A Capability Approach to Understand the Scarring Effects of Unemployment and Job Insecurity: Developing the Research Agenda

Having a poor start in the labour market has a ‘scarring’ effect on future employment and well-being. Indeed, unemployment at any point of the life-course can scar. While there is extensive quantitative research examining scarring effects at the macro- and meso-levels, evidence regarding scarring from the micro-level that provides insights into individual perceptions, values, attitudes and capabilities, and how they shape employment trajectories is lacking. A qualitative approach which avoids the imposition of values and choices onto individuals’ employment trajectories, and accounts more fully for the contextual constraints which shape available options and choices, is argued for. In emphasising people’s substantive freedom of choice, which may be enabled or constrained by contextual conditions, the Capability Approach is proposed as providing a valuable lens to examine complex and insecure labour market transitions. Such an approach stands in contrast to the supply-side focused active labour market policies characteristic of neo-liberal welfare states.


Introduction
Following the 2008 economic recession, unemployment increased across Europe (Eurostat, 2014), with youth unemployment rates a particular concern for governments (European Commission, 2013). Alongside rising unemployment, precarious employment became a prominent labour market status, often leading to later long-term unemployment (McTier and McGregor, 2018). Research shows that future employment, job quality and well-being may be compromised by periods of unemployment and/or a poor start in the labour market (Brandt and Hank, 2014;Gallo et al., 2006), with the latter characterised by low starting wages, inadequate job match, employment instability and/ or repeated unemployment spells (Cockx and Picchio, 2013). While the short-and medium-term implications of unemployment and/or a poor labour market start should not be downplayed (Bradshaw et al., 1983), it is these longer-term implications that are of concern here. There is a well-established knowledge base on the 'scarring effects' of unemployment and/or a poor start in the labour market, characterised by a focus on quantitative measures and headline indicators. Some studies consider more subjective reporting of satisfaction and well-being (Cutler et al., 2015;Helbling and Sacchi, 2014;Strandh et al., 2014) but do not necessarily reflect on individual perceptions, values, attitudes and capabilities, and how they shape employment trajectories. This literature does not provide conceptualisations of scarring that include the heterogeneity of experiences or an understanding of individual choice and agency. The need to develop the evidence base on scarring effects, and the contribution that the Capability Approach (CA), particularly through the application of qualitative techniques, can make, is argued for in this article. The CA emphasises substantive freedom of choice to achieve well-being (Sen, 1999), taking into account 'personal diversities in the possibility of converting primary goods … into achievements of well-being' (Sen, 2003: 27). In differentiating between the availability of choices and the ability to make choices (Bonvin and Orton, 2009), a nuanced consideration of job quality and career prospects is possible. The CA's distinction between 'true' and 'constrained' choices (Egdell and McQuaid, 2016) allows for an examination of complex, insecure and non-linear labour market transitions, looking at why, when and how scarring is experienced. Such nuances undermine arguments characteristic of active labour market policies (ALMPs) in neo-liberal welfare states that 'any work is better than no work' and that sanctions and compulsions to enforce individual choice are justified (Dörre, 2015;Friedli and Stearn, 2015;Wright, 2012Wright, , 2016. However, it can be questioned whether greater real choice in job search prolongs unemployment and deepens disadvantage (Dunn, 2010(Dunn, , 2013Dunn et al., 2014) more than the potential skills and jobs mismatches resulting from taking on any job. Given the stigma attached to unemployment and the long-term negative impact that scarring has, such a focus is crucial and highlights the unrealistic nature of much of the workfare state's demands and compulsions (Beck, 2018).
The arguments made in this article speak to, and extend, the growing qualitative research base that examines state-manufactured precarity and the resistance emerging from that (Dörre, 2015;Greer, 2016), as well as research which uncovers the micro-level experiences of workfarist social and active labour market programmes (Patrick, 2014). To date, the lived experiences of unemployed persons do not receive much attention in social policy research, nor in government and media debates. The ebb and flow of labour market trajectories are not revealed. Thus, there is limited understanding of the complexities of the motivation, powerlessness and agency of unemployed persons (McIntosh and Wright, 2019;Patrick, 2014;Wright, 2016). As demonstrated in this article, application of the CA could serve to address this gap.
The remainder of this article is structured as follows. The knowledge base on scarring effects is reviewed before outlining the very different approach that the CA brings to this subject matter. The potential of the CA is considered before concluding comments are offered.

Researching scarring effects
The key message emerging from much of the work on scarring effects is that experiencing an unemployment spell is associated with increased likelihood of future unemployment (Helbling et al., 2016). The broader labour market conditions in which an individual enters the labour market (Liu et al., 2016;Summerfield and Theodossiou, 2017) and engaging in 'poor' work (e.g. skills under-utilisation, insecure contracts, low-paid work) leave their mark (Baranowska et al., 2011;Gebel, 2010;Mavromaras et al., 2015). Scarring may be experienced into mid and later working-life (Brandt and Hank, 2014), although in some circumstances it is less permanent (Gebel, 2010 Scarring can also occur in terms of career development, income, health and wellbeing. Those who experience early unemployment are more dissatisfied with their career progress (Helbling and Sacchi, 2014) and long-term wage penalties are observed (Arulampalam, 2001;Gregg and Tominey, 2005). Post-unemployment earnings losses are largely permanent and particularly significant for high-wage, older and female workers (Gangl, 2006). In terms of health and well-being, those graduating from any type of education in periods of high unemployment have (alongside lower income) lower life satisfaction, greater rates of obesity, and smoke and drink more later in life (Cutler et al., 2015). Others identify scarring in terms of elevated levels of psychological distress (Daly and Delaney, 2013), with negative effects on mental health exacerbated by double or triple exposure to unemployment (Strandh et al., 2014).
There is heterogeneity of scarring effects between groups, with two standing out as especially significant. First, the duration of unemployment is important. Research establishes that prolonged unemployment decreases the chances of finding employment, although not necessarily starting wages and job stability (Cockx and Picchio, 2013). Others show that in some national contexts, women may be at higher risk of downward mobility the longer the unemployment experience (Evertsson et al., 2016). Second, males tend to be more affected than females, with scarring persistent among males and more short-lived or less pronounced amongst females (Gebel, 2010;Gregg, 2001;Mooi-Reci and Ganzeboom, 2015). In terms of other demographic characteristics, there is less distinction between groups, including by ethnic background (Birkelund et al., 2016) and age groups (Gallo et al., 2006).
Attention is also paid to potentially protective factors such as education or skills. Adverse effects on later unemployment of early-career unemployment for the unskilled, and the reverse for the more skilled, are identified (Burgess et al., 2003). Cutler et al. (2015) highlight that higher levels of education can mitigate negative health and wellbeing effects of completing education during periods of high unemployment. The value that one attaches to work is also relevant, with those with a greater attachment experiencing deeper scars (Laurence, 2015). Experiences of scarring are located within broader (protective) institutional structures. ALMP training, and to a lesser extent ALMP employment participation, mediate scarring effects (Strandh and Nordlund, 2008). Other welfare provision (unemployment benefit systems or strict labour market regulation) also act as a buffer (Gangl, 2006).
Demand-and supply-side factors may be driving mechanisms of scarring effects. However, while there is a good knowledge base on overall scarring effects, and differentiated group outcomes and protective factors, one key methodological challenge is the identification of causal effects (Helbling et al., 2016). Determining the psychological mechanisms through which unemployment affects later well-being presents a challenge (Daly and Delaney, 2013). The emphasis of scarring effects research to date is on headline indicators such as un/employment outcomes and/or quantifiable measures (Arulampalam, 2001;Burgess et al., 2003;Cockx and Picchio, 2013;Mavromaras et al., 2015). Even those reporting on satisfaction and well-being (e.g. Daly and Delaney, 2013;Helbling and Sacchi, 2014) do not explore lived experiences. 1 As such, individual perceptions, values, job quality and capabilities remain under-investigated. This is important because a qualitative judgement of individual situations is crucial. For example, the combination of scarring and a low-paid job may result in cycling in and out of the labour market, involving difficulties with welfare benefits and potential use of food banks (McBride et al., 2018;McTier and McGregor, 2018;Shildrick et al., 2012). In turn, in a dual-earner household with previous spells of unemployment, undertaking a low-paid job may be a valued individual/household strategy for reasons of flexibility that enable care for dependants (Leßmann and Bonvin, 2011) rather than a clear-cut reflection of scarring. Yet the distinctions between a positive or 'true' choice and the absence of choice or availability of 'constrained' choices only, are not considered in the scarring literature. This nuanced level of understanding is not revealed, which could lead to a misclassification of situations as scarring.

The Capability Approach
The CA is concerned, and critically engages, with notions of the individual's 'substantive' freedom of choice to live a life that they value, set against the context of their personal, environmental and social conditions (Sen, 1985(Sen, , 1999(Sen, , 2003. It offers a framework to develop a contextualised understanding of scarring. Initially developed as an approach to welfare-economics by Amartya Sen, it draws attention to individual differences which, as outlined above, are important to understanding the way in which scarring occurs. The CA frames individuals as autonomous persons who should be able to decide what they wish to achieve, based on their own understanding of a 'good life' (Sen, 1985(Sen, , 2003. While the CA and human rights have a common focus on dignity and freedom, the language of capability provides a powerful tool that draws attention to what is involved in securing rights (Nussbaum, 2000;Vizard et al., 2011). (In)equality is understood in terms of the potential to achieve valued functionings. Individuals need to have agency, rather than being told how to think (Sen, 1985).
How do individuals achieve a life that they have reason to value? The CA considers the commodities (the material and non-material resources) to which the individual has access to, alongside their functioning (what they do and are) (Goerne, 2010;Hollywood et al., 2012;Robertson and Egdell, 2018). However, the approach moves beyond resources and functionings to consider an individual's capability-set (all that they can do and be). Capabilities are the combinations of functionings that the individual has and the real (not just formal) opportunity to achieve them (Sen, 2003: 40). In considering an individual's capabilities, the conversion factors (personal, environmental and social conditions) mediating the way that commodities are transformed into functionings are accounted for (Goerne, 2010;Hollywood et al., 2012;Robertson and Egdell, 2018). Finally, and featuring an area in the scarring literature that is not well developed, the CA highlights that an individual's understanding of a 'good life' will shape the choices they make. However, individuals cannot always realise their capabilities, or may be constrained in their choices, because of structural inequalities, low expectations and/or circumstances leading to 'adaptive preferences'. As Nussbaum (1997: 283) details 'we are especially likely to encounter adaptive preferences when we are studying groups that have been persistent victims of discrimination, and who may as a result have internalized a conception of their own unequal worth'. The long-term implications and consequences of adaptive preferences and, ultimately, the absence of choice (Beck, 2018), is not developed in the scarring literature. This is especially important in the context of the psychosocial compulsions inherent in many workfare systems (Friedli and Stearn, 2015).
Sen does not outline in detail how the CA might be applied empirically and maintains that there is no need to define a core set of capabilities. However, Nussbaum (2003) contends that Sen's 'perspective of freedom' stance is too vague and that the complexities of ideas of freedom need to be understood. Nussbaum (2003) argues that an understanding of what core capabilities are is necessary to construct a normative understanding of social justice, and develops a list of 10 central interrelated capabilities: life; bodily health; bodily integrity; senses, imagination and thought; emotions; practical reason; affiliation; other species; play; and control over one's environment (Nussbaum, 1997(Nussbaum, , 2000(Nussbaum, , 2003. These 10 capabilities are in turn categorised into three different types of capability: (1) basic capabilities (the innate equipment necessary in the development of more advanced capabilities); (2) internal capabilities (states of the person themselves); and (3) combined capabilities (internal capabilities combined with external conditions for the exercise of a function) (Nussbaum, 2000: 78-80, 84-85). While the list is abstract and open-ended, leaving room for interpretation depending on the context (Hollywood et al., 2012), some argue against such a list as the CA is operationalised for different purposes (Robeyns, 2005). Indeed, guidance is provided for the development of contextspecific lists (e.g. Burchardt and Vizard, 2007). Therefore, those using the lens of the CA for the study of scarring effects need to decide whether to use a pre-determined or bespoke capability list.

Potential CA contributions to scarring
The CA is an ideal framework for qualitative investigation of scarring due to its emphasis on empowerment, process freedom and choice. McRae (2003aMcRae ( , 2003b demonstrates the need to consider the constraints on the enactment of lifestyle preferences, not confusing voluntary action with genuine or unconstrained choice. Preferences are mediated by circumstances and 'to adduce genuine choice from an outcome (behaviour) is too simple and may produce trivial information' (McRae, 2003a: 586). In this context, authors disenchanted with 'traditional' employment measures propose the CA as an alternative that frames a person as disadvantaged if they do not have the opportunity to achieve what they value (Beck, 2018;Egdell and Graham, 2017;Egdell and McQuaid, 2016;Otto et al., 2015). As such, and of relevance in terms of scarring, whether a transition to work is positive depends on how it is valued by the individual (Bartelheimer et al., 2012). 'True' versus 'constrained' choice scenarios are also revealed, highlighting that agency is usually bounded (Egdell and McQuaid, 2016).
The CA differs from existing supply-side oriented approaches to ALMPs characteristic of neo-liberal welfare states, by revealing the contextual constraints shaping available options and choices (Egdell and Graham, 2017;Egdell and McQuaid, 2016). In operationalising the CA, attention is paid to employment possibilities (not merely openings), social environments, employers and individuals (Leßmann and Bonvin, 2011). The CA is therefore a stark contrast to the neo-liberal underpinnings of workfare states, which purport to raise self-initiative (Dörre, 2015) while attaching value only to income-generating activities, using sanctions and compulsions to compel and enforce individual choice (Friedli and Stearn, 2015;Wright, 2012Wright, , 2016. Such contradictions make agentic and individually valued decision-making impossible, and problematically frame paid work as a panacea (McTier and McGregor, 2018;Shildrick et al., 2012). This is a key contribution of the CA, in that it highlights the role of demand-side issues in an individual's labour market experiences, while ALMPs to date have largely been supply-side focused (Egdell and Graham, 2017;Egdell and McQuaid, 2016).
The development of a qualitative methodology to research scarring effects is necessary to investigate the contextual causes and effects of scarring and to consider individual values and preferences. Such a differentiated understanding of scarring over the life-course leads to a revised conceptualisation which does not exclude scenarios in which positive choices are made that result in individuals undertaking activities that are meaningful to them, but which may not be positively evaluated by the workfare state. In turn, qualitative investigations would also highlight the extent of scarring that occurs as a result of individuals not being in control of their own choices and agency (Sen, 1985). The CA moves away from ALMP approaches that frame paid work as a cure-all and as a way of achieving equality and social citizenship. The CA speaks to calls within the social policy literature for a more nuanced understanding of agency, alongside more variety in the definition of successful outcomes, when examining experiences of workfare states (Wright, 2012(Wright, , 2016. This is not to say that the use of a qualitative, CA-based investigation of scarring is straightforward. Researchers first need to decide whether to apply a pre-determined capability list (e.g. Nussbaum, 1997Nussbaum, , 2000Nussbaum, , 2003 or develop their own context-specific list. Second, finding data that can address the complexities of the approach, with regards to employment opportunities in particular, is difficult (Chiappero-Martinetti et al., 2015). The value of the CA is that it invites researchers to conceptualise and examine labour market transitions beyond quantifiable measures. While it is possible to observe outcomes (functionings), observing an individual's freedoms (capabilities) is much harder (Leßmann and Bonvin, 2011;Verd and Andreu, 2011). Capability variables are often lacking in existing large-scale datasets (Chiappero-Martinetti et al., 2015). There may be indicators of job quality and satisfaction, but in-depth, qualitative work is required to understand differentiated forms of scarring, to establish causal relationships and options available to scarred individuals, and to distinguish between 'true ' and 'no'/'constrained' choices. Finally, as Chiappero-Martinetti et al. (2015) detail in relation to undertaking capability framed research with young people, participants may not be used to expressing their views and may find the open questions required in research applying the CA hard to answer. Using mixed-method approaches when operationalising the CA could be useful with qualitative methods complementing quantitative investigations to provide in-depth insights into the quality of opportunities and job satisfaction, as well as individual experiences of scarring and motivations (Chiappero-Martinetti et al., 2015;Leßmann and Bonvin, 2011;Lugo, 2007).
Qualitative CA investigations of scarring bring empirical and theoretical benefits and innovation. It is empirically useful to understand scarring from the perspective of those who experience unemployment and/or poor work, particularly at the start of working life. Theoretically, there are two potential innovations. First, the differentiated approach could lead to a conceptualisation of scarring effects that includes different experiences and, potentially, forms of scarring. Second, the issue of long-term scarring and life-time effects of early problems of transitioning into the labour market will allow a development of the CA in its focus on choice and lack of attention to 'no'/'constrained' choice scenarios.

Conclusions
In conclusion, this article details that while there is an extensive research base examining the scarring effects of having a poor start in the labour market, and/or experiencing periods of unemployment, at the macro-and meso-levels, evidence from the micro-level is lacking. Quantitative research on scarring effects does not provide insights into individual perceptions, values, attitudes and capabilities, and how they shape the trajectories of 'scarred' individuals. By resorting to quantitative approaches to study scarring effects, we risk falling in line with rigid neo-liberal policies which attach value only to incomegenerating activities (Friedli and Stearn, 2015). Such policies do not allow for individually controlled decision-making, nor do they pay attention to the value attached by the individual to the outcomes they have, or have not, achieved. They do not consider the contextual constraints informing choices, values and motivations.
This article argues for the need to develop the qualitative evidence base on scarring effects and considers the contribution of the CA in applying these methods. CA-informed qualitative approaches are capable of reflecting individuality, choice and values; addressing questions not considered in the scarring literature to date. For example, do individuals who have experienced unemployment feel this has affected their labour market behaviour? Do they perceive that they subsequently experienced poor employment and well-being outcomes? How, and in what ways, do these experiences shape job quality, progression opportunities, and capabilities to engage in valuable work? Does having a poor start in the labour market or experiencing past unemployment, shape an individual's current experience of, and attitudes towards, their own labour market status and career expectations? How are individuals who have a poor start in the labour market able to carve out progressive transitions, even in unpromising circumstances?
While this is not a straightforward proposition, the value of this approach is that in emphasising the individual's substantive freedom of choice within the context of external and internal constraints, it stands in direct contrast to the neo-liberal underpinnings of workfare states. It extends the qualitative research base examining state-manufactured precarity and workfarist policies by focusing on the motivations, powerlessness and agency of 'scarred' individuals. The CA provides a valuable lens to examine the complex, insecure and non-linear labour market transitions, looking at why, when and how scarring is experienced. By emphasising individual choice and the backdrop against which choices are made, it provides a nuanced level of understanding that can reveal scenarios in which positive choices are made, but which may not be positively evaluated by the workfare state, leading to a potential mis-classification of situations as scarring.
transitions on labour market participation, and individual and social aspects of unemployment and underemployment, in particular for those Not in Education, Employment and Training (NEET) and individuals with multiple and complex barriers to employment. Her research has been funded by, amongst others, the Economic and Social Research Council (ESRC), the Independent Social Research Foundation, the European Commission, ACAS, and the Government Equality Office.

Date submitted January 2019
Date accepted January 2020