Conflicting logics of online higher education

The advent of massive open online courses and online degrees offered via digital platforms has occurred in a climate of austerity. Public uni- versities worldwide face challenges to expand their educational reach, while competing in international rankings, raising fees and generating third-stream income. Online forms of unbundled provision offering smaller flexible low-cost curricular units have promised to disrupt this system. Yet do these forms challenge existing hierarchies in higher edu- cation and the market logic that puts pressure on universities and public institutions at large in the neoliberal era? Based on fieldwork in South Africa, this article explores the perceptions of senior managers of public universities and of online programme management companies. Analysing their considerations around unbundled provision, we discuss two conflicting logics of higher education that actors in structurally different positions and in historically divergent institutions use to justify their involvement in public–private partnerships: the logic of capital and the logic of social relevance.


Introduction
Unbundling -the disaggregation of educational provision and its delivery, often via digital technologies -has promised to address inequalities and challenge elitism in higher education (Rizvi, Donnely, and Barber 2013;Craig 2015). Yet how does unbundling develop in contexts with deeply rooted historical inequalities and asymmetric distribution of economic and reputational capital in and among public universities? How does the evolution of technology-enhanced learning intersect with the entry of for-profit actors into public higher education in South Africa, committed to social justice and redistribution, while dealing with austerity? How are different higher education institutions participating in this process, given their historic profiles and current national and global ranking performance? What factors do decision-makers consider when engaging in the practices of unbundling through public-private partnerships?
This article addresses these questions through empirical inquiry. We highlight the competing narratives and emergent patterns of decision-making regarding public-private partnering in this rapidly changing field, focusing on South Africa. During March-October 2017, we interviewed over 30 senior managers of public universities and online programme management (OPM) providers. Here, 'OPM provider' refers to private companies providing digital platforms that host content created and managed by universities or offer services around marketing, enrolment and course design to facilitate online education (Mattes 2017). This article presents the findings of interviews conducted at three universities and five OPM providers. The universities represent different historical trajectories, structural positions and engagement with partnerships. The OPM providers -some current, others considering opportunities in the South African online education market -are internationally active, partnering with public universities to offer massive open online courses (MOOCs), short paid courses or full online degrees. We triangulate the interview analysis with the discussion of the historical university reforms and desk research about the existing partnerships between public universities and OPM providers in South Africa.
Using Luc Boltanski and Laurent Thévenot's (2006) framework of different orders of justification, connecting them to the sociological literature on institutional logics (Friedland and Alford 1991;Thornton and Ocasio 2008), we argue that if market-driven unbundling occurs in a field with pre-existing symbolic and economic inequalities, like higher education, it can further entrench these inequalities. We suggest that more explicit and nuanced national and institutional policies need to be produced around unbundled provision, which are cognisant of emerging trends in and dangers to the evolution of unbundling at public universities.
With some of the most prominent universities on the African continent, and globally (QS 2018), South Africa remains one of the most unequal societies on earth (World Bank 2018). The country's apartheid history, and the current political and economic context (Beresford 2016), has placed public universities at the intersection of historic inequalities and conflicting imperatives to negotiate economic austerity with a commitment to social justice (Swartz et al. 2019). South Africa's historically advantaged institutions (HAIs) are leading players in research, digital innovation and market-led growth, but historically disadvantaged institutions (HDIs) cannot compete equally (Ntshoe 2004). This article shows how the unequal distribution of financial and symbolic capital influences the process of unbundling in South African higher education. Scrutinising decision-making processes around public-private partnerships highlight both critical potentials and dangers of how unbundling could reinforce existing inequalities.
After introducing the South African context, we discuss two different logics informing unbundling in higher education: the logic of social relevance and the logic of capital. Within the former, we suggest, unbundling could provide socially embedded, pedagogically innovative educational opportunities for students disadvantaged in several ways. Within the latter, we argue, unbundling remains an instrument of marketisation of higher education privileging particular universities endowed with reputational capital and reproducing inequality. We show how different institutions and OPM providers position themselves regarding unbundling processes in relation to these two logics that coexist in ever-growing tension with each other, the logic of capital increasingly marginalising the logic of social relevance. We focus on the perceptions of senior leaders in universities and OPM providers, indicating how they make decisions regarding unbundling and the forming of partnerships.

South Africa: unbundling on an unequal terrain?
Post-apartheid South Africa inherited a two-tier public university system. Stratified by race, it drew distinctions between well-resourced universities and technical colleges (technikons) dedicated primarily to white elites and underfunded universities and technikons training smaller numbers of black South Africans. These are known as HAIs and HDIs, respectively. In the early 2000s, South African universities were restructured and reclassified as 'research-intensive' , 'comprehensive' and 'universities of technology' (Jansen 2002;Ntshoe 2004). The reform aimed to create 'a single, coordinated [university] system … without racialised inequalities' , following trends towards 'technology-driven' , 'information-based economies' (Jansen 2002, 5).
This reform coincided with the African National Congress, the ruling party since 1994, embracing privatisation and restructuring public services, including education, in order to galvanise foreign investment (Beresford 2016). Promises of growth-induced redistribution barely materialised. An economic crisis, paralleled by the global recession, left a third of South Africans aged 15-24 unemployed (STATS SA 2018). Meanwhile, government subsidies to higher education fell from 49% to 40% between 2000 and 2013; universities' income from sources other than fees or subsidy also stagnated (Cloete 2016, 3 and 28). Student fees increased 9% annually between 2010 and 2014 to compensate for funding shortfalls (van der Berg 2014).
After a wave of student protests, then-president Jacob Zuma announced in 2017 that higher education would be free for low-income students (Davis 2017), appointing a government commission to investigate the viability of fee-free education (Presidency [South Africa] 2017a). The report's executive summary suggested further investigation into online and blended learning for 'efficiency savings and academic support improvements through the use of ICTs' (Presidency [South Africa] 2017a). At the same time, the report cautioned that 'a massive shift towards online education … could be detrimental to the South African education system' (Presidency [South Africa] 2017b, 547). Such an investigation has not yet been conducted. Meanwhile, the use of education technologies -previously limited to devices and virtual learning environments -had advanced in South Africa (Mahesh 2017). During the #FeesMustFall protests, universities used online platforms to enable the completion of the academic year (Czerniewicz, Trotter, and Haupt 2019), ironically positioning some universities for the Covid-19 pivot online (Czerniewicz 2020).

Unbundling higher education
The concept of unbundling has recently gained traction in higher education policy and research. Unbundling the traditional university 'bundle' affects not only property, services and facilities, but also administration, evaluation, issuing credentials and even teaching (Wallhaus 2000, 22). This process involves separating educational provision (e.g. degree programmes) into component parts (e.g. courses) for delivery by multiple stakeholders, often using digital approaches (Swinnerton et al. 2018). Universities can unbundle on their own, offering individual credit-bearing modules outside bounded disciplinary curricula, or in partnership with OPM providers, offering MOOCs or credit-bearing courses or programmes. Proponents of unbundling suggest that the disaggregation of television and music production and its re-aggregation as on-demand digital content like Netflix or Spotify could represent a template for universities (Craig 2015;McIntosh 2018).
As unbundling is increasingly shaping university systems, institutions and courses (McCowan 2017, 737), the academic literature on it has also grown. Researchers often deal with a specific aspect in which unbundling shapes or informs changes in the curriculum (Cliff et al 2020), the academic profession (Macfarlane, 2011), policy reform (Lewis and Shore 2019) or market-making between public and private institutions (Komljenovic and Robertson 2016). Yet, apart from Komljenovic's (2019) work on trust-building between university management and private companies, there is little indication of the meaning-making of actors involved in the unbundling process. To fill this gap, our article focuses on how decision-makers at public universities and private companies perceive the unbundling of teaching and learning in academic programmes through the introduction of digital technologies.
Enhanced by high-speed Internet, mobile technologies and social media, unbundling promised to tackle key challenges in higher education. Journalists (Economist 2013), policy-makers (Bradwell 2009;Rizvi, Donnely, and Barber 2013) and higher education leaders (Napolitano 2015;Davis 2017) hailed the radical potential of MOOCs to open access to individual units of content to global audiences at low or no cost as a way of challenging the dominance of traditional universities. Individually certified paid unbundled courses have been commended as opening equitable, flexible pathways for disadvantaged learners to become qualified for high-skilled high-paying jobs (Bradwell 2009;Rizvi, Donnely, and Barber 2013). Flexible start, examination and graduation dates enable access to mature students and workers, people with caring responsibilities or those with disabilities (Alevizou 2015). New forms of managing content and responsive analytics could benefit learners. Students could opt out of fixed curricula and select from global content, tailoring their CVs for the job market (Christensen and Eyring 2011, 3-4). Personalised flexible learning forms could address the declining social relevance of university education and offer professional development on the job (Newton 2015).
However, unbundling occurs in the era of marketisation of university education, increasingly presented as an asset instrumental to capital accumulation within a globalising field of higher education (Marginson 2013). The introduction of market logic into the sector happens even if higher education is a stratified positional pseudo-market with scarce excludible resources only available to groups with access to a few prestigious institutions; its outcomes and value are difficult to measure in purely economic terms (Marginson 2013, 359-362). Despite the sector's massification, students from economically disadvantaged backgrounds are still less likely to enter universities, even less so to prestigious programmes and institutions (Boliver 2017, 424). The rise of league tables places students with access to economic and cultural capital at a further advantage while increasing fees produce skyrocketing student debt (Lynch and Ivancheva 2016). Instrumentalising rankings in a battle for international and mobile talent and capital, governments worldwide push teaching-oriented public universities into a competition for efficacy of provision and performance-measured success according to benchmarks set by privately endowed research-intensive institutions (Hezelkorn 2015). Even in countries like South Africa where discussions of ranking are nuanced (Swartz et al 2019), public universities compete for rankings to attract fee-paying international students and third-stream income.
Tomlinson captures these processes in his discussion of two conflicting logics of value in the university sector: higher education for 'empowerment and agency' of graduates, and higher education as offering students 'value-for-money' (Tomlinson 2018, 714). Under accelerated marketisation, Tomlinson (2018, 714 and 724) argues, higher education is reduced to the latter frame and measured in terms of income generation, employability, consumption and performativity. Building on this framework, and relating it to unbundling, we identify the emergence of two organisational logics of higher education: the logic of social relevance and the logic of capital. Their tension, we argue, crystallises when unbundling processes enter the higher education field.

Orders of justification behind the logic of capital and the logic of social relevance
Institutional logics are 'supra-organizational patterns of activity by which individuals and organizations produce and reproduce their material subsistence … [and] symbolic systems, ways of ordering reality… rendering experience of time and space meaningful' (Friedland and Alford 1991, 243). Unlike new institutionalism, which remained focused on processes of institutional isomorphism or the replacement of a static single logic by another, the institutional logics perspective offers a more dynamic multi-level view: a plurality of logics coexist in complex interrelations within organisational fields like higher education (Friedland and Alford 1991;Lepori 2016). This lens helps us to go beyond deterministic readings of how existing institutions (market, society, religion, etc.) shape individual and organisational action. It gives us a meta-theory and method (Thornton and Ocasio 2008) to study how actors justify decision-making and re/align organisational fields to currently available institutional logics, like the logic of capital and the logic of social relevance, outlined in the following as co-existing in higher education.
Unlike older theories of institutional isomorphism, the institutional logics perspective has not been applied productively to studies of higher education (Lepori 2016). This article addresses this gap, and goes one step further. By employing Boltanski and Thevenot's grammar of worth, we do not simply apply existing institutional logics perspective taxonomies. Instead, we operationalise the two logics identified by scholars as active in the higher education field. On this basis, we explore the justification practices of decision-makers with agency to affect change. Interview data on how decision-makers perceive unbundling processes and practices allow us to explain how different justification orders relate to unbundling. As Table 1 shows, and we elaborate further in the following, within higher education these six orders broadly align into the two logics identified by our research.
In contrast to Bourdieu's single hierarchy of economic and cultural capital accumulation (Bourdieu 1998), critical sociologists Boltanski and Thévenot present heterarchies (Lamont 2012) of competing or complementary orders of justification which social actors use to legitimate their decisions. These are the profit-driven market order; the production-driven industry order; the civic order driven by collective values of civility and solidarity; the inspiration order driven by creativity and innovation; the fame order driven by public opinion; and the domestic order driven by interpersonal relations and traditional hierarchies (Boltanski and Thévenot 2006, 163-203). Not in a stable hierarchical relationship, these orders are selectively emphasised or marginalised by groups when attempting to justify actions or positions (Lamont 2012).
Unbundling was initially presented as a way to revolutionise higher education, responding to what Boltanski and Thévenot's term the civic order of social justice and inclusion, the industry order of integration of education in social production and the inspirational order of the pedagogical affordances of digital technologies. In our discussion, we call this the logic of social relevance, a term used in the Venezuela's Bolivarian higher education reform to designate a pedagogically innovative, equitable and socially pertinent higher education (see Ivancheva 2017). Aspects of unbundling also justify the logic of capital in higher education, however: a mixture of Boltanski and Thévenot's fame, market and domestic orders. These are represented respectively through the brand domination of Oxbridge and Ivy League colleges, mimicked in league tables, the marketisation of universities and their dependence on the concentration of economic and cultural capital passed on and legitimated through elite networks. While these two logics are often in tension, one is not necessarily subordinated to the other. However, given the rapid marketisation of public universities, South Africa faces the risk that the logic of capital marginalises the logic of social relevance. We explore how a process like unbundling, promising a complete restructuring of the higher education field, is negotiated and justified by decision-makers vis-à-vis economic austerity and a post-apartheid agenda of social justice.

Research design, sample and methodology
This article draws on interviews with senior decision-makers across South African public universities and international OPM providers. We also use information freely available on Internet websites about university ranking and existing relationships between OPM providers and public universities in South Africa.
Having secured ethical clearance from our own and all institutions involved, during the period March-November 2017 research team members carried out semi-structured interviews with over 30 senior managers of public universities and OPM providers. Our questions were based on research literature topics around unbundling, technology-enhanced learning and marketisation in higher education. We operationalised the perceptions of our interviewees on the institutional experience and position in decision-making processes and on the potentials and perils of unbundled education in the institutional, national and global contexts. Interviews lasted between 45 and 75 minutes. They were transcribed and then coded using NVIVO software. Institutions and informants were anonymised and de-identified in order to strip features recognisable within the relatively small South African academic field.
Codes for the categorisation of interview excerpts were established, reviewed and debated by the project team before final codes were agreed. The coding framework was developed by research team members, identifying key etic themes from the research literature on marketisation, unbundling, digital education and inequality into smaller analytical categories to reflect complex relations. We also coded for emic themes drawn directly from interviewees. We intersected themes on decision-making and institutional culture regarding the public-private divide in partnering, digital learning, teaching and pedagogy, the place of ranking and institutional brand, universities' relation to local and larger contexts, and historical and contemporary inequalities. We also performed meta-coding vis-à-vis the orders in Boltanski and Thévenot's work. While we were open to challenging the framework, we found topics clustering around the six orders of justification to respond to the two logics of higher education, outlined earlier.
We approached leaders and senior managers within a sample of six universities representing different institutional types -according to their histories, positions in world rankings and partnering experience. We found that three cases best illustrate divergent trajectories and patterns of unbundling, and therefore this article presents interviews with senior managers of the following three universities: • University X, an urban elite university historically serving white South Africans, ranks in the Times Higher Education top 400 universities worldwide. It maintains research excellence, attracting foreign funding and private donations. It has a handful of OPM partnerships, offering MOOCs and short courses. Its strong international brand gives it leverage over its choice of partnerships. • University Y is a comprehensive institution based in a former apartheid homeland (a physically and legally segregated area where black people were concentrated and controlled). It caters to a local cohort of mostly black undergraduate students, funded by state subsidies. The university is unranked in global ranking systems and has no partnerships. • University Z is a comprehensive urban institution ranked among the 1000 top universities worldwide. It combines a former HAI with a former HDI. It promotes research and is planning an ambitious partnership with an OPM provider for online degrees.
All three universities offer blended learning through virtual learning environments and some access to digital devices and the Internet for students. Their senior managers have a mature understanding of unbundling processes, although not always using the term 'unbundling' . The institutions' different positions in relation to OPM providers, structural constraints and teaching offerings shape differing engagement with unbundling.
We also interviewed senior managers from five key OPM providers of different sizes, with different programme offerings currently working or planning to work in South Africa. Given the small number of OPM providers active in South Africa, we cannot reveal much about the OPM providers for reasons of anonymity. University and OPM managers were asked similar questions that allowed us to probe central topics with both groups, while also enabling a level of thematic emergence.

Unbundling as viewed by senior managers of OPM providers
Senior managers of OPM providers used a number of orders of justification to provide rationales for partnering with public universities. In interviews, they prioritised the industrial (productivity and employability), market (immediate and mid-term strategies of profit-making) and, to a lesser extent, inspiration (innovative aspects of digital technologies) orders. Yet, upon closer analysis, the domestic (elite networks) and fame (university reputation) orders were most likely to dominate their choices around partnering, alongside the market order. This reveals an emphasis on what we call the logic of capital. While the civic order, challenging social inequalities and the industry logic of embeddedness in production, were stated in their missions, the interviews revealed that these missions were not necessarily being enacted, and that the logic of social relevance was marginalised.

Institutional culture or rankings?
Senior managers at OPM providers had somewhat cynical views of universities' reasons for partnering. 'Most often, they want to just increase the global brand recognition of their university …' , Senior Manager 1 of OPM 1, a medium-size MOOC company entering the online programme market, explained. Given that research bids place increasing emphasis on public engagement, 'engaging thousands of people via an [online course is considered] more successful than traditional public engagement' . Thus, OPM managers believed that even while public universities were engaging with the civic order of impact, they presented a market rationale for doing so.
On OPM providers' choices about partnering, managers underlined a mixture between organisational culture and efficiency: between industrial and inspiration orders, as well as the domestic order of interpersonal trust. Senior Manager of OPM 2 said: 'This is about taking phenomenal universities and partnering with them in the deepest possible way. We become their brand. [Our] team there [is] committed to them day in and day out.' Their company chose university partners according to what they considered shared characteristics of capability, accountability, commitment to delivery and strong leadership. They also chose partners according to their access to technology, capital as well as national higher education strength and availability of 'burgeoning middle classes': 'If Lithuania had a strong sector we would be there!' Senior Manager of OPM 3, another large company offering online degrees, said that getting to know individual university leaders and managers was crucial. Senior Manager of OPM 4, a smaller company offering online short courses, was concerned about trust: 'We are for-profit players in a not-forprofit sector' .
While trust and working culture alignment were important, a high rank in league tables was a necessary condition for OPM providers to consider partnering. Another senior manager of OPM 1 said they would only partner with the top 200 universities globally. OPM 2 invested in 'capital, brand, reputation, time-effort': shared risk and reward required careful selection. The university brand had to be recognisable in targeted regions, and prominent on national evaluations: 'No one knows who [OPM] is, everybody knows who [top ranked university] is' . Thus, simple market order was underlined by Senior Manager of OPM 3: ' A brand can sustain a good price … [Y]ou're putting in millions. You need high-ranking partners' .
This position was also articulated by Senior Manager of OPM 2: 'I don't want to muddy the brand and go into lower ranked universities' . When asked about whether they would partner with a South African HDI, Senior Manager of OPM 4 stated: If this university, which I'm not confident, had the faculty expertise and recognition from a population that wanted to be recognised by this university for skills, then we would certainly assess it. My sense though, is that their opportunity for use of online education is outside the market that we understand.
In this case, the logic of capital also implies that some audiences are not considered a valuable education market.
Senior Manager 1 of OPM 1 emphasised that this was not the choice of OPM providers alone: The world of higher education is, how to say this politely … elitist. High-rank universities like to keep the company of other high-rank universities. If the vice-chancellor of a prestigious university gets wind that, oh the vice-chancellor of another prestigious university is doing something, then they just want to copy each other. Some of our partners joined us was because they don't want to be the university that didn't join.

Course marketability or applicable knowledge?
Senior managers at OPM providers qualified statements about rankings in discussions of content. Ranking, according to some, was 'not everything' but needed to be measured against the actual demand for a profession, profitability of specific disciplinary fields and the content being suited to technology-enhanced learning. Senior Manager of OPM 2 shared that 'Midwifery is not popular online, so I'm not interested in a university's ranking there. I am interested in rankings in courses with big online market: health, psychology, law, business, engineering, and education. ' OPM managers were troubled by universities' lack of vision regarding student outcomes and progression. Senior Manager 1 of OPM 1 lamented: 'seldom do we get a university partner saying, "The reason I want to develop this course is because I know there's a huge skills gap in an area"' . Senior Manager of OPM 5 shared this concern. Here, private companies project a profit-driven market order into public higher education, prioritising marketable areas of teaching where they identified gaps in the market. If universities wanted to partner on less profitable courses, this could affect the partnership, according to Senior Manager of OPM 4: 'We don't say "We're not going to publish that course", but we have to say, that might not be a course that [is viable] from marketing [viewpoint] so we might not support them much' .
Senior Manager 1 of OPM 1 identified a contradiction between a mid-term strategy industry order and a short-term profit-generating market order: a dichotomy that other senior managers also acknowledged. Paradoxically, even when a lower ranked university had the potential to offer job-market relevant courses, OPM providers would still prefer to partner with high-ranked universities: There are likely to be universities outside that top-ranked group who might have the resources, motivation to actually meet our objectives much better than some of our existing partners… Those universities are typically unranked and specialise in technical, vocational skills… [that are] the needs, the demands of a global audience, perhaps more that than any esoteric piece of academic research [produced by high-ranked universities].

Widening access or reasserting the elite 'bundle'?
OPM providers did not conceal that their primary aim was profit. While widening access featured in their mission statements and interview narratives, the civil order was not among their priorities. Senior Manager of OPM 2 reminded us that the cost-free model of MOOCs had shifted to paid access to assessment materials and content after course completion. Senior managers' contribution to the question of widening access through online education was largely about content crossing geographical borders. While issues of access to the Internet, devices and payment currencies were discussed, the high cost of studies, particularly for poor students in South Africa, was acknowledged as a failing: 'We think about it a lot and we haven't really delivered' , Senior Manager of OPM 1 stated. Access then becomes reduced to the accessibility of the platform and the payment currency rather than the affordability of the course. To the weak civil order, there was an inspirational component regarding student support. Senior Manager of OPM 2 specified that they targeted primarily mature students considering a mid-career change. For them, equality referred to more learner-centred approaches: 'bringing 'the sage to the stage' (exposing students to VIP researchers) and 'the coach on the side' , or as Senior Manager of OPM 4 put it 'hyperrelational personnel' .
Yet, in most interviews, the domestic and fame orders were reinforced, when the 'bundle' of residential degrees was compared to online learning. 'Still, there is a non-academic piece … you can't replicate the experience in an Oxbridge dorm online … that is pretty priceless, what those guys are getting from a badge' , Senior Manager of OPM 2 said. Senior Manager of OPM 4 reflected: The job we do for our thirty-eight year olds, working professionals is so different from the job that universities do for their undergrad students. Undergraduate programs are to build networks, to find their mates for life … to immerse oneself in the social networks that build social habits […] Networking face-to-face, I don't think we can win that for our students online.
Networking is also a form of social capital that senior managers in OPM providers draw on to connect to university leaders. As Senior Manager of OPM 2 said: 'as [a member of a number of sector organisations] I end up bumping into friends, going into every VC's office …' Senior Manager of OPM 4 detailed: when you have a human challenge to gain alignment it could be between partners in a relationship, it could be between friendships, it could be between groups of people working in the same company across departments, across public and private lines. [That is how] … we're happy to count [top-tier universities] as some of our partners.
While not all partnerships were built on personal connections, the 'conventional wisdom' of the domestic order reinforced the logic of capital in partnering next to profit and ranking centrality.

Unbundling as viewed by senior managers of three public universities
Our conversations with senior leaders at universities had both similarities and differences to those at OPM providers. Similarities concerned a mature understanding of unbundling via digital platforms. University senior managers focused on unbundled education's potential reach beyond their usual catchment area, region and country, emphasising the civic order. However, it was difficult to tell whether the civic or the market order was leading decision-making regarding unbundling at universities as senior managers were eager to generate income in times of austerity, and treated unbundling as a possible way to address emergent budget constraints. Unlike OPM providers, however, they emphasised the importance of who controls content and innovative online pedagogy.

Challenging or reasserting rankings?
Unlike at OPM providers, senior management at universities perceived their own partnering value more broadly: in terms of their ability to target broader audiences, local relevance of their programmes, student employability and embeddedness with local industry. An intersection between domestic and industrial orders was employed when discussing quality in partnerships. University leaders prioritised domestic fit and comfort with OPM providers. They acknowledged that online education was resource-heavy and required staff training, but hoped that by employing agile OPM providers, risk-averse public universities could generate income without jeopardising their missions.
Regarding rankings, senior managers at University X (HAI) and University Z (HAI-HDI site) saw them as a necessary evil. Senior Manager 1 of University Z said 'Complaining about rankings gets you absolutely nowhere' , adding that this was particularly true for up-and-coming universities. At HAI University X, Senior Manager 1 was sceptical: '[T]he ranking system is evil …, nonsensical, distorted … students mistakenly go to research-focused universities. Researchers often don't teach undergraduates' . Yet, to convince their colleagues to diversify provision through online courses, the same senior manager used rankings: You can't persuade by diktat … [but] bring along respected institutions ahead of the curve. If someone from MIT says something it'll carry more weight than me saying it … they'll think, 'If MIT's doing it, perhaps we should think about it too' .
Yet brand sharing was considered risky at University X: '[OPM providers] benefit much more from the University's brand than we do from theirs' .
Rankings, and thus the fame order, played a less important role at University Y (rural HDI). As Senior Manager 2 of University Y explained: 'Where I'm sitting, the ranking is not important. The most important thing is to improve the quality of education: staffing, infrastructure, curricula, and throughput' . For them, HDIs competing for ranking with HAIs was like 'comparing apples and oranges' . The university should rather ensure their graduates could enter research universities or high-skilled jobs without feeling inferior. Senior Manager 1 of University Z feared rankings were used by the government to push HDIs to 'farm students' for wealthy high-ranked HAIs.
The link of the market order to unbundled provision was also shared at University X, where Senior Manager 1 said 'If income generated [through unbundled courses] improves our ability for teaching, research, or funding our students, I don't mind how it's raised' . Income was a way to reach existing students, not new ones. Senior Manager 2 of University Y shared this Realpolitik market-based view of unbundling-generated income: There is appetite in my university to work with the private sector toward unbundled higher education. We can emerge, if we act quickly as highly commercially successful ventures, trading in intellectual and human capital assets, to make money and bridge the deficit between state subventions and viability' .

Tackling historical inequalities?
Higher education senior managers had divergent concerns regarding student employment, leading to different strategies that combined civic and industrial orders. At University Z, graduates were mostly employed, yet there was no clarity how online or offline degrees correlated with employability. Senior Manager 1 reflected: 'If you've done a Master's in Psychology … working in McDonald's is not necessarily a success. ' At University X, employability in online versus traditional degrees was a concern only as long as the university was responsive to market gaps internationally and nationally. Senior Manager 1 explained: We need massive expansion [in areas] where we couldn't do face-to-face, like data science … Europe needs a hundred-thousand data scientists in the next decade. If we train twenty, each one will leave for Europe, we'll be left with none. So we've got to train hundreds in the hope that some stay.
HDIs faced different challenges. Senior Manager 2 of University Y prioritised the relevance of their programmes with the surrounding community, tailoring their short courses to the local schools and industries, thus widening graduate employability within the civic order. Senior Manager 2 there added that 'We are organising to embrace the industry' , but cautioned: 'There's been an appetite from the industry to prescribe curricula to universities, yet we have our own agenda … We should decide what skills we want to transfer to students, to apply in industry. ' Thus, within the industrial order, HDIs like University Y were eager to get involved in unbundling, with what they considered the most appropriate OPM providers as partners.
OPM providers were not approaching them for partnering. Yet senior managers there saw this as a result of excessive state regulation and budget cuts, as well as competition from urban universities rather than OPM providers' preference for higher-ranking partners. Senior Manager 1 of University Y reasoned: '[HAIs] have a very strong footprint in rural areas where they recruit students and create links with rural industries. so there will be competition between HAIs and HDIs'' . They believed rural universities and HDIs were better positioned to champion rural industries, if urban universities were to follow a civic order by recognising historical disparities, relinquishing competition. They saw unbundling as a way for HDIs to challenge the urban/rural divide: 'Either we push the agenda of unbundling and it fights against the cultural, climatic, and other elements, which mitigate against it and there's an explosion, or we stretch the boundaries and the thing grows naturally. I say let's stretch the boundaries. ' For managers of University X and University Z, however, stretching the boundaries meant moving into the African continent's market. Senior Manager 1 of University X stated: 'Through online education [we] offer access on the continent where higher education participation is even lower than in South Africa, and geography hinders access, or university expansion. ' This expansion strategy, while perhaps extending access and knowledge production, enfolded the logic of social relevance into the logic of capital. The strategy of University Z was less expansive but saw profit-making as related to innovation coupled with the brand expansion: 'We'll move out in concentric circles [from] our urban area -outwards. That's the global pattern -it breaks barriers. '

Reaching out to mainstream or marginalised learners?
In this conjuncture, the civic order of widening access was only an indirect effect of unbundling, conditional upon the market order of income generation. Senior Manager 1 of University X referred to skills required for online learning: '[Y]ou need university experience, mastering writing and reading skills and discipline, taking lecture notes and summarising before you can take advantage of an online course' . Senior Manager 1 of University Z stated: … someone who, regardless of skin colour, is academically competent, eager to get a degree and to move into a profession and they don't have a lot of money but they do have some. That's probably our market and that might be huge.
For Senior Manager 2 of University Z, this was also a matter of targeting different students: those who 'can afford it and would like to study from home … Not just your non-traditional students. We're not talking MOOCs, we're talking credit-bearing, pretty expensive programs. ' Online education was not seen solely as way of profit-making but added pedagogical value. Senior Manager 2 of University Z emphasised this aspect of teaching developed organically: 'The more enthusiastic and committed lecturers were the first people to try out new things' . However, they worried that online courses via OPM platforms did not prioritise pedagogy or access: 'Listen, this online stuff is about money. The party line is we want to expand access and reach more people but that's not what's going on. It's third-stream income because all South African universities are in crisis' .
Senior managers at University X praised online education's pedagogical affordances, but not necessarily coupled with unbundled provision. MOOCs and short courses were seen as beneficial for postgraduate students 'coming into PhD programs, from other universities, not sufficiently prepared in a particular foundation discipline' . Online unbundled provision's potential to surpass offline modes was seen as conditional upon its offer of support. Senior Manager 2 of University X emphasised the support that poorly prepared students could get through OPM providers, which could have an 'extremely efficient support regime … people, real humans by way of phones, emails and chat-rooms, enable them … to get an awful lot of students succeeding, and having the motivation to continue' .

Discussion
Senior managers at OPM providers and public universities spoke directly of key concerns in higher education. Analysing their responses not only through the orders of worth, but also in reference to their specific alignment with the logics of capital and social relevance at play in the field of higher education, allows us to anticipate directions of change in public higher education in the process of unbundling.
Both groups relate to an organisational logic of social relevance, all insisting discursively on the civil order of widening access, the inspiration order of the new affordances of digital technologies and the industrial order of employment and new production-embedded unbundled provision. Yet the market order is foregrounded among both groups. It produces another rationale and reality around partnering that privileges the logic of capital: short-term profits from commercially appealing courses make OPM providers rely on fame and domestic orders justifying high-ranking partners through elite institutional and personal networks. If this development remains unregulated and central to higher education institutional strategies, it could reduce unbundling to the logic of capital and inhibit equitable outcomes.
While senior managers of OPM providers perceive benefits in producing courses that address specific skills deficits, they choose partners according to short-term profit-making and university brands. Ultimately, most admit, they target top-ranked research-intensive universities, rather than vocational universities with industrial links. OPM managers' fame-induced perception that prestigious universities were not willing to partner if companies 'muddied their brand' reinforced this tendency, as did the domestic order of elite networks between senior managers of prestigious universities, policy circles and OPM providers. While most OPM providers' missions pronounced access and equality, senior managers there tended to prioritise the logic of capital over the logic of social relevance. They appealed to university leaders' concern with generating income in times of austerity, rather than their social justice and inclusion agendas.
Senior managers at public universities showed a more heterarchical system of justification of partnering decisions. Their narratives display more clearly how institutional logics can be combined to coexist within the same organisational field, compared to the more singular narratives of the OPM managers. Yet our data show that, under the pressure of income generation through unbundled partnerships, university managers employ the competitive market order of strategising individually rather than as a sector. Universities with different historical trajectories find themselves in competition. Despite their espoused commitment to a long-term unbundling strategy prioritising the logic of social relevance, universities in South Africa are not currently assisted by government regulatory policies to fully enact this logic, reinforcing the dominance of the capital logic.
Bearing in mind the competing tendencies of unbundled provision -the egalitarian promise of unbundling and the effects of marketisation on universities worldwide -the framework of institutional logics operationalised via orders of justification helps us to delineate the logics of social relevance and of capital currently pushing the sector in different directions. Based not on short-term profit and reproduction of prestige, but on the embeddedness of education in production and society at large, and in innovative and inclusive pedagogy, unbundling can be transformative. While this will require serious investment to train staff and students, and to develop online infrastructure within lower-ranked universities, it would also mean engaging with programmes with social relevance and industry relations. Without state investment in policy and resources that strengthen the social relevance discourse emerging from public universities, there is the potential that unbundled provision could well be overtaken by OPM providers' priorities, driven by the logic of capital, as evidenced in our data.

Conclusion
Framed within the logic of social relevance, the initial egalitarian and innovative promise of unbundled education could still address the necessity for university transformation and tackle inequalities. Our findings indicate, however, that programmes which respond better to the social relevance logic and could enhance student experience, employment and skills needed by industries are not a priority for OPM providers. As long as unbundling takes place within dominant marketisation frameworks, it is more likely to reinforce global hierarchies of prestige and locally rooted historical advantage (Boliver 2017, Tomlinson 2018, Swartz et al 2019. Also in this framework, public-private partnerships risk becoming overly reliant on the reputational capital of elite universities as a short-term strategy. The logic of capital, then, feeds off a 'technologically-determinist discourse that elevates efficiency gains, increased productivity and value-for-money, inside the logic of economic growth' (Hall 2013, 53). This could lead to unbundling entrenching old asymmetries in an already unequal terrain. Our findings suggest that, in the light of this, development and pitfalls are more explicit and nuanced national and institutional policies and models need to be produced to enable unbundled provision to realise its transformative potential.
There are therefore two possible scenarios of development within public higher education in South Africa and beyond regarding processes of unbundling. In one scenario, thanks to unbundling, following the social relevance logic, smaller and lower-ranked universities or those in peripheral locations could offer cutting-edge, well-endowed, shorter, flexible, lowcost industry-oriented and socially responsive courses to lower-income students. If successful, this would attract mainstream learners to these institutions, challenging the dominance of highly resourced institutions. In the opposite scenario, the dependence on profit-driven business decisions by OPM providers to optimise immediate gains might compromise the social relevance logic. Here, OPM providers will continue partnering with wealthy, high-ranked, research-intensive universities. Smaller universities, in this scenario, will be unable to participate in unbundling. Thus, unbundling would benefit a small number of students undertaking campus-based degrees, but would fail to fulfil its egalitarian promise. Public education would be instrumentalised for short-term OPM profit, and surplus would be guarded by elite universities for reinvestment into the exclusive elite 'bundle' .
As this is an emergent field of practice and scholarship, we can only infer reflections and intentions related to unbundling. Rather than discussing longitudinally implemented forms of unbundled provision, we draw on a limited number of actual relationships and plans. Nonetheless, the insights arising from the analysis allow us to connect the literature on unbundling with broader discussion of policies tackling inequalities in higher education. National and institutional systems are not permanent and unchangeable but are the outcome of specific institutionalisation processes (Lepori 2016, 248). Showing how two conflicting institutional logics affect the decision-making on public-private partnerships around unbundled provision in South Africa allows us to shed light on a conflict becoming central to public higher education everywhere. Regulatory engagement by governments in South Africa and beyond is urgent, as the outcome of current unbundling contestations could fundamentally alter the shape of the universities of the future.

Disclosure statement
No potential conflict of interest was reported by the authors.