Prepared for the QAA Scotland Enhancement Themes 2020 conference on 3 November 2020
Digital data technologies are at the centre of many current controversies. From Airbnb driving up over-tourism and property prices, to Twitter taking measures to reduce misinformation in the US election, digital data technologies are not merely tools enabling people to do things. They change what we do—how we travel, socialise, communicate, consume, and how democratic societies function.
Similar changes are already playing out in higher education, as part a long history of using measurement and rating techniques to change how universities operate, and the more recent history of datafication and surveillance in the sector. For at least ten years, educational technologies and increased data production has been affecting teaching, learning and decision making in universities—but today more so than ever. Digital technologies and data can productively inform decision making, at institutional, faculty or even individual levels, but should be informed by shared values and purposes of higher education. This is an important point, because digital data technologies may prioritise other values, purposes and interests—such as calculating the value of a degree as a ‘return on investment’ in labour markets, making ‘performance’ measurement the key indicator of teaching ‘quality’, or even by treating HE as a sector to be ‘disrupted’ for purposes of commercial gain.
In this context, important discussions need to be had about the kind of values and purposes we want to support as we begin the difficult process of sectoral recovery from the current crisis.
Higher education has been the focus of a great deal of ‘reimagining’ in recent years. Magazine articles such as the ‘Vanishing University‘ series in Quartz highlight a trend to envision HE systems where institutions, textbooks, and teaching practices are all transformed. Various experts and authorities speak of ‘smart campuses’, ‘unbundling’, ‘outsourced universities’ or the ‘University 4.0’ that is aligned to the ‘Fourth Industrial Revolution’ of advanced technologies, artificial intelligence, and automation.
These imagined visions anticipate a high-tech, digitally-driven, data-intensive, and partly automated university in which as many functions as possible are delegated to or augmented by digital tech, and all processes are recorded and analysed as data to inform decision making. The decision maker in the University 4.0 may not even be entirely human, as these visions highlight how data analytics and dashboards can augment institutional decisions; make decisions on behalf of students, such as ‘recommending’ course texts; or make pedagogical decisions for staff, such as evaluating student performance and assessing assignments.
Imagining and materialising high-tech HE visions
The new ‘visionaries’ of HE can be found everywhere from international consultancies McKinsey, KPMG, Deloitte and Ernst Young, and global organisations such as the World Economic Forum, to national sector agencies and government departments. Their reports stabilise a particular vision of the future of HE, aiming to produce consensus and conviction in others that this data-led digital transformation is desirable and attainable. They provide ‘actionable’ insights into realising such transformations in practice too.
They’re also full of assumptions about the value and purpose of higher education. Those include alignment of the university to the data economy as a producer of the technical talent required to boost innovation and productivity. A higher education becomes centred on hyperspecialised careers development in the science, technology, engineering and maths disciplines, with the related downgrading of the arts, humanities and social sciences.
It’s right to be sceptical of the grand idealised claims of such visions. Their full realisation is doubtful. But, in more limited form they are already materialising in policy proposals, governmental interventions, and on-the-ground technical developments. In the UK, various reports and requirements produced by Jisc, the Office for Students, the Higher Education Commission, KPMG, Universities UK, and the Department for Education have pushed the data agenda into actual developments and uptake. There is considerable multi-sector consensus about the directions higher education should take to ‘satisfy’ digitally-savvy students—innovating the ‘learning experience’—and to deliver economic renewal through innovation in teaching and learning.
High-tech HE visions have materialised most spectacularly in the rapid growth of the educational technology industry and its financial sponsors. The education market intelligence agency HolonIQ recently identified and taxonomised hundreds of edtech suppliers serving every conceivable task or activity of higher education in a soc-called Global Learning Landscape, and estimated edtech to be worth hundreds of billions of dollars.
A market intel agency like this does two things. It attracts customer attention, like an online recommendation engine for ‘what works’ in higher education ‘digital capability’. And it attracts venture capital and private equity in products that can deliver profitable return on investment. In other words, market intel makes edtech markets by convincing customers that digital solutions will solve their problems, and by building up the financial power of the industry that provides technological solutions to them.
We can understand the educational technologies supported by these market-making activities as the computer-programmed instantiation of the vision of a high-tech, digitally-enhanced and data-intensive HE sector. That is to say, the imagined HE of University 4.0 or the smart campus gets operationalised by being written in computer code as the software to be used by organisations.
The learning management system Canvas, for example, was acquired this year by a private equity firm for 2 billion US dollars. This followed its chief executive’s claims that Canvas had the most extensive database on the student experience on the planet, which it would monetise with predictive algorithms and study recommendations. The online learning company 2U claims it provides an ‘operating system’ for hybrid teaching and learning, bringing both the language and the specific mechanisms of computation to education.
There are also new product types that use emerging sources of data. Spotter and others now offer student location monitoring through smartphone Bluetooth or wifi identification. Library system providers like Ex Libris have expanded to become cloud-based library services platforms running granular analytics. They don’t just support library managers to monitor usage, but feed back into course leaders’ decisions about what texts to assign, or even automatically anticipate and recommend what students should read next, based on performance comparison of items included on lecturers’ reading lists.
Computer programming ‘executes’ and operationalises a particular vision of what HE is, can or should be. Once programmed according to these ideals and the business plans of companies, edtech gets into the hands and practices of university managers, educators and students—with the aim of changing processes of teaching and learning, and of affecting organisational behaviours and decision-making practices.
What these examples show is that a vision of the smarter data-intensive University 4.0 is being pursued into being, sponsored by consultancies, put into policy proposals and projects, financed by investors and paid for by institutional customers, and operationalised through programming. The problem, according to many promoters of this vision, is that the HE sector itself has been resistant to such change. Until, of course, Covid-19 struck earlier this year. In the educational emergency of mass campus closures, opportunities for ‘long overdue’ revolution were identified, and higher education entered into a historically unique period of edtech experimentation.
A key question surfaced by the edtech emergency is what values and purposes were pursued in this experiment? Yes, values of equality, access, fairness and quality education for all became key issues, particularly to address digital inequalities and ensure educational continuity for millions. But this was coupled with private interests such as securing market share, competitive advantage, and return on investment. The issue is whether these private interests support the public values, ideals and common goods of education by offering temporary ‘relief’ from the catastrophe, or whether they are actually motivated to ‘reconstruct’ higher education for the long term in ways that reflect private agendas.
The stakes here are high: long-term commercial reconstruction around the visions and programmed templates provided by digital data technologies could fundamentally alter how HE operates, potentially raising major and damaging disputes. The so-called ‘global education industry’ is highly controversial for positioning the private sector advantageously to deliver an increasing array of tasks in universities.
Recently, too, academic researchers have begun documenting the issues arising from digital surveillance during the pandemic—such as governmental dependence on global technology companies or consultancies for epidemiological and contact tracing. HE is at the centre of these two issues—increasing privatisation and increasing monitoring through data.
One indication of the increasing privatisation is the surge in edtech markets. Large quantities of investment have been made over recent months, both as direct venture capital in companies and their products, and as investments in new portfolio edtech funds dedicated to ‘digital transformation’. The wealth and asset management company Credit Suisse ran partner content in the Financial Times on the potential of the ‘Netflix moment’ for education, and attracted investors to its ‘Edutainment Equity Fund’.
These and other funds reposition higher education as a fast-moving market opportunity, and as an ‘edutainment’ sector with the disruptive potential of Netflix and other platforms. One of the main market segments of interest in these funds is consumer edtech—selling products and streaming content to students with all the convenience and subscription fees of streaming platforms.
One of the other main market segments to grow in interest and activity was online degrees and program management services. HolonIQ calculated that the online degree market would be worth 74 billion dollars in 2025. Online program managers have begun diversifying from the provision of graduate programs for distance international students to undergraduate programs, and shifting to being enablers of hybrid programs who partner with institutions.
This model of public-private partnership reconfigures HE as a hybrid between universities and private industry, and is itself of serious interest to investors, companies and institutions alike—for investors, for the ROI; for companies, for the revenue-sharing agreements, where they often take around 50-60% of student fees; and for universities seeking income from enrolments. The edtech commentator Phil Hill recently wrote about the ‘instant global campus’ as the outcome of such deals.
These public-private partnerships anticipate a greater degree of unbundling of core university operations to outsourced private providers. Take Noodle, for example, which offers a full stack of services: marketing and recruitment; business model; data flow between the student info system, learning management system, and third-party integrations; and learning design support for online and hybrid education, as part of its promise to produce an ‘agile campus’ for its partners. Noodle is not just an online learning platform provider, but a kind of outsourced shadow campus that augments many of the tasks of its public partners, right down to the organisation and development of taught courses. It offers a platform template for pedagogy that constrains how HE partners can develop and deliver courses.
Data lakes and cloud campuses
A key feature of online learning is how private providers have configured higher education in terms of ‘job-relevance’. Massive open online courses returned this year as data-driven platforms for online and hybrid education. Coursera for Campus extended the MOOC into formal degree programs, providing back-up for students and staff lacking access to either classrooms or online learning facilities. But Coursera was also all about ‘job-relevant, credit-ready’ online education, including new ‘industry credentials’ offered by partners including Google, Amazon and IBM that might even short-circuit the need for a formal higher education. Google’s own offer on the platform was a 6 month online credential that, it claimed, would count as much as a four year degree for posts in the company.
Coursera even released reports detailing its own ‘impact’ and ‘quality‘, using ‘eight years of learner data and nearly 200 million course enrolments to provide actionable, data-driven insights into how instructors and learners can optimize their digital learning experience’. This is the realisation of the data-driven university that is barely possible on a campus location, with mere thousands of enrolments to study. Coursera has 200 million enrolments to analyse, giving it unprecedented power to quantify and identify ‘what works’ in digital teaching and learning. It’s a kind of cloud university constantly producing outputs and recycling these as data inputs to improve its product.
Coursera indicates the truly global scale of many of the commercial digital suppliers advancing across higher education. The data they extract and analyse requires significant back-end infrastructure for data storage and processing too. Coursera itself uses Amazon Web Services to ‘handle half a petabyte of traffic each month and scale to deliver courses’ to its millions of learners from around the globe. The learning management systems Blackboard and Canvas, too, are institutional AWS partners, meaning a significant proportion of the world’s universities are now plugged in to the Amazon ecosystem of cloud and data infrastructure facilities.
In September AWS announced a price discount scheme for universities to develop ‘data lakes’ of very large volumes of heterogeneous information. The process of ‘architecting a data lake‘ involves tying together multiple AWS programs for data storage, interoperability, management, analytics and machine learning with institutions’ own student info systems and their third-party learning management providers.
The longer-term implications of this remain unclear, but higher education now depends to a substantial degree on AWS for back-end data services. We might even say that the production of data lakes condenses into the formation of cloud campuses that exist in Amazon’s giant data centres—cloud campuses that overshadow the physical spaces of the university while siphoning off their data lakes, and performing the programming actions required to run the partner institution in its idealised University 4.0 format.
The result of the formation of such cloud campuses is the possibility of using artificial intelligence for ‘personalised’ education. Personalisation is the central objective of the University 4.0 imaginary. This is a university that runs constant data analytics, in real time, to diagnose students’ strengths and weaknesses, their progress and problems, in order to prescribe automated interventions or direct educators’ attention.
The University of Buckingham’s proposed ‘Education 4.0 trailblazer degree’, for example, will use AI ‘to create personalised and adaptive course content tailored to each student’s specific abilities and learning methods’. Microsoft’s Power Platform offers similar promises of utilising machine learning for personalised learning. As does Coursera’s recent upgrades for ‘tailored study suggestions’, ‘smart material’ recommendations, and personalised advice towards ‘career-specific skills’.
The University 4.0 vision may have been relatively long in the making, and slow in the uptake, but it is now materialising at speed. It represents the hybridisation of the public role of universities and the private interests of technology providers: promises of institutional relief and recovery from the current crisis are tied to the enrolment of universities on to corporate cloud infrastructures and AI systems at truly huge scale. It anticipates the reconstruction of HE by the tech and data industry.
Reimagining the post-pandemic university
The purpose of detailing these transformations is not necessarily to critique them as negative developments. They demand much more careful empirical study for their implications and effects. And nor is it to dismiss online teaching and learning, which can be and often is done creatively and purposively by educators in a range of contexts. But these developments do of course raise many critical issues. Articles in the Chronicle of HE in the US, and the UK edition of Wired magazine, have questioned the ‘edtech mania’ of ‘utopian-minded tech gurus’, and the use of data analytics as ‘surveillance’ technologies to spy on students through their course ‘engagement’ traces. Even Teen Vogue magazine ran a piece on the student anxieties caused by course surveillance.
In the wake of scandals over the Higher and A level exams this summer, the media is already interogating the other algorithms that might be deciding students’ futures. If we turn over the sector to commercial companies, AI, data analytics and cloud centres—and absorb the problematic language of technological solutions, data-driven decision-making and personalised learning—we can expect this critical scrutiny, and perhaps more drastic actions, to grow in intensity.
A modest solution, perhaps, is to open up a much richer discussion about the possibilities of the post-pandemic university, by drawing on the sector’s own research and expertise. Paul Ashwin’s Transforming Higher Education: A Manifesto returns us to questions about the values and purposes of higher education. He resists the dominant framing of HE as an economic investment, even in countries that do not charge fees, and the valuation of a degree in terms of preparing the future workforce for the benefit of individuals and the economy. The purpose of a higher education, he argues, is ‘bringing students into a transformational relationship with disciplinary and/or professional knowledge’. The book asks readers to consider this transformational relationship to knowledge – rather than transforming education into data-driven career skills training for employability – as the central purpose of a higher education.
The Manifesto for Teaching Online, by Sian Bayne and colleagues, questions the ‘impoverished techno-corporate futures’ for higher education promoted by commercial and government edtech. It asks that we take into account the values and experiences of teachers and students in universities instead. The book embraces the creative, critical and innovative potential of online education driven by teaching values and purposes rather than driven by data, arguing that ‘those of us actually teaching using technology … need to take active control and ownership of digital education’ rather than allowing it to be driven by data-led technological solutionism or commercial business plans.
In these manifestos, digital data still plays a part in decision making, but those decisions are informed by truly educational purposes, by acknowledgement of the social and economic contexts of education, and by a recognition of the public value of a higher education to individual lives and to the wider society.
One model for dialogic negotiation over the shape of higher education to come is the Post-Pandemic University Network—a series of online events and a collective blog aimed at stimulating and sustaining discussions about the purposes of higher education and how to practice them. Working collegially in such ways, the sector needs to take the lead in restating its values and purposes, and making private providers work to advance explicitly educational aims, rather than edtech and even Big Tech making decisions that make higher education work for them.