What makes adults choose to learn: Factors that stimulate or prevent adults from learning

Adult learning policies need to be based on an understanding of the inequalities in the uptake and benefits of learning and why adults might not participate. This needs to go beyond a mere insight in barriers that, once removed, do no longer provide a reason for adults not to participate. This article aims to delph deeper in understanding what makes adults choose to learn. It starts by applying a capability approach perspective to adult learning to evaluate whether adults have freedom to value learning and, whether they can effectively turn this freedom into learning. This conceptual framing puts certain concepts in the spotlight, namely, ‘agency’, ‘conversion factors’ and ‘benefits of learning’, which were further explored through a narrative literature review analysing 109 articles. This resulted in an exploration of these concepts and their interplay feeding into a conceptual model, opening new perspectives for evaluating whether adults have equal opportunities to value adult learning and turn their willingness into actual learning. This model supports future empirical studies aimed to understand participation and non-participation of adults in learning that can in turn feed policy makers with better insights and tools to develop interventions actually provide the right encouragements for adults to learn.


Introduction
To design and develop policies and programmes that respond to adult learners' needs, it is essential to understand the dynamics that underlie adults' participation and nonparticipation in learning.If not fully captured, policies and programmes are not able to engage those adults that need learning the most.Evaluations and studies on adult learning policies consistently report that specifically disadvantaged adult learners can hardly be re-engaged in learning, despite tailoring the approach to solving group-specific barriers.For instance, the Swedish Adult Education Initiative from 1997 to 2002 (Rubenson, 2001) educated half a million adults, but did not change the pattern of nonparticipation among the group of the most difficult to reach adults (Antikainen, 2005).As a result of not addressing effectively persistent non-participation of specific groups of vulnerable adults, adult learning is increasingly perceived as widening differences in opportunities instead of bridging them (Kocór & Worek, 2017), pointing to the often referred Matthew effect of cumulative advantage and cumulative disadvantage (Boeren, 2009;2017;Marcaletti et al., 2018).To prevent these negative effects of adult learning from occurring, and for developing evidence-based and effective policies, it is essential to improve our understanding what stimulates or prevents adults to learn.
The non-participation of adults in learning is primarily analysed in terms of the existence of barriers (see, for instance: Cross, 1981;Roosmaa & Saar, 2017).This analysis concentrating on barriers presupposes that all adults have a natural desire to learn and that if barriers are removed, all adults are provided with equal opportunities to engage in learning and development (Ahl, 2006).Barriers are perceived as impeding individuals that want to engage in learning because they realise they will fall behind their competitors if they do not learn (Regmi, 2015).In terms of policies, this results in putting more responsibility for success and failure on the individual (Boyadjieva & Ilieva-Trichkova, 2021).This results in the policy-theory which states that when policies remove barriers to participation, and lower thresholds to participation, individuals will decide learning is good for them and participate in adult learning programmes.By this barriers-related paradigm, governments can argue that policies provide everyone with the same opportunities to participate but some individuals simply choose not to do so.Hence, in this paradigm, there is no need for additional policy action as individuals make a deliberate choice not to learn.
To avoid this, and support design of better policies and programmes, wider structural barriers need to be taken into account explaining inequalities in the uptake of learning and how people may benefit from learning (Eynon & Malmberg, 2021).Therefore, this article, proposes an alternative perspective to analyse non-participation in learning compared to the approach that presupposes that adults have a natural desire to learn that will emerge if barriers are removed.This alternative perspective is based on the capability approach as promoted by Sen (1993Sen ( , 2009b)).The capability approach provides a framework to evaluate social phenomena.At a micro, individual level, the capability approach asks what people are able to do and what person they are able to be.At a macro, policy, level, the capability approach asks whether the institutions, practices and policies focus on people's capabilities (their opportunities to do what they value) and to what extent they offer equal opportunities.The capability approach aspires to look beyond the effect of social policies in terms of participation, and explores the black box on how policies play out for different people, increasing their capabilities in terms of freedom of what they choose as valuable to achieve.The actual possibilities (or capabilities) instead of outcomes (achievements) are the lens through which policies are designed and evaluated (Lewis & Giullari, 2005).The capability approach provides a framework to evaluate social policies looking at how they enable individuals to turn available resources into capabilities and achievements they choose to value (Robeyns, 2005).
The capability approach is widely applied to evaluate various social phenomena.It is, however, not so widely applied to evaluate adult learning.While the role of education for welfare development is discussed in the capability approach, usually, 'education' is seen as a capability that allows different individuals to pursue beings and doings that they see as valuable (Nussbaum, 2013;see discussion in: Unterhalter, 2003).In addition, Lanzi sees education from the development perspective, defining its value 'by the sum of instrumental values (wages, test scores, certificates, etc..), intrinsic values (achievements in agency, autonomy and well-being) and positional values (established social relations, access to positional goods, etc..)' (Lanzi, 2007, pp. 425-426).Hence, the capability approach is in earlier research applied more at a system level and less so on understanding and evaluating individual freedoms to choose education, something that is essential for understanding adults' participation in adult learning, as this is more subject to individual choice than participation of young people in initial education.
From this perspective, this article focuses on the question: what makes adults choose to learn?Alternatively, to formulate the question more precisely: what makes adults intentionally engage in organised and structured learning or development activities that aim to sustain or improve their existing situation (economically, socially and individually well-being)?Further conceptualising our research question in a capability approach suggests that we should not look at participation in learning alone or lowering specific barriers as the measures for a successful, and equitable adult learning system.Through answering this question, and exploring factors stimulating and hampering learning, this article may provide input for a conceptual model that is further developed through future empirical studies providing a better base for policy development in adult learning understanding how policy initiatives can better stimulate adults to learn.
While in principle all forms of adult learning is subject to our question, irrespective of its form (i.e.formal, non-formal and informal), or purpose (for leisure, work or personal development), to understand better the choice aspect, in this article the focus is on learning for which the adults make a choice to engage with.Hence, the learning we focus on here is intentional from the learner's perspective, organised and structured, for instance, in a formal or non-formal setting (Boeren, 2017;Cedefop, 2014).In the Dutch context, adult learning is conceptualised as lifelong development (Ministerie van Sociale Zaken en Werkgelegenheid, 2020), being '[t]hroughout life, the (pro)active development of qualities based on motives and possibilities for a sustainable contribution to society, one's own work environment, health and happiness, for now and for the future' (translation authors) (Kuijpers et al., 2019;Kuijpers & Draaisma, 2020, p. 1).Learning as development conceptually points to that any learning should result in some form of change or application of what is learned, in whatever form.
Theoretical framework: A capability approach perspective to adult learning Within the capability approach, there are a number of key notions that are relevant for conceptually exploring adults relations to learning.These concern functionings and capabilities, human agency and conversion factors.The capability approach looks at what people can do (capabilities) rather than what they actually do (functionings), together with substantive freedom of choice, taking into account external factors and personal characteristics (Egdell & Graham, 2017;Sen, 2009a).This distinction between functionings and capabilities is a key notion in the capability approach allowing assessing whether people have the freedom to do the things that person has reason to value (Robeyns, 2017;Sen, 2009b).Another key underlying notion, besides the notions of capabilities and functionings, and in fact the starting point for the capability approach is the notion of human agency (Sen, 1993), seeing individuals as autonomous persons who should be able to decide what they wish to achieve based on their own understanding of a 'good life' rather than one imposed upon them (Egdell & Graham, 2017).In the end, individuals make choices from their capability set and transfer resources into achievements they choose to value.But even before that, people can have aspirations as capabilities whereby only some might turn into 'real' aspirations (functionings), because they are indeed considered valuable to be pursued.As clearly expressed by Hart (2016, p. 336)), 'The kinds of aspirations we have influence the kinds of capabilities for which we strive'.A final key notion is that of conversion factors.Persons have all kind of abilities to convert resources into functionings.These abilities are referred to as conversion factors: the factors which determine the degree to which a person can transform a resource into a functioning (Robeyns, 2017;Sen, 1992).A resource, or commodity, might be a material object (for instance, a bicycle); a person might have the ability to turn this resource into the functioning (in our example of a bicycle, to be mobile); the conversion factor is the extent to which this person is indeed able to do so (in our example, someone that learned how to ride a bicycle has a high conversion factor, whereas a person that never learned how to ride a bicycle has a low conversion factor) (Robeyns, 2017).Conversion factors influence how a person can be or is free to convert the characteristics of the resources into a functioning.
From this capability approach perspective, asking the question what makes adults intentionally act towards engaging in organised and structured learning or development activities, requires us to consider whether adults find adult learning a valuable thing to do and if so, whether they have the resources and opportunities to turn this capability into actual learning.It concerns how individuals can use resources (or commodities) to enhance their 'capability set', or combinations of potential functionings, choose what they find valuable to achieve as functionings and actually whether they have the abilities to achieve those functionings.The focus is on the capability set, the choice of what that person find valuable rather than the resources that person has access to (Walker, 2005).
Positioning adult learning in a capability approach perspective leads therefore to further exploring a number of interrelated concepts and questions: 1. Firstly, do individuals have adult learning as a capability within their wider capability set and do they have a possibility to choose adult learning as a valuable life choice?This relates to whether individuals can be expected to have aspirations towards adult learning and whether individuals are provided with equal opportunities to value adult learning.This is closely linked to whether individuals perform agency towards adult learning.2. Secondly, do individuals have the abilities and support to convert resources into the actual adult learning (summarised as 'conversion ability')?Here, the conversion factors that relate to adult learning need to be further explored.3. Lastly, do individuals have equal opportunities to benefit from adult learning participation?Does adult learning and lifelong development lead to the same outcomes and results for individuals?
Hence, we need to investigate the underlying dynamics and structures that either stimulate or prevent a person to aspire learning and under which conditions that person is able to convert this aspiration in actual learning.Reflecting from a capability approach perspective on the question 'what makes adults act towards engaging in learning' brought us to consider the interplay between agency and the context in which an individual is situated and the results and benefits that a person yields from engagement with learning.

Method
A narrative literature review aims at theory development, proposing a novel conceptualisation or theory regarding a specific phenomenon covering diverse disciplines (Baumeister & Leary, 1997;Hall et al., 2021;Snyder, 2019).The literature is explored through an article title search in Web of Science using combinations of search terms related to adult, educat*, learn*, empower*, agenc*, self-efficac*, motivation, autonom*, aspirations, activation, self-directedness, self-determination, empowerment, conversion, capital, welfare, barrier*, capability*.In total, 459 articles were identified.When analysing the articles and coding them in Atlas TI, in the end, 109 articles were identified containing codes related to capability approach, agency, conversion factors, and benefits and results.Articles that were not considered relevant mostly concerned articles in which 'agency' referred to institutions, articles that specifically focused on certain types of adult learning delivery (for instance, online and blended learning, higher education), articles that focused on learning in a specific context (for instance, linked to medical professions), and finally, articles that focused more on initial education and young people instead of adults.
The articles cover both single concepts and a combination of concepts.Of the 109 articles, 54 covered only one of the concepts, 37 covered two, 15 covered three and only three articles covered all concepts (Boyadjieva & Ilieva-Trichkova, 2018;Field & Lynch, 2015;Rubenson & Desjardins, 2009).The following table (Table 1) provides a summative overview of the identified articles and the coverage of the key concepts by indicating how many articles (and the share of the total number of articles) covered the key concepts (end of the row) and how many articles covered two different key concepts.What can be seen is that agency and conversion factors are the most occurring combination of two concepts in the 109 articles (identified in 36% of the articles).Furthermore, in terms of combinations of three concepts, twelve articles (11%) covered the combination of agency, conversion factors and benefits and results (Alkire, 2005;Allmendinger et al., 2011;Booker et al., 2021;Boyadjieva & Ilieva-Trichkova, 2018;Evans et al., 2013;Eynon & Malmberg, 2021;Field & Lynch, 2015;Hachem, 2022;Hammond & Feinstein, 2005;Leung & Liu, 2011;Rubenson & Desjardins, 2009;Yamashita et al., 2022).The outcomes of the analysis of all identified articles are presented in the following sections.

The concept of 'agency'
In the capability approach, human agency is associated with a person's ability to turn resources into functionings.As indicated earlier, human agency forms the starting point for the capability approach.There are however also other approaches to the concept of human agency.The narrative literature review, linked to adult learning identified eleven articles that conceptually discuss 'agency' in a capability approach perspective among the 84 articles that discussed 'agency'.In the following sections, social-cognitive, socio-psychological and sociological perspectives on agency are discussed to identify aspects that can be taken on board in a further conceptualisation of agency related to adult learning.
A social-cognitive perspective on agency.From a social-cognitive theory perspective (Chen, 2006), Bandura defined human agency as 'a combination of human capacity and potential that assists a person to exercise some control over the nature and quality of his or her own life, including aspects such as forethought, self-regulation of motivation, affect, and action through self-influence, self-awareness, meaning, and purpose in life.As these agentic variables interact and interplay as a whole, they shape one's direction in life and its associated course of action' (Bandura, 2001;Chen, 2006, p. 131).The core belief in one's own self-efficacy is the foundation of human motivation, performance accomplishments and emotional well-being (Bandura, 2010).Individuals with low self-efficacy tend to back away from daunting tasks as they often see these tasks as threats.Furthermore, they set lower targets and have a weak commitment to their set goals and focus more on selfdoubts, their deficiencies, the consequences of the failure and give up rather than considering ways to overcome challenges.Individuals with higher self-efficacy levels often do the opposite (Bandura, 2010;Calaguas & Consunji, 2022).
This social-cognitive approach emphasises the role of individuals as intentionally influencing their life circumstances and being self-organising, proactive, self-regulating and self-reflecting (Brady & Gilligan, 2020).This approach to human agency emphasises the person's internal determinants (beliefs of self-efficacy), but places these in an external environment where the person's choices and behaviour lead to confirmation, boosting again a person's self-efficacy beliefs (Alkire, 2005;Bandura, 2001).Closely related to self-efficacy concepts is the concept of 'locus of control' which can be described as 'a generalized attitude, belief, or expectancy regarding the nature of the causal relationship between one's own behaviour and its consequences' (Cobb-Clark, 2015, p. 1;Rotter, 1966, p. 2).Both the selfefficacy and the locus of control concepts refer in their own way to perceived behavioural control (Ajzen, 2002), while the locus of control concept adds to the self-efficacy concept the perceived internal and external control related to outcomes (Cobb-Clark, 2015; Kormanik & Rocco, 2009).The social-cognitive perspective on agency, is primarily oriented to the internal control of a person and emphasises less the social embeddedness as condition for agency.
A socio-psychological perspective on agency.A more psychologically oriented theoretical perspective on human agency, that situates individual agency in a social context, is the selfdetermination theory (Ryan & Deci, 2000).This theory was developed against the background of the question why some people act, and others don't and looks at conditions that foster versus undermine positive human potentials.Understanding the causes of human behaviour supports us in designing social environments that optimise people's 'development, performance, and well-being' (Ryan & Deci, 2000, p. 68).The theory sees a person (self) in interaction with the environment and in this interaction, the self is constantly evolving and forming an inner and consistent representation of the self (Kellenberg et al., 2017, p. 24).The self-determination theory, through an empirical, psychological, inductive approach, identifies needs that form the basis for self-motivation, growth, constructive social development and personal well-being.Self-determination consists of a continuum running from a state of amotivational (lacking any intention to act), to extrinsic motivational states (which combine external and internal regulatory processes), to finally an intrinsic motivational state.'Amotivation results from not valuing an activity, not feeling competent to do it, or not expecting it to yield a desired outcome' (Ryan & Deci, 2000, p. 72).To be placed in a more motivational (either extrinsic or intrinsic) state can be supported firstly, through seeing behaviours of significant others to whom a person feels (or wants to feel) attached or related (relatedness), secondly, through being presented tasks at the right competence level (competent), and thirdly, through feeling autonomous in that the person is doing the task on its own (autonomy).
The self-determination theory describes an autonomous person as 'his or her behaviour is experienced as willingly enacted and when he or she fully endorses the actions in which he or she is engaged and/or the values expressed by them.People are therefore most autonomous when they act in accord with their authentic interests or integrated values and desires' (Alkire, 2005, p. 242).Alkire argues that this concept of autonomy comes close to Sen's concept of agency, 'because it focuses on capabilities that the person values (in contrast to self-efficacy, which identifies capabilities a person understands herself to have-whether or not she values them)' (Alkire, 2005, p. 242).The basic psychological needs, competence, relatedness and autonomy 'must be satisfied across the lifespan for an individual to experience an ongoing sense of integrity and well-being or "eudaimonia"' (Ryan & Deci, 2000, p. 75).In the end, having ownership of your decisions, feeling empowered contributes to overall well-being.
The self-determination theory offers support to design social environments that support the internalisation of motivation.It is therefore a useful theory to apply in the design of adult learning programmes (Kellenberg et al., 2017) and guidance approaches.The self-determination theory is associated with the classical concept of andragogy, emphasising self-directed and intrinsically motivated learning (Knowles, 1984;Wehmeyer et al., 2018).The self-determination theory reveals psychological and social conditions that explain why an agent in a vulnerable position, against the odds, might act and start learning.An agent in a favourable position, having all sociological conditions in place, might not act on the other hand, and refrain from learning.
A sociological perspective on agency.Other agency-perspectives, while positioning agency in wider social contexts, also emphasise the boundedness of agency.In Evans' concept of bounded agency, actors have a past and imagined future possibilities, 'which guide and shape actions in the present, together with subjective perceptions of the structures they have to negotiate, the social landscapes that affect how they act.Bounded agency is socially situated agency, influenced but not determined by environments and emphasizing internalised frames of reference as well as external actions' (Evans, 2007, p. 93).Agencies can differ in their power to act, and this is also influenced by the environment the agent is in.In fact, '(a)gency in adult life operates through engagements in and through the social world; it is exercised through the environments and institutional practices of everyday life in changing social landscapes' (Biasin & Evans, 2019, p. 49).With this concept of bounded agency, Evans is close to wider agency accounts, 'bounding' agency through a life course perspective (Biesta & Tedder, 2007;Emirbayer & Mische, 1998).The bounded agency model points to the idea that agency is not absolute: it is bounded by welfare state regimes, agents' past and imaginary futures, social environment and the accompanying options on which an individual could act.
While reference is made to the capability approach to deepen the interaction of the individual to its context, the bounded agency models make a strong case to situate the individual agency in a wider socially constructed context that determines the potential decisions to take and the structures available to allow individuals to overcome barriers (Radovan, 2012).Therefore, as suggested by the bounded agency models, participation in learning is not as voluntary an act for a person to choose, as it is often regarded, but the 'agency to freely choose to participate could be bound by structural conditions'.(Tikkanen & Nissinen, 2018, p. 617).In this sense, when comparing a person in a vulnerable position with someone in an advantageous position towards learning, even within the same welfare state regime, the agency of the vulnerable person is confined firstly by a less favourable past, secondly, by less favourable present circumstances, and thirdly, by an imaginary future that shows less opportunities that the person feels applicable to its situation.Surmounting these unfavourable conditions and starting to learn would mean taking risks that are incomparable to what a person in a favourable position would have to take.
Agency revisited.Ultimately, a person's agency (self-determination) is determined by relatedness, competence and autonomy or, when referring to self-efficacy, by intentionality, forethought, self-reactiveness and self-reflectiveness.All these aspects, however, are the result of the situatedness and boundedness of the agent in question: whether an agent is self-determined, or has self-efficacy towards learning depends on the social context (namely, having significant others that paved the way towards learning or seeing self-efficacy beliefs confirmed by external determinants); their own competence to learn (namely, prior experiences) and the possessed autonomy (namely, seeing learning as something valuable).Approached from a capability approach perspective, agency is in all theories linked to a wider social context.The wider context can refer to 'relatedness' (Ryan & Deci, 2000) or some form of 'boundedness' (Boeren, 2017;Evans, 2007;Rubenson & Desjardins, 2009).When combining the sociological and social-cognitive approaches to human agency, the following aspects are closely linked to individual agency, in the sense that these aspects explain why a person is in a position to act or not: 1. Aspiration and forethought: This relates to the extent to which an individual is able to set future goals, develops aspirations about well-being and acts anticipating on future events and (changing) circumstances.In the literature, this aspect is well covered in the self-efficacy literature, motivation literature and literature on (Alkire, 2005;Bandura, 2001;Hammond & Feinstein, 2005;Hart, 2016;Ibrahim & Alkire, 2007).As argued by Hart (2016), aspirations are future-oriented, driven by conscious and unconscious motivations and they are indicative of an individual commitment towards a particular trajectory or end point.Bandura includes the concept of intention, being a representation of a future course of action to be performed (Bandura, 2001).

Competency and confidence (autonomy):
To be able to act relies on having competences and having confidence in using them.This comes close to the intrinsic motivation concept in the self-determination theory looking at autonomy and competence (Ryan & Deci, 2000).It also links to the Bandura's aspect of 'selfreactiveness (self-regulation)' being the property where intentions are converted into actions (Bandura, 2001;Boomkens et al., 2019).

Reflectiveness:
Being an agent also relates to being able to reflect on own actions, take ownership of your acts and act intentional.These aspects are all covered in the self-efficacy theory (self-reflectiveness) (Bandura, 2001).'Through reflective selfconsciousness, people evaluate their motivation, values, and the meaning of their life pursuits.It is at this higher level of self-reflectiveness that individuals address conflicts in motivational inducements and choose to act in favor of one over another' (Bandura, 2001, p. 10).

The concept of 'conversion factors'
As indicated earlier, 'conversion factors' are a key notion in the capability approach.
Persons have all kind of abilities to convert resources, commodities or inputs for capabilities (Hvinden & Halvorsen, 2018), into functionings (in other words, achieved beings and doings).These abilities are referred to as conversion factors (Robeyns, 2000): the factors which determine the degree to which a person can convert or transform a resource into a functioning.Conversion factors influence how a person can be or is free to convert the characteristics of the resources into a functioning.Linked to adult learning, the narrative literature review identified 54 articles that discuss conversion factors and associated concepts.Only ten articles explicitly elaborated on conversion factors from a capability approach perspective.There are different ways conversion factors can be clustered.Sen proposed five factors: personal heterogeneities, distributions within the family, differences in relational positioning, varieties in social climate and environmental diversities (Hvinden & Halvorsen, 2018;Sen, 2009a).Robeyns (2017) clusters them in three broad groups.Firstly, personal conversion factors relate to a person's personal (physical/mental) abilities, knowledge, skills and competences.Secondly, social conversion factors stem from the society in which a person lives and relates to the social policies, social norms, practices and hierarchies.Thirdly, environmental conversion factors include physical or built environment in which a person lives.Other scholars provide a more nuanced set of conversion factors, specifically developed to analyse youth unemployment and job insecurity transitions amongst young adults, namely: (1) institutional, (2) social, (3) familial, (4) economic, (5) cultural, (6) political and ( 7) personal (Bøhler, 2019).Taking Bøhler's categorisation as basis and including suggestions from other authors leads to the following set of conversion factors that are applicable to the topic of adult learning: 1. Institutional conversion factors (Boeren et al., 2012;Bøhler, 2019): these ad dress the impact of institutions on a person's sense of capability.This concerns the ways in which schools, unemployment agency offices or sport clubs might alter a person's ability to live a life according to his or her visions and values.2. Social conversion factors (Allmendinger et al., 2011;Bøhler, 2019;Field & Lynch, 2015): these concern the influence of friends, friend groups or social networks on personal capability.(Allmendinger et al., 2011;Boeren et al., 2012;Bøhler, 2019;Chu, 2010;Field & Lynch, 2015): these associate capability with relations within the family.4. Economic conversion factors (Allmendinger et al., 2011;Bøhler, 2019;Tikkanen & Nissinen, 2018): these address an individual's income and wealth in relation to capability. 5. Cultural conversion factors (Allmendinger et al., 2011;Bøhler, 2019): these capture capability's relationships to cultural norms, practices and customs.6. Political conversion factors (Bøhler, 2019;Field & Lynch, 2015): these designate how larger changes in the politics or economy of a nation impact on a person's possibilities for capability.7. Personal conversion factors (Bøhler, 2019;Hart, 2016;Tikkanen & Nissinen, 2018): these address how the active agency of an individual might improve his or her capability, including choices about voluntary versus paid work in relation to an individual's well-being and/or human capital.8. Employment conversion factors (Boeren et al., 2012;Tikkanen & Nissinen, 2018 In addition to the list of Bøhler): these capture work-related factors (job position, company size, working time, work autonomy, sector, skills use at work, skills mismatch and qualifications mismatch).

Familial conversion factors
In addition to those, life events or disjunctures that trigger participation in learning of (older) learners are identified in different studies (Feinstein & Hammond, 2004;Hachem, 2022;Jarvis, 2012).These can also be interpreted as conversion factors turning capabilities into functionings.These life events, for instance, include retirement, illness, loss of spouse, empty nest, leaving caregiving tasks or becoming (financially) independent.Life events can also be described in terms of 'activating event', triggering new attitudes and motivations to learn (Biasin & Evans, 2019;Jarvis, 2012).

The concept of 'benefits and results'
Linked to adult learning, the narrative literature review identified 26 articles that discuss benefits and results of adult learning.Only six articles explicitly elaborated on benefits and results from a capability approach perspective.
Besides someone's agentic power and ability to convert capabilities into functionings (actual learning), another key factor for a successful and equitable adult learning system is that learning and development lead to some form of improvement and benefits, which can also be phrased as preventing negative consequences of not learning.If taking benefits of learning by individuals is not assured, the learning might easily leads to deception and non-continuation of learning.Adult learning is associated with three functions, each having associated benefits of learning; being firstly, economic progress and development (linked to human capital), secondly, personal development and fulfilment (linked to identity capital), and thirdly, social inclusiveness and democratic activity (social capital) (Aspin & Chapman, 2000;Cocquyt et al., 2019;Schuller et al., 2004).Regarding the outcomes, benefits and results of (adult) learning, many (empirical) studies and literature reviews have been conducted and a wide range of potential benefits are referred to (Allmendinger et al., 2011;Balatti & Falk, 2002;Leung & Liu, 2011;Ruhose et al., 2019;Schuller et al., 2004;Yamashita et al., 2022).Based on these studies, the following list of benefits and outcomes of adult learning and development can be provided: 1. Economic position and progress (Allmendinger et al., 2011;Balatti & Falk, 2002;Schuller & Desjardins, 2010): this relates to employment and quality of working life (e.g.finding paid employment), improved job security, job mobility, increased wages and reduced unemployment risks, increased job prestige, pro gression in career.2. Personal development and fulfilment (linked to identity capital) (Feinstein & Hammond, 2004;Hammond & Feinstein, 2005;Leung & Liu, 2011;Schuller & Desjardins, 2010): this relates to gained skills, gained competences, gained qualifications, gained opportunities to further learning and development, gained agency (aspirations, competency, confidence and reflectiveness), increased au tonomy and self-efficacy, but also personal safety (e.g.decreasing crime activity, applying conflict resolution) (Balatti & Falk, 2002).3. Social inclusiveness and democratic activity (social capital) (Leung & Liu, 2011;Ruhose et al., 2019;Schuller & Desjardins, 2010): this relates to improved social support, familiar support, increased participation in society, reduced crime activity, increased cultural capital and increased social capital such as membership in civic groups, political interest, voting, social networks and trust.4. Overall well-being and health (Balatti & Falk, 2002;Feinstein & Hammond, 2004;Leung & Liu, 2011;Schuller & Desjardins, 2010): this relates to improved physical well-being, mental well-being, happiness, better health and life satisfaction.(Balatti & Falk, 2002): this relates to obtained skills, competences, qualifications, acquiring credentials, progression in learning, access to courses and an increased interest in learning.

Connecting agency, conversion factors and 'benefits and results'
The interplay between agency, conversion factors and outcomes of adult leaning can be positioned in a wider social context.From the theoretical explorations, it becomes clear that the social context and environment function as enablers for an agent to convert a desire, through using resources, into learning.The social context and environment determine the likelihood an agent values learning and impact the conversion power of an individual.Without entering too deeply in a much debated topic on structure and agency, there is a clear dependency between an individual's choices and the social structure the individual belongs to (Hitlin & Elder, 2007).This social structure, context and environment might not completely determine a persons actions, but they do inform them.Hence, while certain persons are confronted with a disadvantaged proposition to learn, answers on how to increase their chances to engage in learning can probably be found in the direct social structure, context and environment.This conducive environment for developing and exploiting agency towards learning is a concept that might both influence the agentic ability and the conversion ability, and hence the actual participation in, and outcomes of adult learning.The concept of environment recognises that individual's actions are always embedded in a larger context in which that context both determines the individual's choices and in which the individual's choices in turn determine the context.This concept is hence situated in a wide sociological discussion on 'structure and agency', while avoiding social determinism (Archer, 2003).Figure 1 summarises the conceptual explorations of agency, conversion and benefits of adult learning based on the capability approach and positioned in a wider social context and environment.Each of the three concept discussed (agency, conversion and benefit) can be both the precondition and the outcome of the other as illustrated in Figure 1.Individual agency is both a condition to take action towards learning, but learning in itself impacts the agency of the learner (Hammond & Feinstein, 2005;Rubenson & Desjardins, 2009).First, perceptions of achievement in adult education increase self-efficacy.Secondly, adult education leads to more challenging occupations, which build self-efficacy.Thirdly, resistance to participation in adult education is reduced as self-efficacy increases.Finally, learning on the job can build self-efficacy, and although participation in employerprovided training courses does not appear to play an important role, it reflects engagement in occupations where the value of learning is recognised (Hammond & Feinstein, 2005).In addition, learning taking place in social groups, affects socialisation and agency through feeling related to others (Ryan & Deci, 2000).Increased self-efficacy and agency is in turn also related to higher levels of well-being (Ryff & Keyes, 1995).The outcomes of adult learning hence also refer back to agency-enhancement and support adults, in terms of the capability approach, to expand their perspectives on the life they choose to value.When adults participate in learning, they feel more supported to review what they value in life and this brings benefits beyond individual well-being and economy and a narrow idea concerning the 'return on investment' (Allmendinger et al., 2011;Boyadjieva & Ilieva-Trichkova, 2018).Hence, agency as key outcome of adult learning drives also reaching other societal and economic outcomes.The concepts of agency and conversion factors have a similar relationship (being both a precondition and outcome).Some authors position agency as a result of a combination of conversion factors (Robeyns, 2000), some see a more fundamental role of individual agency (Bandura, 2001;Meyers, 2017) and others see that conversion factors and agency are in a continuous interplay (Evans, 2007;Hart, 2016).Agency is shaped in different ways by the conversion factors.Firstly, the conversion factors influence the set of capabilities from which a person can freely choose to value one.Secondly, these conversion factors influence whether this person can realistically turn the capability into a functioning (Hvinden & Halvorsen, 2018)).Ultimately, given availability of resources, agency, conversion factors and outcomes of adult leaning together determine the likelihood that adults can benefit from adult learning.

Discussion: Towards an evaluative approach for adult learning
The capability approach allowed to apply a more elaborate conceptual approach to evaluate what factors stimulate or prevents adults to learn.It is not enough to only look at participation or non-participation in adult learning to understanding inequalities related to adult learning.It is also not enough to only look at barriers for adult learning to understand why certain adults find it difficult to engage in adult learning.The capability approach on the other hand opened perspectives to look at whether adults are at all in the position to learn, whether they have the freedom to choose adult learning as a valuable capability, and whether they have the abilities to turn resources into actual learning.This conceptual model can be used to evaluate whether adults have equal opportunities to value adult learning and turn their willingness into actual learning.
For evaluating an individual's situation related to any social topic, Sen (1993) argues that we need to look at two dimensions.Firstly, the evaluation looks at whether the person has freedom to achieve what that person values (capabilities) or actually achieved what that person values (functionings).Secondly, the evaluation looks at whether the person is primarily engaged with ensuring the own well-being (health, living conditions) or whether the person can pursue its overall agency goals, in other words, those goals that a person has reason to value (Anand & van Hees, 2006;Sen, 1993).Together these two distinctions constitute four clusters in which people's state can be evaluated.Figure 2 uses Sen's two dimensions and applies them to adult learning.Through this, the figure summarises the conceptual explorations and indicates how the different groups of adults connect to adult learning.
Related to this, four groups of adults can be identified, namely: 1. Adults in a non-realistic position to learn: this concerns persons who are preoccupied with securing primary well-being and are not in the position to project a future in which learning plays a role.In addition, the context (social, economic etc.) does not support any conversion of resources into actual learning.2. Adults in a realistic position to learn, who do not (yet) participate in learning: this concerns persons who have achieved primary well-being and see the value of learning for reaching future goals, but at the moment see no added value to learn.They have agency towards learning, but did not convert this into actual learning.3. Adults who are learning, but not as a result of being in a realistic position to learn: this concerns persons who lack agency towards learning and do not see added value, but as a result of pressing external factors (conversion factors) are in some form of organised learning.4. Adults who are learning as a result of own agentic power and favourable conversion factors: this concerns persons who made a deliberate choice to learn, make future projections in which learning plays a role and have the supportive conversion factors to materialise this learning motivation.
It follows that adults with differences in agentic power and conversion abilities might also need different incentives to start learning, working on the specific factors that stimulate or prevents them to start learning.The actions to increase agentic behaviour towards learning are different from stimulating learning of adults already possessing aspirations to learn.The presented theoretical perspectives signal systematic challenges for adults in a more vulnerable, or disadvantaged position to participate in learning.The capability approach situates the individual in a wider context that should facilitate an agent to feel autonomy to choose what he/she/it values and asks the question whether the person is realistically at all in the position to choose for learning.In other words: consideration must be given to whether learning is an opportunity that an agent could find valuable.In this context, the agency discussions showed the path-dependency of agents.Aspirations are confined by past experiences making it difficult or risky to break away from past and current social context through starting a learning pathway.For those adults having negative experiences with schooling and learning, starting learning falls in the area of a risky imaginary futures and long-shot aspirations, while for those having a better experience with learning, it is less risky.In addition, in terms of social conditions and welfare state factors, more vulnerable adults have a more bounded or confined agency hampering the willingness to take up learning.Furthermore, the capability approach emphasises that conversion factors unevenly affect a person's ability to turn potential learning participation (capability) into achieved learning participation (functioning).Again, those individuals in vulnerable situations and with a disadvantaged background with regard to learning are less likely to have the conversion factors needed to turn a desire to learn into actual learning.As stated by Karin Evans, 'societies need to ensure that the greatest demands to 'take control of their lives' do not fall on those who are the least powerfully placed in the social landscape they inhabit' (Evans, 2007, p. 93).This is, however, exactly what happens in relation to firstly the agentic capacity (freedom) to value learning (Boyadjieva & Ilieva-Trichkova, 2021;Evans, 2007), secondly, the lack of a conducive and supportive environment to start learning (Boeren, 2017) and thirdly, in relation to the take up benefits of adult learning.This means that specific vulnerable adults are, from three perspectives, at a disadvantage to benefit from adult learning (namely, related to reduced agency, reduced conversion ability, reducing benefits of adult learning), making the so-called Matthew principle of accumulated advantage ('the rich get richer and the poor get poorer') even more challenging to overcome.
Instead of looking at barriers to adult learning participation, as shown in this article, policies that aim at increasing participation in adult learning need to take into account a wider set of conditions and factors explaining why adults engage in learning.This implies that policies should shift their conceptual orientation from lowering barriers and offering opportunities towards proactively securing that individuals all have the freedom to view adult learning as a valuable option.This means considering participation not solely as the responsibility of learners themselves, but considering this in 'interaction with broader structural conditions within a country or geographical area' (Boeren et al., 2012), hence maintaining that governments cannot fully shift responsibilities for starting learning to the individual.

Conclusions
As mentioned in the introduction of the article, we argue that looking at sociodemographic background characteristics and barriers to participation is insufficient in understanding why adults do not learn.Providing the evidence base for adult learning policy development and the monitoring of policies requires a re-examining of how adults' connection to learning is positioned in a wider social, economic, environmental context and how this context provides a conducive environment in which adults first of all would value learning and can pursue learning.Based on the capability approach, we explored the interplay between different factors that influence whether adults intentionally act towards engaging in organised and structured learning or development activities.Agency, conversion and perceived benefits of learning are mutually enforcing whether adults see learning as a valuable (life) choice.
This approach opens new perspectives to empirically explore the interplay between agency, conversion and benefits and identify main factors stimulating adult learning.This empirical research will bring us closer to a validated conceptual model on what prevents and what stimulates adults to learn.A model that is very much needed to evaluate and monitor adult learning and lifelong learning policies delivering on their priority status and combating current and future economic and societal challenges.

Figure 1 .
Figure 1.A capability approach-model depicting the interplay between agency, conversion factors and results and benefits of adult learning to explore what stimulates or prevents an adult to learn.Note: Developed by authors.

Figure 2 .
Figure 2. Agency, conversion and actual participation in learning: Four groups of adults and their relationship with learning.Note: Developed by authors.The two dimensions are extracted from texts fromSen (1993),Anand &van Hees (2006), andGangas (2016).

Table 1 .
Coverage of concepts by articles included in the narrative literature review (N = 109).