Part 2: Three visions of a university
This is Part 2 of an eight-part essay, ‘So, You Run a University?’. This essay is authored by Darcy W.E. Allen, Chris Berg, Sinclair Davidson, Leon Gettler, Ethan Kane, Aaron M. Lane and Jason Potts. Part 1: Introduction is here.
To design university strategy, we first need to understand what a university is, or at least what people typically think a university is. This task is more difficult than you think. For some, universities are romantic places—they are for learning and for truth seeking. Others look at the university and see research machines—faculty make the chemicals and lasers that push industry forward. And of course, to many universities are education factories—they take young people and transform them into credentialled graduates, ready for the workforce. Let’s explore these competing visions.
The university is an ancient institution. It is older than the joint stock company or the Westphalian state, dating back to medieval Europe. And, as historian of education Walter Rüegg points out, the university is the only institution of that time and place that has survived largely in the form that it was established. We can see some parts of the modern university in its eleventh and twelfth century ancestors.
The earliest universities were structured as semi-autonomous, voluntary, and even exclusionary communities that brought groups of students together with groups of tutors for mutual gain. Many developed as extensions of monasteries or cathedral schools under the patronage of the church. Whether the university was governed by the faculty, or by the students—such as the University of Bologna, founded in 1088, which was brought together by a guild of students—the medieval university had a ‘federal’ character where multiple groups with distinct organisational structures and hierarchies come together to cooperate.
No institution survives unchanged for a millennia. Through time the purpose of the university needed to be rethought, partly in response to innovation. Chad Wellmon argues that the disruption of the nineteenth century university was partly driven by technological change. The explosion of cheap printed books meant that old clerical and generalist education was looking obsolete. What was the point of the lecturer when so much knowledge was available through reading?
More practically, how would the modern university be organised, governed, and professionalised? In France, the university sector was structured around elite speciality schools—the grande ecoles. Teachers taught a rigid curriculum. Students were subject to strict discipline. For the post-revolution French state, the university was an extension of centralised industrial planning—providing the exact training that the nation needed.
Germany offered a more liberal vision—rather than centralised control, the state stepped back. While the professors were civil servants, they had a wide freedom to determine what they taught and how. The university kept multiple faculties within one institution while at the same time delivering to its students a scientific education in their chosen discipline rather than drilling them in vocational skills. It combined research and teaching under one roof, with the idea that the former influenced the latter.
With variations, it is the German model that has come to dominate the global university sector. It is from this German model that we get two starkly contrasting visions of the university in the twenty-first century: the romantic vision and the industrial vision.
The romantic vision
In the romantic vision, the university stands apart from the other institutions of society. Universities are separate spaces motivated by distinct humanistic values. While this romantic vision can be traced back before the nineteenth century disruptions described above, it was in the philosophical unification of research and teaching as a combined endeavour that the romantic vision flourished. The modern romantic vision adopts the idea of a separate, cloistered institution outside the realm of secular politics and commercial society. This vision recreates the medieval university and combines it with a particular idea about scientific and humanistic learning. The university “embodies an alternative set of values in [its] very rationale”, in the words of the University of Cambridges’ Stefani Collini. Leiden University’s Willem Otterspeer writes that:
a university is a form of social capital, one of Western society’s answers to the dilemma of collective action, an instrument for preserving and restoring equilibrium, and hence for fostering continuity … a university is a confidence-building mechanism that generates solutions to the serious problems facing society.
The romantic vision sometimes takes a more distinctly conservative flavour. For the conservative romantics, the university is a place where outside passions are calmed. John Henry Newman, in his influential 1852 lectures (prior to his appointment as inaugural rector of the Catholic University of Ireland) published under the title The Idea of the University, argued that the true university:
aims at raising the intellectual tone of society, at cultivating the public mind, at purifying the national taste, at supplying true principles to popular enthusiasm and fixed aims to popular aspiration, at giving enlargement and sobriety to the ideas of the age, at facilitating the exercise of political power, and refining the intercourse of private life.
Those outside passions were as much economic as popular or political. Writing in 1950, the philosopher Michael Oakeshott argued that a student should not leave a university with “a qualification for earning a living or a certificate to let them in on the exploitation of the world.” Rather, the student should have “learned something to help him lead a more significant life … [and] had the leisure to replace the clamorous and conflicting absolutes of adolescence with something less corruptible”.
The modern culture wars critique of the university—that universities are hopelessly biased to the left and fail to teach the great ideas of civilisation—is also deeply infused by this romantic vision of the university. Allan Bloom subtitled his 1987 book The Closing of the American Mind, an extended critique of what he saw as the dominance of relativism in the academy, with the explanation: How Higher Education Has Failed Democracy and Impoverished the Souls of Today's Students. For Bloom, the role of the ideal university is “to maintain the permanent questions front and center. This it does primarily by preserving—by keeping alive—the works of those who best addressed these questions”.
In the romantic vision, academia as an ‘ivory tower’ is a feature, not a bug. It should be physically and mentally cut off from the more pragmatic concerns of the world. The demands of the market can only ever be an intrusion on the fundamental role of the university as a romantic institution. As Bloom writes, the university must be “must be contemptuous of public opinion because it has within it the source of autonomy—the quest for and even discovery of the truth according to nature”. There is little room here for the strategic interests of the university as a corporate organisation.
The economic sustainability of the university is a regrettable necessity. Oakeshott, for instance, rejects much of the modern university’s way of thinking about itself: sceptical of mission statements, of skills-based education, of the bewildering array of alternative courses and micro-credentials. We can only guess what he would have thought of the twenty-first century university’s interests in research commercialisation and research funding on the basis of ‘national priorities’. The romantic vision sits uncomfortably with the enormous corporate enterprises that many universities are today.
The romantic vision is deeply infused into academic culture. Scholars and teachers craft their expectations around this romantic vision. It informs how they think about their employer-employee relationship. And it shapes the public debate about the social purpose and obligations of universities. But romantics can only come after the university functions as an economic entity first. Therefore while the romantic vision may be extremely appealing to faculty, it is useless for understanding the future of the university. And ironically, for all the claims about the importance of university autonomy and academic freedom, it provides the university with little strategic room to maneuver and be entrepreneurial. Rather, the romantic vision stacks the deck and demands that university administrators pursue almost nothing but unencumbered funds from the government. In an environment where governments are unlikely to provide such support, the romantic vision provides no guidance for the strategic choices of the modern university.
The industrial vision
The romantic vision of the university appeals to scholars and (some) students. But the romantic vision is feeble because it fails to appeal to those groups who fund universities: governments, students and industry. These stakeholders are more likely to view the university as having a distinct set of economic functions—universities take inputs and produce specific outputs. They produce and credential the productive employees of the future. They produce research that has direct social or industrial benefit. Universities, in this industrial vision, are credential and innovation factories.
The credential factory
In the industrial vision, students attend university to obtain a qualification. Universities endorse that students have met some standards. In some systems those standards are controlled by government regulators, while in others those standards are controlled by professional associations. Ultimately the qualification might give students entry to a particular profession, or at least set them on a path of more highly paid employment than they would otherwise be qualified for.
According to this vision, the university is like a factory. Students go in one side. They are processed over many years according to a standardised formula. Eventually graduates emerge with a certification that proves they have been processed. They then enter labour markets where employers use that certification to assess their employability. And this process seems to work: the evidence is quite clear that university graduates earn more over their lifetime than non-graduates. But why? What does certification symbolise? What is it that the factory produces that is valued by employers?
Most people—including those who study the education system—assume that getting a degree involves the transfer of specific skills that can be used in the labour force. The university factory inputs skills into students that transforms them into something desired by industry. Students learn how to design buildings or read balance sheets. Economists describe this as building human capital. But human capital is too simplistic an explanation for the role of the university as a credential factory.
How can we explain that in many degree programs universities don’t seem to be imparting to students skills that are directly valuable in employment? An alternative explanation for the value of education is the ‘signalling’ theory of education (see Box 1 below). All of those years students spend being processed might not impart skills, but act as a signal to employers that the students have been able to voluntarily commit to education over an extended period of time. It is hard for employers to identify which potential young employees have a good work ethic. The certification is a costly signal that hints at the future productivity of potential workers. University factories still produce certified graduates, but they’re not certifying what they say they’re certifying.
Of course these interpretations are not mutually exclusive. Most undergraduate and advanced degree programs impart both skills and non-skill based activity. But both see the university as a machine for graduate production, whether that production consists of learning skills or testing work ethics.
Two theories of why students do degrees
There are two ways to explain the economics of education: the human capital model; and the costly signalling model.
The human capital model
The economics of education begin with the work of the 1979 economics Nobel Prize winner Theodore Schultz and his work on the human capital model. He introduced the idea of human capital in his mid-century work on agricultural productivity. We can think of a farm as improving its productivity in two ways: invest in a new tractor, or invest in educating the children. Education could be a better capital investment to improve productivity because it enables more knowledge to enter the farm.
In the same way that you rationally assess any investment decision, human capital helps us think about education from the perspective of return on investment. Those investments have both private and public benefits. From this perspective universities produce human capital through investment, and governments ‘invest’ in university education because it is industrial infrastructure for economic growth.
The costly signalling model
Michael Spence, another economics Nobel winner, introduced us to the alternative costly signalling model. In his 1973 paper, Spence invited the following metaphor: there are two ways you can add value to a rough diamond: polish it, or grade it. Polishing adds value because it makes the rough diamond better, shinier and smoother. Rather than polishing the diamond—the analogue of the human capital model above, where educators add work to the student to make them better, shinier smoother—we could simply grade the diamond. Grading adds value to that rough diamond because it provides assurance about what it is—its clarity, colour, flaws, carats.
The graded diamond is the same diamond. But it is worth more because I know more about it. Students are like rough diamonds. They will certainly know themselves what qualities they have—how smart and hard working and conscientious they are—but can only capture that prospective value if they can reliably prove that to an employer. Universities, in this model, are like diamond graders, who closely observe, inspect and then certify an uncertain product. Universities add value by providing assurance about quality.
Why does ‘grading’ a student take four years? Because it must hurt. This was the genius of Spence’s model, which drew on an idea from evolutionary biology about honest signalling called ‘the handicap principle’. Anyone could say that they are smart and conscientious. But only someone who really was that smart and conscientious could afford to waste four years studying something really hard and utterly useless. Someone less smart or hard working could not afford to carry this easily observed ‘handicap’. (Handicap is from racing, as the extra weight added to a fast horse to make the field more competitively even.) Four years ensures that a degree is an honest signal of quality.
The human capital model feels right. It flatters prejudices of educators and scholars about the societal value of the work they do. The costly signalling model seems scandalous. It emphasises that the university's main business is certification, and that adding human capital through teaching is just a means to that end. Years and billions are spent getting students to do pointless things at great expense to figure out whether they’re lying. And once we find out they’re not, we celebrate that with glorious pageantry. That sounds much less noble. It sounds faintly ridiculous.
Spence’s ‘costly signalling’ model of education is the foundation of Bryan Caplan’s 2018 book The Case Against Education. As a tenured economics professor at a top US university, Caplan puts the human capital and costly signaling model to the test. Caplan’s conclusion is pretty stark: the evidence, from a wide variety of sources and tests, overwhelmingly supports the … costly signaling model. The clue was in the subtitle of his book ‘Why the education system is a waste of time.’
But, whether by human capital building or costly signalling, the university factory isn’t solely focused on labour market outcomes and graduate earnings potential. While students may seek an education in their own self-interest, education has a positive externality that accrues to society as a whole. In the aggregate, a more highly educated population tends to be a richer population. University education is correlated with non-market benefits accruing to the individual student, such as better health, social tolerance and lower risk of imprisonment. Indeed, higher education can support other social goals as well. Education might create more informed voters, enhancing democratic decision-making. Educated university graduates might be more conscientious, leading to improved social and environmental outcomes. Education might also encourage the adoption and experimentation of new technologies and ideas, propelling innovation.
The university is a machine that takes the raw material of students as inputs and produces educated and socially responsible individuals as outputs. Governments subsidise education to produce more of these socially desirable outputs.
Education as a credence good
One reason that the economics of education is interesting is because education is a peculiar type of good called a credence good. There are three ways we can categorise economic goods and services:
Information goods: all the features of the good can be provided before consumption.
Experience goods: some (perhaps all) of the features of the good can only be provided by consuming the good.
Credence goods: all the features of the good cannot be known even after consumption of the good.
Why do these categories matter? We can turn an experience good into an information good through repeat purchases. For instance, you know what a glass of your favourite wine tastes like. It was an experience good the first time you tasted it, but over time it became an information good. But we cannot turn a credence good into either of the other two categories. Once a credence good always a credence good. And education is a credence good: you cannot know all of the features of education (including its quality) even after you have consumed it.
Let’s clear up some misconceptions. A credence good doesn’t imply no knowledge about the good. It is possible to advise a potential student what they will be doing at university. It is possible to tell them what courses or subjects they would undertake and the content of those courses or subjects. It is even possible to suggest that they will enjoy their time at university. In that sense being at university is an information good. That sort of information (classes, courses, subjects, assessment tasks, etc) relate to university processes and procedures. Universities are able to describe those things to all university stakeholders.
Universities supply education services. Students demand education services. But to become ‘educated’ is a joint production between the student and the university. Now the consequences of education being a credence good are problematic. What can a university claim about any given graduate? The uncomfortable answer is less than we would like and less than what we do. We can claim that graduate X was exposed to information Y and assessed in process Z. We cannot claim that any given graduate has knowledge or a particular skill due to our efforts. This is because we cannot exclude prior knowledge, innate ability, or self-study from any calculus of ability or performance.
Universities tend to select their cohort on proven past performance and then graduate them upon compliance with university processes and procedures. We assume and claim that value-add occurs while students are at university. Anecdotally we can point to the fact that a third year student is more knowledgeable than a first year student in general and on average. But a third year student may be older, wiser, and more knowledgeable of university processes and procedures than a first year student.
Another uncomfortable consequence of education being a credence good is that students and employers know this to be the case (even if they cannot articulate their understanding of education being a credence good). Students generally do not care if lectures are cancelled, or finish early, or if the lecturer is unable to cover the entire syllabus in term time. There are some exceptions to this but as a rule students are indifferent to these events. Students might complain that they had to ‘come in for nothing’ but even that is a complaint about their own opportunity cost of time and not a complaint about missed information or knowledge.
Nor do employers care about such things. Prospective employees may be asked which university they attended, which course or subjects they studied, and what grades they achieved. Employers do not generally ask to see course or subject guides, nor assessment tasks, nor individual pieces of assessment. University education is an initial screening mechanism to employment. It is not a key decision variable in offering employment.
The innovation factory
The modern university factory isn’t just in the credentials business—universities are industrial powerhouses of research. Research has both intrinsic value by deepening society’s knowledge about the world, and value as an input into economic and social innovation. Eric S. Lander of MIT and Harvard and Eric E. Schmidt of Google are explicit about this vision of research as industrial production:
For more than a half century, the United States has operated what might be called a “Miracle Machine.” Powered by federal investment in science and technology, the machine regularly churns out breathtaking advances.
Research can be fundamental or practical. It runs the gamut between investigations with little clear practical applications—such as the study of pure mathematics—to research directly targeted at solving specific commercial problems. In some fields, such as engineering it is common for researchers to go on to build innovative companies with inventions they developed as part of academic research. Indeed, many universities have commercialisation teams that assist forming spin-out or start-up ventures.
Again, the industrial vision of the university as an innovation factory is not a cold one. The benefits of university research need not be just productivity and economic growth, but can be directed to tackle social problems. Since the mid-twentieth century research policy has been driven by a belief that fundamental research and socially-beneficial research would not be provided by the market, and that governments have to step in with research funding.
The industrial vision gives us an uncomfortable relationship between teaching and research. To teach is to impart skills and a general body of current knowledge within a given field. It is not to romantically inculcate the moral mind. To research is to make discoveries at the frontier of science. Given research specialisations those advances are found in increasingly niche areas. Frontier research is not for the university’s students. Research is for the industry outside the walls of the university. Indeed, Michael Oakeshott made this point:
As teachers, they may be either better or worse than those elsewhere; but they are different because they are themselves learners engaged in learning something other than what they undertake to teach. They are not people with a set of conclusions, facts, truths, dogmas, etc., ready to impart or with a well-tried doctrine to hand out; nor are they people who make it their main business to be familiar with what may be called “the current state of knowledge” in their department of study; each is a person engaged in the activity of exploring a particular mode of thought in particular connections.
It is unclear why in modern universities employees are both scholars producing research and teachers producing education. The skills for research and teaching seem vastly different. In a large factory we would expect these roles to be performed by different people. This seems like a rejection of the benefits of specialisation at the highest levels of intellectual life. And of course, for many faculty teaching is seen as secondary to research, despite the fact it is the main revenue source for the organisation. The primary qualification for a typical faculty is an advanced research degree, which requires the candidate to make a research contribution to frontier knowledge and contains little to no training for the teaching that will dominate their career if successful.
While the allocation of teaching and research roles might not align with the industrial vision, the presence of specialised non-academic management does. Michael Shattock, the former registrar of Warwick University, shows how university governance has moved from the traditional spaces of the faculty board and university senate into empowered ‘executive offices’—typically staffed by non-academic staff members. Corporate management is a specialised role. If the university produces graduates like a factory produces widgets, we would expect to see the university under the control of professional management. In the words of Glyn Davis, a former Vice-Chancellor of the University of Melbourne, the university now uses the “language of management” and deploys “administrative processes once the preserve of commerce”.
Why do universities do research?
Universities do research because that’s what everyone—parents, students, politicians, donors, faculty, trustees—expects them to do. That’s why universities in general do research. Research sets a university apart from other colleges and institutes of post-secondary education. But why does your specific university do research? Research can be astronomically expensive. Of course some of this may be defrayed through government grants or private contracts, but that is generally only some of the cost.
Universities don’t do research for the general benefit to humanity and society. That’s just propaganda. They do research for reputation. A university’s reputation is tied to the scope and prestige of its research achievements. The reason is tight business logic: a university faces a quality uncertainty problem that it solves by doing a hard costly thing that only a quality institution could do, namely high quality research. As we saw in Box 1 above, universities are signalling.
The research production function takes research inputs (machines, experiments, labour) and producers research outputs (papers, books, knowledge).
Research outputs are translated into reputation (solving a quality uncertainty problem) through a system of journal gatekeepers and industry impact.
Reputation is translated into market power (further research grants and industry funding, more and better students, attracting superstar faculty).
Market power is competitively translated into university objectives that are allocated across stakeholders.
This four-stage mechanism converts research inputs into university benefits. Research inputs (academics) produce research outputs (publications) which produce reputation (seen by other academics, but also students, employers and benefactors). This reputation enables the university to exploit a competitive position because people see it as higher quality.
Research spending is a costly quality signal that a university will strategically choose to invest in, and competitively with respect to other universities, to maximize the benefits from a business model that is based around teaching. The fact that research also produces benefits to the economy and society is an unintended (but happy) consequence.
If the university is a factory producing credentials and socially valuable research, then it looks like a strikingly inefficient one. One of the most remarked-upon trends in the university is the growth of administration—that is, bureaucracy—in the last few decades. In the United Kingdom, administrative and professional staff now outnumber academic staff in the majority of universities, according to a 2015 analysis. In his book The Fall of Faculty, Benjamin Ginsberg argues that academic staff are being displaced by
armies of functionaries—the vice presidents, associate vice presidents, assistant vice presidents, provosts, associate provosts, vice provosts, assistant provosts, deans, deanlets, deanlings, each commanding staffers and assistants—who, more and more, direct the operations of every school.
This view that there has been a huge jump in middle and senior management roles is widely held within the university sector (and particularly amongst academic staff). While in the next part we further consider the function of this administration, it is important to foreshadow here that the conventional view of shifts in administration may be misleading.
While academic staff have declined as a proportion of the total employment, in the United States the sharpest growth has been in non-managerial professional staff—that is, those who have operational and administrative roles, technical and paraprofessional roles, student-facing roles and sales and marketing roles. There has been growth in the administrative centre of the university—but it has not been only, or even solely, driven by the elaborately-titled, well-paid ‘Deputy Associate Pro-Vice Dean (Research) (Acting)’ type roles. We return to this in Part 3.
The industrial vision and the romantic vision of the university aren’t necessarily mutually exclusive. But they exist in an uneasy truce. Neither can ever be fully realised.
The romantics see the industrialists as an uncomfortable intrusion. At its best, the romance exists in the calmer side areas of the factory—embodied in the students who want more than just a credential, but an induction into a community of learning, embodied in the research motivated by less by the prestigious publication or target, but by pure curiosity.
The industrialists see the romantic ideal of university life as an ideology only to sometimes pay lip service to. The romantic vision offers little guidance to administrators, appeals to only a small fraction of the student base, and is of no interest to national agencies that fund the university sector (whether through education subsidies or grants).
Donald Kennedy, in his book Academic Duty, makes clear the fundamental conflict between what the academic staff imagine the social function of the university to be, and how society at large views the university as an economic institution:
Whereas those within the system generally believe that their mission is to produce graduates who can think well and work effectively, and who are able to understand, analyze and reflect upon their culture and upon the natural world, much of the world outside sees higher education as a credentialing device: a way of estimating, for employment or other purposes, the comparative worth of individuals.
As the sociologist Steve Fuller writes:
The fate of the university depends on what we make of its history up to this point. Two views are prominent nowadays, neither of which helps the case for promoting the institution in the 21st century. The first is that the university represents an impossible ideal that has never been realised, and indeed has been invoked to cover a multitude of sins, especially a velvet glove approach to the perpetuation of rule by elites. The second, perhaps more charitable view is that the university, unsurprisingly for an institution that has been around for 900 years, has become obsolete, such that today it tries to serve too many functions, each of which is better handled by a separate organisation.
Here we have our dilemma. On the one hand, the university is accused from within of failing its social role. Universities fail to provide a separate domain of disinterested learning outside the market and outside politics. But on the other hand, the university is accused by society at large of diverting resources away from its core economic role. The university is expansive and bloated.
There is a good reason for this dilemma. Generations of university leaders, employees and observers have failed to understand the basic economics of higher education. But this is not their fault—until the twentieth century we did not have the theoretical tools available. But now, by using the insights from the economics of platforms—the economics that is used to understand technology-first businesses like credit card networks, social media, and gaming consoles—we can see not only why universities look as they do, but how they will look in the future.
The university as platform
The way that we think about universities needs a clean break. We must escape from our romantic and industrial visions. To do that we need new theory.
Choosing a theory that fits the complexity of universities is hard. We begin with a simple and intuitive insight: universities are organisations that match different groups to trade. They match students with teachers, researchers with industry, donors with socially valuable research, high schools with placements for their students, national governments with a skilled workforce and a national innovation system. We have a name for these modern matching services: ‘platforms’. Through the study of other platforms like ridesharing, shopping malls, newspapers and credit card networks, scholars have developed ‘platform economics’ to understand these new business models. In this essay we use platform economics to understand the modern university.
Platforms are often associated with technology companies, but they have existed ever since we began to trade. Every small town and every great city has had a central place where producers and consumers could meet and exchange. The great medieval trade fairs in Champagne were central places where merchants from across Europe could trade. These fairs had distinct business models: fees were charged in exchange for services like security and safety. These were the first platforms.
The newspaper industry is one of the most significant and consequential platform business models. They embody some typical platform business model dynamics. They are intermediaries that match readers with advertisers. Advertisers pay to show advertisements to a readership that is attracted to the platform for its journalism. That advertising revenue subsidises journalism activities. The more advertising revenue there is, the more journalism, which attracts more readers, in turn attracting more advertising revenue, and so forth.
Platforms as matchmakers are major parts of our modern economy. Credit card and payments platforms such as American Express, PayPal and Square match merchants, banks and consumers. Shopping malls match retail stores and consumers in a physical location. Operating systems like Microsoft Windows, Google’s Android, and Apple’s mobile iOS and macOS match application developers, manufacturers, and consumers. Video game platforms such as Nintendo’s Switch, the Sony PlayStation and Microsoft’s Xbox match developers with gamers. Event ticketing firms like Ticketmaster and StubHub match event venues with audiences.
Digital businesses have proven to be highly scalable matchmakers. They do this by leveraging key technologies of the digital economy: ubiquitous mobile internet access, stable digital identities which allow for reputation management, and large scale data analytics. These features make platforms the dominant business model of the digital economy. Drivers and passengers are matched by Uber and Lyft. Retailers and consumers are matched by platforms like Alibaba and Amazon Marketplace. Holidaymakers and renters are matched with home owners through Airbnb. Social media users and advertisers are matched by Twitter and TikTok. Recruiters and jobseekers are matched through LinkedIn. Audiences and copyright holders are matched by Netflix and Spotify. There are different cross-subsidies in each of these business models, supporting matches in different ways.
While platforms such as ridesharing might match two distinct sides, other platforms can include many more sides. An operating system, for instance, has at least three distinct sides. Users seek the most user friendly operating system with the most applications for their needs. Application developers want to work on a system that is developer-friendly and has the most users. Hardware builders (from the original equipment manufacturers to consumer level retailers selling pre-built systems) want to build for the operating system with the largest market. Operating system markets can get even more complex. With its Android operating system platform, Google operates in a four-sided market, matching a distinct advertiser side with its users, hardware, and software sides. The dynamics of platforms get complex quickly as more sides are added. Cross-subsidies need to be managed and pricing structures become opaque.
Multi-sided have distinct economic dynamics because they strongly benefit from network effects, tend to feature seemingly peculiar pricing strategies, and often appear to dominate their markets in a monopolistic way.
Each side of the market benefits from having more possible counterparties on the platform. Every additional user of the platform increases the value of the platform to other users. Existing users do not directly pay for the additional value they receive from each new user who joins the platform—but they receive that value nonetheless. Technically we can say that platforms tend to have increasing network externalities.
The existence of these increasing network externalities makes platform pricing strategy peculiar. In a factory market model, pricing is determined by an assessment of both production cost and how much consumers are willing to pay. Generally the goal is to maximise the difference between production cost and sales price (constrained of course by the competition of other factories who are trying to do the same thing). In a platform market model, access is priced differently because any charges for access to the platform have to be traded off against the possibility that those charges might reduce the number of users of the platform. If the access price is too high it will reduce the value and attractiveness of the platform to new users.
While it is users of the platform making trades, it is the platform itself that sets prices. Platforms make strategic decisions about pricing to boost network effects. The result is that platforms tend to charge different prices to different sides—one side of the market cross-subsidises another. In the traditional newspaper industry, for example, advertisers pay large sums where readers (who want content, but would rather avoid advertisements) pay small sums. Under the ‘free’ newspaper business model, the papers themselves were provided at zero-price to readers and distributed at train stations and other public spaces.
Matching is costly and requires work. If the sides of the market could find each other easily to trade then there is no need for a platform. Typical platform services include payments systems, identity services and reputation management. Someone has to pay for these services. As a whole these services need to be profitable or the platform will collapse. Decisions about where to charge fees, and which side of the market should bear the burden of the fees, is a core strategic decision for any platform business.
Platform pricing is further complicated by a ‘chicken-and-the-egg’ problem—why would you join a platform if there aren’t enough users on the platform already to trade with? Few people are keen to drive for a ride-sharing network with no riders, and few riders will bother registering on a platform with no drivers available. There is a clear need for bootstrapping to expand the size of the platform market. Attempts to solve this problem often involve a dynamic strategy: initially offer services at a discount—using early adopters as loss leaders—and apply fees later as the market matures. Eventually, the pricing strategy of platforms, and the multi-sided markets they support, need to be financially sustainable. The process to get there is fraught with danger.
Platforms don’t only compete on price or services. They also have to decide the rules for participation on the platform—that is, platforms provide governance. Who is allowed to join? What standards are participants expected to maintain? What sorts of market exchanges will the platform allow? For example, Uber requires drivers to pass a police background check, and removes drivers and riders from the platform if their reputation rating drops below an acceptable level. These standards are imposed to ensure the platform keeps up its own reputation for high-quality service. But if the strict rules for participation come at a cost of reducing the number of potential participants on the platform.
At least since Adam Smith, economic analysis has been typically focused on one particular market organisational structure in which goods are produced by producers and purchased by consumers. Smith had his pin factory, where producers optimised the production of pins, from which he explained the central ideas of the division of labor and specialisation. The supply and demand diagrams that are taught in introductory economics courses around the world show how the prices of goods and services are arrived at by the interaction between buyer’s willingness to pay a given price and seller’s willingness to produce a given quantity at that price.
But in the 1970s and 1980s the rise of a school of economic thought—transaction cost economics—made it clear that this simple blackboard model didn’t explain many of the most important features of the economy. The transaction cost economists pointed out that there were many more costs in the economy than just the cost of producing goods. It can be expensive to find goods (search costs), to come to agreement about a price (bargaining costs), to verify that those goods were of a certain quality (monitoring costs), or to ensure that agreements were fulfilled (trust and enforcement costs). These transaction costs act as frictions in mutually beneficial trade, preventing us from coordinating.
Seeing the world through the lens of transaction costs—and the various ways that we can mitigate them—helps us to explain the way that economies are structured. It gives us a way to think about why firms, markets, platforms and governments exist. Different transaction costs of searching, of bargaining, of preventing opportunistic behavior in a counterparty, shift how we might organise our economic activities.
This transaction cost economics understanding has been applied to understand the different structures of firms and markets:
Economists like Ronald Coase (who won the economics Nobel prize for his work in 1991) and Oliver Williamson (who won his Nobel prize in 2009) looked inside the ‘black box’ of the corporation. They taught us why companies exist and why they might choose to merge with other companies.
For Armen Alchian and Harold Demsetz, for instance, the company form exists to monitor the production of goods (or services) that need to be produced by teams of people.
For Oliver Hart (whose ideas are at the foundation of what we call the Hart asset, see Chapter 5), the fact that business-to-business contracts cannot fully describe every possible contingency means that some production is better insourced than outsourced.
Then, in 2003, two French economists, Jean-Charles Rochet and Jean Tirole (the 2014 economics Nobel Laureate), published a groundbreaking paper after they observed that many firms don’t look like factories. They introduced us to multi-sided markets. Some firms don’t have clearly identifiable inputs and outputs—not all firms bought from primary producers and sold to customers. Rather, many firms look like they perform a matching function. The primary role of these firms seemed to be reducing the transaction costs of finding partners to make exchanges with. These platform business models support multi-sided markets, where two or more participants who want to make a trade use a ‘platform’ as a common meeting ground to do so.
In these tools of multi-sided markets—as well as the broader insights of transaction cost economics—that we find our new vision of universities. Our vision is neither romantic or industrial. Universities are not factories—universities are platforms. Universities don’t produce widgets—they perform a matching function. To understand how universities work, and how they will change in the future, we need to understand what the sides of the university platform are and how these matches and trades are managed. This is our task in the next chapter.
Part 2 takeaways
There are many visions of why universities exist and how they should operate. We covered three main visions in this chapter.
A romantic vision is deeply entrenched within the culture of academic life. It sees universities as a place of humanistic values, seeking truth and enlightenment outside of business and politics. The romantic vision provides us little direction in understanding the modern university.
The competing industrial vision is more common in government and industry. It sees universities credential or research factories.
As credentialing factories, universities take students and transform them into graduates through years of lectures and examinations, badging and certifying them with degrees. It remains hotly debated whether the average graduate has grown their human capital, or simply signalled some underlying traits they already had. Nevertheless, they are credentialed.
As innovation factories, universities produce fundamental and practical research. These are inputs into a highly productive economy. Research enables universities to build their reputation, expand their market power, and draw in more revenue.
The romantic and industrial visions of the university are in tension; universities are criticised from both directions. They fail to grasp the core function of the university.
The thesis of this essay is that universities are platforms. This platform vision understands universities as matchmakers that reduce the costs of different sides of the market contracting. Universities facilitate matching of teachers with students, industry with graduates, governments with researchers, and so on.
We can use platform economics to understand why universities are structured the way that they are, and ultimately to devise university strategy in response to COVID-19 and technological change.
Part 3 is coming soon.
 de Ridder-Symoens, H., & Rüegg, W. (Eds.). 2003. A History of the University in Europe: Volume 1, Universities in the Middle Ages (Vol. 1). Cambridge University Press. p xix
 See Gabriel Compayre (1893) Peter Abelard and the Origin and Early History of the Universities. (republished 2002, The Minerva Group)
 Woods, T 2012. How The Catholic Church Built Western Civilization. Regnery Publishing.
 Jacques Verger, A History of the University in Europe, v1, p40
 Wellmon, C., 2015. Organizing enlightenment: Information overload and the invention of the modern research university. Johns Hopkins University Press, p4.
 Collini, S., 2012. What are universities for?. Penguin UK.
 Otterspeer, W. 2008, The Bastion of Liberty: Leiden University Today and Yesterday, p 9-10.
 Newman, JH. 1873. The Idea of a University Defined and Illustrated, p 177-178
 Oakeshott, M. 1950, The Idea of a University, The Listener.
 Oakeshott, M. 1950, The Idea of a University, The Listener.
 Bloom, A. 2008, The Closing of the American Mind: How Higher Education Has Failed Democracy and Impoverished the Souls of Today's Students. Simon & Schuster.
 ibid. p 254.
 Davis, JA and Farrell, MA 2016. The Market Oriented University: Transforming Higher Education, Edward Elgar.
 For general lifetime earnings, for instance, see Carnevale, A. P., Cheah, B., & Rose, S. J. (2011). The College pay off. https://vtechworks.lib.vt.edu/bitstream/handle/10919/83051/TheCollegePayOff.pdf?sequence=1; Carnevale et al 2015 https://cew.georgetown.edu/wp-content/uploads/Exec-Summary-web-B.pdf.
On a specific profession (law) see Simkovic, M., & McIntyre, F. (2014). The economic value of a law degree. The Journal of Legal Studies, 43(2), 249-289.
Also see (discussed below): Caplan, B. (2018) The Case Against Education. Princeton University Press: Princeton.
 Spence won the Nobel Prize in 2001.
 Caplan, B. (2018) The Case Against Education. Princeton University Press: Princeton.
 Glaeser, Edward L., Giacomo A. M. Ponzetto, and Andrei Shleifer. 2007. "Why Does Democracy Need Education?" Journal of Economic Growth 12 (2): 77-99.
 For an overview of non-market private benefits see Macmahon (2009: ch 4).
 Macmahon, WM (2009) Higher Learning, Greater Good: The Private and Social Benefits of Higher Education, Johns Hopkins University Press, Chapter 5.
 The basic case was made by Richard Nelson: ‘were the field of basic research left exclusively to private firms operating independently of each other and selling in competitive markets, profit incentives would not draw so large a quantity of resources to basic research as is socially desirable.’
 Oakeshott, M. 1989. The Voice of Liberal Learning, Yale University Press.
 Michael Shattock
 The Fall of Faculty
 Kennedy D 1999, Academic Duty, Harvard University Press 1999
 http://www.oecd.org/education/imhe/42348745.pdf p28.
 A good overview of the broader implications of platform economics is Tirole, J. (2019) Economics for the Common Good. Princeton University Press: Princeton.
 See Rochet J and Tirole J 2003, Platform competition in two sided markets, Journal of the European Economic Association, June 2003, Vol.1(4), pp.990-1029. Also see Berg, C, Davidson, S., Potts J. (2018) "Outsourcing vertical integration: introducing the V-form network," Medium.
 Earl P and Potts J 2000 Latent Demand and the Browsing Shopper, Managerial and Decision Economics 21, p. 111-122.
 Takakuwa K 2015. A Survey on the competition between the multi-sided platforms: Mobile platform competition between Google and Apple, Annual conference of The Infosocionomics Society, At Tokyo, Japan, Volume: Vol. 10, No. 1.
 See Coase, R. 1937. ‘The Nature of the Firm’, Economica, vol. 4(16), p. 386-405; Coase, R. 1960 ‘The Problem of Social Cost’, Journal of Law and Economics, 3(1), p 1-44; Williamson, O. 1975 Markets and Hierarchies: Analysis and Antitrust Implications. The Free Press; Williamson, O. 1985 The Economic Institutions of Capitalism: Firms, Markets, Relational Contracting. The Free Press.
 See Alchian, A., Demsetz, H. 1972. ‘Production, information costs, and economic organization’ American Economic Review, 62(5): 777–795.
 See Rochet and Tirole 2003, ibid.