Whose target group is it anyway? The messy business of enacting widening participation policy

ABSTRACT The widening participation (WP) policy agenda directs higher education institutions (HEIs) in England to deliver measures aimed at increasing participation rates among under-represented groups. A central aspect of this is the idea that targeting must be used to reach individuals belonging to these groups. However, the detail of how to go about putting this into practice has largely been left to the interpretation of HEIs and the staff working within them to deliver WP. Drawing upon in-depth interviews with WP practitioners to gain insight into how this policy is being enacted on the ground, this paper will demonstrate how external and situated contexts shape the interpretation of targeting; in particular, the closely interlinked contextual factors of an institution’s position in the HE market, makeup of its student body, and nature of its relationship to the regulator. Furthermore, examining how social class, ethnicity and gender feature within targeting approaches, it will show how different contextual factors make certain target groups possible while effectively precluding others.


Introduction
The widening participation agenda has been a prominent aspect of the English higher education (HE) policy landscape for the last twenty years. Initiatives delivered under the banner of WP have been framed as redressing historic and ongoing inequalities in HE participation; inequalities that have been especially prominent in relation to social class or socio-economic status (SES) (Boliver 2013;Chowdry et al. 2010), ethnicity (Bowes et al. 2015;Crawford and Greaves 2015), and gender (Broecke and Hamed 2008;Crawford et al. 2016). Through government policy and regulation, HE institutions have been directed to put in place measures to widen participation and to target these efforts at those from under-represented groups. This has been especially true of activities aimed at children and young people -outreach programmes such as summer schools and mentoring schemes -chiefly delivered by discrete WP departments within universities. Over the last two decades, the staff working in these departments (commonly referred to as WP practitioners) have been subject to shifting messages about targeting, with the conception of 'WP target groups' reconfiguring over time (Holland et al. 2017;Stevenson, Clegg, and Lefever 2010). At the same time, while policies put in place parameters for action, they 'rarely dictate or determine practice' (Ball, Maguire, and Braun 2012, 3). In the context of WP, the particulars of how to go about putting targeting into practice have been left to HEIs and the staff working within them to deliver this agenda. This combination of autonomy and constraint has resulted in a multiplicity of approaches being used across the sector which share a 'family resemblance' but vary considerably in the details.

The widening participation policy agenda in England
While the concept of increasing HE participation among under-represented groups has been around for far longer (Ross 2005), the publication of the Dearing Report (NCIHE 1997) marked a step change in policy focus on WP and a new era of activity aimed at realising New Labour's goal of a 50% participation rate (Thompson 2019). This included significant additional funding being distributed to institutions through the widening participation allocation (also known as the 'postcode premium') according to the number of recruited students likely to be in greater need of support (HEFCE 2003). The 2003 White Paper The Future of Higher Education announced the introduction of variable tuition fees and, along with these, the requirement for HEIs wishing to charge higher fees to produce access agreements (AAs) outlining their plans for improving participation among under-represented groups (McCaig 2018). Fair access was framed around the 'unacceptably wide' gap in participation between social class groups and increasing 'the proportion [of students] coming from lower-income families' (DfES 2003, 8), although this would shift over time with an increased focus on student outcomes alongside access (BIS 2014). The Higher Education Act 2004 legislated these changes and established the Office for Fair Access (OFFA), an independent regulatory body which would support the statutory duties of the Director of Fair Access, including the approval and monitoring of AAs (BIS 2016a). More recently, the Higher Education Research Act 2017 established a new HE regulator, the Office for Students (OfS), combining into one body the duties of the Higher Education Funding Council for England (HEFCE) and OFFA, and replacing AAs with access and participation plans (APPs). At different times each of these bodies have played an important role in shaping the conception of WP target groups through regularly published guidance documents.

WP outreach targeting
The first centrally funded national WP outreach programme, Aimhigher, launched in 2004, running until abolished by the Coalition government in 2011 (Attwood 2010). Operating through regional hubs, Aimhigher primarily delivered 'aspiration raising' information, advice and guidance (IAG) activities to 14-19 year olds (Atherton 2012). HEFCE directed Aimhigher partnerships to target activity at young people who were 'under-represented in HE at the national level or in certain types of institution or course' (HEFCE 2004, 12) and at his time identified a long list of groups including those from lower socio-economic backgrounds, those living in deprived areas and areas with low rates of HE participation, 'first in family' students, minority ethnic groups, disabled students, those identified by schools as 'gifted and talented', and groups under-represented in specific subject areas, 'for example women in engineering' (HEFCE 2004, 12).
A few years later, HEFCE produced a dedicated guide to targeting HE outreach (HEFCE 2007), which used the language of a singular target group and equated this firmly with socio-economic disadvantage at both an individual and area level, specifically highlighting the National Statistics Socio-economic Classification (NS-SEC), the Indices of Multiple Deprivation (IMD) and free school meals (FSM) eligibility as possible proxies. Among the identified target group, the guide further set the expectation that WP activities should 'particularly target those whose parents/carers have not had previous experience of HE ' (HEFCE 2007, 8). Students identified by schools as 'gifted and talented' were again highlighted as an additional target group, as were disabled learners. Conversely, ethnicity was referenced only to clarify that while 'there may well be a need to take account of the specific needs of particular minority ethnic groups' within the target group of socioeconomically disadvantaged learners, 'this does not make them a specific target group for general WP activity in their own right' (HEFCE 2007, 9). Targeting on the basis of gender was not mentioned. In the years following, annual guidance published by OFFA advising HEIs on the production of AAs referred institutions to this 2007 HEFCE guide (see for example : OFFA 2012, 16, OFFA 2013. In 2016, the White Paper Success as a Knowledge Economy included two new goals for WP: 'to double the proportion of people from disadvantaged backgrounds entering university in 2020 compared to 2009, and to increase the number of BME students going to university by 20% by 2020' (BIS 2016b, 54) -'disadvantaged' here referring to those living in a low participation neighbourhood (LPN) according to the POLAR indicator. In addition, OFFA was directed to 'focus access agreements on some other key challenges including increasing participation among young white males from lower socio-economic groups [. . .] and supporting participation by students with disabilities' (BIS 2016b, 54).
The Higher Education and Research Act 2017 continued the requirement that institutions take action 'to attract applications from prospective students who are members of groups which, at the time when the plan is approved, are under-represented in higher education' (section 32) but did not stipulate these groups. The following year, guidance from the Secretary of State to the OfS (DfE 2018, 7-8) identified these priority areas and groups: • students from less advantaged backgrounds, students with disabilities, students from some ethnic minority groups, and care leavers; • access for young white males from disadvantaged backgrounds; • outcomes, such as the attainment gap, for students from black and minority ethnic backgrounds; • access and success of students with specific learning difficulties or mental health needs.
Recent APP guidance (Office for Students 2020) largely reflected this, adding a longer list of specific groups including refugees, young carers, and estranged students. In addition, the guidance emphasised the importance of considering intersections of disadvantage, noting that 'the OfS includes male and female students in combination with the characteristics above' and specifically highlighting that 'white British men and women from lower socioeconomic backgrounds are among the most underrepresented groups in higher education' (OfS 2020, 17). However, despite the long list of target groups and characteristics identified in this document, only the POLAR indicator features in a national key performance measure for participation rates (Office for Students 2020, 19).

Summary
Clearly, policy framing of the target of the WP agenda has shifted over the past twenty years. A focus on socio-economic inequalities has been consistent under different governments and regulatory regimes, whereas ethnicity and gender inequalities have been more of a mixed picture and comparatively side-lined. While a long-running guidance document explicitly distanced ethnicity from WP targeting, at other times it has been present as: a focus on general 'BME' (Black and Minority Ethnic) participation (later attainment); in the idea that 'some' minority ethnic groups are underrepresented and should be the focus of WP; and in a more recent focus on white students from socioeconomically disadvantaged backgrounds. Gender inequality has been a less prominent policy focus, appearing early on in reference to under-representation of female students in science, technology, engineering, and mathematics (STEM) disciplines and resurfacing more recently as a dimension to be applied in combination with other targeting criteria. Finally, all three of these social categories come together in the intersectional group 'white males from lower socio-economic backgrounds', who receive repeated attention from 2016 onwards.

Institutional responses to national policy
Since the introduction of AAs, institutions in England wishing to charge higher fees have been required to outline, deliver and report on the activities they plan to undertake to widen participation to under-represented groups. Early guidance from OFFA elaborated on broad government policy, setting the expectation that these activities would take the form of outreach initiatives, student retention and success measures, and financial support (OFFA 2011). In general however, national policies do not provide specifics on how to go about realising the aims of WP; rather, as in other education policy fields, 'they create circumstances in which the range of options available in deciding what to do are narrowed or changed, or particular goals or outcomes are set' (Ball 1994, 19). Instead, policy actors in autonomous HEIs are required to make decisions about how to translate these national directives into their local context.
Research exploring the ways that HEIs have responded to the WP agenda has generally understood different institutional approaches as reflecting the 'contextually contingent' (Trowler 2002, 1) nature of policy enactment; universities respond to such policy directives in ways that take into account contexts including historical, geographical and material factors as well as those relating to their relative position in the HE sector nationally and internationally (Cleland et al. 2015;Evans et al. 2017;Greenbank 2006). This analysis has often identified points of difference between the so-called 'post-92' HEIs (former 'polytechnics' that became universities through the Further and Higher Education Act 1992, and other institutions awarded university status after this time, which are typically more teaching intensive) and 'pre-92' HEIs (older, typically more research intensive, institutions). It has been argued that both kinds of institution, but especially pre-92s, have used the enactment of WP policy as a means of reinforcing their image and shoring-up their position in a marketised HE sector (Bowl and Hughes 2013;Evans et al. 2017;McCaig 2015;McCaig and Adnett 2009;Rainford 2016). Evans et al. (2017), through interviews with policy actors at different levels in HEIs, find evidence of a process of WP policy enactment which is 'intimately bound up with the construction of institutional narratives about how an institution works and what it does' (p.111), playing a key role in the '(re)production (and self-perpetuation) of a hierarchically structured HE system' (p.113). In pre-92 HEIs, they identify this in the delivery and targeting of outreach activities aimed at 'raising aspirations' among 'high calibre' school students from disadvantaged backgrounds, in contrast to post-92s who they find respond to WP policy through, for example, activities aimed at raising awareness of HE among mature learners.
Other studies have noted this pattern in AAs. McCaig and Adnett (2009) argue that pre-92s enact WP policy in a way that maintains their position as 'selecting institutions' by focusing on activities that promotes access to the highest academically achieving students from under-represented groups. By employing WP targeting criteria that includes prior attainment, selective HEIs can engage in 'cream-skimming' (Coates and Adnett 2003) these students while still positioning themselves as committed to social mobility and fair access (McCaig 2015;Rainford 2016). Conversely, post-92s enact WP policy in a way that best serves their position as 'recruiting institutions' by using it to increase potential student numbers through generally raising awareness of HE in their locality (McCaig and Adnett 2009), in combination with significantly more flexible academic entry requirements. While these institutional documents 'offer insights into the way in which universities interpret and 'speak to' government policy' (Bowl and Hughes 2013, 10), their nature as high-level policies means that they reveal little about the processes of contextualised interpretation taking place on the ground within WP teams when it comes to targeting -a gap that this study seeks to address.
Another strand of research has focused on the issue of translating WP targeting policy into practice. Hunter, Hewings, and Suddards' (2018) exploration of challenges encountered in responding to the policy steer to target 'white working-class boys', alongside two sector reports by Atherton and Mazhari (2019) and Baars, Mulcahy, and Bernardes (2016), provide some insight into how HEIs are working with this group. These studies highlight some of the difficulties of targeting intersectional groups, including uncertainties about which data to use to identify students, worries about stereotyping and stigmatisation, and concerns about the ethics of singling out a specific sub-group of low SES students in a way that excludes others. Relatedly, Stevenson et al. (2019) find that ethnicity targeting of Black, Asian and Minority Ethnic (BAME) students is almost always done in combination with other measures of disadvantage (most commonly low income), with institutional survey responses citing uncertainty about how to address racial inequality and a lack of evidence of what works as key reasons for ethnicity targeting not taking place, alongside data issues and geographic context. Overall, this suggests that HE market position, while particularly influential in shaping WP policy enactment, is not the only contextual factor at play.

Theoretical lens
Policy enactment is an alternative to conventional top-down, linear models of policy implementation which 'captures the multifaceted ways in which policies are read alongside/against contextual factors, by different sets of policy interpreters, translators and critics' (Maguire, Braun, and Ball 2015, 487). Theorising policy as enacted, rather than straightforwardly implemented, brings into focus the activity that takes place between text and practice, understanding policy 'as a process, rather than a product' (Ozga 2000, 2); one which involves struggle, mediation and recontexualisation (Ball, Maguire, and Braun 2012;Ozga 2000). Although developed in relation to secondary schools, this theorisation offers a useful lens for understanding how policy gets 'done' in the context of widening participation to HE. As previously noted by Rainford (2021), WP practitioners occupy a similar position within the policy enactment process -or 'staircase' (Reynolds andSaunders, 1987, as cited in Trowler 2002) to teachers in the school setting. Importantly, within this understanding of policy, these education workers are understood not only as subjects but as key actors in the process (Braun, Maguire, and Ball 2010).
In order to understand how the significant diversity within the English HE sector (including institutional mission, size, student demographics, geography, and academic selectivity, to name just a few) shapes the enactment of WP policy around targeting and target groups, the decision was made to focus on the role of context. To facilitate this, the research utilised a framework developed by Braun et al. (2011) in which different contexts are conceptualised as belonging to one of four dimensions: situated, professional, material and external. While HE institutions and schools are not a direct comparison, they share enough similarity -and the dimensions are sufficiently generalisedthat this framework was able to capture the range of contextualising forces acting upon the enactment of targeting policies. In order to create space for a fuller presentation of the findings, this paper will focus on the frequently intertwined dimensions 'situated' and 'external'.

Methods
The highly localised and contextualised nature of policy enactment requires a research approach that is able to account for and draw out this complexity. In the present study, this was achieved through the use of in-depth qualitative interviewing of 'information-rich cases' (Patton 1990, 181); staff employed in WP departments in English HEIs. A purposeful sampling approach was used to recruit participants through social media, relevant mailing lists, and professional networks aimed at WP practitioners. Selection was informed by the literature review, which suggested that institution type would be an especially pertinent factor (Evans et al. 2017;McCaig 2015). A strong response to the call for participants also enabled the inclusion of both larger and smaller institutions and a good geographic spread. Ultimately, the sample included nineteen practitioners from eighteen different HEIs, ranging in size from fewer than 1,000 to over 25,000 undergraduate students. This included three HEIs specialising in a disciplinary area with the remainder being multidisciplinary.
Semi-structured interviews of between 45 and 90 minutes were conducted using the video conferencing platform Zoom. A guide of (primarily open-ended) questions was used to approach the issue of policy interpretation indirectly, asking participants about different aspects of targeting in WP -their understanding of targeting policies and practices in their institution, the ways that they are involved in targeting, their experiences and perceptions of targeting different groups -to allow for the contexts shaping policy enactment to surface through their responses. Transcribed interview data was analysed using the software NVivo (v.11), applying an iterative, thematic approach, guided by a framework of the four contextual dimensions developed by Braun et al. (2011) [ Table 1].

External contexts
Interviewing WP practitioners from a variety of pre-92 and post-92 HEIs highlighted the particularly significant contextualising force of an institution's position within the marketised English HE sectorthat is, where it is located relative to other institutions in terms of perceived status and success in competing for student 'customers'. In line with previous findings by McCaig and Adnett (2009) and Evans et al. (2017), the enactment of WP targeting policy appears to reproduce the hierarchical organisation of this HE system. However, where Evans et al. (2017) find a tendency for research intensive universities in Wales to favour pre-entry activity and teaching intensive universities to favour more flexible pedagogical approaches post-entry, these interviews suggest that in England both kinds of institutions engage in a significant amount of pre-entry activity, with the external context of their position in the market shaping the interpretation of targeting as schools-focused or individual-focused.
WP departments in HEIs which occupy a 'recruiting' position, primarily post-92 institutions with lower entry tariffs and more diverse student bodies, were found to favour targeting schools over targeting individual students. Where individual targeting was used, this tended to include a broad range of targeting criteria and be less demanding in terms of the number of criteria a participant must meet. This inclusive approach to targeting, which allows the team to work with the greatest number of students possible, interprets national policy in a way that meets the needs of the institution's market position. For these HEIs, the external policy pressure to meet APP access targets is reduced as a factor shaping the enactment of policy. As one WP Manager noted, their institution had already earned 'a reputation of being a very WP university' through accessible entry requirements, further shaped by their geographic context: [. . .] we're really lucky, as I said, like we meet every single one of our access targets and some. And I'm not saying that's because we're brilliant. It's just because we're really lucky, our kind of local area and the university [. . .] it's just easier for us, we've got an easier job. (WP Manager,London) As argued by McCaig and Adnett (2009), in this market context, the recruitment function of WP is about general HE awareness-raising to increase the pool of potential applicants locally. When the target student is one who is potentially not yet engaged with the idea of attending university, an interpretation of targeting framed around building strong relationships with schools makes sense. Maintaining, as one WP practitioner in a small post-92 arts institution phrased it, 'a solid relationship with the schools' is critical in order for the WP department to get the 'buy-in' necessary to deliver projects as desired. As such, school partnerships become another key external context shaping the enactment of targeting policy: I think because our work is so focused on the relationship we have with schools, and the kind of situation that schools are in terms of their performance, and their outcomes, that the schools are really our main outreach groups in a way. (WP Manager,East) At the same time, an approach which partially or totally transfers control of individual targeting to school or college partners can also generate particular constraints on a WP team. School staff are subject to their own policy pressures and may have priorities that do not align with WP priorities. One practitioner, who had previously worked in an academy school in a deprived London borough known for sending a high number of young people to selective universities, felt there was 'definitely a mismatch' between the students that schools would like to give opportunities to and the students that WP departments are trying to engage. She attributed this to schools bolstering their chances of meeting their own HE participation targets by selecting the students they considered to be the most likely to progress, not 'the ones who, you know, have whatever criteria or whatever going on at home [. . .] why would we waste the opportunity on them when we know whoever else is like a high flyer?' (WP Practitioner, pre-92, East). Another practitioner described targeting for a white workingclass boys' programme being derailed by a school prioritising its own discipline policy: One year, a school decided that boys who were down to come to the summer school had done something in school the week before or something and so they were then no longer allowed to come and they sent ten girls instead. (WP Manager,Midlands) Finally, this approach is dependent on schools being willing and able to take part, something that several practitioners raised as a barrier: 'the most eligible schools [. . .] who are very likely to have the highest proportion of eligible students, are the most difficult to engage' (WP Practitioner,Midlands). For schools facing the most challenges due to their location and the deprivation level experienced by their intake, WP activity is 'kind of an added extra really . . . it's not a priority' (WP Practitioner,Midlands). In contrast, those in high-tariff institutions were more likely to use an individual targeting approach over and above schools targeting, employing points-based systems to prioritise and shortlist applicants meeting the most targeting criteria. While these WP teams are based in HEIs that are broadly 'selecting' rather than 'recruiting', WP serves an explicit recruitment function in the sense that increasing the proportion of under-represented groups that enter high-tariff institutions is a national objective ([OfS] n.d.). Since institutions are assessed on progress towards this target in terms of their own intake, it is perhaps unsurprising that their main interpretation of targeting is one which uses attainment based targeting criteria to identify students from under-represented groups who are likely to meet the higher academic entry requirements of these HEIs. At one highly selective London university, a practitioner described the ways that this kind of attainment-based criteria, combined with the situated context of their geographic location, constrained the students that the team could work with to the extent that some target groups were effectively excluded: I guess that's really difficult because of where we are, and the pool of applicants that we get in London is so multicultural, and absolutely there are pockets of white working class communities within and around the outskirts of London, even within central London, but it's really difficult to engage with . . . for the stage at which [WP programme name] takes place, a lot of the white working class students that need to be on those programmes wouldn't be on track to be on a programme like [name], in many cases. (WP Manager,London) Alongside an institution's position in the HE sector, the external context of the HE regulatory system and pressure to meet the expectations set out by the OfS was a clear force shaping targeting policy enactment. A number of practitioners bought up the shift in approach from OFFA to the OfS and a sense that they were operating in 'a very highly regulated environment' with potential 'penalties' that were now 'much more severe and significant' (WP Manager, pre-92, London). Following the guidance set out by the OfS and meeting the institution's APP targets were considered by many to be the main influence on decision making about targeting: I want to say it's evolved since I started, towards being much more policy and government focused, but maybe that's not totally fair. Certainly at the moment it's definitely Office for Students influenced. Obviously if it's a funded programme like Uni Connect it is, but even our kind of other institutional outreach is very much focused on areas that the Office for Students are prioritising. (WP Manager,East) External policy pressures were especially obvious in relation to the POLAR indicator. While POLAR was a feature of targeting approaches across different institutional contexts, it occupied a particularly salient position for WP teams in pre-92 institutions, and especially those located in geographic regions with fewer target areas. One practitioner at a southern pre-92 described having to 'put a case forward', to senior management, to justify the rejection of any summer school applicant meeting this criterion. Similarly, a London-based practitioner recounted a discussion about contextualised admissions at the institution which gave an indication of the dominant positioning of POLAR over and above other targeting criteria (as noted by Harrison and McCaig 2014): [. . .] I'm afraid it was one of those pragmatic decisions where we needed to say well, our target with the OfS is actually anyone from a POLAR quintile one background. That target would include anyone who happened to be at a private school [. . .] So on that occasion, that did trump socio-economic background. (WP Manager,London)

Situated contexts
The situated context of a WP department, including factors like the size of the institution, the academic disciplines it offers, student intake, and geography, influence how targeting policy is enacted in a number of ways. Where an institution is situated geographically can apply a shaping force both because WP departments typically deliver most of their activity in their local region, and because many institutions recruit a significant proportion of their student intake locally. Therefore, the demographic of the population around the university, and other factors such as population density, can have a direct impact on which target groups the WP department decides to prioritise. In one post-92, geographic context had steered the team to develop work for Gypsy Roma and Traveller (GRT) students as a target group, after finding the city contained a 'large population where we could be making a difference' (WP Practitioner,North). In another part of the country, a practitioner at a pre-92 described how the location of the institution, and its STEM-based subject offering, combined to ensure a 'very diverse cohort in terms of ethnicity, on the BAME side' (WP Manager,London). Similar to the position of WP departments operating in socioeconomically diverse 'recruiting' institutions, this context removed any pressure to attract a more ethnically diverse cohort through outreach work, and ethnicity targeting had therefore never been 'actively pursued'.
For institutions favouring a schools-based approach, the inclusion of ethnicity as a targeting criterion can be closely related to geographic context. Where an area is more ethnically diverse, data about the ethnic breakdown of schools can be used to passively target ethnicity without further targeting at the individual level: 'if the school has a high proportion of those students in it, then we'll just automatically be getting a higher proportion of those students on our activities' (WP practitioner,North). However, if the institution is located in a less diverse geographic area, a school-based approach can make ethnicity targeting challenging. One practitioner described what they saw as the practical and ethical issues involved in ethnicity targeting in this context: We don't do I think a real lot of conscious targeting around ethnicity. That's a tricky one. Partly it's because our county is not diverse at all, really. So that kind of presents some challenges in the sense of . . . if there's very few individuals from minority ethnic background in classes, do you try and pull them out? You know, that doesn't really feel like the right approach to sort of single people out in that way. (WP Manager,East) As well as being linked to geographic context, the intake of an institution is a strong contextualising factor in its own right. The demographic of an HEI's student body is directly related to the nature and degree of scrutiny it receives from the OfS and relates closely to its APP targets. Gaps in a student body can therefore shape the targeting undertaken by the WP department, including on the basis of ethnicity; one practitioner referred to 'four specific ethnic minority backgrounds' identified in the institutions' APP that the WP team prioritised in targeting 'because they're way underrepresented in the student body' (WP Practitioner,South). In general terms, the more diverse an institution's intake, the less focus on access related targets and less attention on the work of the WP department: We currently have 98% WP students or students who meet at least one WP criteria. So in a way targeting, although we absolutely prioritise students from particular backgrounds on particular programmes, we'll find the majority of our students meet multiple WP criteria anyway [. . .] So in a way some of the pressure is off us as a team in terms of how we treat targeting and how specific we have to be [. . .]. (WP Manager,London) Conversely, at one small specialist arts institution, the interpretation of targeting was contextualised by its particularly small intake, highly competitive admissions, and the large proportion of its student body from higher social class backgrounds, manifesting in a stricter targeting approach focused on socio-economic status. In the case of applications to a summer school bursary scheme, not only was targeting individual-based, it was taken a step further in requiring participants to evidence their circumstances (such as proof of receiving means-tested benefits): 'it's one of our biggest investments [. . .] so I have to be able to demonstrate that those young people meet the criteria in order to benefit from that' (WP Practitioner, small specialist, London). At the same time, similar to Baars et al.'s (2016) finding that some WP practitioners are concerned that targeting specific intersectional groups such as white working-class boys could 'lead to other marginalised groups being overlooked' (p.28) any further narrowing of targeting was viewed as inappropriate: We just have a real lack of any sort of diversity [. . .] So I wouldn't want to run a project that ruled out a workingclass person who was white because at the moment we aren't good enough at even just like socio-economic diversity or school type diversity or regional diversity. We're not great at that yet. So I think that, yeah, pinning down even more into different levels of diversity would be very challenging. (WP Practitioner, small specialist, London) At this particular HEI, targeting criteria were also inflected by the situated context of the institution's focus on performing arts disciplines and in particular its music provision. The prior training required for these courses is highly skewed towards more socio-economically advantaged families and the privately educated (Gutierrez 2018), creating a context in which state school students in general become a relevant criteria for measuring progress in WP. More widely, disciplinary context shaped targeting approaches in relation to gender. Though gender is rarely an explicit standalone targeting criterion for WP teams, several interview participants commented on academic departments driving forward gender targeting in relation to recruiting girls into STEM. One participant described the conflicts that can arise from these disciplinary interpretations of WP targeting policy: [. . .] STEM subjects in particular . . . the departments will often run programmes with us where they prioritise females because they have a serious representation problem. In fact a subject like computer science [. . .] they say they're not particularly interested in our other indicators that we use because their problem is females. So they would rather take females that met far fewer criteria in terms of socio-economic, you know, WP criteria, than males who meet more. (WP Manager,East) This example demonstrates what Braun, Maguire, and Ball (2010) identify as an often overlooked aspect of the policy process -that individual policies are not enacted in a vacuum but rather must be read alongside, and prioritised amongst, an array of other policy demands. In this situated context, the WP agenda competes with another aspect of the wider policy landscape (an external context): a national agenda addressing female underrepresentation in STEM (Ro, Fernandez, and Alcott 2021). WP activity offers policy actors in these departments an opportunity to respond to a policy pressure with particular local saliency and the interpretation of targeting is therefore further recontextualised in a way that foregrounds gender as a priority.

Conclusion
Building upon earlier research into WP policy enactment, these findings suggest that an institution's position within the marketised English HE sector, in combination with the closely related contexts of its student intake and nature of its relationship to the regulator, play a significant role in shaping the enactment of WP targeting policy. While the specifics of how targeting is put into practice vary widely from institution to institution, it is these contextual factors that most clearly influence whether an institution will adopt an approach which can broadly be described as simple and relaxed or strict and complex, and position either schools or individual young people as the primary target group. At the same time, these interviews suggest that several other external and situated contexts, including geography, school partners, institution size and disciplinary offering, also inflect the interpretation of targeting in meaningful ways, overlapping with and modifying an approach underpinned by institutional positionality. Furthermore, although not the focus of this paper, other 'material' contexts of the staffing and resources available to the WP team, and 'professional' contexts of the values, beliefs and experiences of WP practitioners, were found to be at play. The impact of these factors will be explored in-depth in a separate paper.
National WP policy has narrowed the range of options open to institutions when it comes to target groups; both in relatively consistent ways (such as SES), and ways that have been inconsistent and arguably confusing (such as ethnicity). However, this study has shown that the contexts a WP team operates within can make the pursuit of some target groups possible or desirable, while locating others as redundant or impractical. Schools partners, subject to their own pressures, can subvert attempts to target defined groups such as boys or FSM recipients. Local demographics can facilitate, limit or negate the targeting of specific ethnic groups. Delivering WP in the context of certain academic disciplines can reprioritise less common WP target groups, such as girls (in STEM), or state schooled students (in music conservatoires). An institutional context of high academic selectivity, in a particular geographic context like London, can result in the de facto exclusion of many 'white working-class boys'. These findings suggest that, within the broader parameters set by national policy, contextual factors play an important role in determining which individuals will have the chance to take part in WP initiatives at a particular institution.
As this research highlights, translating into local institutional settings a policy of targeting underrepresented groups requires WP practitioners to navigate a complex range of constraints and enablers, often involving a degree of struggle -between different policy actors, between coexisting policies, between institutional priorities. Drawing upon an understanding of policy as enacted brings to the fore these processes of recontextualization and helps to explain the diversity of targeting approaches seen across the sector. This has implications for HE policymakers who may be informed by traditional accounts of policy implementation; rather than framing this diversity as mere 'disruptions in practice' (Maguire, Braun, and Ball 2015, 486), acknowledging the messy, contextually-contingent reality of policy enactment could lead to greater honesty and transparency about who is being targeted with efforts to widen participation, and why. specialising in project evaluations and theory of change development. She completed an MRes in Social Sciences at Sheffield Hallam University in 2019/20 and holds an MA in Arts Policy and Management from Birkbeck, University of London.