University under siege?: Trust and accountability in the contemporary academy

What is at stake is the erosion of the Humboldtian notion of the university, a conception that dates back to the nineteenth century. We call it ‘Humboldtian’ because the German humanist Wilhelm von Humboldt (1767–1835) wrote about his ideal of a university in a way that anticipates the developments in the nineteenth and twentieth century and that captures what we want to discuss (von Humboldt 1956, 1990; see also Weischedel 1960; Tschong 1991). In its ideal-typical form, the Humboldtian notion describes the university as an autonomous body of self-governing professionals, accountable to and monitored by itself. Lecturers and professors operate as a freischwebende Intelligenz (Mannheim 1997: 153–64; see also Hoeges 1994), supposedly autonomous from class or other interests, and speaking and writing with an authoritative, critical voice. These academics have a stable sense of professional identity that is embedded in a specific discipline and a tradition. There is a clear boundary between the academy and the outside world, with the latter subjected to market forces and the former immunised against these mechanisms (compare with Bernstein 1996; Beck 1999). This differentiation between the university and its external environment ties in with a juxtaposition between the sacred and the profane. The intellectual activities of the academics supposedly have value in themselves as they are directed towards the central ideals of society, such as Bildung or self-edification. These academic endeavours might have practical spin-offs for society, but that is precisely what they are, unintended effects. There is no expectation for the Humboldtian Professor to create knowledge that has immediate and visible practical value.

Several contrasts with organisational ‘best practice’, especially as recently rewritten by economic theory and business research, are immediately apparent. As organisations, universities continue to break many of the rules of efficient business organisation put forward by consultants (e.g. Naisbitt and Aburdene 1985; Peters and Waterman 1995) and economists (e.g. Porter 1985; Teece 1992). Rather than choosing one area of specialisation, they identify broad engagement in and mixing of subject disciplines as a principal source of strength. Defying confinement to one ‘core competence’, they generate two very different types of product, teaching and research. These, although in some ways synergistic, make competing demands on universities’ human and physical resources, and force them to deploy two very different types of academic producer to serve very different categories of purchaser and sponsor. The products are offered across a broad range of disciplines, even though most universities have areas of strength and weakness, increasingly well highlighted by research and teaching assessment exercises.

The client base for teaching – school-leavers, mature students and executives on refresher courses – shows little overlap with that for research, whose ‘purchasers’ are generally institutional rather than individual. The ‘purchasers’ of research divide into two contrasting constituencies: the state, which traditionally sponsors ‘pure’ research for the perceived public good, and private businesses, more concerned with ‘applied’ research that can promise commercial payback. Similarly, the teaching product is offered at three distinct levels – undergraduate, taught postgraduate, doctoral – each of which targets contrasting clienteles, and the last of which often merges with the research product.

Universities also perpetuate – and claim to draw strength from – ‘cultural’ differences of a breadth, which in private business would be written off as the product of misguided merger or mismanaged expansion. For example, many long-established social science departments have, in recent years, set up ‘research centres’ whose members have different aims and working methods. Staff of these centres tends to focus on research, with minimal or purely voluntary administrative duties. They generate income through specific, time-limited research grants rather than continuous remittances for teaching duties. They tend, in consequence, to be employed on fixed-term contracts rather than tenured. Their output is assessed on the delivery of specific research results, rather than of teaching hours. This output takes a wider range of forms than the traditional journal article, since it may extend to consultancy reports, briefings, audiovisual packages, and various other forms of presentation that are not necessarily refereed or placed on public view. Ineligible for tenure, contract researchers’ status tends to be lower, and their career progression less certain, than those of university lecturing staff – even though universities’ ranking and funding relies increasingly on the quality of their research.

Of course, the Humboldtian notion of the university never existed in its pure form. It is an ideal-type in the Weberian sense of the word (Weber 1949: 87–112; 1964: 109–12), and as such captures the essential features of a model. It was instantiated differently from one country to another (Clark 1995). Closer to home, for instance, John Henry Newman challenged Humboldt's emphasis on learning and scholarship. Although motivated in large part by a specific lifelong mission to revive and update Roman Catholic religious instruction, Newman's vision of the university as primarily a teaching institution had a profound impact on the history of British academic institutions (Newman 1873). There has, sometimes, been a discrepancy between what academics aspire to and what is actually taking place, and some commentators indeed talk about the ‘Mythos Humboldt’, a myth of the Humboldtian vision that, according to them, has permeated academia for a long time (Ash 1997b; Lundgreen 1997). Many excellent universities have been able to remain in existence precisely because of the existence of professional schools and applied research, so the Humboldtian model relies to some extent on an ‘applied’ one. Notwithstanding these reservations, this ideal type is useful for sketching historical changes and demonstrating the extent to which we have indeed been moving away from it in the last couple of decades.

Equally important is that a lot of academics, especially in the humanities, even now adhere, refer or aspire to the Humboldtian notion. Establishment or expansion of research function was one of the adaptations prioritised when UK polytechnics and higher-education colleges were converted to universities during the 1990s. The strength of resistance to the concept of a teaching-only university, even though narrowly economic calculations might count in its favour, is matched by universities’ continued insistence that teaching is a useful complement to research, even if it detracts from research time. The training of future academics reflects this view of what university life is about: working on a specific subject within a particular tradition, with a long-term perspective, solitary, independent (at least from pressures outside the university) and (with the exception of the viva) unaccountable. A similar academic genre is still somehow assumed in many social studies, including the sociology of science, the sociology of knowledge and educational science. Most importantly, many academics conceive of this concept of the university as a utopian vision, as something worth striving for, and the universities that resemble most this Humboldtian notion are ranked amongst the most prestigious.

This is not to say that the Humboldtian vision has remained unchallenged throughout the twentieth century. In the American context, some commentators have argued that, whereas liberal academics used to address a wider audience, there is, in practice, little place for the public intellectual in the modern university system (Jacoby 1986). But closer to home, the main criticism of this concept of the university was that it relied upon and promoted an elitist view of society and of education. This accusation became increasingly prominent in the course of the twentieth century, and took two forms. The first allegation was that the university, as such, contributed to, reinforced and legitimised the cultural forms that were associated with a particular class or segment of society. There was progressively more talk about the essential arbitrariness of cultural distinctions and about questioning the opposition between high and low culture. The second issue, more pedestrian but also more tangible, was that, in practice, higher education was only available to a small fraction of the population. This was not just an ethical quandary, but it also became a practical problem for an economy that relied increasingly on the input of university-educated professionals. This dilemma would eventually lead to state funding and intervention.

Whereas previous sponsors of research – aristocrats, the church and early entrepreneurs – tended to bring selectivity to the choice of topic and bias to the conclusions, high hopes were held out for the newly emerging nation-state governments – in Europe and elsewhere – as systematic and disinterested sponsors of academic enquiry. Humboldt, who served for a time as education minister in the emergent German government, envisaged an institution that received state funding but conducted its affairs with no political influence. He believed it possible for state to exercise the same degree of disinterest as the historically exceptional aristocratic sponsors identified as central to Europe's post-enlightenment knowledge creating prowess by David (1998, 2001). But the state would work on a much larger scale, allowing for the emergence of scholarly communities pursuing more comprehensive and systematic research initiatives. The state had self-interest in sponsoring autonomous universities, since academic freedom was a precondition for research that could generate productivity-enhancing knowledge and technology, and teaching that could equip (a section of) the population to use it.

In the Humboldtian prospectus, higher education would generate economic advance, and the social improvement (and enhanced state power) that went with it, while engaged in apparently unfocused knowledge creation. Public authorities could thus expect a payback on their university investment, even if this were a by-product rather than intentional product of the pursuit of discovery and learning for its own sake. These benefits had been only haphazardly exploited when scholars had to seek private patronage; and the narrow range of subjects that attracted such patronage, plus the isolation in which its recipients tended to work, inhibited the formation of the scholarly community so highly prized by Newman (1873). Coordination and funding by the state seemed the most effective way to sustain a community with sufficient size, diversity and freedom from political and commercial pressure to create the learning environment favoured by Newman, as well as Humboldt's knowledge-creation environment.

This partly explains why after World War II, in most West-European countries, the state became a significant financial contributor to the growth of institutions of higher education (Seville and Tooley 1997: 22 ff.). In return for increased state-funding universities opened their doors to various sections of the population that had effectively been barred from entry hitherto. The broader intake was expected to bring about more of a meritocracy (see also Nemetz 1961), but this quickly proved a misplaced hope. Rather than focusing on how the growth of higher education might bring about a more equal or just society, social scientists started to study the mechanisms by which the educational system has become implicated in the reproduction of social inequality (e.g. Bourdieu 1996). So, ironically, attempts to deal with the second allegation, the issue of intake and participation, fuelled further interest in the first criticism, in the content of the cultural forms that are being produced within universities.

To widen involvement in their activities while still assigning it on merit, universities became dependent on the state. This dependency was not particularly visible until the late 1970s, mainly because of the post-war consensus, which assumed compatibility between the needs of academic institutions and the requirements of society. But from the early 1970s, neo-liberal academics and advisers questioned the efficiency and equity of state intervention in the economy (e.g. Joseph and Sumption 1979; Brittan 1988). They extended their criticism to state support for education, and especially higher education. In a situation where even ‘natural monopoly’ industries have been privatised and deregulated as far as possible, schools and universities have been identified by many economic liberals as the next producer-dominated, subsidy-dependent, unaccountable and (on selected measures) under-performing state-owned monopoly on which to focus its attack (e.g. Hague 1991; Tooley 1995).

Part of the current neo-liberal argument is a rejection of the assumption that contemporary higher education works for the greater benefit of all. Academic institutions were now increasingly seen differently, no longer unequivocally benevolent, not necessarily deserving the levels of trust that we had bestowed in them. Suddenly, a whole range of new questions emerged. Do all academics work? Do they work effectively? Is what they do worthwhile? Or might they be using their privileged status to subvert the very same society that is feeding them? Do ‘we taxpayers’ have to keep supporting them, and if so, under which conditions? Note the last question. The dependency of the system of higher education on state funding made universities vulnerable to the assertion that they needed to be scrutinised so that only the deserving would receive financial support. If the academy is funded, then it ought to be accountable. This simple argument captures four pivotal clusters of changes, which would eventually undercut the Humboldtian vision: the pitfalls of state funding as governments democratised, the emergence of an audit culture; the erosion of trust; and market-driven modifications in the nature of funding.

As recipients of public financial support, universities came under pressure to adopt the same forms of public accountability to which governments themselves were increasingly subjected. In particular, universities’ traditional elitism was increasingly called into question as their financing drew more heavily on public budgets to which most people were contributing as taxpayers. The inaccessibility of university courses to most people had to be justified on the basis that only a few were able (and willing) to absorb and produce knowledge of the quantity and quality required – and that their doing so generated sufficient external (social) benefits that the activity deserved to be continued at social expense. The inaccessibility of university research output to most ‘lay’ listeners and readers had to be justified on the basis that knowledge grew more abstruse as it became more refined, so that only other specialists could fully understand it and endorse its validity in the original. The rest of the community had to settle for being kept informed by ‘popularised’ restatements, or just by consuming the benefits of academic knowledge breakthroughs without fully comprehending how they came about.

Initially, universities tried – and largely succeeded – to deliver the necessary accountability through self-regulation, following the practice of other ‘professions’. These maintained their elitism (and its economic advantages to members) by exclusionary recruitment procedures, justified on the basis that society gained if membership were restricted to those of proven competence (Perkin 1990). Self-regulation was argued to be in the public interest: it spared governments the expense of running their own regulatory institutions, insured against political interference in professional activity, and guaranteed a better quality of regulation, because only other members of the profession held the necessary knowledge and experience to judge the work of their peers. The autonomy and peer assessment advocated by Humboldt dovetailed with the growth of other self-regulating professions. The ‘professionalisation’ of academia entered mutual reinforcement with the ‘academisation’ of the professions, since most of these (notably law, accountancy, engineering, school-teaching, physiological and psychological medicine, economics, religious ministry and public service) were seeing their basic training, and the preparation for it, brought inside or more fully inside the university system.

But professions in general, and academia in particular, have found their autonomous and self-regulatory traditions progressively more difficult to defend (Matthews 1991). This principally reflects dissatisfaction with the restriction of membership on the basis of demonstrated quality or personal attainment, and the insider (peer) assessment of that quality. Once public support for peer-assessment is lost, professions’ membership selection processes begin to resemble an exclusionary practice. They can appear designed to keep out candidates from unfavourably regarded social backgrounds, and to inflate members’ fees through the artificial restriction on the supply of their services that results. To the extent that this is true – resulting in high pay to professional members whose non-meritorious selection reflects in their performance – it becomes reinforced by periodic displays of professional incompetence or misconduct. The British army and civil service after the 1853–6 Crimean War, European and American banks after the 1982 Latin American debt crisis and American accountancy firms after early-2000s corporate failures (exemplified by Enron) are among many examples of professions which lost public support after widely publicised procedural failure. In all cases, self-regulation of entry to and conduct of the profession was shown to be inadequate, and moves to tighter external regulation followed.

The academic world, and especially academic social science, has found its professional privileges particularly difficult to defend. Much of the explanation can be traced to its success, in promoting social change that alters people's attitudes in a more critical direction. Academia's entry restrictions have been challenged by the expectation, communicated by or on behalf of an increasingly well educated public, that all who perform competently in their final years at school should have the chance to go to university. The perception encouraged by academics, that even knowledge creation ‘for its own sake’ brings economic and social benefits, has given politicians an additional reason for wanting to see university access expanded, whether or not more people want to take it up. European universities’ exclusivity has been progressively broken down over the past five centuries – by the switch from Latin to vernacular languages of instruction and dissemination, the multiplication of affordable vernacular texts through printing, deliberate efforts to popularise science-based learning to counteract the previously exclusive scripture-based learning (Boorstin 1983: 480–538), and the promotion of self-teaching and self-help as a means to raise industry's skills ahead of its tax bills during early industrialisation (Green 1993).

At the same time, academia's internal (peer-reviewed) validation of knowledge has been challenged, by the same combination of popular and political forces (e.g. Gilland 2002). An increasingly educated and informed public is ever less content to leave academics as the sole judges of what it valid or reliable knowledge – and ever more suspicious that any inability to understand academic discourse is down to its unnecessary abstruseness, rather than their obtuseness. Politicians, similarly, are inclined to feel both their private pride and their public purpose defeated if they cannot reach their own opinion on the validity of universities’ teaching and research output. That opinion is, because of the pressures on elected representatives, increasingly prone to equate validity with value – ways in which academic output can feed into better public policy. Even if the process of natural-scientific research is often too complex for non-specialists to understand, its products can readily be evaluated by their usefulness, for instance, in yielding new consumer goods, building cheaper or better public infrastructures, making ‘greener’ technologies economically viable. When methods are in dispute, it is still possible (and, for time-pressed decision-makers, always tempting) to judge new social-scientific results and ideas by their ability to solve problems, improve the effectiveness of choices, or expand available resources so that fewer tough choices need to be made.

In the course of the 1980s, we have witnessed the appearance and rapid spread of what became known as the ‘audit society’ or ‘audit explosion’ (Power 1994, 1997; see also Munro and Mouritsen 1996), which aims at greater transparency and accountability for state-funded organisations such as hospitals, schools and universities. Some commentators have described this phenomenon in the context of a ‘new managerialism’ or ‘neo-Taylorism’ (Pollitt 1993; Delanty 2001: 106–8; Shore and Wright 2002). In the case of the academy, the irony is that science has tended to pride itself on its transparency, not only in that it discloses and makes public the external world, but also that it does this by using procedures that are available and clear to all scientists involved. It is, however, an altogether different quality of being pervious that has been imposed on universities. It is not about shedding light on the external environment or producing knowledge that is accountable to peers, but it is about making transparent levels of productivity, work content and quality to government agencies.

In Britain, neo-liberal intellectuals and Tory politicians initially introduced the ‘declinist argument’ to justify some kind of monitoring of higher education: the fact that universities were not geared towards industry, in the way in which they were in other advanced countries, was regarded to be one of the main causes for the British industrial decline. The Institute of Economic Affairs, a key source of ‘new right’ ideas, championed the ‘denationalisation’ of school education (through its ‘Black Paper’ series) and of higher education in (Seldon and Harris 1968). With historical hindsight, this could be seen as a return to non-state solutions which inappropriate public intervention had overridden (Wiener 1981; Barnett 1986; Green 1993).

Under New Labour, the belief in the necessity of an intricate relationship between education, research and industry has only intensified, as can be inferred from the numerous references to the knowledge economy. New Labour referred to this sociological concept to indicate the extent to which the economy today is dependent not only on a mass of highly educated professionals, but also on the expansion of industry-related research (Jarvis 2001a,b). This idea is, for instance, central to Anthony Giddens and Will Hutton's attempt to rethink progressive politics in the context of globalisation (Giddens and Hutton 2000; see also Giddens 1998, 2000, 2001). It crosses the whole political spectrum. The Dearing Report very much emphasises the need to direct research and teaching in higher education in the UK towards a flexible knowledge-based economy (Dearing 1997), and other countries are developing a similar perspective (Etzkowitz et al.1998; Geuna 1999; Jongbloed et al.1999; Neave 2000).

Despite this initial impetus of the declinist argument, however, the ‘audit explosion’ in universities has focused on levels of productivity and efficiency, on meeting stringent academic standards, rather than on being economically useful (Shore and Wright 1997, 2002; Kogan and Henkel 1998). It is mainly about measuring how good an academic you are, whether you publish in the right journals or with the right publishers, whether you have been able to make a mark in your field (see also Nolan et al.1996). Blake et al. (1998: 67–88) calls this the ‘normalisation’ of research and this Foucauldian term is aptly chosen in that quantifiable benchmarks are set, and output is graded and ranked in relationship to those criteria. It is here that a curious symbiosis has taken place between the arrival of a new regime and the institutional and professional culture that was already in place, a compromise between the new discourse of accountability and the remnants of the old ethos. But the question for the politicians and policy makers was who would monitor the academics if the standards had to be academic. The answer was obvious: academics would do it themselves (HEFCE 1992). So the system of auditing became heavily reliant on the ongoing cooperation of senior university teachers from various universities and who devote a large amount of their time on judging and ranking the output of colleagues in their profession.

For the institutions under scrutiny, a very different question emerged: how to adjust itself to cope with the new requirements. The answer lay in organisational restructuring: management structures were brought in at various levels in order to allow the monitoring of its personnel in relationship to the standards set by the external auditing committees. Here we observe a particular instance where systems of inspection replicate themselves because one level of scrutiny leads to another one. How does one know, as Head of Department, how the Department will fare in the research assessment exercise? By checking on its members. How does a Dean of a Faculty assess its viability? By monitoring the performances of the various Departments. Etcetera. In this case, the system of auditing tends to create a culture of expectations and pressures that makes for the reproduction of auditing at various levels. This raises the question about the precise relationship between auditing and trust.

External regulation and performance ‘audits’ are usually introduced to rebuild public trust in a profession, when this has been eroded by members’ misconduct or monopoly abuse of their entry restriction. Questioning of academics’ elite authority was at the basis of their exposure to the audit culture (O'Neill 2002: 43–59; Shore and Wright 2002). But the measurement and ‘accountability’ of performance achieved by auditing may not be of the type that creates trust in professionals, and may even serve to erode it, by at least three mechanisms (O'Neill 2002). They do a less good job, because assessments and form-filling reduce the time they spend on it, and target-chasing narrows the tasks that occupy that time. They allocate the remaining time less efficiently and equitably, because accountability promotes a form of attention-seeking, analogous to the ‘rent-seeking’ observed by economists (see, for example, Buchanan et al.1980; Buchanan and Musgrave 1999) in which those who campaign or complain loudest get more and better treatment than those who most deserve it. And they lose the esteem that superior knowledge previously gave them. In the doctor–patient relationship, for example, ‘mutual respect precludes rather than requires across-the-board openness’ (O'Neill 2002: 69, italics in original).

The upshot of recent attacks on higher education's research and teaching assessment exercises, especially in the UK, is that academia ranks among the professions where accountability designed to raise the quality of ‘output’ may actually cause it to decline. The performance can be impaired because of the disruption of the process of knowledge production and validation in pursuit of more (and possibly more valuable) products (Nisbet 1971; Lee and Harley 1997, 1998; Strathern 1997, 2000). Suspicion, rivalry and non-cooperation among specialists charged with monitoring one another's work is often central to observed breakdowns of the peer review process (see e.g. McCutcheon 1991). The new pressure on academics to out-publish their colleagues, to be first to put their name to new discoveries, to compete for ever scarcer jobs and research funds, and to use their time originating new results rather than replicating and corroborating those of others, seems to erode collegial trust and the peer-assessment it used to cultivate (Parker and Jary 1995). And decline of trust among colleagues apparently has more damaging consequences in academia than in most other professions, for the reasons set out by Humboldt and Newman in their vision of the community of scholars as the core of the university idea.

More specifically, the academic experience provides an illustration of ways in which external regulation can destroy public trust in professions that rely on high levels of internal trust – that is, trust that resides in the social capital sustained by inter-collegial sharing of knowledge, and in collective effort to create and apply this knowledge. The recent economic ascendancy of low-trust, highly litigious societies (notably the US) over those regarded as high-trust and minimally litigious (notably Japan) has encouraged the view that the substitution of audit-based for trust-based systems of obligation may be socially progressive. Communities which lessen their dependence on trust for successful commercial and social exchange, instead developing impersonal ways of enforcing contracts, may be better equipped to extend their exchanges to new groups of people, to the extent that systems of rules and sanctions for rule-breaking are easier to extend across community borders than trust, which requires a longer history of successful interaction (Axelrod 1984; Ostrom 2000). This greater capacity to widen exchange is economically beneficial to the extent that trade becomes more profitable the more diverse the trading partners (El-Agraa 1996); and more socially beneficial to the extent that ‘weak ties’ between people who hardly know each other tend to produce more lucrative interaction than ‘strong ties’ between those who know and trust each other well (Granovetter 1973, 1983).

Low-trust societies may be able to make up in private capital, accumulated through more profitable exchange with widely dispersed strangers, any deficiency they suffer in social capital relative to tightly-knit communities, whose exclusionary character makes it harder for outsiders to engage in exchange (Clay 1997; Macy and Skvoretz 1998; Lovett et al.1999). If this argument is correct, then the destruction of trust among colleagues in a profession, when external accountability is imposed, need not necessarily undermine its performance. The group may be able to continue doing its job with well-enforced rules substituting for previous professional ethics. Provided performance remains unimpaired, trust with clients is not eroded. The audit culture is not inherently destructive. It only becomes so if cumbersome or poorly designed regulation destroys trust between professionals and their clients.

As well as university activity in general being especially prone to loss of public trust when external audit erodes its internal trust, some specific branches of university – mainly in the humanities and social scientists – have encountered a second source of eroded public trust, traceable to changing concepts of knowledge validation. The loss of academia's professional mystique, ‘donnish dominion’ in the words of Halsey (1992), is especially damaging to social sciences, where the status of ‘knowledge’ relies more on the credentials and authority of those who proclaim it, and peer-review it, than in any practical success in solving problems or passing tests. It is worth making a distinction between the corrosion of foundations of knowledge in general on the one hand, and the specific status of the social sciences and humanities on the other.

Let us start with the general case. There has, over the last couple of decades, been a dissipation of consensus regarding standards of knowledge, compared with (for example) the first half of the twentieth century. There is now less clarity as to what distinguishes superior and inferior knowledge, or valid and non-valid assertions, and there is more scepticism than ever about claims regarding objectivity or truth. We are not simply referring here to the ‘anti-foundationalism’ that underscores most of the writings influenced by the postmodernist bandwagon (Lyotard 1979), but also to developments within Anglo-Saxon philosophy. The writings of W.V. Quine and Thomas Kuhn initiated a sceptical position, which found a most persuasive formulation in neo-pragmatism (Rorty 1980, 1991, 1999). Bauman puts it very nicely when he writes that, in a post-modern context, intellectuals can longer be legislators and become interpreters instead, simply helping to make sense of various cultural traditions (Bauman 1987, 1992). In a similar vein, Foucault presented himself as a ‘new intellectual’ as opposed to the ‘traditional intellectual’ (Baert 1998). Whereas the latter preach from above and try to impose a worldview onto people, the former simply provide tools so as to allow people to take distance from their taken-for-granted world and see things differently.

These self-doubts eventually find their way to a wider public and must have affected the way in which intellectuals are perceived. In this light, it is not surprising that universities have received competition from a whole range of knowledge-producing institutions such as management consultancies and think thanks. More and more governments turn to these organisations for expert advice. Whereas universities traditionally rely on codified, theoretical, discipline-based knowledge, these competing organisations employ hands-on, practical, multi-disciplinary knowledge (see also Gibbons et al.1994; Gibbons 1999; Scott 2000b). So we are witnessing ‘the end of donnish dominion’ not simply as a decline in status and work conditions of academics (see Halsey 1992; Farnham 1999), but also in a more profound, structural way: though Universities have still a monopoly of validating knowledge and accrediting knowledge-producing individuals, but no longer on executing knowledge.

Trust in the natural scientists is endangered by researchers’ inability to reach clear conclusions (e.g. over the safety of GM crops and mobile phones), or sudden changes of conclusion (e.g. over cholesterol and cold nuclear fusion); but trust is on balance maintained by the many problems and opportunities that science seems successfully to resolve or open up. Trust in social sciences confronts a more pervasive difficulty, arising from a disjuncture between the breadth of their claims and the depth of their foundations. The popular perception of social scientists’ contemporary role could well be characterised as one of power without responsibility.

On their ‘applied’ side, the social can rival the natural sciences in terms of impact on lives and livelihoods. But they are not so readily subjected to the assessment of validity by usefulness. Communities and, sometimes, whole generations and nations can have their lives changed (sometimes for the better) by such social-science innovations as high-rise housing, comprehensive schooling, monetarist inflation-control and post-traumatic stress counselling. Yet if such measures fail to deliver on the original promise, they do not suffer the immediate rejection (as erroneous) of underlying calculations, as would designs for a spaceship that explodes or washing powder that eats clothes. Disgruntled ‘users’ (subjects) of misfiring social science application are instead told that failure was due to background factors not remaining stable; or to architects’ and policymakers’ well-conceived actions having ‘unintended consequences’ for which social scientists must now be trusted to produce new follow-up solutions; or (especially in economics) to actors not behaving rationally, as the policy models assumed they would, derailing a project which might have succeeded if people could stick to the choices that maximising models suggest they ought to make (see also Eichner 1983; Backhouse 1994).

Social sciences’ theoretical foundations, although in principle as far removed from ordinary concern as those of theoretical natural science, also tend over time to interact with the social experience they are intended to depict (Richards 1983: 93–196). People's views and actions become shaped by, as well as shaping, the views and prescriptions that social researchers form about them, setting up a feedback between ‘observer’ and ‘subject’ not encountered in classical natural science (and encountered only at the subatomic level in post-Newtonian natural science). People's ability to comprehend the theories about them, and then act to reinforce or overthrow those theories, increases as improved communication brings social theories to wider attention, and as improving education equips them to assimilate and analyse what is said about them. Social scientists’ views on social reality become a part of that reality, affecting the way that people regard themselves and the individuals and institutions around them. Some commentators rightly observe that this phenomenon is becoming more prominent in contemporary society (Giddens 1990, 1991).

Auditing is recognised by many academics as the price they have to pay for remaining ‘elitist’ – in the traditional (non-pejorative) sense of offering advanced education to only a minority of the population, and in the (more equivocal) newer sense of pursuing specialist knowledge to levels of complexity that surpass the understanding even of an intelligent lay audience. Before the widening of access criteria designed to open access to a majority, popular criticism focused on the university's perceived exclusivity: people found themselves becoming more dependent on its products (needing a university degree for the work they sought, or needing others with a degree for those they sought for work), better informed about its activities, and more aware of the costs of supporting it (which was tending to rise over time, as evidenced by the growth in higher education budgets, before the rapid increase in student numbers overtook it). After the recent expansion of the university sector, and consequent widening of access, new criticisms have been directed at academics’ continued desire for ‘producer led’ research and teaching in areas they perceive as most socially relevant and/or regard as most personally satisfying, rather than accepting a ‘consumer led’ switch to the research and teaching regarded as most socially and personally ‘useful’ by the students and research sponsors who more directly have to pay for it.

To many academics, the new auditing regime now feels like a fate worse than ‘privatisation’. Their work is exposed to market forces, to the extent that research activity must secure its funding through a new emphasis on ‘wealth creation’, and students must pay their own tuition fees, leading them to demand courses geared to raising their post-graduation marketability to employers (and biasing them against staying on for a poorly-paid academic career). At the same time, a large proportion of their financing continues to flow from the government, with all the exposure to politically shaped agendas and outcome-preferences that this can now entail. Auditing and external regulation appear more as complements than as alternatives to market forces.

Many academics were persuaded to accept the new, assessment-based accountability regime in the belief that a market-driven system would be even more disadvantageous. Instead of a market, the UK government has – with academics’ reluctant consent – subjected higher education to a ‘quasi-market.’ It has applied similar procedures to other state-funded organisations such as the health system and broadcasting (Le Grand and Bartlett 1993; Bartlett et al.1998). Consumer choice and producer competition are introduced into a state-financed system, in order to boost its ‘allocative efficiency’ (by putting resources where consumers most want them) and its ‘productive efficiency’ (by minimising the resources needed to deliver satisfactory results in these areas).

The quasi-market trades two main products – degree-course places (demanded by students), and research results (demanded by businesses, by government departments for their own policy formation, and by government in general for the perceived benefit of society in general). The intention is that these products are still assigned to the most deserving customers, rather than (as in a normal market) to the consumers most able to pay. But customers’ willingness to pay is brought into the equation, the aim being to assign (on merit) their own budgets, so they can show exactly which products they want to buy from which producers. These producers, in turn, are expected to compete so as to meet customers’ demands in the most efficient fashion. The new auditing systems’ main function is to let customers ensure that competition is on the basis of quality and cost-effectiveness, rather than simply on price. New funding arrangements’ main function is reorganise producers into the structure most receptive to customers’ needs.

On the teaching ‘demand side’, those academically qualified to become students are enabled (by a generous loan system, and occasional grants) to ‘buy’ their degree courses from a university of their choice. On the research ‘demand side’, Research Councils have been developed as ‘expert customers’ who will assign state funding to the individuals and institutions judged most likely to deliver high-relevance and high-quality results. On the ‘supply side’, universities and their staff are thus required to compete for the custom of these fund-holding students and research sponsors. The quasi-market for allocation of public funding combines easily with private markets’ parallel allocation of private funding – self-financing students choosing which university to buy their degree-course place from, and private-sector research sponsors choosing whom to commission for their research. But the quasi-market is intended to promote a distinct form of what economists call ‘consumer sovereignty’ (e.g. Lancaster 1998), with places assigned to students and research to sponsors on the basis of academic merit rather than ability-to-pay; and a distinct form of competition among producers, once again designed to maximise the quality of research for a given budget rather than drive down its cost by rewarding those who promise the same results for less.

Against past attempts to present systems of legal regulation and accountability a as alternatives to the market system, especially by economists (e.g. Lane 1985; Zwass 1987; Williamson 1992), the dependence of markets on accounting and other social regulations has now been well documented, especially by historians and sociologists (e.g. Polanyi 1991; Callon 1997; Shipman 2002). Although quasi-markets underpinned by quality audits were intended as an alternative to commercial market pressures, their intention is to generate the same cost reductions and productivity gains that real markets are conventionally credited with producing (Le Grand and Bartlett 1993). Among the signs that higher education's acceptance of the audit culture has not insulated it from market pressures, there have been major changes to the structure of university funding. Public financial support per student has substantially fallen, partly because of the growing parsimonious nature of the state, partly because more educational and research organisations are eligible for funding. Still more significant are the changes taking place in how the financing of universities is organised and how the money is distributed. Increasingly, we see the introduction of quasi-market forms of distribution of funding, not just because of the regular research assessment exercises, but also because of the increasing possibilities at research councils where huge sums money are available for successful bidders.

As in other professions, quasi-market pressures have been no less effective than real market pressures in deflecting academic work towards the fulfilment of set targets, undersupplying outputs that these undervalue and eliminating outputs whose value they ignore. (Lee and Harley 1997, 1998). Students are ‘taught to the test’, and discouraged from acquiring any knowledge that is surplus to exam requirement. Research proposals are narrowly specified around themes known to interest sponsors, and results shaped (in terms of presentation if not content) around the themes and conclusions those sponsors are known to prefer. Research assessment exercises find the scale of output easier to quantify and compare than the value of output, despite the well-known risks of premature and duplicate publication (see e.g. Griffith 1995). Courses are designed to maximise student numbers, entailing (in the view of some internal critics) selection-out not only of topics that prospective students views as irrelevant to their future work, but also those seen as too difficult to ensure the grades that will secure that work for them. Students’ need to think beyond their immediate coursework – to the marketability of the qualification they will gain and hence their ability to repay the credit on which they gain the qualification – reflects the dependence of their purchases in the academic quasi-market on simultaneous or subsequent engagement in real markets – for loans before, part-time work during, and full-time work after, the studies they undertake.

The academy's exposure to market pressures, through the malleable if not penetrable shell of peer-pressured audit, can be seen to have at least three important consequences. Firstly, it leads to a double commodification of knowledge. Not only are research projects bound to be more successful with grant bodies if they somehow promise cash-value at the end of the road, these projects then become a commodity within the academic unity. They allow the academics involved to buy themselves out of teaching, to employ research assistants, and so on. Secondly, senior academics are forced to adopt managerial models to run big projects, to supervise the people who are working for them and to control the finances involved. Success in academia has become progressively more dependent on qualities that were traditionally associated with business: administering projects, managing people, motivating teams. Thirdly, there are clear ramifications at the level of professional identity. Whereas in the Humboldtian notion academics had a relatively stable notion of selfhood, centred round a vocation towards a discipline and tradition, working with research bodies is all about constantly redirecting research to the ever-changing demands of society. The notion of who you are and what kind of research you do is continually restructured around these shifting needs.

We have argued that the visions associated with Humboldt and Newman led universities into professionalisation and state subsidisation, which over the past century brought their elitism under fire and forced them to accept greater external regulation. The ‘auditing’ form of external discipline was chosen as the lesser of two evils, the other option being re-privatisation and a more explicit submission to market forces. The new audit culture compromises traditional autonomy and self-regulation, which the academy (in common with other professions) finds impossible to defend once it acquires a central role in economic and social development, so that participation in its work and consumption of its results is extended beyond the traditional elite. But the recently developed measurement and accountability regimes have, at least in Europe, retained enough of the old self-regulatory features for most universities to accept them, if only as the least-worst viable option. The new ‘quality assessment’ processes retain an emphasis on letting academics peer-review their own work, on assigning student places according to intellectual rather than material resources, and on assigning research funds according to intrinsic merit rather than ability to generate wealth.

However, we have also argued that the new regime has promoted restructuring of the university system as a ‘quasi-market’, which imperfectly shields it from the pressures of the commercial market. Social sciences have been shown as especially vulnerable to the new research assessment procedures, because the absence of any clear method for validating and evaluating their results raises the temptation to resort to market valuation, selecting students and research projects on the basis of ability to pay and repay.

These transformations may bring about other modifications, and there is indeed a growing literature on the various undesirable effects of what has been described so far (e.g. Blake et al.1998; Jary and Parker 1998; Shore and Roberts 1995; O'Neill 2002: 43–79; Shore and Wright 1997, 2002; Strathern 1997, 2000, 2002). But it is useful to make a distinction between possible and actual shifts. The former refer to changes that are perfectly feasible, that make sense logically speaking, the latter to changes for which we have empirical evidence that they are already taking place. The former allude to what we could reasonably imagine will happen, the latter to what we know is happening. In current highly charged debates about contemporary academia, the two are often conflated.

Four main questions are raised by this conclusion. Firstly, there is the question to what extent the erosion of the Humboldtian notion of the University brings about the pre-dominance and spread of those forms of knowledge that exhibit a practical dimension, an instrumental rationality. In the case of the system now spreading across Europe from the UK, it is not obvious that the auditing system has had this effect so far within the universities, although a department with huge grants gets rewarded in the research assessment exercise and ‘practical’ projects are more likely to receive these financial awards. The emergence and flourishing of ‘useful’ knowledge seems to be happening outside University departments, for instance, in think thanks, management consultancies or government agencies. If anything, there seems to be a growing division of labour, with universities involved with highly theoretical knowledge and the other knowledge-producing institutions tied to various forms of social policy and human intervention. These forms of knowledge have, at the same time, gained in status and importance.

Secondly, there is the issue what kind of effects the fall of the traditional university has for the intellectual pluralism with the profession, whether it increases or decreases the number of intellectual genres and paradigms that are recognised, tolerated and promoted. There is a growing anxiety that the auditing system of the kind we have witnessed in the UK can be employed to impose a particular framework, to guide research in a particular direction and root out intellectual rebellion, although it could as well be used to bring about the opposite and produce intellectual diversity. In which direction the system is employed seems to vary from one academic discipline to another. In general, if a discipline has already operated with one leading paradigm, the auditing system tends to be used in ways that reinforce the consensus. Economics, for instance, only recognises publications in a small number of mainstream journals and operates with a clear rank order of the journals involved. Other subjects, which have tended to lack a dominant paradigm, seem to use the research assessment exercise in a more liberal fashion, but competition among approaches according to their research ‘productivity’ may create conditions where a similar ‘mainstream’ will emerge.

Thirdly, there is the question what effect the audit system has on the professional culture. There are indeed serious concerns that the auditing culture may undermine the academic ethos, the idea being that the act of surveillance may well destroy what it is supposed to bring to light. With its replicating systems of monitoring at various levels (supra), the fear is that the audit system may destroy the moral fabric that underlay university life and held it together. The surveillance techniques might have atomising and alienating effects on the individuals concerned, aware as they are of the commodification that is taking place, the extent to which publications are more than ever a lifeline in an increasingly competitive market.

Fourthly, there are concerns about the quality of what is being produced. Whilst it is true that, as result of the government-orchestrated assessment schemes, the quantity of output has increased, we do not know what has happened to the average quality of the books and articles that are written. There are certainly more possibilities for the publication of insignificant writings than there used to be: there are an ever-growing number of journals on the market such that articles of varying qualities manage to get into print. Regardless of the conditions of work, background and ability, all academics are under a similar pressure to produce a certain number of books or articles, so we might expect some to produce lesser material or replicate the content of previous publications (Graham 2002: 74–9). They might not have published these superfluous writings under the pre-audit system. But whether this means that the overall quality has declined remains an open question. All we know is that the next research assessment is probably not a good reason for writing.

Patrick Baert is University Senior Lecturer in Sociology at the University of Cambridge and Fellow of Selwyn College, Cambridge. He wrote Time, Self and Social Being and Social Theory in the Twentieth Century, and he edited Time in Contemporary Intellectual Thought.

Alan Shipman is Affiliated Lecturer in Social and Political Sciences at the University of Cambridge and Bye-Fellow of Selwyn College, Cambridge. He wrote The Market Revolution and its Limits and Transcending Transaction.