palier 2021).

Other perspectives, which examine social policies starting from their concrete delivery, complement this macro analytical framework emphasising other intervening factors. More specifically, bottom-up and micro-level analysis demonstrate that cross-country and regional diversity also depends on local policy implementation (Johansson and Panican 2016). In this respect, a significant contribution is offered by the street-level bureaucracy (SLB) theory which explains the critical role played by the actors responsible for the translation of social objectives into actual service delivery (Brodkin 2011; Lipsky 1980). The SLB literature describes the working practices and beliefs of civil servants in public services and acknowledges them as the last but most important link in the policy-making chain because it is only through their action, and interaction with welfare recipients, that social policy takes concrete form (Lipsky 1980; Rice 2013). Even in countries that are highly centralised and homogeneous in terms of domestic welfare policy, several organisational and cognitive factors influence the behaviour of civil servants when they are in direct contact with welfare recipients and have the task of assigning them benefits or sanctions. Lipsky (1980) stresses the discretionary power that is exercised in determining the nature, amount, and quality of benefits or sanctions provided by the agencies in which front-line workers operate. Discretion is implicit in their role, dealing with specific cases starting from regulations, interpreting and adapting them to the conditions with which they are presented on each occasion. Civil servants’ pragmatic choices ultimately become the policy of the organisation, which may contrast starkly with its official stated aims (Cooper et al. 2015).

Part of SLB literature has focused on how civil servants have responded to the demands arising from the New Public Management wave of reforms enacted in Europe since the ‘90s and to the progressive adoption of managerial types of accountability system, performance indicators and outcome-based orientations, showing that despite the spread of routine strategies and standardised procedures, front-line workers still end up having a significant impact on the outcomes of the policies they are called upon to implement (Hupe 2019; Soss et al. 2011; Cohen 2018). Discretion occurs not only because technical limits are in place to monitor their work, but also because professionals are required to develop a workable policy in practice (Evans 2011).

The public administration literature addresses discretionary decisions as a coping strategy that civil servants adopt to adjust to policy change, to lessen external or internal conflicts and demands they face, or to downscale the tension between the goals to be attained and the means provided for doing so (Kosar 2011; Thomann 2015). Different administrative cultures also get a reflection in the organisations and practices carried out in public entities and shape different styles of public management. Administrative ethics and understandings of the public role intertwine with formal and rational choices, influencing how service providers perform when delivering social policy (Osborne 2010; Lavee and Strier 2018).

What the SLB perspective stresses is the existence of additional areas of decision-making independence and action; the presence of unforeseen and perverse effects of policy implementation; and the interaction of macro, meso and micro factors that come into play in influencing civil servants’ decisions. The discretion they enjoy and their agency are embedded in an institutional and socio-cultural context which conditions their motivation and behaviour. Pressures and constraints deriving from the legal, organisational and economic environments, as well as motivations stemming from political, social, cultural, ideological and emotional conditions impact on civil servant intervention. Yet, these multiple aspects behind civil servants’ motivation of discretionary action are extremely difficult to recognise and disentangle, so much that the understanding of the nature of discretion in their daily practice is not fully achieved.

With the aim of exploring additional socio-cultural factors which may play a role in affecting civil servants’ agency and discretion, we focus on the informal aspects of relationship structures in society that have been highlighted by the theory of social capital. More specifically, we refer to the types of relationships based on shared social norms, trust, reciprocity, and solidarity that can shape different forms of social capital (Coleman 1990). As the following section explains, we assume that this concept can be useful in highlighting some of the socio-cultural aspects of human agency and individual discretion which may intervene when wider public benefits of social welfare are served or sanctions are not applied based on a sense of solidarity, social obligation or shared identity (Christoforou 2013).

One of the reasons why the concept of social capital (SC) in the ‘90s reached notable popularity is that it is particularly valuable in exploring the motivations for different ways of acting within individuals, communities and societies. SC theory focuses on the potential of individual action that derives from relationship structures and it has been developed to show how different types of relationships, based, for instance, on expectation, dependence, or group unity, shape different forms of SC (Coleman 1990). These types of relationship are constructed on the informal norms present in a given social context, on forms of obligation, control, or on the sharing of values and principles that are engendered in the processes of socialisation to match a community's culture, societal values and moral provisions (Bagnasco 1999).

According to Coleman (1990), SC is a neutral resource that facilitates any manner of action. It stems from different situations and is embedded in a set of social relationships on which a subject or a community depends. It is a source of cognitive resources, such as information, or normative resources, such as trust or solidarity, which can be ‘activated’ in certain conditions for the achievement of goals. The forms of authority, solidarity, cooperation or reciprocity which take shape within the structure of social relations can become a resource for action, for exchange or control between two or more people. The more individuals depend on each other, the more obligations, bonds and SC are available for use.

SC has an individual dimension, which can be used for subjective purposes, and a structural or systemic dimension, which may promote collective action or cooperation within specific communities or a society (Bagnasco 1999). The concept has been developed and rendered operational in different ways by several scholars in order to demonstrate that individuals, communities, and societies have diverse endowments of SC. More specifically, it has been adopted to reveal, on one hand, phenomena of social stratification, inequalities and occupational mobility (Bourdieu 1980; Granovetter 1985); on the other, mutual trust, expectation of certain behaviours and different forms of ‘civicness’ have been used to explain the different routes towards modernisation and the varying levels of performance in institutional and economic development (Putnam 1993; Fukuyama 1995). One of the best-known works from the latter strand of research is that of Putnam (1993), who focused on interpersonal relations and actions for mutual advantage in dense social networks which favour the diffusion of norms of reciprocity and foster internal habits such as trust and solidarity. Putnam refers to civic engagement networks (e.g. neighbourhood societies, purchasing associations, cooperatives) which are a source of social commitment projected beyond the group and treats SC as a public good which favours social, economic and institutional functioning.

Some of these theories have received both criticisms – particularly for their broad definitions, methodological shortcomings and multiple applications – and appreciation. Nevertheless, one of the strengths of SC, which cannot be neglected, is that it is particularly effective in highlighting the constitutively interpersonal character of action. Developing Coleman's thought, Pizzorno (1999) demonstrates the normative nature of action and the relationship between rational action and recognition, providing alternatives to the explanations offered by instrumental rationality and methodological individualism. In Pizzorno's interpretation, interests are always embedded in a setting in which the relationship with others, the evaluation of others and the recognition of others have a significant weight. Rationality itself is a matter of imputation, judgement and recognition by participants and observers of the action. What is key in his approach is a focus on what he calls the ‘reception’ of the action, since the ways in which the action is received, identified and recognised by those who participate in the action and the context in which the action takes place, call into question a set of conventions, expectations and beliefs that shape its evolution.

Action is defined by the relationship between the players involved and by the settings in which it is implemented: the resulting social bond is articulated through recognition by the other, identification of the other, and sharing, for instance, a common belonging, values or the rediscovery of a common past. The dynamics of creation of provisions and processes, whose measures are used to recognise, ascribe and formulate judgments, are triggered by relationships. The identification of the circles who recognise the players’ choices and express judgements on them is what SC theory enables us to see.

As the SLB literature highlights, street-level workers’ decisions are influenced by the relationship established with welfare recipients and by the context in which it takes place. How can we exclude that SC dynamics are completely absent in this relationship? Can SC be an intervening factor arising from the relationship itself and influencing its outcome? If we combine these two approaches, we can assume that it may be possible that civil servant action is influenced by some forms or operating principles of SC, which may be rooted in the circles to which civil servant and welfare recipient belong or in the institutional, organisational and socio-cultural contexts in which their relationship is embedded.

Pizzorno identifies, as bearers of SC, those relationships in which it is possible to recognise the lasting identity of the participants, and that are characterised by certain forms of solidarity or reciprocity (1999). The SC of solidarity develops through the intervention of a third party (namely a social group, an agency or an institution) which ensures that the relationship between two persons is conducted without fraud or opportunism. It is based on a type of social relation that arises through, or is supported by, the presence of a cohesive group whose members are connected to each other with strong and lasting ties. In this case, it is to be expected that they will act according to principles of group solidarity. This relationship constitutes SC when a person expects that the other complies with a request for help, because they both belong to the same group – acting under obligations of solidarity and internal trust – or because one of them belongs to a cohesive group which is able to reward or punish the other, who consequently acts under the obligation of bonds or external trust.

The SC of reciprocity is constituted by the relationship between two parties, in which one of the two provides some sort of assistance to allow the other party to pursue its objectives, assuming that a dyadic relationship of mutual support is established. This is based on weaker ties and can be seen in relationships of cooperation, or situations in which there is a common goal for the two parties. In addition, in this form of relationship, one person helps another without being asked, or without receiving anything in return, and this can happen for several reasons: because this reciprocity will be repaid in the future, to establish a relationship of gratitude – as occurs in the processes generating collective SC – or to increase the prestige of a community (for example, a compatriot, ethnic or religious group). In these cases, there is no cohesive circle of recognition capable of identifying a single member; there is an ideal reference community, to which one imagines belonging to, and from which one cannot expect explicit rewarding or penalising acts. The action depends on the subjective evaluation of those who help, and could be motivated by universalistic principles, internalised to become part of the individual's identity.

These two ideal typical forms of SC are also found in the typology by Portes and Sensenbrenner (1993), who distinguish four main social processes that lead to SC. They refer to a form of bounded solidarity which takes place in adverse circumstances that cause the emergence of principled group-oriented behaviour and a ‘we-ness’ sentiment that steer forms of altruistic conduct; and to reciprocity exchanges, which consist of the accumulation of ‘chits’ based on previous good deeds to others, backed by the norm of reciprocity. This typology also consists of value introjection, a behaviour characterised by an element of moral obligation because individuals have been socialised to certain values, norms and moral imperatives that inform their action; and finally, enforceable trust, which is a source of social capital based on the internal sanctioning and monitoring capacity of the community that ensure compliance by individuals with social expectations and norms.

In this perspective, the theory of SC recalls Polanyi's three forms of integration and the central idea of embeddedness (Carroll and Stanfield 2003). Indeed, Coleman (1990) uses SC to emphasise how social institutions influence individual choices and that individuals do not act independently of each other. In Polanyi's analysis (1957), economic processes are always embedded in wider social institutions, upon which their dynamics are dependent. To delineate the basic forms of economic integration, Polanyi underlines the complex networks of shared obligations that motivate individual behaviour, arguing that alternative forms of exchange function, differently from market exchange, could be reciprocity and redistribution. These are characterised by different sets of motivations, alternative to principles of rational behaviour, resting on foundations of kinship connection or on political obligations. These reciprocations can be informal or they can be formalised into redistributive welfare states and are based on specific institutions that support them and sanction those who do not comply with them in various ways (e.g. family networks, the state). As Pizzorno's SC of reciprocity, Polanyi's ‘thick’ reciprocity is not an expectation of exact equivalence, is not strictly dyadic and its timing is open-ended (Block 2003).

Types of relationships based on norms of solidarity or reciprocity, on forms of redistribution, on value introjection, or enforceable trust, can be used to forge SC shared with multiple others and to take actions that could never be justified by purely rational calculations or routine and formalised practices. Building on these considerations, we assume that forms of SC may be present in the relationships between civil servants and welfare recipients, which could be shaped by the social norms and values shared in the communities, organisations or society in which these relationships are embedded. Therefore, the above-mentioned mechanisms of SC generation can represent a meaningful theoretical framework which can contribute to explaining the multiple reasons behind discretionary behaviours in welfare delivery.

Building on the SLB literature which emphasises the socio-cultural factors influencing street-level workers’ behaviour, and on the theories that identify different kinds of SC arising from specific social relationships, this article hypothesises that some forms of SC – together with other intervening variables – may contribute to affecting the discretionary decisions of street-level bureaucrats. To demonstrate the plausibility of this hypothesis, we conducted an exploratory analysis on the implementation of one specific social policy tool, the principle of welfare conditionality. The research question which has guided the analysis is whether some operating principles or mechanisms of SC can be identified in the social relationships and practices carried out in the implementation of welfare conditionality. The aim is to examine whether, among the variety of structural, institutional and socio-cultural factors that shape how conditionality is implemented in different contexts, ‘signals’ of SC can be traced.

In Table 1, we operationalised the concept of SC to identify the mechanisms of SC generation, assuming that different types of SC are not mutually exclusive.¹

The exploratory analysis aimed at identifying SC operating principles which can be associated with the SC ideal-types, and to understand whether these mechanisms may have contributed to motivating individual action in the relationship between civil servants and welfare recipients.

The analysis has focused on the implementation of income benefit conditionality, a typical active labour market policies (ALMPs) instrument which links welfare rights, namely the access to income support benefits, on an individual first agreeing to meet particular obligations or patterns of behaviour, usually the acceptance of job search activities and training. As it is known, since the end of the 1990s there has been a shift towards ALMPs across Europe driven by the expansion of the European Employment Strategy and induced by its financial and technical incentives (especially the European Social fund and the Open Method of Coordination) (Bonoli 2013). ALMPs are typical of Nordic welfare states and began to gain ground through the 2000s in some continental, Anglo-Saxon and Central-Eastern countries while Southern European countries followed more slowly. Even so, the multi-faceted and variegated nature of activation policies remains to be fully discovered (Barbier and Ludwig-Mayerhofer 2004). We do not discuss here the wide range of activation measures (namely income support, incentive reinforcement, skills training, employment assistance and job matching), but we focus on one of the core-principles of activation: the notion of conditionality. This has been identified as a key factor in improving the effectiveness of labour market policy and has become central in policy reforms. Conditionality means making entitlement to income support conditional on job search and participation in training programmes. The link between activation and income support can be more or less strong and punitive, and the conditionality principle can be applied at different levels, ranging from participation in activation programmes (more specifically, training or personalised assistance in job-searching), to mandatory job search, right through to the application of sanctions onto those who fail to fulfil entitlement conditions.

Our analysis focuses on discretional spaces existing in the phase of social services delivery to investigate whether some forms of SC may be found in the social relationships established in public employment services (PES) in the implementation phase of ALMPs, if SC can influence the practical approach of civil servants, and the consequences that this may have.

The exploratory study follows a most dissimilar country selection with respect to national welfare regimes – Nordic, Continental and Mediterranean models – but similar with respect to the local context. To this end, we have selected three regional cases in Western Europe: west Sweden, Catalonia and Rhône-Alpes with a strong focus on the three comparable post-industrial cities of Barcelona, Lyon and Gothenburg. The aim is not that of generalising the cases examined, but rather of understanding them in their individual characteristics, complexity and specific socio-economic context (Yin 2008).

At the time of the fieldwork, all three cities were governed by centre-left coalitions – with assertive social justice and equal opportunity normative orientations – and were receiving the European Social Fund. They were selected to represent economically dynamic contexts (GDP per capita and male and female employment rates are above or in line with the national and European averages) meaning that there is a need for a labour force and there are more resources to develop labour market and social policies compared to less dynamic areas in Europe.

Despite this, the three contexts have been hit hard by the 2008 economic crisis. Catalonia, in particular, had not yet recovered by the time of the interviews (end of 2015) and was still affected by acute unemployment (in 2014 in Barcelona the unemployment rate was 20% and 47.1% among residents aged 16–24 according Eurostat). Since 2008, the context has long been characterised by chronic high unemployment, acute segmentation, a large irregular economy, a cheap, unskilled, temporary workforce and insufficient provision of services. The crisis, which went on for almost a decade, intensified social inequalities, bringing about the general impoverishment of the middle classes and increasing the vulnerability of citizens. In addition, deep fault lines were exposed between the region and the rest of the country.

The crisis also exacerbated income inequalities in Rhône-Alpes. Both unemployment and poverty rates increased in the region between 2008 and 2010 and the area was profoundly shaped by the economic polarisation and poverty segregation. Metropolitan Lyon became characterised by a social and territorial fracture, with precarious workers, the unemployed and low-income families concentrated in certain geographical areas of the urban community, the Southern and Eastern districts, where job uncertainty and poverty rates are higher than the regional average. The unemployment rate increased by 2.3% between 2008 and 2009, reaching 8.6% in 2009 and 9.8% in 2013. Fiscal and social redistribution has had an important re-balancing effect on income inequality here, but the poverty rate has increased over the last few years, displaying its close links to the employment situation.

Finally, in West Sweden, economic recovery after the 2008 financial crisis proved to be faster than in the other two contexts. Unemployment in the region fell after 2014 and, for the first time in two decades, reached a lower level in Gothenburg (6.3%) than in Stockholm (6.7%). and since then it has continued to decrease. Certain groups, however, continued to find it difficult to get a firm foothold in the labour market: young people with a low level of education, over 50 unemployed and, in particular, immigrants. Gothenburg has a sizeable immigrant population and socioeconomic and housing segregation has become one of the city's major problems.

The three countries denote slightly different subnational administrative autonomy and spending margins. According to OECD data, in 2016 subnational expenditure represented 25% of the GDP in Sweden, 20.8% in Spain (both above EU and OECD countries averages), and 11.1% in France. As a unitary and decentralised state, the Swedish central government holds exclusive legislative powers, but the Constitution recognises local self-government. Counties have competence in the field of growth and development and municipalities hold powers in the fields of social welfare, economic development and education. In Spain the central state has competence in labour legislation and social insurance, but the Catalan autonomous community manages education and has full legislative power on social assistance, non-contributory pensions and unemployment benefits. The competences of the municipality of Barcelona are mainly related to the promotion of social reinsertion and education. On the contrary, French subnational institutions do not have legislative power but exercise their functions by means of regulations and through the execution of their budget. Regions, departments and municipalities share responsibility in education, vocational training and apprenticeship, regional economic planning, inclusion and social welfare. In all cases these activities are financed through municipal taxes, government grants and charges: how financial resources are to be distributed is a political decision at local level.

The exploratory research was based on a combination of qualitative tools: in-depth process tracing and analysis of national, regional and municipal policy programmes, schemes and regulations collected both online ‒ on the websites of Ministries of Labour, regional and municipal labour and social units, agencies and PES ‒ and directly in the three contexts; and content and thematic analysis of 34 semi-structured interviews conducted in Gothenburg, Barcelona and Lyon between November 2015 and October 2018 (see list of interviews in Annex) with local relevant key informants, transcribed and organised with MaxQda software into clusters of conceptual categories and themes. Interviewees were selected to collect different perspectives on how ALMPs are concretely delivered and especially how the principle of conditionality is applied. To this aim, we interviewed civil servants at different hierarchical positions – directors of employment and social services, municipal policy officers and street-level employees – local labour and social policy experts and representatives of the social partners involved in local labour market regulation.

The analysis of policy documents confirms that the legal requirements of jobseeker behaviour monitoring and income benefit conditionality have been introduced in all contexts considered and sanctioning rules are similar in the three countries: despite Sweden has the stricter reporting and monitoring procedures, in all the contexts the refusal to participate in ALMPs leads to temporary disqualification from benefits and repeated refusal of a suitable job causes a complete disqualification. Yet, the fieldwork reveals very different forms of implementation of these rules. The next sections describe the three cases in detail, showing that in Barcelona, conditionality embodies a very weak principle which is not applied in practice; in Lyon it has a symbolic function representing a formality providing access to an unconditional right; while in Gothenburg it is performed as a rigorous and effective proof of activation. Structural and institutional factors, and especially the different inclusive capacity of the diverse welfare regimes, are certainly key in explaining these differences. However, different forms of social relations, mutual recognition and trust appear also to have an influence on civil servants’ motivation for discretionary action.

The requirement to be actively seeking work was incorporated into the Spanish unemployment benefit statute in the 2000s. As mentioned, it required monitoring of jobseeker behaviour, and if conditions were not met, it entailed sanctions in relation to benefits. However, as explained by some interviewees, the concrete application of conditionality made its implementation very weak in practice.

Conditionality criteria were not implemented by street-level workers for several reasons. Interviewees revealed that conditionality was interpreted by public servants as an ‘unfair measure’ in a context where resources were already scarce. They considered it as a way of increasing ‘the risk of social exclusion in the community’, poverty and stigmatisation of non-compliant recipients and it was also intended as a selectivity and targeting tool. ‘It decides who are the people entitled to benefits’, because they are acting in ways that are desirable, ‘who is in need’ and it ‘assumes that these people need to be steered to make decisions in their own interest’ (officer, Social rights unit, Municipality of Barcelona, November 2015).

Many factors come into play in explaining the decision not to apply conditionality. Firstly, organisational and economic reasons have a weight, since the public services responsible for developing employment-related strategies and their implementation lacked the human and economic resources to deliver efficient services. Interviewees reported that the insufficient number of public servants in local administration prevented them from guaranteeing an efficient orientation service and allocating training and job offers to the unemployed, and from maintaining an effective control of the requisites.

Secondly, the agency of some actors such as third sector organisations (e.g. Caritas, the Red Cross) – who have always been important in this country with its familial social policy and Catholic imprinting – influenced conditionality application, having ‘a growing role in brokering and trading locally to tackle long-term unemployment and the risk of social exclusion’ since the onset of the crisis. These players, which represent a source of group identity and recognition, as defined by Pizzorno, together with street-level bureaucrats ‘have limited the impact of conditionality criteria’ (Policy advisor, Autonomous University of Barcelona, November 2015).

Indeed, the lack of implementation of conditionality and sanctions was also driven by the perception of a compelling activation which was dominant in this context following the economic crisis. ‘Austerity weakened the role of the state and made it difficult to create effective and capacitating policies, while a strategy that aimed to make labour market inclusion as fast as possible prevailed’ (trade unionist, Barcelona, November 2015). Measures primarily oriented towards the removal of obstacles to labour market participation were promoted, and work incentives were strengthened, without considering the limits of employment demand and real capacities for access. This increased the potential risk of placing workers in unstable job situations, and possibly led to new requests for social assistance.

Therefore, the adverse economic circumstance, combined with the limited and fragmented system of income support schemes and weak training and job matching programmes, typical of the Mediterranean welfare model, contributed to motivating civil servants’ discretionary action. Yet, these conditions have also represented the foundations for the emergence of Pizzorno's group-oriented behaviour and we-ness sentiment that steered a form of altruistic conduct. In fact, civil servants’ decision was also motivated by a sense of solidarity towards the local community, reinforced by a sort of bottom-up reaction to austerity and a sense of community resistance to conditionalities imposed from above. In this case, the notion of appropriateness, which in Pizzorno's theory is a criterion of action, and of judgement on the action, which the subjects involved in a given context utilise in order to provide recognition and continuity, seems to emerge, as stated by one interviewee: ‘certain measures are not applicable in a context such as ours, which has already been severely challenged by the crisis (…) and in which the local community has united, in the face of widespread difficulties’ (Policy advisor, Autonomous University of Barcelona, November 2015).

Civil servants in Barcelona considered the inappropriateness of conditionality and sanctions and acted according to principles of group solidarity. The choice of not implementing conditionality does not appear to be determined solely by structural variables, but it is also linked to the sharing of a common ground, the common experience of a difficult condition, the presence of internal trust and strong social ties among this community. Some operating principles of Pizzorno's SC based on solidarity appears to be in action in this context, which is characterised by an intrinsic commitment to strengthening the community, a shared sense of belonging and the connection with others, the recognition of the identity of others and the fact that the community are ‘all in this together’.

The conditionality principle had also been introduced to the French system as early as 1988, as part of the first Minimum Insertion Income (Revenu Minimum d’Insertion – RMI) provision, which contained a placement clause requiring all benefit claimants to formally engage in activities that had been organised to ease their integration into the labour market. Activation was based on a reciprocal engagement between claimants and the state, but this definition was rather ambiguous as regards the degree of obligation required of claimants. The RMI was replaced in 2009 when the Active Solidarity Income (Revenue de Solidarité Active – RSA) and Law 758/2008 on Rights and Obligations for Jobseekers (Loi relative aux droits et aux devoirs des demandeurs d’emploi) came into effect. This introduced another ambiguous concept, the ‘reasonable offers of work’ that benefit claimants have to accept in exchange for income support, and the sanctions in case of non-compliance with the Personalised Project for Access to Employment (PPAE).

Interviewees reported a stronger emphasis, compared to the past, on claimants’ commitments and on mutual interactions with this new law. Yet, in practice,

Although the objective was to accentuate the potential of benefit conditionality as an instrument for getting people back into employment, according to interviewees, the conditionality principle seems to have had a largely symbolic function. Benefits are formally linked to the active search for jobs and to engagement in activities promoting social inclusion, but the failure to respect these conditions does not lead to sanctions, because they are not applied. ‘The activities required from claimants, such as searching for employment, are used to assess their compliance to receive benefits’ (Civil Servant, PES in Lyon, January 2016). Conditionality is interpreted by civil servants in terms of general proof of means in order to enjoy an unconditional right to an income. The contractual dimension is intended more as an agreement allowing claimants to maintain their income support, than as a tool encouraging re-entry into the labour market. This interpretation of conditionality represents a very loose form of constraint and does not follow a coercive approach.

In this case, however, the civil servants’ actions do not seem to be motivated by a sense of belonging to a specific group, nor by a shared idea of inappropriateness of the measure. In describing the social entitlements of unemployed people, civil servants refer to deep-seated universalistic values resulting from the normative structure of society. The implementation of ALMPs is tied to the traditionally French attention to the collective dimension of social exclusion, which appears deeply embedded in the public administrative culture and civil servants’ value setting. Activation is merged with the logic of insertion sociale, a way ‘of fostering participation in the local collective life of those who are in need, based on citizenship rights and on the strong responsibility of the public authority to address them’ (Policy officer, Rhône-Alpes Regional council, January 2016). This shared view gives shape to a ‘protective’ activation based on reciprocity, and primarily oriented towards security, redistribution and assistance, which evokes Polanyi's reciprocations formalised into redistributive welfare states.

Certainly, this approach is deeply rooted in the French conservative welfare regime and reflects the Bismarckian tradition of social insurance. Despite French social programs and institutions have been progressively reformed to adapt to the new economic and social environment, France seems to preserve its social model and welfare system based on encompassing passive policies and programmes aimed at increasing the possibility of entering the labour market through a combination of job creation schemes, training and job assistance. Indeed, the shared ideals of collective responsibility, and the obligation of the state towards its citizens that emerged from interviews are also supported by the use of state-aided jobs which continue to have considerable relevance within the activation programmes.

As mentioned, interviewees reported the introduction of new activation tools which are associated, just as in the Barcelona case study, with the period following the economic crisis, and the incorporation of EU employment goals in the assessment of trends in the domestic labour market. Interviewees identified the contrast between, on one hand, activation measures, and on the other, traditional social provisions, stressing that ‘in France they are based on the principle of solidarité’ according to which ‘all citizens are dependent on one another by law’ (policy advisor, University of Lyon, January 2016).

Though the general perception of interviewees at the time of the fieldwork was that the primary objective of state institutions remained providing protection against social risks, and that the role played by passive policies was still considered to prevail, increasing inequality and recent protests and tension regarding the latest social and labour reforms in France are calling the values of social equality and reciprocity into question.

ALMPs lie at the foundations of the Swedish welfare model's historical configuration. This is a model based on low unemployment and a high rate of labour market participation for all groups, in which the conceptualisation of social exclusion is closely intertwined with the condition of not working, as confirmed by interviewees who described work as ‘the only entrance ticket to the universalistic welfare state’ and placing activation at the core of this system: ‘I don't think a society can function without every person active and working’ (Trade unionist, Gothenburg, April 2016).

Interviewees in Gothenburg referred to the local system of activation, developed since the 1990s, which allows the municipality to require anyone who receives social assistance to participate in training programmes and other activation measures. The principle of income benefit conditionality is described by civil servants as a stringent, effective proof of activation. This means that, differently from the situation in Lyon, it is not just a mere formality providing access to income subsidies, because the requisites for activation are demanding, recipients are strictly monitored, and subsidies are rigorously bound to training, active job-searching and job offer acceptance.

Civil servants explained that they adopted ‘considerable autonomy in linking activation requirements to social assistance benefits’ (Officer, Gothenburg City, April 2016). Yet, while in the Catalan case they minimise the effects of conditionality, in Gothenburg their counterparts adopt firm rules to incentivise participation in activation measures and impose sanctions, by reducing social assistance payments. They exercise substantial discretion in deciding on the level of benefits and their duration, and in demanding participation in job search and upskilling measures in individual cases.

In this case, rather than SC mechanisms based on solidarity or reciprocity, therefore linked to the relationship between subjects in the community/society, a type of generalised trust emerges. Civil servants’ action seems driven by a strong sense of faith in the institutions, confidence in the effective cooperation and commitment between the state and welfare beneficiaries, and in the welfare returns guaranteed by institutions.

As in the case of Lyon, social inclusion is reinforced by means of a broad-based social policy. However, in this model social policy is conceived as productive factor and citizenship becomes equated to being in paid work, to a far higher degree than is commonly seen elsewhere (Halvorsen and Jensen 2006). Active policies have priority over passive measures, and the role of reciprocal commitment between the state and those in need is much more intense. ‘Active measures have always been privileged, rather than passive income support’, according to the premise that ‘no people should be granted long-term public income support until all ways of making them self-sufficient through employment have been tried’ (policy officer, Labour Market Unit, Gothenburg City, April 2016). This approach – certainly also influenced by the fact that in recent years Swedish governments have followed an agenda of cuts and privatisations that is reshaping the Swedish social model – is the result of a shared vision amongst local policymakers, stakeholders and civil servants, who agree that inclusion and economic growth can be boosted by working together on upskilling programmes. The emphasis on increasing the number of people entering the job market is based on completely different assumptions compared to the work-first orientation that emerged from the Barcelona case-study, where any job was good. ‘The principal tool of making people self-sufficient in Gothenburg is vocational training oriented towards human capital investment’. The aim is not to find the first possible job, but to promote independence, freedom of choice and also ‘to offer the chance of good-quality jobs to those who have obsolete skills’ (Trade unionist, Gothenburg, April 2016).

Civil servants have internalised this emancipating idea of activation and the principle of empowering behind ALMPs, aimed at guaranteeing a skilled and flexible labour force, and making the best use of human capital through the optimised matching of skills and job roles.

In this case, a prerequisite of generalised trust seems a necessary value, forming the foundation for civil servants’ action and the construction of social relationships in this context. It is trust in the institutions, more than interpersonal trust, which allows them to act with confidence in this system.

As we have shown, multiple and interconnected structural, institutional and cultural factors contribute to explaining the results of the comparison presented in this article. Differing institutional organisations, administrative capacities and cultures, welfare regimes, and socio-economic conditions are independent variables that shape the divergent patterns of how conditionality is implemented, or not implemented, in the three contexts examined. As illustrated, economic constraints have been particularly important in shaping the results of the Spanish case. Yet, the agency of the actors involved in the concrete implementation of policies is also an important explanatory piece in this puzzle. The differences highlighted cannot be explained without referring to the interplay between structure and agency. Together with explanations based on differing institutional arrangements, specific attention needs to be given to the informal aspects of relationship structures in society and the variety of strategies, as well as formal and informal practices of the actors involved, their scope for action, and their capacity to intervene and make choices.

This article aimed at contributing to the literatures on social policy delivery and SLB discretionary action. The exploratory research intended to gather preliminary evidence to support the hypothesis that the theory of SC represents an additional analytical lens for the interpretation of the logic behind different modes of operation and the different forms of social relations that motivate civil servants’ choices. In fact, the results of the analysis reveal diversified forms of recognition behind civil servants’ interventions, which in some cases may represent a source of SC. In particular, the Barcelona case study seems to be characteristic by some operating principles of SC based on solidarity, founded on a strong sense of belonging to the local community and group unity, which tends to encourage the non-implementation of income benefit conditionality.

In the Lyon case, instead, a combination of Polanyi's notion of redistribution and some features of SC based on reciprocity seems to emerge, defined by the universalistic values that underpin the actions of civil servants and the ideal of social justice intrinsic to the French social model, which make conditionality a symbolic tool providing access to entitlements that are generally considered as universal rights of citizenship.

Differently, the situation revealed in Gothenburg relies on a sense of generalised institutional trust, confidence in society and reciprocal commitment between the state and its citizens, which represents the historical legacy of the Nordic welfare state and administrative cultural tradition internalised by civil servants, shaping their expectations and actions.

Given the exploratory orientation and the limitations of this study – the difficulty of isolating all potential intervening variable, the comparison based only on Western European cities with common socioeconomic characteristics – this article calls for further research on the relation between SC and SLB in the use of conditionality and on the interaction with other institutional, cultural and socio-economic factors. For instance, new insights might emerge from a comparison with countries with highly individualistic and liberal welfare states, which would make it possible to investigate discretion in contexts where less room for autonomous action is expected.

Yet, this analysis corroborates some key assumptions of SC theory which prove that this is a valuable theoretical framework and can have relevant implications for policy makers as well. On the one hand, the empirical analysis confirms that when a situation of ‘social deficit’ arises, initiatives aimed at reconstituting SC in new forms develop. This hypothesis principally regards solidarity-based forms of SC, such as that illustrated by the Barcelona case study, which relies on a cohesive group. On the contrary, for the creation of SC based on reciprocity, individuals have to be able to form links outside their immediate group, for example with their national community.

On the other hand, despite the fact that over the last few decades studies on identity in contemporary society have shown that individuals have weaker social ties compared to the past and demonstrate more individualistic types of behaviour, the stability of ‘circles of recognition’ is still a relevant factor giving coherence to personal identity, and shaping individual and collective action.

A preliminary version of this article has been selected to be presented at the International Conference in Honour of Alessandro Pizzorno, organised by the Department of Social and Political Sciences of the University of Milan, in collaboration with Fondazione Feltrinelli, in November 2019. The author would like to thank Marino Regini, Gabriele Ballarino and Luciano Fasano for the discussion and comments received during the conference; Alberta Andreotti for her insightful suggestions made on a later version of the article; the editor and the three anonymous referees for their significant and constructive comments.

No potential conflict of interest was reported by the author(s).

Gemma Scalise is assistant professor of Economic Sociology at the University of Milan-Bicocca. Her research interests include labour market and welfare regulation, ideas and institutional change, EU multi-level governance. She is the author of 'The Political Economy of Policy Ideas: The European Strategy of Active Inclusion in Context' (Palgrave, 2020).