Problematising flagship ‘disadvantage’ policies in English schools: agenda setting and incoherence in the absence of an over­arching theory of change

This paper draws on research conducted in four state schools with sixth forms to problematise two flagship ‘disadvantage’ policy agendas in the English context: the Pupil Premium (focusing on the narrowing of attainment gaps) and widen- ing participation (focusing on fairer university access). While such ‘priority’ policies necessarily incorporate the promise of change, it argues that multiple forms of incoherence militate against this, including: their relative agenda-setting power in a highly marketised system; the contested, constitutive power of different proxy indicators; competing policy preferences that under-attend to the intersections between educational opportunities and material disadvantage. In contrast, interviews with school staff highlighted the transformative potential of lived experiences of disadvantage and reinforced the importance of adopting a wider lens than that of the school. The paper concludes that the absence of an over-arching theory of change is part of ‘the problem’ and it suggests where the mapping of one might begin.


Introduction
This paper problematises two flagship policy agendas in the English context that share a common focus on 'disadvantage': the Pupil Premium (focusing on the narrowing of attainment gaps) and school-based widening participation interventions (focusing on fairer university access). While these are separate agendas with their own logics they are of course related, as the lower prior attainment of less socially advantaged young people is known to limit higher education access not only in the English context (Gorard, Boliver et al., 2019;Pickering, 2019) but internationally (Jerrim & Vignoles, 2015). Flagship policies such as these need to be situated in the context of both place and time (Maguire et al., 2015) as well as in the particularities of national education systems (Frandji, 2012, p. 15). They also operate in the context of multiple intersections, including between the educational, social, economic, and political . Drawing on data collected in four purposively selected 11-18 schools with sixth forms in the South-East of England the paper identifies various forms of policy incoherence that militate against change, including: the relative agendasetting power of such policies in a highly marketised system; the contested, constitutive power of the proxy indicators used; competing policy preferences that under-attend to the intersections between educational opportunities and material disadvantage. In contrast, staff interviewed highlighted the transformative potential of lived experiences of disadvantage and the importance of adopting a wider lens than the school. While Ball (2021, p. 388) correctly argues that it is important not to 'over-estimate the sensemaking capacity of policy actors,' the paper concludes that the absence of an over-arching theory of change is part of 'the problem' (Bacchi, 2004, p. 131) and it ends by suggesting where the mapping of one might begin. Hay (1997, p. 48) highlights the 'agenda-setting' function of policy, noting that it is both 'context-shaping' and 'conduct-shaping' and drawing a distinction between its direct and indirect effects (Hay, 1997, p. 51). As policy involves 'multiple circuits of power' (Rose, 1999, p. 5) school staff are not only positioned policy subjects but also powerful agents of policy Levinson et al., 2009). They encounter a plethora of policies (Braun et al., 2010) linked to agendas of varying strategic importance , some of which are also in tension Dale, 1989). Braun et al. (2010, p. 558) highlight the importance of the 'policy, practice, positioning' nexus and the research included the perspectives of staff leading the Pupil Premium and widening participation agendas in each school. While it aimed to utilise these different strategic positionings to better understand how 'disadvantaged' young people were being (differently) identified and supported under the widening participation agenda in each school, the data suggested the opportunity for a more wide-ranging problematisation: one that integrates an interest in maximising the potential for change with an understanding of the importance of the 'who' and 'what' of policy (Ball et al., 2011, p. 611) and issues of power. Hay (1997) who lays down the challenge to 'disarticulate the analysis and identification of power from its critique' (Hay, 1997, p. 50) as part of a discussion of Lukes (2004) 'three-dimensional conception of power (that is, one sensitive to power as decision-making, agenda-setting and preference-shaping' (Hay, 1997, p. 48) has been particularly useful here. While the resulting analysis is necessarily specific to the English context, it is more widely relevant as 'disadvantage' has come to dominate education policy agendas internationally (Tormey, 2010) as these have become increasingly 'homogenous' (Ozga, 2021, p. 301) and because issues of power transcend national boundaries.

Schools, disadvantage, and change
While education policy agendas designed to address disadvantage necessarily incorporate the promise of change, in England, as in many other contexts, it continues to be strongly associated with experiences of schooling. Peter et al. (2010) argue that differences in the strength of this intersection stem in part from differences in state welfare policies. However, Mertanen et al. (2020) in a discussion of Finnish youth policies note that even policies conducted within a single policy agenda may incorporate conflicting ideological positions. Variations in the intersections between schools, their intakes and local area add another level of complexity (Jensen, 2013) as do the very particular spatial and temporal processes around which disadvantage becomes intergenerationally tied (MacDonald et al., 2020). Ainscow et al. (2012, p. 12) attribute the failure to disrupt the association between schools and disadvantage to 'the contradictions between different agendas' and they identify the effects of marketisation policies as a particular cause for concern. McInerney (2007) writing about the Australian context argues that these, when combined with strong accountability mechanisms, have tended to accentuate rather than ameliorate between-school differences. Forsberg (2018, p. 901), in an analysis of the local education market in Stockholm (Sweden), also notes that for more 'socially exposed' schools, 'competencies' such as support for 'disadvantaged' (in this case immigrant) communities become a risk rather than a marketable commodity. Notwithstanding, successive governments in the English context have consistently adopted performance-based approaches to raising attainment levels in schools located in 'disadvantaged' communities (Burn & Childs, 2016;Lupton, 2011). While these kinds of 'imperative/disciplinary' (Ball et al., 2011, p. 612) agendas achieve more traction, Raffo et al. (2009, p. 351) argued more than a decade ago that they exaggerate the contribution that schools can make as the issues are wider and more structural, leading them to conclude that such policies need to be underpinned by a more explicit and coherent theory of change. Hay (2014, p. 476) notes, however, that a level of policy incoherence is to be expected as: 'different parts of the state do different things in different ways with different degrees of autonomy to yield, in all likelihood, a great variety of contradictory effects.'

Operationalising 'disadvantage' within flagship policy agendas
As the term 'disadvantage' incorporates an understanding of relative/hierarchical social positioning its use is strongly associated with stigma (Dyson & Raffo, 2007;Lukes, 2004;Pemberton et al., 2017). Lukes (2004, p. 28) argues that where power is being used to build consensus it is at its most insidious and Tormey (2010, p. 189) notes that, despite its political and contested nature, the term tends to be used uncritically in policy and practice discourse. In contrast, debates about the proxy indicators preferred within flagship agendas such as these tend to focus on more technical questions such as the failure of area-based measures to adequately distinguish between individual, institutional, and local characteristics (Dockery et al., 2016;Harrison & McCaig, 2015;Pickering, 2019). Such concerns reflect the difficulty of capturing the multidimensionality of disadvantage (Smith & Smith, 2014) but also a desire to see finite resources targeted more accurately (Boliver et al., 2015;Dockery et al., 2016;Harrison & McCaig, 2015). Boliver et al. (2015) argue that in the English context those persistently eligible for Free School Meals are amongst the most disadvantaged because this measure is strongly associated with low-income benefits. However, Gorard, Siddiqui et al. (2019, p. 5) note that not all eligible young people are registered and that higher rates of in-work poverty have increased the number experiencing material disadvantage but not included within this measure. Indicators such as these have the power to construct the thing that is being measured (Ladd & Fisk, 2011;Piattoeva & Boden, 2020;Rose, 1999), with differing political preferences making it possible to be identified as 'disadvantaged' for some purposes but not for others. Peter et al. (2010, p. 260) argue that it is important to connect education with social policy analysis as an understanding of both is required 'if transformative social change is to develop across schools.' In England, the extent to which education policies designed to address 'disadvantage' have been aligned with wider social welfare agendas has varied over time (Burn & Childs, 2016;Dyson & Raffo, 2007;Featherstone & Manby, 2006;Ridge, 2013). The election of the Coalition government in 2010 is frequently identified as a turning point (Craske, 2018;Lambie-Mumford & Green, 2017;Ridge, 2013), with the global financial crisis of 2008 acting as a catalyst for austerity policies that have re-drawn the boundaries between the individual and the state, contributing to a sustained decline in the 'material conditions for low-income households' (Pemberton et al., 2017(Pemberton et al., , p. 1160 and increased recourse to crisis services such as food banks, including in rural areas (May et al., 2020). Concerns have been raised about the disproportionate impact of changes initiated under this agenda on children living in lone-parent households and larger families (Lambie-Mumford & Green, 2017;O'Connell et al., 2019). That experiences of material disadvantage impinge directly on experiences of schooling is suggested by a study conducted with Primary teachers (ages 4-11) in the North-East of England that highlights the effects of hunger on learning and behaviour (Gooseman et al., 2020). Policy changes directly relevant to the two flagship agendas that are the focus of this paper include the replacement of the Education Maintenance Allowance for sixth form students with a narrower school-based bursary system (Ridge, 2013) and an increase in university tuition fees that has shifted much of the cost of university from the state to the individual (T. Clark et al., 2019;Johnston, 2013).

An introduction to the two agendas
Both flagship agendas can be located amongst 'priority education policies' defined by Frandji (2012, p. 3) as being 'designed to have an effect on educational disadvantages through systems or programs of focused action. ' Bacchi (2004, p. 131) notes that policies contain not only 'representations' of 'problems' and assumed 'solutions' but also an indication of what might change and what might not. What follows is a necessarily brief overview of each.

i. The Pupil Premium agenda
The Pupil Premium agenda continues to evolve but was first introduced by the Coalition government in 2011. 'Disadvantage' is largely constituted as those eligible for Free School Meals at any point in the last six years and those adopted from care or 'looked after' by the local authority, with the addition more recently of children living in families with no recourse to public funds (DfE, n.d). The actual provision of Free School Mealsa subject that has become increasingly a focus of political controversy -is not covered under this agenda. The Pupil Premium agenda covers all statefunded schools and contains two core components: • Funding targeted at school level on the basis of the proportion of 'disadvantaged' young people in its intake • Accountability measures focusing on performance gaps between this 'disadvantaged' group and 'all' others in each school. Ladd and Fisk (2011) writing about the weighted school funding system operating in the Netherlands emphasise that its socially progressive nature reflects its very particular contextual and political dimensions. Lupton and Thomson (2015) note that the announcement of the Pupil Premium funding stream coincided with shortfalls in school income linked to budget cuts initiated under the austerity agenda. Granoulhac (2017, p. 436) further argues that this funding was relatively small in the light of savings made in areas such as youth services. While it was originally hoped that linking the Pupil Premium to school funding might reduce the tendency for young people eligible for Free School Meals to be concentrated in particular schools the evidence on this is inconclusive Morris & Dobson, 2021). The Pupil Premium is a high-stakes agenda, being reported on in mandatory inspections (Craske, 2018;Morris & Dobson, 2021). Schools are encouraged to adopt evidence-based approaches to intervention but have flexibility over how the funding is used. It can include such things as covering the cost of additional staff and providing targeted academic and welfare support (Craske, 2018;Yaghi, 2021). Craske (2018) argues that the need to evidence how this funding has contributed to improved performance outcomes has incentivised increasingly reductive approaches.  conclude that the effect on attainment gaps is not yet clear. Fleming and Harford (2021) writing about the Irish DEIS (Delivering Equality of Opportunity in Schools) programme argue that because educational disadvantage is also an economic issue, too much emphasis is being placed on the role of schools. te Riele (2011) in a discussion of the Compact with young Australians notes a similar focus on raising minimum attainment levels at school-leaving age but also additional requirements around work and welfare. The Pupil Premium agenda can be seen to share these broader economic and social rationales.

ii. The widening participation agenda
The widening participation agenda pre-dates the Pupil Premium and it aims to change unequal patterns of university access, including by socioeconomic status, gender, ethnic group, course of study, university status and characteristics of school attended. Osborne (2003) highlights the long legacy of this agenda internationally, noting that it contains a number of strands, of which university-led, school-based widening participation outreach activity is only one. In the English context, such activities are offered increasingly intensively between the ages of seven and 18 and they include such things as campus visits, residential summer schools and subject 'taster' sessions. Those emphasising 'aspiration raising' have been critiqued for incorporating deficit assumptions (Evans et al., 2019;Rainford, 2021). Within this agenda 'disadvantage' has been variously constituted as preferences for particular indicators have changed. An early emphasis on prospective first-generation entrants reflects the assumption that differences in patterns of access are closely connected with family knowledge and experience of university. However, Patfield et al. (2022), writing in the Australian context, highlight considerable within-group variation and suggest something closer to a continuum of relative disadvantage. A subsequent shift in favour of area-based proxy indicators has been contested because this is seen to have increased the risk of resources being misdirected towards the relatively more advantaged (Dockery et al., 2016;Harrison & McCaig, 2015).
In England, the widening participation agenda is less powerfully positioned in school-level performance systems than the Pupil Premium agenda as it is the universities offering outreach activities that are held to account through recruitment targets linked to fee levels (Evans et al., 2019;Rainford, 2021). While Osborne (2003, p. 54) concluded more than two decades ago that there might be better ways to leverage change, Pickering (2019) concludes that higher status universities in England have been incentivised to focus their support more narrowly on young people living in areas where university participation rates are lower and/or attending schools with higher rates of eligibility for Free School Meals. Nevertheless, national data continue to show large and continuing differences in the university participation rates for 'disadvantaged' young people, including when measured by Free Schools Meals eligibility, care experience and ethnic group (GOV.UK, n.d). Although there has been a lot of focus on the increased cost of university post austerity, it is not yet clear whether this has deterred participation or simply changed pre-existing patterns of stratification (Callender & Mason, 2017;S. Clark et al., 2015;Harrison, 2019). While there continue to be clear social justice rationales for this agenda, there are also economic motivations (no wasted 'talent') and a potential for it to be used in ways that justify the status quo (Gazeley, 2018).

Research design
Four non-selective, state funded schools catering for young people aged 11-18 were purposively selected to take part in the research. This ensured some variation in terms of: the proportion eligible for Free School Meals; performance in national examinations; local context. This variation has been reflected in the pseudonyms used to preserve their anonymity: Coastal, Rural, Urban, and Town. Each school had its own sixth form provision, something that is now less common in the English context as young people often move to larger and/or more specialist sixth form colleges or Colleges of Further Education at age 16. This leads Manley and Johnston (2014) to argue that young people attending smaller school sixth forms are disadvantaged by a more restricted curriculum offer and more limited support with university admissions processes.
Ethical approval for the research was obtained from the University of Sussex Social Sciences & Arts Research Ethics Committee in May 2016 [ER/ LHG20/3] and all participants were asked to provide informed consent. Head teachers were approached by email and once four had confirmed institutional willingness to participate, arrangements were made to collect data in each setting. School staff were purposively selected to ensure a range of perspectives on the two agendas (n = 16). In each case, they included: • The institutional leads for Widening Participation (two teachers and two non-teaching staff) • The institutional lead for Pupil Premium (all teachers) • A member of the school leadership team • A long serving member of support staff, ideally living locally.
Individual semi-structured interviews 1 included: discussion of how young people were being selected for widening participation outreach activities; the characteristics of the institutional and local context; reflections on any implications for policy and practice. The institutional lead for widening participation was also asked to identify a sample of prospective firstgeneration entrants accessing school-level interventions to take part in focus group discussions that explored how they were being supported in developing their knowledge and experience of university. A total of 74 were recruited across all four schools, covering three time points: • Age 13/14 (when widening participation interventions often start) • Age 15/16 (just prior to sixth form study, when high stakes national examinations are taken) • Age 17/18 (sixth form study, when final examinations are taken and university choices often made). 2 The school-level data were collected over the course of a single day. While this afforded good access to situated understandings of both policy agendas from a range of perspectives, it precludes a focus on policy enactment although both interviews and focus groups afforded some insights into aspects of this. At one school, four parents also agreed to be interviewed by telephone. Additional data were collected from academics 3 and stakeholders 4 living locally who were invited to participate in a consultative seminar at which the school-level findings were presented. This included an opportunity to discuss the wider implications for policy and practice, with reflections captured in a short questionnaire completed anonymously at the end of the day (12 out of 20 returned). Data analysis for this paper has been an extended, abductive process, incorporating the following elements described in Richards (2015): • Emergent theories (what you bring in, in this case a critical perspective on both disadvantage and change) • Opening up the data (questioning, interpreting and reflecting, supported by annotations and visual mappings) • Building ideas into themes (including through iterative engagement with the data and relevant literatures).
Given the particular focus of this paper data from the parent interviews have not been cited although they have informed the analysis.
Locating the four schools Thrupp and Lupton (2011, p. 290) suggest that different institutional contexts can be located on a continuum that stretches from the 'cumulatively advantaged to the cumulatively disadvantaged.' Although the South-East is one of the more affluent areas of England, there are schools where performance in public examinations taken at age 16 is below the national average (Social Mobility Commission, 2017). The area also has one of the lowest university participation rates for 'disadvantaged' young people (DfE, 2019). Nevertheless, none of the four schools would be considered amongst the most disadvantaged nationally on the basis of the proportion eligible for Free School Meals. Town had the lowest levels of attainment as measured by performance in national examinations taken at age 16 as well as the highest proportion eligible for Free School Meals. Its intake was described as: A core of white British kids who seem to come from the same families over and over again . . . and then this transient bit that tends to change with the fluctuating political whims of the world (Town 3) Coastal and Urban schools were described as having more socioeconomically mixed intakes and Rural as having 'quite an affluent catchment' (Rural 1). However, Robertson and Dale (2013, p. 429) note that education systems operate as 'spaces of socially-differentiated positions.' Although Rural school had the lowest proportion eligible for Free School Meals and was the highest performing in examinations taken at age 16, it also had the joint largest within-school attainment gap between 'disadvantaged' young people and 'all.' This lower level of attainment was also no different from that of 'all' young people attending Town. Consideration of institutional performance in examinations taken at the end of sixth form study 5 further complicate the picture as all four schools were now performing below the national average. While average attainment at Rural and Coastal was broadly similar and only just below the national average, performance at Town and Urban was again broadly similar but slightly lower. While it is possible to position Town and Rural at opposite ends of a continuum of relative disadvantage, over-reliance on single performance indicators and/or time points can mask more nuanced intersections between institutional and individual disadvantage.

Discussion of findings i. Locating flagship agendas within a policy hierarchy
This section focuses on the relative power of these two flagship policies in a marketized and highly performance-orientated system. Interviews with staff across all four schools highlighted the 'agenda setting' (Hay, 1997, p. 48) power that comes with a preference for 'imperative/disciplinary' (Ball et al., 2011, p. 612) approaches to education policy: So we now report and talk about retention and destination figures -as you wouldn't be surprised to know -since the Department for Education introduced that measure (Coastal 3) Their 'context' and 'conduct shaping' (Hay, 1997, p. 51) power could also be seen to have stimulated an incipient interest in building connections between the two agendas: We've never really targeted Pupil Premium students [for widening participation interventions] whereas now it's a case of I think we should be looking at supporting those students (Town 2) Nevertheless, there was still some uncertainty about the extent to which this was happening, in part because the two agendas were led by different staff but also because both 'problems' and 'solutions' (Bacchi, 2004) tended to be constituted discretely rather than holistically: We provide food and uniform for some [Pupil Premium] children when necessary . . . and some of those are clever children and I suppose we would target them (Coastal 2) Maguire et al., (2015) note that some policies can become very tied to an individual actor and Pupil Premium leads appeared to be more centrally and powerfully positioned. In contrast, widening participation leads expressed feelings of professional isolation and reflected on the difficulty of getting whole school buy-in to the agenda: Staff see it as being additional work, additional time, additional responsibility and no additional money and it's making sure that I communicate to staff what's already going on and the importance of it (Town, 2) As also found by Gazeley (2018), this less powerful positioning could make it difficult to get permission to take young people out of school to attend activities hosted on university campuses: 'because of lost learning time, because of the focus on academic progress' (Rural 4). The existence of a 'disadvantage' policy hierarchy was further suggested by one participant who described having been able to use the Pupil Premium agenda to leverage permission to take young people out of school to attend a widening participation activity when this had previously been refused: Discussion of the composition of the sixth form at Rural suggested that closer alignment of the two agendas is important as it could offer a means of better scaffolding the educational trajectories of some of the most disadvantaged young people overtime: We have no looked after children 6 in the sixth form at the moment . . . and yet there are students lower down in the school . . . so what barrier is there to coming to our sixth form for those students? And very few Pupil Premium . . . A handful who have had free school meals in the past five years (Rural 3).
Staff interviews pointed to the strength of the connections between local education markets and the composition of school sixth forms. At Town, 'the most socially exposed' (Forsberg, 2018, p. 901) of the four schools one participant noted that 'If you're coming here you're coming here with the perception of having failed' (Town 3) and another used this intake to explain a 'very low uptake of university' (Town 2). In this instance institutional 'disadvantage' can be seen to be being actively constituted through the policy preference for a marketised system that incorporates strong accountability systems (Ainscow et al., 2012;McInerney, 2007). In this case, incoherence can be seen to stem from preferences that indirectly contribute to the need for widening participation interventions.

ii. The constitutive power of 'disadvantage' indicators
This section affords insights into how the preference for particular 'disadvantage' indicators both opens up and closes down access to the additional educational opportunities offered under each agenda. Based upon intake and attainment-levels Town best met the profile of a 'widening participation' school. Consequently, staff were working with several local university outreach teams and therefore able to be selective about which opportunities to take up: 'We kind of dip in and out of what we need which is great for us' (Town 2). In contrast, Coastal and Rural had much more limited access to widening participation interventions and staff indicated that this placed them in an uncomfortable position of power: So life is a competition. If I've got more students to take up the number of places available for anything it's first come first served . . . That's the bit I don't like and I'm in a way playing god (Coastal 4).
While staff at Town saw nothing problematic about interventions being targeted towards schools with 'less than average university goers', some at Coastal and Urban questioned the validity of the approach. One also suggested that the 'context' and 'conduct-shaping' power (Hay, 1997, p. 48) of the current political preference for area-based indicators within this agenda might have contributed to the withdrawal of a local university from a long-standing partnership: I don't know whether they're just looking at our school's postcode which, as I say, puts us into an affluent society that progression [to university] is good but our kids aren't from that postcode (Coastal 4) Piattoeva and Boden (2020, p. 14) note that the need to turn complex phenomena into something measurable involves a process of 'decontextualization' that is distant from the 'complex living texture of the world' and in general staff suggested a preference for more nuanced approaches to the identification of disadvantage, with access to interventions informed by personal knowledge of individual circumstances: Some of your Pupil Premium students may . . . actually have been on Free School Meals for a very short period of time when someone in a reasonably affluent job was made redundant and had perhaps six months out of work . . . There will be other children who will go right through their schooling . . . needing that sort of financial support (Coastal 2) Staff at Urban also reflected that there was sufficient stigma around these kinds of identifications to limit open discussion of their rationales. One described this as an 'ethical dilemma' while also noting that these rationales might be known even if not disclosed: [The young person said] 'You know, I got chosen because I'm from a poor background.' . . . I thought I might challenge her and say: You're not from a poor background, your background's very wealthy in lots of ways (Urban 4) Discussion of these issues led one seminar participant to reflect on the need for: 'Training with teachers around their role as "gatekeepers" -the responsibility of this.' Another reflected on the challenges involved in: 'Striking a balance between meeting the needs of the most needy and avoiding a competitive discourse around disadvantage. ' Robertson and Dale (2013, p. 429) note that such 'decision-making power' aids 'structural processes that create and maintain privilege and disadvantage.' However, in the context of these two flagship policies, decision-making power can also be seen to be actively constituting hierarchies between the disadvantaged. Incoherence is suggested by Rural's relative lack of access to widening participation outreach interventions despite the young people contributing to the focus groups there demonstrating comparably low levels of knowledge and experience of university.
iii. Keeping the materiality of disadvantage in Lukes (2004, p. 29) argues that a three-dimensional view of power aids understanding of 'the many ways in which potential issues are kept out of politics.' In England, successive governments have tended to foreground cultural, familial, and attitudinal explanations of disadvantage and to downplay its material and structural dimensions (MacDonald et al., 2020). Staff could be seen to constitute the 'problem' of 'disadvantage' differently, with reflections on the likely causes also shaping their understandings of possible 'solutions' (Bacchi, 2004): Obviously Pupil Premium a lot of that is based on free school meals. So we are looking at income as a family. But whether that holds students back? I guess just general absence rates mean they don't achieve what they can do (Town 1) Despite their contextual differences, staff in all four schools recognised that material disadvantage was a part of the context in which they were working: Some of our children we give vouchers for the Food Bank, for example, we've got children who are as hard up as that. (Coastal 2)  suggest that income threshold levels are set so low that not all young people affected by material disadvantage are able to access support under the Pupil Premium agenda, a point also noted by some staff interviewed: There's this huge margin who miss out on support and I think that's the main concern really . . . the children whose parents don't just qualify for Free School Meals, but really struggle to make ends meet (Urban 2) Staff across all four schools also identified material disadvantage as directly impacting on opportunities associated with the widening participation agenda. For example, a member of staff at Town reflected ironically on the expectation that families would be able to meet the cost of travel to events held on university campuses. One young person attending the sixth form at Urban also described how the cost of travel had prevented them from visiting any universities to inform their final course selection. Staff accounts highlighted the inter-dependent nature of material constraints, noting that for some young people just staying on to study in the sixth form would be too difficult: This was the first school I ever worked at where I had students in the sixth form leaving school to work because their parents didn't have any money (Town 3) A member of staff at Coastal suggested that the departure of young people at this crucial juncture was a 'problem' that had been exacerbated by the withdrawal of the Education Maintenance Allowance. However, the most frequently raised concern was the increased cost of university. Staff across all contexts reported having to work harder to reassure young people around the benefits of a degree: I've been teaching a long time and [student debt] wasn't a conversation I would ever have had to have years ago but some of them are concerned about that (Rural 1) There is some evidence to suggest that this increase in cost has changed prospective first-generation entrants' university decision-making (Henderson et al., 2020) and staff clearly recognised that differently positioned social groups would be differently affected: It's going to be more of a worry for low-income families than it is for those that don't have to worry about it (Town 1) One participant explicitly identified this change as a policy preference associated with the austerity agenda, while at the same time evidencing their understanding of the inherent connectedness of these two flagship agendas: We could reduce how much it costs to go to University, I think that would see participation rates rise for Pupil Premium students, but I don't think that'll happen anytime soon, in the financial climate we're in (Urban 4) Although neither agenda is intended to address the material dimensions of disadvantage directly, staff accounts evidence the impossibility of keeping the material out. At the same time, they point strongly to the 'contradictions' that stem from competing policy preferences (Hay, 2014, p. 476).

iv. The transformative potential of lived experiences
Despite a tendency to emphasise the intractability of the relationship between disadvantage and schooling (Raffo et al., 2009), change is possible and teaching affords particularly good opportunities for social mobility (Braun, 2015). More than half of staff interviewed self-identified as having come from a disadvantaged background themselves and they indicated that these experiences directly informed their responses to both agendas: [I] was the first person in my family to go to University; the first person in my family probably to go to college, and all those sorts of things. So I think for me . . . its making sure that students know that there's opportunities (Urban 4) Several staff highlighted the need to look beyond the period of formal schooling as disadvantage is often associated with delayed and more recursive educational trajectories (Hanrahan et al., 2020): I went out to work and eventually went on to do a degree . . . So it's a lifelong learning type of message that we need to get across (Town 1) A member of staff who recognised that the period of formal schooling could be difficult for some young people emphasised that they explicitly communicated that there would be second chances: They need to know that you don't have to necessarily get it right at school, you can still be successful (Coastal 4) Another participant drew on first-hand experiences of growing-up on one of the poorest local estates to highlight the importance of education services located in the community but noted that austerity policies had affected access to these: The importance of an understanding of the local context was further suggested by a member of staff at Town who described how they were using the example of a teacher who had grown up on one of the most 'disadvantaged' local estates and gained a doctorate to change life chances: These kinds of small political acts (Winter et al., 2020) can challenge deficit assumptions (McKay & Devlin, 2016) but they also reinforce the importance of the positioned actions of school staff working in the context of flagship policy agendas such as these.

Conclusion: towards an over-arching theory of change?
This paper adds to literature seeking to problematise policy (Ozga, 2021) through its specific focus on the association between schools, disadvantage, and change. It evidences the analytical purchase that can be generated by juxtaposing flagship 'disadvantage' policies such as these rather than focusing on them singly. It focuses attention on the risks inherent in assuming that simply sharing a common focus on disadvantage will ensure that policies are aligned and mutually supporting. At the same time, it highlights the impossibility of divorcing flagship policies such as these from the national contexts that produced them and the local contexts in which they subsequently operate. Its focus on the 'agenda-setting' power (Hay, 1997) of these policies highlights numerous forms of incoherence that militate against change, adding weight to the claim that there is a need to disturb the consensus that has arisen around 'disadvantage' as a policy priority (Tormey, 2010) which as Lukes (2004, p. 28) notes is power working at its most insidious. While Hay (1997, p. 51) warns that it is not the job of the critical researcher to suggest what forms of change are desirable, the paper supports the contention that the absence of an over-arching theory of change is part of the 'problem' (Raffo et al., 2009). The following sections therefore begin to map some possible starting points. Peter et al. (2010, pp. 260-261) argue for better understandings of 'the linkages between education and broader social policy contexts' and any overarching theory of change would therefore need to recognise these intersections and anticipate their direct and indirect effects. Dale (1989, p. 32) argues that the inherent contradictoriness of state education policies and systems is exposed 'at times of major crisis of accumulation' and staff accounts clearly evidenced how a preference for small state approaches to social support (Pantazis, 2016) at a time of national retrenchment can undermine flagship policies such as these. A statement published following the visit to the United Kingdom of Professor Philip Alston, United Nations Special Rapporteur on extreme poverty and human rights (Alston, 2018, p. 22) included the assertion that: 'The experience of the United Kingdom, especially since 2010, underscores the conclusion that poverty is a political choice.' While the UK government contests this analysis, any over-arching theory of change would have to engage with the multiple ways in which material disadvantage threads through school experiences over time, either constraining or opening-up opportunities as they are shaped by competing preferences that extend well beyond the remit of individual schools. Harrison and Waller (2017) emphasise the importance of clarity about each component in any theory of change. The need to produce an overarching theory of change might raise awareness of the many areas of contention that surround the use of 'disadvantage' indicators but also their constitutive effects (Rose, 1999). Neither of these flagship agendas provides support on a universal basis and Van Lancker and Van Mechelen (2015, p. 73) note that when it comes to poverty reduction: 'Targeting as such might not be the problem; rather it is important how targeting is done.' However, such questions are not simply technical (Piattoeva & Boden, 2020;Tormey, 2010) as the associated decision-making processes have real-world effects. This includes leaving some young people who are objectively inscope for intervention, out-of-scope for reasons that suggest a degree of arbitrariness in the exercise of power. Reay (2020) argues that particular policy approaches are taken up because they resonate with the preferences of policymakers and the need to make these preferences explicit and reconcile them might perhaps contribute to change through the more nuanced distribution of resource.
Both flagship agendas focus solely on the period of formal schooling and an over-arching theory of change would also need to be informed by the understanding that educational trajectories develop differently, in context and over time. Lister (2002, pp. 44-45) argues that incorporating the perspectives of those with lived experiences of poverty would contribute to better policymaking. Staff accounts clearly suggested the value of including the perspectives of those who combine an understanding of the everyday workings of flagship agendas such as these with lived experiences of disadvantage. Although such an exchange of perspectives seems less likely in the context of the more oppositional relationships developed between teachers and English policymakers post 2010 (Braun, 2015), it is all the more urgent in the face of continuing national and global challenges.

Notes
1. At one school, two staff elected to be interviewed together. 2. Despite efforts to do so, it was not possible to interview young people aged 17/18 at Town. 3. Researchers and teacher educators. 4. School staff including widening participation leads plus widening participation staff working in universities and third sector organisations to deliver this flagship agenda. 5. Typically taken at age 18 and important for university access. 6. Children with care experience.

Acknowledgments
I would like to thank Dr Tamsin Hinton-Smith and Dr Jacqui Shepherd for their contribution to the research on which this paper draws. The analysis presented in this paper is, however, my own.

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

Funding
This work was supported by the Sussex Learning Network.

Notes on contributor
Louise Gazeley is a Senior Lecturer in Education and Director of the Centre for Teaching and Learning Research (CTLR) in the School of Education and Social Work at the University of Sussex, England. Louise's research focuses on the intersection of educational with social disadvantage. It combines a focus on inequalities in university access with how education policies and practices in schools can be developed to improve future life chances.