Using the Lens of ‘Possible Selves’ to Explore Access to Higher Education: A New Conceptual Model for Practice, Policy, and Research

Abstract

The concept of ‘aspiration-raising’ has been ubiquitous in the discussion of differential rates of participation in higher education in England for many years. Potential students from disadvantaged backgrounds are constructed as setting their sights too low and therefore not considering higher education or ignoring elite universities that they could access. However, it is increasingly understood that aspiration-raising is unable to explain patterns of participation and that it risks ‘blaming the victim’ by failing to appreciate the structural constraints forged through their sociocultural context. The purpose of this paper is to present an alternative lens in the form of ‘possible selves’. This is drawn from the discipline of psychology and aims to explain how we all conceive and develop visions of ourselves in future states. These images create a motivational impetus for actions in the present in order to achieve a like-to-be self—or evade a like-to-avoid self. Notably, the theory takes specific account of the individual’s expectations and the importance of having a clear pathway towards a long-term destination. This paper provides an overview of the foundational theory and empirical evidence for a general readership, before presenting a new conceptual model focused on access to higher education. This is then used to explore the principles that might underpin interventions to support participation from disadvantaged groups within highly stratified systems, as well as suggesting a new policy agenda and priorities for future research.

1. Introduction

The question of equitable access to higher education has been of ongoing policy concern in England for at least twenty years (e.g., National Committee of Inquiry into Higher Education 1997; Department for Education and Skills 2003; Department for Business, Innovation and Skills 2014; Department for Education 2016). There has been an acknowledgement that the system—along with the similar systems in the other constituent nations of the UK—has historically been highly stratified across social class, gender, and ethnicity (Raffe and Croxford 2015; Harrison 2017).

In particular, there have been significant and persistent differentials in participation between socioeconomic groups, with 24 percent of young people from low-income households entering higher education, compared to 41 percent of those from higher-income households (Department for Education 2017); the pattern is even more marked for elite universities, where the most advantaged group is nearly ten times more likely to secure a place than the most disadvantaged (Universities and Colleges Admissions Service 2017). Other countries with mass systems of higher education report similar forms of stratification (Marginson 2017), including those in Asia (Yamamoto 2017) and post-communist Europe (Smolentseva 2017). For simplicity, this paper will use a basic dichotomy between ‘advantaged’ and ‘disadvantaged’—the field is, of course, significantly more complex than this, with multiple and intersecting forms of inequality.

In recent years, the stratification of higher education has most commonly been explored through a Bourdieusian lens (see Webb et al. 2017 for a meta-analysis) that stresses the role of social structures in reproducing inequality (e.g., Bathmaker et al. 2016; Burke 2015; Reay et al. 2005). It is generally argued that disadvantaged young people’s access to higher education is limited as they possess forms of social and cultural capital that are not recognised or valued by universities and lack sufficient economic capital to participate. Furthermore, their habitus is in dissonance with the hegemonic power embodied in higher education, making universities appear alien, unwelcoming, and antagonistic (Reay et al. 2010). The net result is that the pathway into higher education is seen as ‘natural’ for advantaged young people, but ‘barriered’ (Gorard et al. 2007) and ‘risky’ for their disadvantaged peers (Clayton et al. 2009). These patterns are stronger still for elite forms of higher education (Bathmaker et al. 2016; Reay 2017), leading to persistent forms of stratification that may be further reinforced by discriminatory practices that serve to undermine access for disadvantaged young people (Boliver 2016; Jones 2012).

An alternative, but not necessarily contradictory, perspective has been offered by writers drawing on Boudon’s (1974) work on structural inequality (e.g., Breen and Goldthorpe 1997; Thompson 2017; Thompson and Simmons 2013), which conceptualises stratification as being the result of disadvantage acting on attainment at various educational stages (designated as ‘primary’ effects) and subsequent decisions whether to extend education and in what form (‘secondary’ effects). In other words, the ability to participate in higher education can be undermined both by the accumulation of disadvantage in their early lives that leaves young people without the required entry qualifications, while some of those that are appropriately qualified will be dissuaded from higher education (or elite universities) because they are more risk-averse, less likely to expect to succeed and less able to meet real or opportunity costs than their advantaged peers (Breen et al. 2014).

This paper approaches the question from a broadly critical realist perspective, with its concern for a conceptual balance between structure and agency in understanding complex social fields and its focus on ‘middle-range’ theory (Merton 1968) that is meaningful for policymakers and practitioners. In particular, it explores the agentic elements that underpin access to higher education and how these are linked to the young person’s wider sociocultural context—the nexus of social structures that shape opportunities, expectations and inequalities across social class, ethnicity, gender, disability, and their intersections. To achieve this, it draws on theory from the discipline of psychology and places it within a wider psychosocial context informed by Boudon’s primary effects and Bourdieu’s habitus.

Policy and Practice Context

England has seen a largely consistent policy aim over the last twenty years to increase the proportion of students from disadvantaged groups in the sector as a whole (often known as ‘widening participation’) or in elite universities (‘fair access’), albeit that the guiding principle and policy discourse has arguably shifted from social justice based on equality of opportunity to social mobility based on ideas of meritocracy and national competitiveness (Waller et al. 2015). The majority of this focus has been on young people; this paper will follow this focus, although many of the principles will be applicable to potential mature students too.

The widening participation and fair access agendas are thus charged with attempting to disrupt the historic stratification of higher education and encourage disadvantaged individuals to transcend the structural disadvantages with which they are faced. While this has a noble prima facie aim of extending educational opportunities to groups that have historically not had access, it has been critiqued from a Freirean perspective (e.g., Burke 2012) for failing to address the macrohegemonic structures that create inequality, such that it risks simply benefiting a few more ‘fortunate’ individuals without wider societal change (McCaig 2018). In other words, these agendas risk failing to recognise the paradox that higher education as it is currently configured can act to both challenge and reproduce structural inequalities. Nevertheless, there is widespread agreement that supporting disadvantaged individuals to be able to access higher education is a worthwhile activity and around £175 million is spent annually to this end (Office for Fair Access 2016), as well as work by consortia through the National Collaborative Outreach Programme (Higher Education Funding Council for England 2016) and by third sector providers working in collaboration or in pursuit of their own organisational missions. One feature that has been largely constant in the English policy discourse since the early 2000s has been a strong reliance on the idea of ‘aspiration-raising’ as the principal means of encouraging people from disadvantaged groups to apply for and enter higher education. For example, the 2003 White Paper (Department for Education and Skills 2003) focuses on families with low aspirations, while the current national strategy document (Department for Business, Innovation and Skills 2014) showcases several examples of ‘good practice’ that are asserted to be predicated on aspiration-raising. Under this conceptualisation, these groups are considered to have lower long-term ambitions for their career (or wider lives) and therefore less likely to view higher education as relevant or more likely to be satisfied with ‘lower’ forms of higher education—even if they have the qualifications enabling them to access elite universities.

This discourse came under sustained theoretical critique from the outset (e.g., Leathwood and O’Connell 2003; Jones and Thomas 2005; Archer 2007a), with writers arguing that it risks ‘blaming the victims’ of structural disadvantage for failing to transcend the circumstances with which they are faced. Bok (2010) and Smith (2011), writing in an Australian context with similar policy drivers, argue that the capacity to aspire (Appadurai 2004) is not equally distributed and that the young person’s sociocultural context and the availability of information are key in understanding patterns of participation.

More recently, the concept of aspiration-raising has come under additional pressure from a series of large-scale empirical studies that have demonstrated that disadvantaged young people do not have markedly lower aspirations for education, careers, or adult life than their relatively advantaged peers (e.g., Archer et al. 2014; Baker et al. 2014) and that there are many more disadvantaged young people who aspire to higher education than who participate (e.g., Croll and Attwood 2013; St. Clair et al. 2013), suggesting that aspirations play a very limited role in influencing educational and wider life outcomes (Green et al. 2018). Rather, empirical data suggest that a focus on expectations rather than aspirations might be more meaningful as they conceptually include an element of the individual’s estimation of whether a particular life outcome is likely or not (Harrison and Waller 2018); importantly, expectations among disadvantaged young people tend to be considerably lower than both their aspirations and the expectations of their relatively advantaged peers (e.g., Boxer et al. 2011; Khattab 2015).

Furthermore, recent analysis of national administrative data has demonstrated that nearly all of the variation in participation rates for different social groups can be attributed to differences in the qualifications accumulated at the age of 16, including access to elite universities (Crawford 2014; Crawford and Greaves 2015). In other words, young people tend to participate at the rates predicted by their level of attainment, with little difference by social class, gender, or ethnicity—more broadly, patterns of participation have tended to closely echo improvements in attainment (Harrison 2017). This is perhaps not surprising given the highly credentialised system in England, where access to future opportunities are strongly tied to qualifications accumulated. This is not, of course, to deny the vital importance of structural inequalities, but rather to recognise that they manifest early in a young person’s life and are thus embodied through the attainment itself, as predicted by Boudon (1974). This again challenges the discourse of aspiration-raising, as it strongly suggests that aspirations can only meaningfully influence participation in higher education for disadvantaged groups to the extent that they influence attainment in school. However, there does not appear be a strong link between raised aspirations and higher attainment. Focused reviews by Cummings et al. (2012) and Gorard et al. (2012) concluded that aspiration-raising was not a good means of influencing school outcomes, with the latter further arguing that any correlation between aspirations and attainment was most likely due to improved attainment driving higher aspirations and not vice versa.

Despite these theoretical and empirical challenges to aspiration-raising, it continues to permeate both national policy documents (e.g., Department for Business, Innovation and Skills 2014) and practice among professionals charged with widening participation and/or ensuring fair access (Harrison and Waller 2018). It has also proved surprisingly resilient within the academic literature (see the review by Younger et al. forthcoming) and the public discourse about educational disadvantage (Burns 2018). This may be due to its conceptual simplicity and congruence with ‘common sense’ explanations for why young people have different outcomes—or, more critically, as it deftly evades bigger questions about inequalities in society by locating responsibility with the individual.

This paper proposes an alternative conceptual model for understanding how young people make choices about school and higher education. Specifically, it draws on the theory of ‘possible selves’ (Markus and Nurius 1986) to explore how individuals view themselves in future states, both positive and negative. It will argue that this provides a more theoretically coherent and persuasive means of understanding how visions of the future impact on here-and-now actions, including both proximal decisions about schoolwork and more distal ones about participation in higher education. It also works from the premise implicit in Crawford (2014) and Crawford and Greaves (2015) that the most effective means of raising the participation of disadvantaged groups is to support them to improve their attainment at 16. This paper comprises five main sections. The first introduces the theory of possible selves through its foundational literature, while the second explores subsequent empirical studies applying this lens to issues of young people’s decision-making, motivation, and engagement with school. The third section applies the insights from possible selves to the field of access to higher education to propose a new conceptual model. The fourth section develops out a specific agenda for practice to address the ongoing challenges of widening access, with the fifth providing a rationale for an alternative policy agenda that is not predicated on aspiration-raising. The paper concludes by outlining some potential avenues for future empirical research.

2. The Theory of Possible Selves

A paper by Markus and Nurius (1986) forms the foundational work for the theory of possible selves. Based on empirical data about people’s perceptions of what their life might hold in the future, they theorised that we all envisage a range of possible identities for ourselves, framed by factors both within and beyond our control: ‘what [people] might become, would they would like to become, and what they are afraid of becoming’ (Markus and Nurius 1986, p. 954). These possible selves are part of building a wider narrative that we use to make sense of our lives in our own social context and ‘represent the individual’s persistent hopes and fears and indicate what could be realized given appropriate social conditions’ (Markus and Nurius 1986, p. 965).

This theorisation forms part of wider work in the discipline of psychology around ‘self-concept’—the ways in which we understand ourselves in the context of our lives. Possible selves therefore form a future tense for the self-concept, representing our current perceptions about where our lives might lead through the construction of multiple representations of ourselves as we might be days, months, or years hence. In particular, possible selves are a key component of the ‘working self-concept’ of how we see ourselves within a given moment; familiar and vivid images that are cognitively accessible and guide our everyday decision-making and the processing of personal experiences:

‘The working self-concept derives from the set of self-conceptions that are presently active in thought and memory. It can be viewed as a continually active, shifting array of available self-knowledge. The array changes as individuals experience variation in internal states and social circumstances’.

(Markus and Nurius 1986, p. 957)

Over time, changes in the mutable working self-concept feedback into more persistent forms of self-knowledge that frame our identity. In this way, a possible self can become progressively more embedded in the ways in which we see and understand ourselves. This process of becoming then works to determine what forms of present action are legitimate and comfortable:

‘Possible selves […] can be viewed as cognitive bridges between the present and the future, specifying how individuals may change from how they are now to what they will become. When certain self-conceptions are challenged or supported, it is often the nature of the activated possible selves that determines how the individual feels and what course the subsequent action will take’.

(Markus and Nurius 1986, p. 961)

Markus and Nurius (1986) argue that individuals accumulate a pool of possible selves and that it is possible to broadly categorise these. At the most basic level, some will appear to the individual to be positive images to be pursued while others will be negative and to be avoided. Each self will also appear more or less likely to come to pass. These two dimensions (of desirability and probability) are key to the argument being developed within this paper. Firstly, a possible self might be more (or less) desirable than others, encapsulating future feelings of happiness, success, security, or wellbeing. Secondly, we employ our judgement to determine the likelihood of each possible self coming to pass, determining both probable and improbable selves within the wider pool. The intersection of desirability and probability is a potentially powerful component of the self-concept, containing positive visions of the future that are felt to be attainable through good decision-making, persistence, and hard work. This theoretical model is illustrated in Figure 1.

Of particular interest here are the mechanisms by which these assessments of desirability and probability are made. Markus and Nurius (1986, p. 954) argue that they are constructed through a combination of personal experiences and the wider context in which the individual lives:

‘An individual is free to create any variety of possible selves, yet the pool of possible selves derives from the categories made salient by the individual’s particular sociocultural and historical context and from the models, images, and symbols provided by the media and by the individual’s immediate social experiences. Possible selves thus have the potential to reveal the inventive and constructive nature of the self but they also reflect the extent to which the self is socially determined and constrained’.

The sociocultural context thus provides an initial starting point of what sort of selves are known about within the family and/or community, and therefore viewed as possible in the broadest sense—e.g., ‘me as a solicitor’ or ‘me as a parent’. It also normatively shapes the individual’s values about what selves are to be considered appropriate or attractive through a stream of ‘social communiques’ (Nurius 1991, p. 246) that are strongly influenced by class, gender, and ethnicity. Our sociocultural context also has an influential role in determining the types of personal experiences to which we are exposed—where we go, who we meet, and what we see within the physical spaces that we inhabit (Prince 2014), creating a situation where young people in different social groups will have very different views of what selves are possible for them, which are desirable, and which are probable.

However, ‘individuals [also] have considerable freedom to define and redefine their significant possible selves’ (Cross and Markus 1991, p. 233). For example, Markus and Ruvolo (1989) and Ruvolo and Markus (1992) argue that the act of envisaging a possible self is itself a form of personal experience that shapes assessments of desirability and probability, as well as providing some of the positive experiences of being that self. Thus, our possible selves are constantly in flux as we imagine and reimagine the selves that we could become through the lens of our everyday lives. In addition, and crucially to this paper, there is scope for planned interventions to enable individuals to envisage and assess new or developing possible selves.

Later work by Markus, Nurius, and their colleagues has built additional elements into the original theoretical framework. Perhaps most importantly, mentioned in Markus and Nurius (1986) and developed in more detail in later works, is that of ‘elaboration’. This describes the extent to which possible selves are fully-formed and detailed, with a vivid vision of what that self would be like and the intermediate steps needed to get there. It is argued that elaborated possible selves have a stronger impact on the individual’s choices than those that are vague or that lack a plausible pathway to achieve them (Cross and Markus 1991).

In particular, elaboration is seen as vital in influencing and motivating current actions as ‘clearly elaborated possible selves and the strategies of realizing them decrease the psychological distances between one’s current state and the desired end-state’ (Ruvolo and Markus 1992, p. 119). As a result, ‘the more compelling the possible self, the more vividly it can be elaborated in the present, and the more it will command attention and structure one’s current activity’ (Ruvolo and Markus 1992, p. 229). Possible selves thus ‘provide a conceptual link between the self-concept and motivation’ (Markus and Nurius 1987, p. 157), providing an impetus to act in the present:

‘Possible selves give specific, self-relevant form, meaning, and direction to one’s hopes and threats. Possible selves are specific representations of one’s self in future states and circumstances that serve to organize and energize one’s actions’.

(Ruvolo and Markus 1992, p. 212)

Oyserman et al. (2004) typify elaborated possible selves as ‘roadmaps’ while Erikson (2007) sees them as internal ‘narratives’ that allow us to plan out what steps are needed to move from the present to a like-to-be self—there are clear resonances here with Archer (2007b) ‘internal conversation’ and Giddens (1991) ‘project of the self’. An elaborated possible self (whether like-to-be or like-to-avoid) thus legitimises current actions and provides a specific form of motivation with short-term and long-term elements. Ruvolo and Markus (1992) argue that this motivation is stronger where there is a ‘balance’ in the same domain (e.g., employment) between like-to-be and like-to-avoid selves (e.g., ‘me having a job with promotion prospects’ vs. ‘me being unemployed’), combining positive and negative stimuli.

A key element in the motivational power of possible selves is argued to be their accessibility to the working self-concept—the ease with which they are brought to mind during everyday experience to initiate or legitimise actions in the short-term (Ruvolo and Markus 1992; Markus and Nurius 1987). Reinforced and repeatedly reactivated possible selves are more likely to be accessible and influence actions in the here-and-now, especially where they are supported within the sociocultural context:

‘possible selves are not likely to become elaborated and thereby either motivationally or behaviourally effective unless valuing them and believing in them are supported or encouraged by significant others’.

(Nurius 1991, p. 246)

However, ‘if a compelling possible self can be constructed, the outcome is valued and simultaneously the expectancy of attaining it is increased’ (Ruvolo and Markus 1992, p. 218). Thus, the process of elaboration not only makes the future vision appear more desirable, but also more likely to pass, which in turn also increases its desirability, providing a powerful form of motivation:

‘an individual’s estimate of the probability of certain possible selves, both positive and negative, considerably augmented our ability to explain current affective and motivational states’.

(Markus and Nurius 1987, p. 167—original emphasis)

To summarise, the theory of possible selves argues that we all create visions of ourselves in the future and that these have a motivational power, especially where they are vivid, detailed, congruent with prevailing norms, and supported by significant others. These selves vary in their perceived desirability and probability, with possible selves that are readily brought to mind being particularly powerful. Importantly, possible selves are highly individualised, but constructed within a sociocultural context that shapes which selves appear possible, desirable or probable, as well as through the personal experiences that challenge or reinforce our palette of possible selves (see Henderson 2018 for a more sociologically focused treatment of the theory). More broadly, Erikson (2007, p. 354) argues that the construction of possible selves is an act of meaning-making where we find self-relevant ways of exploring what it is possible to be: ‘what we expect about the future is dependent on what we are now (and vice versa)’—Rossiter (2009) and Dittmann and Stephens (2017) also emphasise how we use possible selves to make sense of our developing identities. We now turn to review the empirical evidence as it relates to educational choices and outcomes, before returning to apply the theory to build a new conceptual model addressing the specific question of access to higher education.

2.1. Possible Selves, Educational Outcomes and Intergroup Differences

The theory of possible selves has enjoyed substantial purchase across diverse fields within psychology and psychosocial studies, including risk-taking behaviours, dementia care, and housing choices. There has been very little application to higher education to date, with the majority of the literature exploring student experiences and entry into the graduate labour market (Henderson et al. 2018; Papafilippou and Bentley 2017; Stevenson and Clegg 2011). However, there is a small corpus of work using possible selves in school contexts and this section will review the empirical evidence available. It focuses particularly on how a possible selves approach can encourage engagement with schoolwork among disadvantaged (and often disengaged) young people and the resulting impact on attainment at school, bearing in mind the premise that the latter is key in influencing whether (and how) an individual can enter higher education (Crawford 2014; Crawford and Greaves 2015).

Leondari et al. (1998) found a correlation between positive possible selves, planned persistence with schoolwork and academic attainment among 14- and 15-year-olds, although the direction of causation was unclear. In addition, they concluded that the extent to which luck or hard work were felt to be responsible for success was also an important element of how young people perceived their futures. Similarly, Oyserman et al. (2004, p. 133) found that twelve-year-olds in a disadvantaged setting were significantly more likely to have improved academic outcomes where they had well-elaborated possible selves and could articulate strategies for achieving them, with harder work being put into homework and classroom participation, concluding that ‘these “self-regulating” possible selves can preserve positive affect, maintain behavioural focus, and ultimately propel the self toward the goal’. Interestingly, Leondari and Gonida (2008) did not find a direct relationship between possible selves and attainment, but young people with elaborated selves were more likely to report persistence and a deep mastery approach to learning. Through later laboratory work, Oyserman et al. (2015) argue that this motivational effect from possible selves is contextually mediated by both whether the individual feels likely to be successful and whether their possible self is positive or negative in nature.

In addition to these correlational studies, there is a small, but compelling, body of experimental research based around interventions in real-world settings. One intervention, implemented as a multiweek after-school activity in a disadvantaged urban setting in the US sought to help young people to articulate possible selves, develop strategies for achieving them, link short-term school-based possible selves to long-term possible selves in adulthood and build positive supportive relationships with peers and adults (Oyserman et al. 2002, 2006, 2007). This was found to have a positive impact on motivation for schoolwork and on academic outcomes, sustained over a two-year period after the intervention (Oyserman et al. 2002, 2006), as well as working to compensate for lower levels of parental engagement in school (Oyserman et al. 2007). Notably, the intervention ‘focused on changing possible selves by giving youth the opportunity to see the connections between present and future at their own pace, through activities shared with a group rather than “insight” based discussion’ delivered didactically by adults (Oyserman et al. 2002, p. 323). Hock et al. (2006) describe three trials of a longitudinal intervention with US adolescents and university students (based around the six steps of discovering, thinking, sketching, reflecting, growing, and performing), finding significant positive results around sustained future motivation and outcomes when compared to generic careers-based guidance. Finally, in a small study, Landau et al. (2017) demonstrate that priming 11- to 14-year olds with a ‘road’ metaphor for their possible selves increases intentions towards academic engagement, although seemingly only for girls.

Other studies have focused on the vector by which a possible selves approach might operate and the practices that might support this. Packard and Nguyen (2003, p. 261) argue for the importance of mentoring programmes and similar interventions to give young people the opportunity to ‘create, develop, or imagine possible selves through interactions with the significant adults in their environments [where] the durability of desired possible selves relies heavily on the validation of [these] significant adults’. Hardgrove et al. (2015, p. 169) also stress the importance of social support in shaping possible selves, drawing on data from young men in precarious employment to argue that possible selves need to be seen as viable and ‘expectable’ and tied to their situated lived experiences: ‘Young people cannot work towards outcomes that they have no ability to imagine in the future, nor will they be motivated to direct their efforts towards imagined futures if there are no (perceived) viable routes to get to them’. Oyserman et al. (2011) report that while young people from different socioeconomic groups have a similar propensity to construct possible selves, those from disadvantaged groups are less likely to have articulated strategies for achieving them, concluding that the issue is not with different destinations, but the paths to achieving them—this is supported by the review of Dittmann and Stephens (2017). Yowell (2002) concurs, finding that the difference between (lower) expected selves and (higher) desired selves was often the absence of procedural knowledge about how to succeed.

While it is not the focus of this paper, several writers have focused on intergroup differences in how possible selves are constructed and elaborated. For example, young men tend to report selves stressing uniqueness and individual mastery, while women are more likely to elaborate those based around interconnectedness and shared goals (Knox 2006; Oyserman and Fryberg 2006; Leondari and Gonida 2008). Oyserman and Fryberg (2006) report that there are notable differences between ethnic groups in the US, while Leondari and Gonida (2008) found a contrast between urban and rural young people, with the latter being more likely to elaborate selves around happiness and wellbeing compared to the academic and career goals of the former. These differences once again speak for the importance of sociocultural context in framing how possible selves are constructed, valued and pursued and would benefit from future exploration.

Finally, while research to date has primarily focused on the possible selves of young people, Leondari (2007) and Rossiter (2009) argue that the same principles can be applied to adult learners, who may particularly see education as a means of achieving new career-based selves or as an opportunity to ‘try out’ potential new occupational identities.

2.2. Summary: Possible Selves and Aspirations

This paper began by outlining the pervasive role that aspiration-raising has had in discussions of access to higher education and it is worth reflecting briefly on the ways in which the theory of possible selves differs from this. They do share a future-orientation whereby the individual envisages possible outcomes that have meaning for them. However, as Hardgrove et al. (2015, p. 164) argue, aspirations are ‘vague’ and lack ‘specificity’, with this lens having ‘failed to produce helpful insight’ in understanding young people’s decision-making, while Landau et al. (2017, p. 277) argue that simply ‘conjuring up an image of a desired possible identity is not sufficient to motivate goal-directed action in the present’. Crucially, ‘vague, general possible selves lacking behavioural strategies cannot function to guide self-regulation because they neither provide a specific picture of one’s goals nor a roadmap of how to reduce the discrepancies between the present and one’s future possible selves’ (Oyserman et al. 2004, pp. 133–34). St. Clair et al. (2013) bluntly dismiss aspirations as a ‘red herring’.

Four elements of the theory of possible selves distinguish it from the way in which aspirations are general conceptualised. Firstly, possible selves embody an element of expectation about what might happen rather than simply what the individual wishes would happen (Markus and Nurius 1986) and this has a strong sociocultural element (Nurius 1991; Prince 2014). Secondly, it is vital that the possible selves are not simply envisaged, but that they are richly elaborated and strategised, with a clear pathway to realising short-term goals and a long-term destination (Oyserman et al. 2004, 2011; Hardgrove et al. 2015); there is evidence that this is best achieved through work led by the individual but supported by significant adults (Oyserman et al. 2002; Packard and Nguyen 2003). Thirdly, there is clear empirical evidence for possible selves impacting on motivation for schoolwork and thence to educational decision-making (Oyserman et al. 2002, 2006, 2007; Hock et al. 2006), which is lacking for aspirations. Fourthly, negative like-to-avoid selves are seen as having equal motivational force as aspirational like-to-be selves (Markus and Nurius 1986), especially where the expectation of failure is high (Oyserman et al. 2015); paired like-to-avoid and like-to-be selves are seen as being especially efficacious as they make manifest the implications of decisions (Markus and Ruvolo 1989). We will return to these principles later.

It is important to remember, however, that the empirical work to date supporting the application of possible selves to education is quite limited. Its use in the remainder of this paper is therefore generative rather than explanatory, in that it makes assertions about how the theory might be used within a particular context—further empirical work will be needed to test these assertions.

3. Building a Conceptual Model

This section attempts to build a conceptual model for how the theory of possible selves might be applied in the context of access to higher education; this is presented diagrammatically in Figure 2. The purpose of this model is to provide a basis for deriving hypotheses about which practices and policies are likely to be effective in terms of impacting on participation in higher education for disadvantaged groups. Other appropriate theoretical perspectives are woven into the model and shown in italics.

The model effectively begins with the relationship between sociocultural context and decision-making for young people, which has long come under investigation by social scientists resulting in similar theoretical perspectives including ‘opportunity structures’ (Roberts 1968, 2009), ‘horizons for action’ (Hodkinson 1995; Hodkinson and Sparkes 1993, 1997) and ‘bounded agency’ (Evans 2007). These share the basic tenet that our sociocultural context shapes—without simplistically determining—what futures appear possible and probable for us, as well as the limits to the individual agency that we are able to exercise within our prevailing structural constraints. For example, Hodkinson and Sparkes (1997, p. 33) derive their theory of horizons for action from Bourdieu’s concept of habitus as ‘a person’s beliefs, ideas, and preferences [which] are individually subjective but also influenced by the objective social networks and cultural traditions in which that person lives’. They assert that this frames the options that appear possible and desirable to young people through an inexorable link to the social structures that surround them, such as the nature of the local labour market, educational provision, public transport, housing stock, crime, and the normative values of the family and wider community. This context serves to constitute both the opportunities that are readily available and the barriers that exist to realizing them:

‘Horizons for action both limit and enable our view of the world and the choices we can make within it. Thus, the fact that there are jobs for girls in engineering is irrelevant if a young woman does not perceive engineering as an appropriate career’.

(Hodkinson 1995, p. 6)

In the proposed model, the sociocultural context firstly works to shape the initial palette of possible selves that might have meaning for a young person through the normative values that make some adult outcomes acceptable or unacceptable (Markus and Nurius 1986). This is likely to include the ways in which various educational or career paths might be perceived, in relation to the young person’s social class, gender, or ethnicity, as well as the ways in which physical spaces mould what appears normal—e.g., through the educational settings and workplaces that an individual encounters in their day-to-day life (Prince 2014). For young people from advantaged backgrounds, these norms are likely to promote education, including higher education, as a route to entering professional occupations and the affluent possible selves to which they give access (e.g., ‘me as a homeowner’ or ‘me as an international traveller’). However, for their disadvantaged peers, the norms are likely to be based around traditional—and potentially gendered—occupational pathways into manual labour. There may be scepticism or even hostility about the role of education in influencing life outcomes, and for young people, the role of adults is particularly important in transmitting sociocultural expectations—especially parents and teachers with whom they spend the majority of their lives (Winterton and Irwin 2012; Fuller 2014; Archer et al. 2014).

Secondly, it exerts influence on what an individual might expect of their future lives. This is a complex assessment that contains objective, subjective, and personalised elements. For example, a young person might have a good understanding about the opportunities available through the local labour market (objective), feel that they will be disadvantaged due to their accent or where they live (subjective), and be unwilling to move away for work due to strong community ties (personal). The likelihood assigned to certain outcomes may be wholly realistic, based on the economic history of the area, previous family experiences and their assessment of the qualifications that they expect to achieve. For example, the possible self of ‘me as a homeowner’ may be adjudged unlikely where social housing predominates, professional jobs are scarce, and older siblings have failed to find work.

The other influence on expectations comes from the individualised beliefs and dispositions that the young person has about their ability to influence their own future. While there may be many elements to this, research to date on possible selves suggests that ‘self-efficacy’ (Bandura 1977, 1982) and ‘locus of control’ (Ajzen 2002; Lefcourt 2014) are likely to be most salient. The former describes the individual’s contextualised belief in their ability to successfully complete a specific task or achieve a specific goal, while the latter describes how confident they are in being able to shape their future lives through their own agency—that deliberate actions towards a goal will be rewarded with success.

In combination, these reflect how confident that the young person feels that their actions will result in the intended outcomes. In other words, these two elements shape whether a young person expects to be able to influence their future through their own endeavours or whether they expect to fail—either through their own lack of ability, bad luck, or structural constraints. Leondari et al. (1998) found that the juxtaposition of hard work and luck contributed to how possible selves were constructed and elaborated, while Kirk et al. (2012) focused on the link between self-efficacy and expectations. More generally, Ruvolo and Markus (1992, p. 96) stress the role of ‘feeling[s] of efficacy, competence, control, or optimism’ as part of the way in which possible selves have a ‘powerful impact on behaviour’. Chevalier et al. (2009) report that young people with greater reported self-efficacy are significantly more likely to expect to enter higher education once ability has been controlled for, while Chowdry et al. (2010) conclude from an analysis of national data that both lower self-efficacy and locus of control contribute to lower attainment among disadvantaged young people; it should be noted, however, that Gorard et al. (2012) and Cummings et al. (2012) are sceptical on this point.

While individual in nature, it is hypothesised that these beliefs and dispositions are also shaped through sociocultural contexts. Young people from disadvantaged communities can become conditioned to expect failure through negative stereotyping, self-fulfilling prophecies, and an objective scarcity in opportunities (Prince 2014), with various studies finding that they have, on average, lower levels of self-efficacy (e.g., Boardman and Robert 2000) and internal locus of control (e.g., Kraus et al. 2009). Space precludes a fuller discussion of this tentative relationship, which is thus represented through a dotted line in Figure 2.

This process of considering possible selves, their desirability, and their likelihood of coming to pass through the lens of the individual’s personal beliefs and sociocultural context incrementally leads to the creation of one or more like-to-be (and like-to-avoid) selves that are adjudged to be reasonably probable. These enter the working self-concept, but may be more or less elaborated depending on the amount of detail that the vision contains and the presence of a clear ‘roadmap’ (Oyserman et al. 2004) to achieving (or avoiding) them. Elaboration may be supported by everyday experiences, contact with influential adults (Hardgrove et al. 2015), or the ‘trying on’ of new ways of being (Packard and Nguyen 2003)—through this process of reinforcement, the like-to-be self moves from the working self-concept to a more permanent form of future identity.

As discussed previously, the possession of well-elaborated probable selves form ‘cognitive bridges between the present and the future’ (Markus and Nurius 1987, p. 159) that ‘organize and energize one’s actions’ (Ruvolo and Markus 1992, p. 212). They inspire and legitimise activities to achieve short-term goals and support meaning-making in the longer-term. However, as Oyserman et al. (2011) note, it is the development of strategies that tends to distinguish relatively advantaged young people from their disadvantaged peers, potentially due to a greater pool of educational knowledge and experience within the family. Kirk et al. (2012) also found that strategies for achieving success were correlated with future expectations. This suggests that having strategies for achieving possible selves may be an important element in why advantaged young people have (and expected) higher attainment than their peers.

There are many potential like-to-be selves that are predicated on accessing higher education. These might, for example, be occupational (‘me as a doctor’), lifestyle-related (‘me with a dependable salary’), subject-based (‘me pursuing my interest in physics’), emancipatory (‘me challenging myself’), or socially-engaged (‘me contributing to my community’). These might be related to attending higher education (i.e., being a student) or completing (i.e., being a graduate); an individual may also have a like-to-avoid self that involves being unemployed or in low status/prospect work (Harrison forthcoming). In any instance, the conceptual model predicts that there will be two separate, but interlinked, forms of motivation.

Firstly, and most obviously, it will provide a direct demand for higher education as the vehicle to becoming a student and/or a graduate. Secondly, and more importantly, it will stimulate everyday actions that work to boost attainment as a means of acquiring the qualifications that permit access to higher education, including the higher levels required by elite institutions. As noted, this has been demonstrated experimentally for disadvantaged groups (Oyserman et al. 2002, 2006; Hock et al. 2006), with greater levels of effort and higher attainment being reported. If the premise derived earlier from Crawford (2014) and Crawford and Greaves (2015) is valid, increased participation should result.

Increased attainment makes it more likely that the individual will have the qualifications required for entry, including to elite universities, but, within the model, it is also held that academic success in school will reinforce the demand for higher education by making it seem possible and then expected (Gorard et al. 2012), potentially also feeding back into the personal beliefs of the young person about their ability to achieve goals and exercise agency (as represented by the dotted line in Figure 2).

For the sake of simplicity, the conceptual model is represented here as linear, but, in reality, the process of identifying, evaluating and elaborating possible selves is continual and iterative. Everyday personal experiences will make young people aware of different futures they might consider or change their expectations about ones that are already in their thoughts. More generally, and as with all new conceptual models, there is inevitably simplification which can only be unpicked through future empirical study testing the premises, processes and relationships that are asserted by the model.

4. Implications for Practice

As discussed earlier, much practice in this area remains grounded in aspiration-raising as its guiding conceptual model (Harrison and Waller 2017). The purpose of this section is to suggest alternative principles for practice derived from the theory of possible selves. These may, in some cases already, be reflected in current or past practices and, in this sense, are not novel. However, this section derives them afresh from the conceptual model presented in the previous section, informed by the empirical evidence discussed earlier, following Oyserman et al. (2002, p. 315) who argue that ‘structured activities occurring in everyday settings can have great impact on we think we are and what is possible for to achieve, because subtle contextual shifts can powerfully change the sense made of daily experiences’.

Figure 2 identifies four specific points at which interventions might be planned with young people and the nature of these interventions; these are denoted by thick grey arrows. These are points within the conceptual model rather than fixed points in time—indeed, it is possible that a single intervention might be able to cover all four or that this might achieved through an integrated, longitudinal programme. This section explores these conceptual points in more detail and discusses how higher education providers and others might design appropriate activities grounded in the theory of possible selves:

• Intervention Point 1 relates to the palette of possible selves that is available to the individual: what is the pool from which they are able to pursue like-to-be selves or identify like-to-avoid selves? While the size of the pool may not differ markedly between advantaged and disadvantaged young people (Oyserman et al. 2011), the latter are likely to envisage fewer possible selves that are predicated on requiring a degree. This is in part due to their horizons for action, which inform their concepts of what it is possible to be in their own sociocultural context. An intervention at this point would seek to expand the pool of possible selves available that have a relationship to higher education—or to education more generally. These may be occupationally-driven or focused on demonstrating how wider possible selves (e.g., ‘me as homeowner’) are reliant on educational success. In particular, such interventions might seek to strengthen the perceived relationship between education and wider life outcomes. For example, Oyserman et al. (2002, pp. 317–18) describe an activity where young people choose from a selection of images of adults as a vehicle for group discussions around ‘work, family, lifestyle, community service, health, and hobbies’. This sort of activity provides an experiential opportunity to engage with possible selves that have not previously been considered and to potentially add them, perhaps only in outline, to the palette of futures that might be available; what Archer et al. (2014, p. 77) call ‘diversifying’ aspirations. Importantly, such activities would not over-emphasise possible selves as ‘me as a student’ or ‘me as a graduate’, but rather the wider selves to which these states give access. Perhaps more controversially, these activities may also seed conversations about like-to-avoid selves (Ruvolo and Markus 1992; Oyserman et al. 2015).
• Intervention Point 2 relates to engaging with the young person’s beliefs about their ability to exercise control over their future and their ability to succeed at tasks that are important to them. As discussed in the previous section, these are hypothesised to be important vectors in determining the likelihood of a possible self coming to pass and, whereas the wider sociocultural context cannot readily be influenced in the short-term, interventions that challenge these personal beliefs are likely to be successful in shaping what selves appear probable. This is perhaps most important where the probable selves identified by the young person are negative (i.e., like-to-avoid), but where they expect that they will not be able to avoid them due to structural constraints and their own inability to exercise effective agency over their future. Successful interventions are likely to focus on reinforcing the young person’s perceived ability to be successful through supported short-term tasks and a process of reflection that actively demonstrates their potential for more sustained forms of success. Such interventions are likely to be longitudinal in nature, focusing on self-efficacy and/or locus of control or be more academically-focused on the development of a ‘learning orientation’ (Watkins 2010) and metacognitive skills that help young people to understand how they learn; what St. Clair et al. (2013, p. 736) call a ‘day to day process of supporting students to learn how to attain what they want’. They may also engage particularly with parents and teachers as key influencers to ensure that their own expectations are positive, realistic and transmitted to young people (Cummings et al. 2012; Harrison and Waller 2018).
• Intervention Point 3 comes when the young person is beginning to elaborate like-to-be (or like-to-avoid) selves that they feel are probable in their context. This process involves translating their vision of themselves in the future into something that is vivid and detailed in order to provide the motivational impetus that results from integrating this vision into their working self-concept. Oyserman et al. (2002) argue that it is important that young people are allowed to elaborate their own possible selves, rather than passively receive insights from adults about how they should visualise them and what their roadmaps should be. Instead, based on their own experiences of devising and evaluating an intervention, they advocate for a process of providing supported space for young people to identify why their like-to-be selves are important to them and how they might be achieved; this is cognate to the reflecting and growing steps advocated by Hock et al. (2006). Specifically, this is less directive than traditional approaches to careers guidance, with a wider scope beyond the occupational. Activities might include workshops where young people are encouraged to produce actions plans and opportunities to engage with adults embodying the like-to-be selves or to ‘try on’ these selves—for example, through work experience programmes (St. Clair et al. 2013; Waller et al. 2014) or mentoring (Packard and Nguyen 2003; Cummings et al. 2012). The conceptual shift here is away from directive guidance that seeks to coerce young people and towards guided individualised activities that enable them to explore their own futures and devise self-relevant strategies to align their like-to-be selves with their probable selves (Yowell 2002).
• Intervention Point 4 specifically comes as the individual is considering higher education and comes closest to echoing existing aspiration-raising practices with their focus on making higher education appear desirable and realistic. Typically, this includes exposure to a campus environment, involvement in inspirational experiences, collaboration with current students, and information about graduate careers and other opportunities to envisage oneself as a student and/or graduate (Harrison and Waller 2017, 2018). These activities still retain value under a possible selves approach as they form part of the process of elaboration and reinforcement that embeds like-to-be selves involving higher education within the self-concept. However, these activities are unlikely to be transformational for disadvantaged young people without the wider context and individualised strategies provided by the earlier interventions.

In summary, these four intervention points collectively provide an opportunity to widen the palette of possible selves known to a young person, to reinforce their ability to exert control over their own future and to support them in making these visions of the future real, motivating and linked to current actions—they are mutually reinforcing, with single interventions being unlikely to be successful. This differs substantially from most current approaches based on aspiration-raising as the latter generally lack an engagement with expectations about the future and strategies for how these futures might be realised. Nevertheless, Lumb (2018) reminds us that even interventions based around possible selves retain dangers around positioning disadvantaged young people in deficit and leading them into performative and uncomfortable engagements.

5. Implications for Policy

The current policy objective in England is ‘to double the proportion of people from disadvantaged backgrounds entering university in 2020 compared to 2009’ (Department for Education 2016, p. 14). This section argues that approaches based around aspiration-raising are unlikely to achieve this and that the theory of possible selves provides a compelling alternative lens for policymakers. While there is insufficient space to develop and justify a full policy agenda, this section asserts four principles for future consideration.

Firstly, the language of aspirations should be avoided—and actively dissuaded among practitioners. As noted above, the evidence base for aspirations for higher education impacting on attainment or participation is very weak. This is consistent with the theory of possible selves, where vague and depersonalised images of the future are seen as ineffectual in motivating current action. Simply exposing young people to higher education and extolling its virtues through encounters with university staff and students is unlikely to be effective if they do not elaborate a like-to-be self that is deemed probable, with a clear strategy for achieving it and a belief in their ability to do so. In particular, Crawford (2014) and Crawford and Greaves (2015) demonstrate the very limited role that aspiration-raising activities with over 16-year-olds can possibly have, with advantaged and disadvantaged young people having very similar participation rates at a given attainment level. However, while there are some efforts to help younger disadvantaged children attain more highly, the majority of activity with under 16-year-olds remains focused on aspirations (Harrison et al. forthcoming), presumably in the belief in a virtuous circle that aspiration for higher education will motivate them to work harder and achieve higher. Again, the evidence for this is very weak (Gorard et al. 2012; Cummings et al. 2012).

Secondly, policy needs to pay more attention to the expectations of young people, the ways in which they are shaped by sociocultural contexts, and how they impact on motivation and decision-making. This is not a sleight of hand that merely shifts the responsibility from the disadvantaged individual to the disadvantaged community. Certainly, there will be normative elements that work to assert what selves might be acceptable or valued and these may not be congruent with policy goals around access to higher education. However, these expectations may be wholly realistic given the history of structural disadvantage within a community. For example, Raphael Reed et al. (2007, p. 14) studied four urban communities where participation in higher education is uncommon and where ‘many young people grow up in environments where they rarely encounter educational or economic success—or the relationship between the two’. Young people in these communities may not see the link between education and life chances, emphasising instead the role of luck (Leondari et al. 1998) or family connections in a world where they do not feel that their educational efforts will necessarily be rewarded. Appreciating the role of expectations in motivation is not new (e.g., Vroom 1964), but it has generally been neglected with respect to access to higher education.

Thirdly, the theory of possible selves provides policymakers with a vector for how to increase attainment in Key Stages 3 and 4 (11- to 16-year-olds), which is both positive in and of itself, as well as helping to narrow the gap in higher education participation and access to elite universities in a stratified system. Rather than focusing on the ‘carrot’ provided by higher education, policy could promote a more general focus on individual futures and how young people can be supported in articulating strategies for achieving them (Kirk et al. 2012). Importantly, Oyserman et al. (2011, p. 488) conclude that ‘economically disadvantaged children care about school, but they are less likely than more advantaged children to have salient behavioural strategies to make their school-focused possible identities come to fruition’.

Finally, and following on from the previous point, policymakers need to create ‘space’ for young people to conceive, elaborate and plan their like-to-be selves—to create their ‘roadmap connecting the present to the future’ (Oyserman et al. 2004, p. 132). This includes temporal space within a pressured curriculum, as well as access to adults that are able to support them, whether this is through higher education outreach programmes, charitable organisations, or staff in schools. It is important to assert the distinctiveness of possible selves work relative to traditional careers guidance, although it may well be that careers professionals are well-placed to engage with this agenda despite recent funding cuts and upheaval in provision (Watts 2013). A possible selves programme within schools would seek to increase young people’s confidence in their own ability to succeed educationally (self-efficacy) and positively affect outcomes (locus of control), working with key influencers like parents and teachers to challenge ingrained expectations, and providing them with metacognitive skills, micropractices and learning orientations (Watkins 2010) to help them realise their like-to-be selves.

In summary, the evidence supporting the effectiveness of a possible selves approach is considerably stronger than that available for aspiration-raising and there is therefore the need for policymakers to shift the discourse onto more fertile ground.

6. Conclusions

The stratification of higher education is complex, with its roots in persistent structural inequalities that defy ready solutions. Indeed, despite concerted policy attention and considerable financial investment, improvements over the last twenty years in England have been modest (Harrison 2017). This paper has aimed to address one possible piece in the much wider puzzle, while recognising that it cannot resolve the systemic issues that mean that the chances of benefiting from higher education are closely aligned with the circumstances of one’s birth.

This paper has sought to draw on the theory of possible selves, which has enjoyed significant success in terms of understanding motivation and behavioural change, but which has not been extensively used within education—or specifically with respect to access to higher education. It has argued that it provides a rich lens for critiquing the recent policy and practice focus on aspiration-raising and its questionable impact on school attainment and participation patterns. The conceptual model embodied in Figure 2 provides a means for understanding how the identification and elaboration of possible selves can influence the demand for higher education through helping disadvantaged young people to develop strategies to work towards desirable possible selves, as well as identifying four conceptual points at which planned interventions might usefully be made. This has aimed to provide an alternative agenda for outreach activities and policymaking that addresses the shortcomings of the prevailing focus on aspiration-raising. While this has been discussed within the English context, it is likely to be relevant in any country with a highly-stratified higher education sector where identifiable social groups are markedly under-represented.

As there is a relatively small body of empirical evidence to support a possible selves approach, this paper also provides a framework for future research. Studies might explore, inter alia, whether elaborated possible selves increase the motivation for school work outside the US context, the relationship between disadvantage, self-efficacy, and locus of control, effective practices to influence young people’s expectations about their future, the impact of possible selves on elite higher education or the mediating effects of gender, ethnicity, and rurality with respect to possible selves and their relationship with higher education. Certainly, the field needs a new vitality that reaches beyond the simplistic deficit model provided by aspiration-raising.