UCL Anthropology


The Virtual Moment

Daniel Miller

It is perhaps not surprising that anthropology is increasingly involved in the study of audit, partly one imagines as a reaction to the degree to which anthropologists find themselves increasingly subject to audit of their own teaching and research and respond by trying to objectify this relationship as an academic problem. Indeed the core of the response has been a critique of the audit of academia itself (e.g. Strathern 2000). But within that same edited volume there is another response characteristic of the core principles of anthropology. This is to engage in a more general ethnographic encounter through undertaking fieldwork on the process of audit on the basis of an empathetic engagement. That is to try and understand the language, perspectives and structure that constitute such audits.

Such an ethnography now has the advantage of engaging with a developing anthropological literature found not only in Strathern (2000) but also such exemplary texts as Riles (2001), as well as earlier work on bureaucracy in anthropology e.g. Herzfeld (1992). In particular, it puts the researcher in debt to the critical contribution of Power (1997) in his book The Audit Society. The empathetic stance of ethnography also implicated an engagement with the recent work by Du Gay (2000). His book In Praise of Bureacracy makes the point that an engagement with audit has to confront the current contradictions within bureaucracy. In several respects audit appears to be a form of bureaucracy, but one that has the effect of actually constraining the ability of prior forms of bureaucracy from fulfilling the role for such institutions within society set out originally by Weber and reiterated by Du Gay. So an empathetic encounter by no means implies an uncritical analysis. It is this anthropological trajectory that I wish to exemplify in the following paper. That is an ethnographic study that is based around the terminology, stated aims and observed practice of audit, but that is focused on the contradictions of structure that may lead to results that are quite antithetical to those stated aims.

In the conclusion to this paper I will link the study to a larger theoretical structure termed Virtualism (Miller 1998). It was the desire to ground this theory and to develop a further study of the use of the concept of value in contemporary life that led to the choice of a specific example of audit, termed Best Value (Audit Commission 1999, BV from now on). BV, which was established in the UK in 1999 is a colossal operation, said by Martin (2001) to cost £600 million a year. It is based on a new statutory requirement that every service provided by every council in the UK must provide a BV report and be subject to a BV inspection once in every five years. At the time of writing about 1,600 of these have been published ranging from cleaning buildings and security services; maintaining playgrounds, highways or libraries; through to planning and environmental controls - in short every aspect of what local government does (for reviews of BV see Boyne 1999, Martin 2000, 2002, Raine, 2000, Snape 2000).

The inspections seem to follow the general sense of New Labour and Third Way politics in the UK. That is it claims to have transcended any ideological preferences for either the state or the public sector in favour or a re-orientation of all services to the user as the ultimate consumer of the service. It follows what has been termed the modernising agenda (du Gay 2000) of government with respect to all aspects of the civil service, and there are parallels to BV across the board of government. The system of audit in the UK has its own particular history and context (see Power 1997), but is parallel to movements in other countries and is certainly affected by the drive for EU and international standards. There is also a clear influence from the US both in terms of the movement to privatisation and the whole philosophy of cost-benefit analysis as it has developed within the US (MacLennan 1997). Indeed internationally it may be viewed as part of a `New Public Management' regime that is being applied worldwide (McLaughlin, Osborne and Ferlie 2002), within which this kind of performance auditing has been taken much further in some countries, such as New Zealand, Sweden the UK and the US than in others, such as France or Germany (Pollitt and Bouckaert 2000: 70, see also Pollitt et. al. 1999).

The premise of inspections is that local councils have tended to develop historically around their own concerns as providers without properly consulting users and without taking on board modern management priorities such as proper comparisons, emulating best practice, and striving for what is enshrined in BV as `continuous improvement'. While BV was developed prior to the last general election (2001) in the UK it certainly reflects the centrality in that election campaign of arguments about the delivery of public services. Given the political opposition's critique that the government has wasted money thrown at the public services without discernable improvement, the issue became centred on the degree of acknowledgement by the public at large of any claimed improvements in these services. The vision behind BV and its siblings in other sectors of government is then a kind of Copernican revolution in the self-conception of government. Up to the present local government is viewed as something that grew up almost organically, gradually emerging with its own sense of itself as a provider with a concomitant development of its own self-interest, such that the public remain on the outer periphery as merely the ultimate justification for its existence. Under this new vision the public become themselves the shining power at the centre of this universe and local government circulates around it only to the degree that it actually serves the desires and needs of that public.

` The `Best Value' initiative combines several strands of New Labour's approach to performance: an emphasis on what counts is what works,…There is no presumption that private or public provision is superior….This puts local stakeholders in a stronger position to influence services, moving the focus from cost-effectiveness to community-effectiveness'. (Rouse: 1999: 90). By contrast, the view from local councils themselves has been formulated from a quite different perspective, largely the result of their experience of the previous Conservative governments formula of CCT (Compulsory Competitive Tendering). This came to be seen as in essence an excuse to privatise services to the cheapest bidder at the expense of conditions of employment for workers in the public sector. Their suspicion was that BV would be a continuation of this prior political agenda. Although BV officers take some pains to deny such continuities and the government claims to be `simply agnostic' about who delivers services (Martin 2002: 132), others see New Labour as actually more zealous in promoting performance management and expecting more involvement of private organisations than previous government (Horton and Farnham 1999: 255). Given the considerable political advantage to having the public acknowledge improved conditions I see no reason to doubt the sincerity of a program that is devoted to finding improvements in services that potential voters will acknowledge, although as I will suggest below the misgivings of local government do not lack foundation either.

It is not intuitively obvious that the primary instrument for ensuring modernisation should be forms of audit and inspection. The problem faced by government, and the problem that has launched a trajectory towards virtualism in the public sector, is that apart from actual elections it is hard to deal with entire populations. Institutions need to be found that can stand in their stead as though they were representative of `users'. As such the BV regime was created as an inspection revolving around `reality checks' where inspectors test out the service as through they were users. This is intended to differentiate them from other parts of the audit service, which may be largely restricted to checking boxes on paper to confirm proper procedures. In practice they remain part of a vast increase in the activity and power of audit and especially the Audit Commission throughout this sector (Shore and Wright 2000: 63-7 Power 1997: 46-9, 116-7)

BV is formulated around what are called the four C's. More or less every time inspectors make a public presentation, reports are written or anything else is done under BV the four C's will be explicitly repeated - they are the manta of this regime. Councils must Challenge how and why their service is provided, Compare with others, embrace fair Competition and Consult with taxpayers, and customers. Behind the four C's lies a fifth, which is the commitment to Continuous Improvement. In turn all of these are supposed to be carried out according to the three E's which are to work by the most Economic, Efficient and Effective means available. These are the bedrock terms - almost anything that happens is justified in relation to them.

Each council develops a schedule such that all services are inspected over a five year period. In order to be ready for inspection the service has to produce a Best Value Report (BVR), which is of itself a major task taking many months and often requiring the appointment of specialist BV officers. This must include statutory requirements, such as collecting the data for specified performance indicators, and other comparisons with other councils. Evidence must be given for considerable consultancy with the public. The BVR that is based on this data, in turn constitutes the long term plan for `continuous improvement' which demonstrates that this will be achieved on the principles of the four C's. Once the BVR is completed an inspection will be arranged. Two inspectors, mostly but not all, employees of the Inspectorate, will make a preliminary inspection of the BVR and a considerable amount of other documentation.

They will then spend a week on site, interviewing most senior and some junior officers of the service. There may be `pin-point focus groups' with users and/or staff. In order to see the service from the point of view of the consumers they also engaged is what are called `reality checks' (although officially the term refers to all on-site work). This revolved around a kind of `mystery shopper' ethos and might involve anything from inspecting the toilets that are provided in each of the public parks within the borough, to going out with the truck that picks up the rubbish from homes, to asking users of swimming pools how they found the changing rooms, to checking the public library to check the quality of information provided for people who want to make complaints about the council, to comparing the quality of pedestrian crossings of roads in the wealthier and poorer sections of a borough.

In one case we actually did go and walk around every single example of a particular facility in the council area. In other cases, given the short time available, there was an emphasis on the facilities closest to the offices we worked from. The inspectors would also make `fake' phone calls to see how quickly services responded to complaints or enquiries as though from a resident, such as reporting some rubbish tipped on a street, or that the computers in a library were not working. Since inspections are planned in advance I assumed that special efforts would be made to impress for that week, which sometimes happens, `this inspection, it was autumn and we were panicking about autumn leaves all over the place - Oh my god , make sure the street sweepers are out, we've gotta get the leaves up.' Overall though this kind of special response seemed surprisingly rare.

In practice reality checks occupy a much smaller fraction of the inspections than the analysis of internal materials, for example the paper trails of car parking workers reporting sick, or the follow through on recorded complaints about museums being closed on the wrong date as well as the councils own quantitative surveys of `users' opinions, such as on the quality of road crossings. What dominates the inspection, however, is analysis of the paper work devoted to the larger picture such as the BVR and other audits and comparative surveys for example by the Mayor of London on the overall transport situation. The final week will be spent sifting through the material. The inspectors enact what Harper (2000: 47) calls a moral transformation, changing qualitative data and quantitative figures into that which is used to justify the final grading. The inspectors are subject to two quality control meetings by the senior officers of the inspectorate. One based upon what will become their initial presentation to the council before the time spent on site, and the other of the final report before it too is presented to the council. The use of powerpoint to make such presentations is one of a number of factors that mediate the construction and aesthetics of the final document (Harper1998, Riles 2001: 114-142)

The inspectors report allocates one of four grades to the service in respect to two questions. The first is concerned with how good the service is, and is based mainly on their inspection of the service, their reality checks and so forth. The second with respect to its likelihood of improvement, is mainly devoted to the BVR itself and tends to follow the four C's to see how far it is based on BV principles. But it will be also concerned to see if the intended improvements are properly costed and that the politically elected council members actually back it. The results are then published on the Best Value website, which is a sub-section of that of the Audit Commission.

I was fortunate to be granted excellent access by the Audit Commission. I was able to accompany the two inspectors in every aspect of their inspection, including the discussions that took place at Millbank Tower (the headquarters of the Audit Commission as well as of the Labour Party) before and after inspections. I could sit in and record all meetings with local council officials, and accompany inspectors on their `reality checks' of actual instances of these services. I participated fully in three inspections (which I will keep anonymous) and in addition discussed BV with council officers elsewhere that had been inspected or were preparing for inspections. My research both as participant and interviewer took place in and around London during 2001. I also analysed ten reports as well as other documents associated with BV. Such documents are freely available on the websites of almost all the councils in the UK. Although I suspect my presence was sometimes felt to be `the third that makes a crowd' given the intimate nature of the interview process, I do not think I had a significant impact on any of the results or procedures. I did not speak during interviews but usually discussed the interviews with the inspectors after the interviewee had left.

While some other audits were characterised by officers and inspectors as notoriously aggressive, such as the OFSTED inspection of schools in England, almost all these inspectors had previously worked within local government and expected to do so again in the future. As such they tended to desire a collaborative rather than confrontational style. As with most contemporary management they are well aware of what they see as performantive and cultural aspects of their work (see Salaman 1997, Thrift 2002). My impression was that the majority strongly accept the BV agenda, and firmly believe that it will produce benefits for local government and that it would be possible to persuade local government officers of this benign view. Their main reservation was their awareness that BV was merely the largest of a growing number of similar audit requirements and other schemes that between them were imposing a very considerable burden upon the time and resources of local government. They were aware, however, that if BV developed a poor reputation within local government this would not be particular good news for their own future career prospects. They worked conscientiously and extremely hard to deliver their reports given that only three weeks were allocated to each. In turn most local government officers seemed to see BV as at least an improvement upon CCT and despite resenting the labour involved seemed prepared to give it the benefit of the doubt. Indeed my argument in this paper is that the evidence I present that the results do not match the desires of the inspectorate occurs despite not because of the intentions of those that created and those that carry out BV, who are largely well intentioned towards both local government and its impact on welfare more generally.

As an ethnographic treatment I will try and stay as close as I can to the language and concepts used by BV itself. For this reason I will use the 4 C's themselves to structure the argument. Within those terms I will highlight the nature of those structural contradictions that lead to the undermining of the stated aims. It will become evident that in each case these contradictions take the form of an increasingly abstract version of these aims that eventually works against the more grounded and specific pursuit of BV's intentions. It is such abstractions that will be re-cast in my conclusion in the mould of a theory of virtualism. Each of the four examples is used to illustrate a different version of that phenomenon connoted by the general term `abstraction'. The first is concerned with increasing abstraction in language. The second with the creation of externalities that take the process away from its own substantive concerns. The third examines a case where the focus shifts from substance to representation, and the fourth with the rise of quantitative over qualitative assessments.


The first C is concerned to ensure that the council uses BV for a thorough re-consideration of its goals and means. Implicit is the possibility that the service has merely grown through administrative evolution. With the Copernican revolution announced in BV this is the time it should take stock and re-orientate itself to its fundamental raison d'etre, which is service to the community. It should be open to the possibility that it makes more sense to break the service up or re-connect with other services. More pragmatically it is essential that it spells out its goals in clear language, that these include a commitment to continuous improvement and to how this will be achieved. Technically it must show how the service performance will reach the top quartile for all councils in terms of improvement over the next five years. In addition the inspectors look for a kind of vertical integration where one can see how the service's goals relate to the mission statements of the council as a whole and in turn cascade down to particular strategies which are linked to these goals and not simply plucked from the blue. Ideally they should also follow the needs of users as discovered through survey.

I was struck by two commitments within challenge, which I wish to explore. I felt there was a genuine desire that BV should reduce jargon and create clear objectives, and that it should reduce bureaucracy by focusing upon the goals themselves rather than the internal concerns of `process'. These seemed to me clear goals and I was able to watch, for example, a lead inspector carefully weeding out acronyms and other aspects of jargon from a preliminary report by the inspectors on site.

Of course there are terms that an outsider must learn to cope with such as `are you properly incentivised' `have you progressed yourself' `what are your primary drivers' `where are the quick wins', but as long as these remain meaningful within the institution are not necessarily obfuscating. Far more damaging is a process which emerges when one examines BV as a whole. it soon becomes apparent that this commitment to oppose jargon gradually becomes overwhelmed by the production of a new BV language, such that a great deal of what the council ends up doing is trying to find ways of correctly translating past and current practice into the correct BV terminology. In practice councils who failed to do this were penalised, especially when inspectors tended to make it an explicit criterion to ask if the report was sufficiently `driven by BV,' which in practice can come to mean expressed through the 4 C's. As one moved into the larger world of BV and talked to council's newly appointed BV officers, or examined the websites dedicated to BV itself, then it becomes apparent that the higher one moves from the grounded work of actual inspection to the inter-council discussion of how to deal with BV in local government, the more one is faced by increasingly abstracted and specialised language.

For example on a local government website called I&DeA (www.idea.gov.uk), there is a four page address on performance management. The extraordinary quality of these pages is that they seem to say nothing at all except express the most mundane tautologies (i.e. re-writings of the meanings of the words performance and management), which also reiterate the basic rhetoric of BV itself. A typical paragraph reads `For local government, effective performance management requires co-ordinated planning and review systems that enable key decision makers, both political and managerial, to take action based on facts about performance'. Or `Effective management consists of `a clear vision, purpose and focus on outcomes - `Why are we here?' There is not a line that isn't a repetition of something that people working in local government will not have heard hundreds of times. Such language comes across as entirely performative, a kind of incantation. There are hundreds if not thousands of similar empty documents in local government.

Understood comparatively, we can start with our own anthropological academic jargon, which I suggest has the following characteristics. While an ethnographic monograph used to be largely descriptive and clear, increasingly there has been a desire to incorporate jargon in the form of a theoretical language that is dense, opaque and highly abstract. Often this is based on citation of French intellectuals (each decade has its model progressing from Althusser (70s) to Derrida (80s) to Baudrillard (90s). Typically the users claim to be radical and inclusive though as Bourdieu (1984: 485-500) pointed out with respect to Derrida this language ensures that they are in effect elitist and exclusive. They work as a performance of cleverness often led by younger scholars seeking to overthrow an established generation.

This can be contrasted with Riles (2001) recent examination of the language of NGO's in Fiji. She suggests there are two main concerns here, firstly the materiality of documents that are produced, and secondly the pattern of language that constitutes these. She follows UN discussions for hours over the pairing of words and whether tentative brackets should make it to the final text. She is perhaps rather generous to academia in claiming that this contrasts to our work in its privileging form over meaning, since I would say this was also true in our own use of jargon. But behind her claim is her evidence that NGO use of this language is more pervasive than ours. What is distinctive is that NGO language is replicated more or less unchanged from the highest echelons of UN global debates to all its local equivalents as, for example, in debates in Fiji.

By comparison BV as jargon shares some of our characteristics and some with NGO's. As with academic anthropology the abstraction develops at the higher echelons and in contrast to the clearer descriptive language of many of the foot soldiers. Unlike NGOs there are clear differences between the global and local levels. On the other hand the language creates obfuscation by its bland generality while academia does so through the increasing obscurity of its terminology. Instead of performing cleverness it performs transparency. It is so clear as to be essentially empty, retreating to the banality of tautologies. It is generated by the sheer scale of BV. As a huge operation it creates specialist BV officers and experts whose discussions become distilled and abstracted versions of itself, quite distant from the day to day world of inspections. Ultimately these higher abstract versions come to undermine the intentions of the regime as a whole to create clarity. So although we may discern a rise in obfuscation in many fields of contemporary discourse there are grounds for discriminating between the different variants that have arisen.

The contradiction I have just described is not usually discussed in terms of language. There is, however, a close parallel in the discussion of bureaucracy. While I discerned a genuinely anti bureaucratic intention within BV, many newspaper articles (e.g. The Guardian 3/01/01 26/3/01 28/6/01) dismiss the regime for being precisely overly bureaucratic or `intensive', implying an emphasis upon process rather than goals. This is partly a result of the sheer effort that needs to be expended in creating a BVR. The process starts with the ideal of a foundational re-think of the service. Perhaps environmental concerns that for historical was in the same department as parks and leisure would really be better teamed with housing since today they share more in common. But this ideal of challenge is undermined firstly by the way units of inspection are constructed (the `scope' of the inspection) but more by the heavy burden of time and resources which mean there is even less time available for such macro analysis. As in the case of language, we find instead that specialist officers are appointed who ensure that the emphasis is ultimately upon getting the representation of the service right rather than changing it. As experts in using the correct language and phraseology they ensure the proper translation into acceptable BV format. As a result a service that is `driven by BV' will very likely have poured more resources into process and bureaucracy itself, even though the inspectors in good faith see themselves as promoting and then having the opposite effect. So abstraction in this case is in large measure a reflection of scale.


Competition expresses the ideal of BV in repudiation of the previous regime of CCT. It is argued that it is value itself that should determine the argument between private and public provision. Whichever can demonstrate a better return on taxpayers money in terms of performance (Strathern 2000: 289 should be responsible for provision. The term value is seen to internalise questions of the quality of service, factors said to have been external to CCT which local government saw as simply an attempt to cut costs, and which thereby tended to reduce the quality of the service. I would argue that this idea that factors in an evaluation may shift from being external or internet is critical to the issue of abstraction. It represents an anthropological equivalent to the economists' concept of externalities and a parallel to Goffman (1975) and others concept of the frame. I have made this argument this recently with respect to the analysis of the car (Miller 2001a), Callon (1998) with respect to the concept of the market.

The premise of BV is that of market testing, in which services must demonstrate they should remain in-house through a direct comparison with private alternatives. Underlying this is the premise of a `level playing field', such that the odds are not stacked up against either private provision or public provision. This premise is undermined by the treatment of factors as externalities that ought to be part of such assessments. The result gives the appearance of a level playing field, but actually distorts the result to favour the private sector. For example, consider the issue of conditions of employment. Clearly a company that provides a service more cheaply by paying less to its employees is only providing better value if one discounts this factor from the equation (as clearly occurred under CCT see Reimer 1999). There are many arguments about what are called TUPE (regulations regarding the transfer of workers rights) conditions that are supposed to prevent this under BV, but there are ways private companies can work around such protection. While this issue was often raised by the service, the inspectors did not see them as central to such testing.

One of the costs imposed by BV derives from the insistance of market testing in and of itself. For example I was fortunate in finding that one of my ex-students now works for a council service called `Agenda 21' which is supposed to ensure that all council services follow current environmentalist programmes. He was tearing his hair out trying to work out what would satisfy the demand for market testing and pointed out `Its terribly terribly difficult. We tried to find a way to argue we were assessing the competition but it's a corporate process, its sort of the philosophy of the council'.

Similarly there is an important imbalance in the provision of finance. Even though public sector institutions can theoretically obtain finance more cheaply than private sector corporations, if they are not allowed to do so, then the balance is swung heavily in favour of the private sector that has access to capital. Investments required to improve a local government service such as a new incinerator plant can be huge, making the private sector the only option, even if the public sector would otherwise have been better value. An example of this which has been subject to considerable scrutiny has been the recent critique of PFI (Private Finance Initiative) which has become heavily involved in the outsourcing of local government services. Critics such as Monbiot (2000) see this as an example of political corruption on a colossal scale, with the government in effect subsidising private contractors. Research at UCL on PFI in the health service (Gaffney, Pollock, Price and Shaoul 1999) certainly suggest that the odds are stacked anything but fairly.

Another example of such externalities has become the cost in time and resources of BV itself. The public sector is now subject to a vast number of similar auditing and inspection regimes that take up a significant part of its resources. Even when they are employed to provide public services private companies tend to be comparative free from much of this labour. What I found incredible was that inspectors could make reference to an advantage of the private sector being its relative freedom from `red tape', without a trace of irony. Indeed there is a deeper contradiction emerging here. Given the sheer weight of effort required by BV the instrument of judgement can simultenously becomes the instrument of punishment and reward. The best councils are promised a light auditing touch in the future, while failing councils are promised even heavier audit to come (Martin 2002: 133).

The overall point that emerges is that BV understands itself to be a new attempt to create a genuine criteria of value on behalf of the consumers of local government in their other guise as tax payers. But so much of this relies on the discounting of externalities that appearances have become highly deceptive. Value is reduced to an equation in which most of the major factors have been abstracted to appear outside of the evaluation process. So externalisation has been identified as the second vehicle of abstraction


This is the C that takes us to the heart of the issue of virtualism. If workers in local government knew one thing about BV it was that they were supposed to consult with the public. Indeed much of the effort in preparing for BV consisted of carrying out questionnaire surveys to demonstrate awareness of the public's pre-occupations. The problem is that this can lead either to a deeper consideration of how public concerns should be both monitored and integrated, or instead it can become largely an exercise designed to best represent the service to the inspectorate as being based on consultation. This shift from an emphasis on practice, to that of representation is a common property of virtualism.

The inspectors constantly had to face up to some of the complexities and contradictions underlying an ideal of consultation. They would hold pin point focus groups with panels of residents, but in most cases found, not surprisingly, that residents did not prioritise the `integration of the service's forward planning with the Mayor's strategy for London' but rather whether there had been an increase in dog poo on the streets. Challenge insists that the priorities of the service must be based on consultation, but many of these were highly technical and esoteric questions that the public would clearly much rather delegate to professional expertise. Inspectors would sometimes try and adopt a position of populist empathy claiming `The issues that we want to know need to be more about what I describe as Johnny tax payer more than the technicalities. So I'm constantly saying yeah yeah it is all very interesting but what does it mean to the man on the street'. But this encounter with the man on the street has its problems. On one occasion the inspectors tried to elicit opinions from two teenage girls about an aspect of the design of a road crossing they were using. They were met with the kind of haughty indifference that only teenage girls can express with full force.

This left council officers who were trying to determine whether it was cost effective to invest in high quality infrastructure for a commercial high street somewhat bemused. In the face of the sheer impossibility of putting responsibility for decision making back onto the public in practice, we find that as Du Gay (2000: 109) noted, the ideal of democracy as the delegation of necessary but possibly unpopular action gives way to a more facile and possibly unconstitutional acquiescence to public discourse as reduced to questionnaires, since this is the kind of data that can be used to fully represent the act of consultation.

The other contradiction sensed by officers arose when the priorities expressed by the public did not match the priorities expressed by government. The council's first obligation are to changes designated as statutory regulations. So, if for example, such regulations require a major expansion of re-cycling beyond anything the public seem to be bothered to enact, the role of the council turns into providing sufficient `education' in the right places to ensure that next time they consult the public, the public `correctly' priorities the changes they are supposed to be in support of.

On the basis of a combination of questionnaires which might have only a 10% response rate and self-designated `concerned' groups that in areas with extremely low level of active politics tend to be a tiny fringe, it was very hard to claim that consultancy is actually a form of fair representation. Methods such as questionnaires and focus groups also favour public discourse over private opinion, which ethnography can suggest are quite the opposite of each other (e.g. with respect to opinions about corner shops in Miller 2001b: 66-106). Work with activist groups tends to privilege the more self-confident and the middle class that construct themselves as watchdogs over local government. While these may sometimes be highly positive whistle blowers, they could also be merely tiresome and self-important voices. Indeed local government finds its resources increasingly strained by the rise of what might be termed the `rogue consumer' that is the activist who theoretically might make a threat of litigation, against which eventuality institutions have to `cover themselves' often at huge expense to the public at large. This is not just a matter of costs, but of finding out that for example, what the public sees as the common sense delegation of decision making to officials responsible for looking after their children is no longer allowed for fear of such litigation.

Consultation as orientated to specific problems looks an effective way of locating problems , for example `I'm on the panel of the road safety group. Where we see there's an accident problem and we can't influence the buses. So we are trying to draw the public's attention to it and those that can influence it to improve the situation, particularly elderly women hanging onto Sainsbury's shopping bags - I see it every day and there's no accountability of the driver.' One could imagine that taking the issues of buses pulling out too quickly to a high level such as the overall London transport system could prove effective. But in watching the inspection develop once finds that something that starts with the best of intentions to re-orientate services to consumers problems, is gradually turning away from the needs of actual users to become instead an increasing concern with the active representation of a performance of consultation that exists in order to satisfy those who have become the de facto consumers of this exercise - the BV inspectorate itself. What the council is required to do is to write up in considerable detail how its extensive consultation has now isolated the issue of dog poo as the predominant public concern, how this provides the true foundation for policy, prioritisation and in turn for strategies of improvement, and is the new foundation for its BVR. If the logic by which all these are related is clearly stated in the requisite BV language then the council is deemed to be functioning effectively on behalf of its residents. A logic of abstraction which is at the heart of virtualism. The problem is clearly not that of representation itself. Anthropological theory would tend to treat representations as just as `real' as that which they stand for. The problem is to discriminate representations that clarify, and support, from those which, as here, come to suppress and supplant that which it represents.


The final form of abstraction is one of the most common in contemporary institutions and has been reflected in the previous discussions. It is the rise of quantitative over qualitative forms of representation. During the inspections, it was very common, following an interview, for inspectors to reflect upon what they had just heard in terms of their own experience within the various councils where they had previously worked. In general their experience meant they were able to produce plausible and effective comparisons that gave them a good sense of how this particular council measured up against others. I became convinced that, left to themselves, the inspection could have produced a fair system of evaluation (well, at least as fair as academic anthropologists marking exam scripts). The problem was that when it comes to representation in a formal report this counts as `soft' evidence, while comparison based on quantitative forms such as performance indicators counts as hard evidence.

While I would admit to my own prejudices about quantitative comparison, even I was shocked at how the more knowledgeable inspectors were able to take apart almost every quantitative indicator they possessed. As a result of their own experience at attempts to make local and meaningful comparisons (in what are called benchmarking clubs), they could see how this data was in most cases flawed. This might be because they knew different councils were actually presenting data based on different understandings of the unit collected. Or it could be because councils couldn't actually be homogenised simply because the one with a motorway through it, or the one that was mainly affluent, or the one that had a teeming night-time culture, were all special with respect to the data being considered. Inspectors themselves, could, however, exacerbate such problems by their own abstractions. For example I watched one attempt to try and be `fair' in giving the council a grade, on the basis of trying to calculate the number of factors where the council in question appeared better or worse than others previously given the proposed grade. Trying to make this secondary data into a calculation meant losing the specificity of what was compared. This reduces to the same scale the entire evidence for public opinion on the service's performance on the one hand and the quality of its sign posting on the other.

Once again, however, the primary problem was not just that inherent in quantitative representation of qualitative forms, to which we are all subject. There were further contradictions in the process itself. For example the officials all saw quantitative data such as Performance Indicators as having high setting up costs, but becoming more useful over time, thanks to the internal consistency generated over the years. Unfortunately they strongly suspected that the drive to `continuous improvement' that they were subject to, would inevitably also effect a regime such as BV, which would change its own goalposts. Most likely it would be replaced by the next political regime, as had tended to happen with past changes in political regimes. The result was a sense of resignation to a fate where all such work was ultimately wasted labour. As one official put it `This complete obsession with well, we've got to find someone with the same demographics as us, who does it exactly the same way as us, collects exactly the same information as us, has got the same policies and objectives. What's measured is not necessarily what is a local priority. And what the national priorities may be, may not be the same for us. I mean the thing about PI's is how the hell to construct them. And they change them every year so they are not comparable over time or between boroughs. It's a nightmare. There's no consistency in them.' So the abstraction created by the desire for apparently objective comparison becomes the fourth example of the virtual moment.


It is quite common for an ethnographer to search for particular phrases or practices that seem to crop up frequently in conversation and yet are not obviously germane. These are seen to provide possible clues to a better understanding of what they have been participating in. In this case there was one such phrase that I think shed light on both the larger context of and effect of virtualism. Quite often during conversations participants would speak of the issue of `who owns the process'. I take this to be a common expression in modern management, but it was not at first obvious why it should arise so often within this particular exercise.

The term ownership in the specific context of management seems intended to implicate the relationship between having responsibility for and feeling responsible for. So the phrase `to own the process' means that a group has been given sufficient autonomy to control a particular organisational process to feel that the responsibility for carrying it out is unambiguously theirs. The implication is always, however, that only under such conditions will they feel committed to the process itself. Ownership as a concept then seems to be trying to make clear that in order to expect people to be fully engaged in what they do, one has to give them sufficient autonomy and responsibility. Why then should this term come up constantly in discussions about BV?

The irony of BV is that it addresses, as though it was an absence, that quality in which local government is probably uniquely rich, outside of the voluntary sector. This is the sense of being a public servant. A sense that there is a reward for this work quite apart from wages and the internal activity of work. As a service head put it `You know, if you think about somebody who's running a service if they're doing it properly, a public servant isn't just there because they're there. If I wanted to be like that I wouldn't come to work in the public sector. I believe I come to do a job that makes a difference in people's lives.'

This is what distinguishes all public sectors such as medicine and education from private sector employment. But while education and medicine have more immediate positive goals in the form of health and education itself, local government officials are often dealing with less obvious benefits such as leisure facilities or planning. As a result the ethos of public service tends to be generalised around the identity of being in local government itself rather than say being a teacher or a nurse.

I do not want to naively suggest that all workers share this ethos, any more than assuming cynically that none of them do. But what is important is that to the extent to which they do, this is denied by a BV regime that represents its Copernican revolution as though it is only thanks to BV that a council might start to have regard to the public that it serves. As individuals, inspectors who mainly come from local government, were sympathetic to the ideal of public service as inherent to bureaucracy. But this was suppressed by their formal status as inspectors. In turn council officials seemed to acquiesce quite readily in the idea that a service should be privatised or at least castigated if it genuinely failed the public. They were surprisingly open to outsourcing on the basis of a genuine level-playing field. But they insisted that a factor that should remain inside, and not external, to such evaluation was that asset represented by the ethos of service itself, which they felt tended to be lost in privatisation.

There is, however, a deeper issue here, raised by the process of virtualism itself. The immediate threat posed by BV was that of virtualism. I have argued (Miller 1998) hat a property of virtualism is that it operates through a fetishism of the consumer, as opposed to that of the producer. We start with the authority that we attach to the consumer by virtue of our recognition that it is the consequence of services for them that legitimate the existence of that service (see Keat 1994). Under virtualism the mantle of this authority is then stripped from actual users and worn by those bodies that stand in the stead of the public at large. The inspection was created by the government to act as virtual consumers of services, that is to stand in the direct stead of users, who would ultimately represent themselves in their other capacity as voters. So the job of inspection is to ensure that the service is creating improvements that would be acknowledged by the users of that service. As such the key to the authority of the inspectorate is that they possess the authority of the consumer.

There is, however, a significant difference between the service and the inspectorate. This is simply that the council is capable of providing a service to the residents as consumers, while an inspectorate cannot of itself do any such thing. It can only inspect and make recommendations. This then raises the question of `who owns the process', since the effect of this fetishism is that responsibility to the public is being taken away from those who are responsible for carrying out this work and given instead to those who are only responsible for inspecting this work and who cannot provide a service. Furthermore this relates to the single most sensitive aspect of local government identity. That is the very ethos of public service, which gives those who carry out any such work the moral satisfaction that their work is in some way of benefit to the people who they serve. My impression was that this was not often stated as directly as might have been expected. Instead it had been re-cast in the language of management that had become the `proper' vehicle for discussing such matters, to appear frequently as the issue of `who owns the process'.

It follows that the ideal inspector is one who realises that the process of BV has to be carried out extremely sensitively, since the process of BV can only work well if at the end everything that BV has done is given back to the service, i.e. the service comes to retain ownership of what they do. But there are all sorts of elements of BV that militate against this. For example, a more strident inspector can easily turn BV into an instrument of alienation, such that service workers simply feel estranged from their own work. My sense was that on the whole, not surprisingly, the more experienced inspectors tended to have developed an implicit sensitivity to this issue as the heart of the exercise they were engaged in.


The term `best value' implies the government's sense that taxpayers will insist that they should obtain the maximum benefit for the amount spent on any service. It is directed against the accusation that private provision is more cost effective, by putting the onus upon the public sector to demonstrate that this is not the case. There is also a desire to repudiate the traditions of high taxation and high spending by using this audit to actually cut the cost of local government and thereby to keep the tax burden low given the public pledges not to raise taxation. My concern here is not with the politics of such a position but how the institutions structures transform intentions.

Formally then the councils have been told that the inspections are expected to identify areas where costs can be cut and such efficiency savings should amount to an overall cut of around 2%. Most important of all is the desire that the public should acknowledge this increase in `value' and the rooting out of council inefficiencies, and therefore support the government in its actions. It is not surprising, however, that given the contradictions that have been encountered throughout this process the overall consequences of the inspection are once again quite the opposite of its intentions. First of all there is the huge cost of the inspection itself. In monetary terms this comes to around £600 million a year, but perhaps even more important is the time that councils have to devote to preparing the BVR. Most heads of services I talked to estimated that dealing with BV represented a cost of around 5% of their overall staff time during that year, must of this is time spent by their most senior and effective staff. Combined with the attack of morale represented by this question of who owns the process, the overall cost of the inspection itself, which is massive. Far more than is likely to be covered by costs that have been cut as a result of efficiency savings.

Looking at the larger context the contradiction is even more stark. The intention of BV is for the public to acknowledge the improvement in value of such services. But in practice there is almost no public interest in such matters. Unless the service receives a disastrous report, the media tends to take very little interest in the results of the inspection. Indeed other than those who are in some way or other involved in these activities, I have yet to find a friend or colleague who has heard of BV at all. Instead the main impact of BV is internal to local government itself. The effect (in many ways interestingly irrespective of the actual judgement made during an inspection) is that each service exploits the inspection as an opportunity to present its case for further funding to the council as a whole, knowing that all other services will have their turn of councils attention during this period. Inevitably a service comes to prominence within the council for the duration of its inspection. It is a criteria for judging the BVR that there is evidence for council support of the improvement the service claims it wants to make, which have to demonstrate not only that they have been costed but that the council is going to allocate the funds for such improvements. Furthermore the logic of BV means that the service will maximise its argument that it matters to consumers and will impact directly upon them, thus in effect implying to councillors that there is a political dimension which will impact upon the popularity of the council. It cannot help but focus upon the service as a means for influencing the votes of residents.

So in direct contrast to the official and I think unofficial intentions of the programme the effect is to make the best possible case for increasing the costs of each service in turn by committing the council to support improvements partly through persuading them that there is consumer interest and potential voter gratitude which thereby politicises the service itself. Council members were constantly making statements such as `Our council sees this as a vote winner', `we have now said there will be additional money for this area. We have to put our money where our mouth is'. It is hard to see how any of this is compatible with the original plan that BV should produce a 2% cut in the overall cost of the services inspected in a given year. Once again this is not because of any subterfuge or clever manipulation. The contradiction that has subverted intentions follows from the logic of BV as played out within the institutions structure. Clearly all of my arguments imply that were audit itself judged by results in the manner that it argues that which is audited should be, it might not have developed to the extent that it has. Larger surveys of these shifts in performance management suggest the same thing, as Pollitt and Bouckaert (1999: 132) note `The international management reform movement has not needed results to fuel its onward march'.


I have used the regimes own four C's to illustrate four versions of the abstraction that seem to amount to the virtualism I set out to understand. These are:-

1) Abstractions that arise from the scale of the institutions.

2) Abstractions that arise from the practice of externalisation out of the frame of reference during evaluation.

3) Abstractions that result when the representation of the act of consulting comes to replace the purported aim of consulting as a re-orientation to users. 4) Abstraction that follows from a exaggerated respect for hard quantitative over soft qualitative data.

These are only examples of the contradictions that may be excavated within the BV regime. Some of the most important, such as the issue of cost or the issue of politics, are too complex to be included in this brief survey. Contradictions are inherent in the political system, in this case they were exacerbated by the political desire to accomplish many tasks in one system, for example they represent both a desire for greater local accountability but at the same time a distruct of local electorates to make local government financial responsible (Weir and Beetham 1999: 248). Indeed in an excellent survey of the more general internatonal trends that BV is symptomatic of, Pollitt and Bouckaert (1999: 170-171) more or less end up with a list of such trade offs and cycles, including that of the trade off between the desire to increase audit and the desire to release government agencies from too much paper work (ibid: 164). But the argument of this paper is that there is a deeper issue here that rests on a wider sense of abstraction than that employed in the more conventional literature on this subject.

The question as posed by the juxtaposition of the ethnography with the larger question of the place of abstraction in the development of these contradictions is `why should a particular audit that I found (quite possibly in contrast to some other example of audit, management consultancy and the like), to be in general an attempt to be well intentioned, benign and sympathetic to those being inspected, should be subject to the contradictions uncovered. As is common in anthropology we cannot answer this question merely by reference to internationality, subjectivity or any more individual manifestation of processes. Rather the instance needs to be considered within a comparative context, that includes other institutions and not just other examples of management reforms. I have argued elsewhere that there have been a whole series of institutions that have undergone similar shifts towards a debilitating abstraction (Miller 1998). For example if we only think of these issues in terms of audit and the public sector, then we miss the parallels to be found in the simultaneous growth of management consultancy in the private sector.

By placing this material within a comparative context we can discern some common tendencies. At the centre of BV has been a shift from a concern with the actual consumers and users of the service through a re-orientation to audit itself which stands in their stead, and which claims legitimacy by usurping the authority of the consumer to determine what is provided for them. In the equally rapid rise of management consultancy in the private sector, over the same period that audit has expanded in the public sector, we see indigenous management being replaced by consultants that claim to stand for the users of these companies whether, customers or shareholders, and thereby come to stand in their stead as an external review. During the same period neo-classical economics has claimed authority to determine the direction of most governments and international bodies by claiming that its model of the market will be to the ultimate benefit of consumers, but in reality taking this authority and giving it to economists. Even in our own academic domain we have seen the rise of a theory of post-modernism that claims to be about a new consumer society, but has eschewed any engagement with actual consumers to replace them by academic gurus such as Baudrillard and Bauman. Another example might have included the rise of litigation again based on making claims on behalf of consumers.

This suggest that there exist common properties found over a wide range of apparently unconnected phenomenon whose power has grown to an astonishing extent over the last twenty years. It was this observation that led to the creation of a more general theory which I have termed virtualism (Miller 1998). Again this could be argued to be characteristic of anthropological concerns that the detailed investigations developed through ethnography escape from a descent into parochialism by being linked comparatively and theoretically. In turn such a theory demands its grounding in the evidence accumulated from ethnography. So the purpose of undertaking this ethnography was to demonstrate that the theory of virtualism can be grounded in the close observation of a particular instance. That it might be possible to locate `the virtual moment' - that is the actual instances where different forms of abstraction allowed for this construction of distance from the supposed beneficiaries of these institutions and their subsequent replacement by groups that accrue power by coming to stand for them.

Given the breadth of the theory of virtualism it does not presuppose that this is intentional or conscious. Rather it suggests there are qualities to such institutions and discourses at this historical period that leads them to contradict in practice that which they claim to stand for in principle. As such it is possible to discern these contradictions from the stance of an empathetic encounter couched in the terms and expressions such as `owning the process' that are central to observed practice. Nevertheless the interest of anthropology lies in constantly reconnecting such ethnographic observation within macro theorising whether this concerns the nature of humanity, social institutions or as in this instance an attempt to discern a general historical trend.


Audit Commission 1999 Seeing is Believing: how the Audit Commission will carry out best value inspections in England. London: Audit Commission.

Audit Commission (2001) Changing Gear; Best Value annual statement 2001. London: Audit Commission

Bourdieu, P. 1984. Distinction: A Social Critique of the Judgement of Taste. London: Routledge and Kegan Paul.

Boyne, G. 1999 Introduction, processes, performance and Best Value in local government. Local Government Studies. 25: 1-25

Callon, M. 1998 Introduction. The embeddedness of economic markets in economics. In M. Callon Ed. The Laws of the Market. Oxford: Blackwell1-57

Du Gay 2000 In Praise of Bureaucracy. London: Sage

Gaffney, D. Pollock, A, Price, D and Shaoul. J. 1999 PFI in the NHS is there an economic case? British Medical Journal 319:116-119 ( 10 July )

Goffman, E. 1975 Frame Analysis. Harmondsworth: Penguin

Harper, R. 1998 Inside the IMF: An ethnography of documents, technology and organisational action. London and San Diego: Academic Press

Harper, R. 2000 The social organisation of the IMF's mission work. In M. Strathern, Ed. Audit Cultures: Anthropological studies in accountability, ethics and the academy. London: Routledge pp21-53

Herzfeld, M. 1992 The Social Production of Indifference: exploring the symbolic roots of western bureaucracy. Chicago: Chicago University Press

Horton, S and Farnham, D. 1999 New Labour and the Management of Public Service: Legacies, Impact and Prospects. In S. Horton and D. Farnham Eds. Public Management in Britain. Pp. 247-258 Palgrave; Basingstoke.

Keat, R. 1994 Scepticism, authority and the market. In R. Keat, N. Whiteley, N. Abercrombie Eds. The Authority of the Consumer London: Routledge

MacLennan, C. 1997 Democracy under the influence: Cost-Benefit analysis in the United States. In J. Carrier Ed. Meanings of the Market. Oxford: Berg

Martin, S. 2000 Implementing Best Value: local public services in transition Public Administration 78 (1): 209-227.

Martin, S. 2001 What have we created? The impact of inspection on local government. LRGRU Discussion Papers: www.cf.ac.uk/carbs/research/lrgru/publications.html

Martin, S. 2002 Best Value: New public management or new direction? In

McLaughlin, K, Osborne, S and Ferlie, E, Eds 2002 New Public Management: current trends and future prospects. Pp. 129-140 London: Routledge

Miller, D. 1998 A Theory of Virtualism. In Carrier, J. and Miller, D. Eds. Virtualism: A New Political Economy. Oxford: Berg

Miller, D. 2001a Driven societies. In D. Miller Ed. Car Cultures. Oxford Berg. Pp1-33

Miller, D. 2001b The Dialectics of Shopping. Chicago: University of Chicago Press.

Monbiot, G. 2000 Captive State, The Corporate Takeover of Britain. London: Pan Books.

Pollitt, C and Bouckaert, G. 2000 Public Management Reform. Oxford: Oxford University Press

Pollitt, C, Girre, X, Lonsdale, J., Mul, R. Summa, H. and Waerness, M. 1999 Performance or compliance?. Performance audit and public management in five countries. Oxford: Clarendon Press

Power, M. 1997 The Audit Society: Rituals of verification. Oxford: Oxford University Press

Raine, J. 2000 Five lessons from the first year of Best Value reviews. Local Governance 26:3 151-158

Reimer, S. 1999  "Getting by in time and space: fragmented work in local authorities" Economic Geography 75, 157 - 177.

Riles, A. 2001 The Network Inside Out. Ann Arbor: The University of Michigan Press

Rouse, J. 1999 Performance Management, Quality Management and Contracts. In S. Horton and D. Farnham Eds. Public Management in Britain. Pp. 76-93 Palgrave; Basingstoke.

Salaman, G.1997 Culturing Production. In P. du Gay Ed Production of Culture/Cultures of Production. London: Sage

Shore, C and Wright, S. 2000 Coercive accountability. In M. Strathern, Ed. Audit Cultures: Anthropological studies in accountability, ethics and the academy. London: Routledge pp 54-89

Snape, S. 2000 Three Years On: Reviewing Local Government . Local Governance 26:3 199-126,

Strathern, M. Ed 2000 Audit Cultures: Anthropological studies in accountability, ethics and the academy. London: Routledge

Strathern, M. 2000 Afterword. Accountabilty…..and ethnography. In M. Strathern, Ed. Audit Cultures: Anthropological studies in accountability, ethics and the academy. London: Routledge pp 279-304

Thrift, N. 2002 Performing Cultures in the New Economy In P. du Gay and M. Pryke Eds. Cultural Economy. London: Sage

Weir, S and Beetham, D. 1999 Political Power and Democratic Control in Britain. London: Routledge.


1) Transport and transcribing costs for this study were paid for by a grant from the ESRC. I would like to acknowledge the helpful comments of Paul du Gay, Nigel Thrift and an anonymous reader from the Journal.

2) I feel that anthropology tends to lose respect if it is seen as only capable of criticism and marginality, and as part of the larger report I will try to present my own vision of an alternative `light touch' regime. Similarly I have focused on negative features in order to highlight issues of virtualism, while the larger report will also cover positive features such as learning from best practice, for which also see Audit Commission 2001.


Following the traditions of ethnographic analysis a report is given here of Best Value, the UK government's inspection of Local Council services, based largely upon the categories and terms used by the inspectorate itself. It is argued that in each of the main areas of inspection the results appear to contradict the intentions set out for the inspection. This appears to be the result of processes of abstraction, which arise during the inspection. Such abstraction was located in the rise of jargon and bureaucracy as a result of the sheer scale of this audit, in the externalisation of critical factors during market testing, in the emphasis on the representation of consulting over the results of consultation and through the privileging of quantitative over qualitative information. Finally the phrase `owning the process' is analysed to show how the central ethos of serving the public is negated by these same forces. All these forms of abstraction are shown to occur despite rather than because of the intentions of the inspectorate. In the conclusion they are analysed as contradictions which arise as result of forces that are described by a theory of Virtualism.