Governmentality,power and organization

A marginal figure?

In 1988 Michel Foucault was such a marginal, unfamiliar figure in organization theory that Gibson Burrell (1998) was able to write an influential introduction to his work in Organization Studies. Over the next decade, however, Foucault’s work became a touchstone for debates about work, organization and identity, especially in the burgeoning field of ‘critical management studies’: Foucault had rapidly moved from the margins to the mainstream of organization studies (Carter 2008). In part, the adoption of Foucault was symptomatic of a widespread disenchantment with structuralism (Carter, McKinlay and Rowlinson 2002; Rowlinson and Carter 2002), a process that had been under way from the late 1970s onwards. Similarly, the exhaustion of the labour process tradition, inspired by Harry Braverman’s (1974) Labor and Monopoly Capitalism, meant that management strategies could no longer be ascribed solely to more or less adequate readings of market dynamics: new wave management’s fascination with Japanese production systems through to employee empowerment were irreducible to economic forces alone. Rather, ‘management’ was now to be invested with a social and symbolic reality that could not be dismissed as mere ideology. Equally, the dull categories of ‘consent’, ‘compliance’ and ‘resistance’ were inadequate to register the complex, confused and contradictory meanings workers ascribed to these new organizational realities. Foucault’s insistence that power was not so much about prohibition of certain actions as the attempted production of, for instance, new forms of identity was crucial to the consolidation of his influence in organization studies. This open-ended reading of Foucault was, however, far from the most common. Many writers from all points of the theoretical compass criticized Foucault for his dystopian imagery and his gloomy prognosis for any form of resistance. Much of this criticism was misplaced; only the most literal of readings would look for the existence of the perfect Panopticon as the only way of validating Foucault. This is no less true when one considers the question of subjectivity. In truth, when pressed, Foucault offered a thoroughly conventional account of the ways that individuals make their identity using their experience and the cultural symbols at hand. For Foucault, individuals actively construct their identity but within the constraints ‘imposed on him by his culture, his society and his social group’ (Foucault 1994: 11). This broad formulation would surely find favour with the most stringent Marxist or value-neutral Weberian.

Histories of the present

There is a deep irony in that Foucault is cited for his grand theories when his own scholarship is dense with minute data, close reading of texts and detailed allusions. Truth lies embedded in the ordinary details of everyday life: organization charts, the ways that doctors diagnose patients, or managers seek to transform themselves into leaders. Foucault opens one of his most explicitly methodological essays, ‘Nietzsche, Genealogy, History’, with the observation that ‘genealogy is grey, meticulous, and patiently documentary’ (Foucault 1977a: 139). Foucault was not immune to the strange allure of archival dust (Steedman 2001). Here, Foucault is not simply suggesting that we need to contextualize specific texts or practices in order to understand them. Rather, his point is that texts and practices are so loaded with various meanings, ambiguities and contradictions that any adequate reading must locate them in the conditions of their existence, the categories and practices that are their necessary preconditions. Equally, although he is adamant that the genealogical historian must be relentlessly empirical and attend to the specificity of events, by this he (Foucault 1977b: 154) means:

not a decision, a treaty, a reign, or a battle, but the reversal of a relationship of forces, the usurpation of power, the appropriation of a vocabulary turned against those who had once used it, a feeble domination that poisons itself as it grows lax, the entry of a masked “other”.

For Foucault, the need to dispel any hint of teleology or progressivism did not imply the need to read the past in its own terms.

From some of his earliest work, such as The Archaeology of Knowledge, Foucault (1969) commented that the introduction of a new discursive field necessarily has unintended and unpredictable consequences for other fields. The archaeological method was deployed in Foucault’s study of medical knowledge in The Birth of the Clinic. His main concern was to show how new paradigms of knowledge emerged and formed new ways of thinking about the world, rather than only specific illnesses or treatments. To this stress on the expansiveness of discourse and knowledge was added a concern with power and its historicity. While this was where Foucault acknowledged his debt to structuralism, he departed from its concern to isolate the ahistorical rules that govern discourses. Foucault followed the structuralist concern with uncovering the rules that govern discourse but these were conditions of possibility that were rooted in particular historical conjunctures. Again, during a famous exchange with Derrida over Foucault’s reading of Descartes’ ‘First Meditation’, Foucault (1972: 267) stresses the historicity of the text and that ‘discursive practices’ must not be reduced to ‘textual traces’. This notion of power, knowledge and practices was the central and consistent theme that was to run through all his work and politics. Here Foucault’s most important theoretical innovation is that power should not be thought of as singular, identifiable with an individual or even a single institution. Power, then, was not concentrated at clearly identifiable places, and was not susceptible to revolution, of being taken and wielded. Power is not, therefore, a possession or a commodity; instead modern power, Foucault insisted, was altogether more sinuous and insidious than this. The great secret of modern power was that it is hidden in plain sight, visible to all, its various strategies and techniques openly discussed by all. Of course, this is the defining paradox of modern power: at once, continuous, mundane, open, visible and liberating, while knowingly producing subjects whose freedom is monitored, measured and managed. To be sure, much of this paradox was submerged by the overpowering images of surveillance and discipline, especially Jeremy Bentham’s Panopticon prison. Less celebrated than Foucault’s use of Bentham, the dystopian prison designer in Discipline and Punish (1975), was his extensive discussion of sovereignty and classical liberalism in his College de France lectures in 1978–9. Where sovereignty was the dominant issue for absolutist monarchies grappling with reconciling their waning legitimacy in the face of increasingly broad-based democracies, this was not a central concern of the liberal state. Sovereignty cannot describe the disciplinary forms of authority that built nineteenth-century industrial capitalism (Foucault 1976). Rather, questions of sovereignty and legitimacy are displaced by those of success and failure. Where sovereignty placed the monarch at the centre of politics, liberal democracy now targeted the fate of the multitude, the population. The new target of the population could not be comprehended or dealt with by the state alone. Governmentality, Foucault concluded, was a much more expansive – and expanding – notion that involved all manner of organisations, routines and strategies which make knowledge powerful and power knowledgeable.

It is easy to dispose of the hoary canard that Foucault was a sloppy historian, prepared to sacrifice accuracy for style or theoretical purchase. Indeed, Foucault himself acknowledged that Discipline and Punish was a history of a way of thinking – and acting – about criminality, institutions and reform and not a social history of the prison. Bentham’s dream of the perfect prison would not be the starting point for a social history of the prison. Nor did he regard this as an admission of a failing, simply to register a quite different theoretical and historical purpose. To start with Bentham was not to write an idealized history of penology as a sequence of more informed and insightful theoretical understandings and institutional practices. By the same token, to regard his rendering of Bentham as only a history of ideas, insisted Foucault (1991: 81), would require ‘a very impoverished notion of the real’:

It is absolutely true that criminals stubbornly resisted the new disciplinary mechanisms in the prison; it is absolutely correct that the actual functioning of the prisons, in the inherited buildings where they were established and with the governors and guards who administered them, was a witches’ brew compared to the beautiful Benthamite machine. But if the prisons were seen to have failed, if criminals were perceived as incorrigible, and a whole new criminal “race” emerged into the field of vision of public opinion and “justice”, if the resistance of the prisoners and the pattern of recidivism took the forms we know they did, it’s precisely because this type of programming didn’t just remain a utopia in heads of a few projectors.

Here Foucault stands against teleology in that the past cannot be read as the precursor to the present. Of course, this starting point is neither novel nor unique to Foucault. It is this manoeuvre that allows him to juxtapose stark images of different forms of power and knowledge, from the ship of fools, through the execution of Damiens, to the minute detail of a life in the day of a young army recruit. In each case he is not holding up a moment from the past for our entertainment or so that we might reassure ourselves about how much improved we are. Rather, each of these events exemplifies a particular way of understanding power and its mechanics that is – or would be – ridiculous, unthinkable or undoable at any other time: no moral or intellectual superiority is implied. There is one further point that Foucault makes here that is particularly relevant to business history. That is, the futility of searching for origins, a founding myth, from which the genesis of the present can be traced, or being satisfied with uncovering the motives lying behind a particular decision. Equally pointless is the search for deep structural causes of shifts towards similar ways of thinking about other institutions. Foucault, similarly, does not fail to demonstrate the dissemination of institutional innovations across diverse institutions. It is no less apt to reverse the logic of this criticism: it would be amazing if disciplinary techniques did not seep between institutions. Bentham, to take Foucault’s most famous example, was a consummate self-publicist. His appeal to Foucault was, then, not just the Panopticon, but the zeal with which he propounded its logic to all manner of buildings, institutions and constitutions. After all, Bentham was not just a designer of prisons and battery hens but the author of legal codes and liberal democratic constitutions.

Much of Foucault’s most influential work centres on the impact of the Enlightenment and the rise of classical liberalism as a political philosophy. Foucault forgets that liberalism was a much broader intellectual and political project than can be captured with reference solely to Bentham at his most utilitarian. Indeed, even Bentham acknowledged that rational instrumentalism was, as other liberal philosophers pointed out, the counterpoint to ‘sympathy’. The distinction between Bentham and, for example, David Hume and Adam Smith, was that ‘sympathy’, the individual’s awareness of the needs of others, was an impossible basis from which to design rational institutional structures or to make political, self-disinterested choices (Otteson 2002). Contrary to Bentham – and Foucault – the enfranchised citizen and the enterprising self do not confront the state and the market directly. The enterprising self’s relation to the market is mediated by organizations. One task, then, is to reread liberalism in the broadest sense in order to go beyond Foucault. Now, it is not just that ‘pure’ neoliberalism is unsustainable in practice, although that is certainly the case. Neoliberal assumptions and precepts inform practices and metrics, but rarely do so alone: ‘sympathy’ in terms of procedural and substantive justice almost always intrudes into what is measured, rewarded, deplored and punished.

Governmentality

The concept of governmentality was not coined by Foucault nor, one could argue, was he to make the most significant developments in its use. Governmentality was a term first used by Roland Barthes in 1957 to link processes of government with efficiency, an association that he regarded as completely unremarkable. Barthes went no further in refining governmentality. Foucault and Barthes were vital members of the literary, philosophical and political salon that grouped around Tel Quel, a new space indifferent or hostile to the constrictions of academic discipline and literary genre. Although he was almost certainly aware of Barthes’ concept, Foucault did not use it until his lecture series on biopolitics in 1978–9 (Foucault 2008). Foucault’s experience of the Tel Quel group was to prove a seminal moment in his intellectual, literary and political development. Intellectually, he was exposed to ideas that were to take him beyond his early structuralism. For example, Barthes argued that texts should be read not in terms of their author’s singularity and their intention, but in the ways that signs are disseminated and infinitely dispersed. There are clear echoes of Barthes’ reasoning in Foucault’s notion of power as not univocal, unambiguous and located in a specific place but as ubiquitous and infinitely productive. Indeed, many of Foucault’s most important insights came from his reinterpretation of others. Foucault inverted Gaston Bachelard’s (1994) meditation on the meaning of domestic space. Where Bachelard reflected on the ways that the simple geometries of the home belied the infinite complexity of its interior life, Foucault rather looked at how the equally simple architecture of the Panopticon transformed chaos into simplicity and order. In Foucault’s hands, Bachelard’s intimacy of the home becomes the impersonality of the prison. Phenomenology is rendered redundant in a building form that is predicated on the visibility of its purpose and mechanisms.

Tel Quel confirmed Foucault’s move to understanding philosophy as literature and writing as a practice, a departure which then expanded his capacity to read the diagnoses of institutional reformers and the material practices they advocated as texts. Foucault’s involvement in Tel Quel also confirmed his shift to writing philosophy as a form of literature, not merely as a technical genre (Ffrench 1995; Marx-Scouras 1996). Politically, Tel Quel’s move to Maoist politics after 1968 exposed Foucault to a new type of politicized cultural struggle. His involvement in the prison reform movement or, rather, using prisoners’ reports to question the legitimacy of penology itself, marked the beginning of his career as an ‘engaged intellectual’. Where Marx followed ‘Monsieur Le Capital’ into his secret hiding place, the factory, Foucault sought to penetrate the no less mysterious, closed world of the French prison. Foucault delayed the publication of Discipline and Punish to avoid any question about the ethics of traversing the – now extremely porous – boundary between his political and philosophical practices (Starr 1995; Wolin 2010). Of course, Foucault intended ‘governmentality’ to be a political concept, but was aware that it had an applicability beyond the state. Governmentalist projects did not begin and end with the state. Foucault’s use of governmentality was a sharp disavowal of state-centred theory but it also provided glimpses of a new politics that focused on life as a political object (Gordon 1987).

Governmentality was the conceptual bridge between the disciplinary moment of Discipline and Punish, with its relentless stress on bleak individualization, and the notions of the individual’s capacity to make themselves with, through and against these practices and institutions, themes that dominated his final work on the history of sexuality (McCarthy 1994). Governmentality involved a double manoeuvre: at once to construct a population and the possibility of certain individualities. In the interview with a ‘Collective Historian’, ‘Questions of Method’, Foucault (1991) was quick to distance himself from any suggestion of theoretical or methodological similarities between his work and that of Max Weber. Contra Weber, Foucault was uninterested in rationalization as a historical theme and saw no merit in ideal types, although it may be that he was thinking of them as benchmarks against which reality could be judged rather than as heuristic categories. It was only towards the end of his life that Foucault acknowledged the close affinity between his work and that of Weber. Foucault’s reading of Weber began in 1978 and, Arpad Szakolczai (2000; Clegg 1994) goes so far as to suggest that his return to the term ‘governmentality’ was an acknowledgement of their overlapping theoretical and historical work. This was the moment that ‘governmentality’ moved from being a term of uncertain provenance towards being a concept, both abstract and analytical. During these years Foucault reflected that he had paid too much attention to systems of domination and that he intended to redress this imbalance by researching how individuals increasingly came to govern themselves (Bonnafous-Boucher 2001). This was a risky conceptual shift and there is a certain hesitancy evident in Foucault’s lectures and essays (Dupont 2010; Paras 2006)

The liberal state battles to reconcile an open-ended concern with the population’s well-being with a determination that the state should be frugal, constantly seeking to curtail its activities in the interests of cost and liberty. The tension at the centre of governmental rationality was captured by Bentham’s couplet ‘government-interest’. Foucault stressed that utilitarian liberalism was obsessed with delimiting public authorities. After all, Bentham first conceived the Panopticon not for prisoners, ‘to grind rogues honest’, but to allow citizens to observe continuously and control civil servants, particularly to eliminate favouritism. Panoptic control was, therefore, about efficiency, transparency and legitimacy as much as surveillance. Increasingly, from the late eighteenth century, the state becomes responsible for increasing the population’s welfare. The state’s ability to protect and increase social welfare becomes the acid test for any administration in terms of achievement and means. The growth in the state’s administrative knowledge of its population – census, mortality, education, productivity – becomes both a source of knowledge about welfare and a measure of the efficacy of specific state interventions and governmentality in general. In turn, the collection, comparison and interrogation of all kinds of social statistics became not just an anonymous administrative matter but central to all political debate. For Foucault, the liberal state carries an impossible burden. Every liberal state is a failing state since it can only fail to satisfy ever-rising expectations of economic growth, security and well-being and individual freedom. All institutions, including corporations, are subject to the same dynamics as the state. The same logic that determines the state’s search to balance ever-greater security and welfare with efficiency and transparency applies to the corporation. Despite the inevitable failure of systems of governmentality, there is no sense in Foucault – or, indeed, his followers – that such moments of collapse are of theoretical or historical importance. Foucault’s discussions of the rise of systems of governmentality are not paralleled by any discussion of the ways that they fail. Ironically, then, there is an air of inevitability in most accounts of governmentality: that which is appears to be that which had to be, a teleological fatalism that is oblivious to the very possibility of contingency.

Foucault provided the first sketch of a concept of governmentality. However, the most sustained theoretical development and innovative empirical research has been undertaken by a group of scholars clustered around Peter Miller and Nikolas Rose – the London governmentalists (McKinlay et al. 2010). The structural Marxism of Louis Althusser was vital for the London governmentalists in that it was – like Foucault – based on a sophisticated anti-humanism. Equally, Althusser’s sharp break with economic determinism also allowed not just breathing space for culture but acknowledged theoretical practice as a form of political struggle. This opened up space for non-Marxist Tel Quelians and, in much the same way, the London governmentalists developed their own for intellectual space, complete with an in-house journal during their formation. Politically, through the mid-1970s, the British Communist Party was increasingly divided between placing the workplace and the industrial working class at the centre of its theory and strategy, and an emerging sense that this was no longer viable. The alternative lay in some kind of wider cultural politics that stretched well beyond the traditional skilled, male working class (Andrews 2004). In this context, the ‘London School’ made a critical intervention: that governmentality – as discourse and practice – cannot be restricted to the state (Rose and Miller 1992). Here they echo Barthes and Foucault: in Barthes’ notion that texts cannot be read as coherent and self-enclosed but involving signs that are disseminated and infinitely dispersed; and of Foucault’s notion of modern power not as univocal, unambiguous and located in a specific place. Ways of thinking about the individual or the organization are not – cannot – be confined to specific institutions for too long. Ideas of the knowing, sovereign consumer are echoed in the active citizen and the empowered employee. More than this, the practices used to mobilize, measure and manage each of these figures have strong family similarities of incentives, self-direction and an absence – often the withdrawal – of centralized top-down control. Despite the major theoretical insights of the London governmentalists, there remains a tendency to analyse discourses and apply this to practices. Foucault (2008) is clear that he is studying concepts, not practices in action. His unusual clarity on this point suggests that the distinction is important and that we should be careful not to confuse one with the other, far less to elide or ignore this distinction. That is, Foucault’s lectures were a discussion of governmentality, the ways governing is conceptualized, rather than governing, the practices of rule.

The task for governmentalists is to go beyond the manifestoes and manuals that construct new images of individuals, organizations and societies. That is, the task is to draw together the politics that inform the making of particular governmentalist regimes with the ‘witches’ brew’ of everyday practices. There are a small but growing number of studies in sociology and anthropology that have begun this process. Where Foucault’s notion of discipline referred to particular institutions targeting specific groups for rehabilitation or improvement, governmentality involves improving entire populations. This necessarily requires a centralization of key measures and procedures to indirectly impact on individual behaviours. A second necessary condition is that the problem is defined in a way that makes possible or, rather, plausible, that it is amenable to improvement by expert interventions. This process of rendering an issue technical, as Tania Murray Li (2207: 7) puts it, always involves a politically deadening circularity:

. . . the bounding and characterization of an “intelligible field” appropriate for intervention anticipates the kinds of intervention that experts have to offer. The identification of a problem is intimately linked to the availability of a solution. They co-merge within a governmental assemblage in which certain sorts of diagnoses, prescriptions, and techniques are available to the expert who is properly trained. Conversely, the practice of “rendering technical” confirms expertise and constitutes the boundary between those who are positioned as trustees, with the capacity to diagnose deficiencies in others, and those who are subject to expert direction. It is a boundary that has to be maintained and that can be challenged.

This process of defining a problem in terms of what works also makes the issue one of administration, not politics. By such processes, matters as diverse as health, poverty and productivity are depoliticized, the proper province of experts not active, questioning citizens. And this sets the limit to governmentalist projects: to the institution’s inherent frugality and uncertainty over its own legitimacy is added not just resistance but the contradictions and perverse outcomes generated by any policy intervention. The central ambition of any governmentalist policy triggers the definition of a set of intermediate objects, each of which has to be imagined, organized and managed. Such interventions may originally be imagined as a grid in which there are functionally related, predictable chains of effects. On the ground, however, they are certainly compromised, negotiated and improvised to a greater or lesser degree. Governmentality research has focused on those technologies that allow people to be known, to know themselves, and the social world to be acted upon. Of particular importance are applied psychology, accounting and marketing because they are forms of knowledge and techniques that most intimately target the individual while also constructing particular populations.

Foucault’s notion of governmentality rests on the proposition that liberal government involves maintaining or extending individual freedom while being held to account for the welfare of the population. For Foucault, governmentalist projects are always indirect, even oblique: clear targets may be set but delivery requires the willing, active participation of citizens, employees or consumers. What Foucault refers to as the ‘conduct of conduct’ always entails a certain lightness of touch. Again, this insight was far from novel. The notion of detailed local control exercised by a distant headquarters through general rules and standardized procedures was a staple of ‘classical’ organizational sociology. Alvin Gouldner’s (1955) seminal study of a gypsum mining company, Patterns of Industrial Bureaucracy, notes that snap inspection complemented general rules that applied to all plants: ‘In part, then, the existence of general rules was a necessary adjunct to a “spot check” system; they facilitated “control from a distance” by those in higher and more remote reaches of the organisation’ (Gouldner 1955: 167). Indeed, in an important sense, Gouldner did not simply anticipate the insights of Foucault and the London governmentalitsts, but went beyond the formal rules and procedures that constitute the limit of governmentalist research. The power and authority of general rules, Gouldner (1955: 10–11) argued, also derived from their public nature and local negotiated patterns of order:

because the rulers were publicly known, an “enemy” could be used to control and “ally”. In sum, the public character of the rules enabled deviance to be detected by the out-group. This enlarged the information channels open to the heads of the in-group, in turn, enabling them to keep their own junior officers in line.

Governmentalist research need not remain at the level of the schemes of system architects but look at how broad concepts are translated into practices on the ground. This is an invitation to ethnographic and historical research that remains largely uncharted territory.

Pierre Labardin’s essay on the origins of a division of labour in accounting examines the manuals written over the course of almost two centuries from the late eighteenth century to the eve of World War II. Before industrialization, accounting services were not subdivided and were delivered on the basis of personal trust rather than management, organization and profession. By the mid-nineteenth century, however, accounting manuals were cautiously advocating some degree of task specialization. Within a few decades, there was a clear consensus in the practitioner literature that accounting labour should be subdivided and organized hierarchically with a clear separation between design and production activities. It is difficult, Labardin observes, to attribute the formation of the accounting labour process to a search for greater efficiency, profit or control. Quite the reverse: the accounting practitioner literature had a dynamic that was quite independent of – and irreducible to – the economic imperatives of specific firms. Indeed, Labardin suggests that office rationalization and mechanization in the first decades of the twentieth century were attempts to stem spiralling overhead costs in French business. Certainly, the practitioner literature was never simply a description of how accounting processes and labour were organized, even where the authors were senior figures in large-scale organizations. Nor was this a literature given to wishful thinking or wholly abstract. Rather, we should regard the practitioner literature as performative in that it sought to describe organizations not yet quite formed, using a language that moved from hesitant suggestion to a professional self-assurance and clear instruction.

The role of accounting numbers in the making of new organizational forms is also the subject of Alan McKinlay and James Wilson’s essay on the development of the moving assembly line in Ford between 1908 and 1921. Here again we encounter the impossibility of sustaining an argument that is wholly reliant upon an economic rationale. Ford’s introduction of the moving assembly line was a seminal moment in the making of modern mass-production. However, this was a process that cannot be easily explained in terms of market signals or a search for management control. The market for mass-produced utility vehicles simply did not exist before the Model T; while the assembly line was devised through a series of trial-and-error experiments in industrial engineering, not derived from an imaginary of subjugated labour. Initially, Ford compiled a wide range of accounting numbers of great sophistication and ingenuity. Indeed, innovative information flows were central to Ford’s articulation of demand, production and logistics. Accounting measures, particularly the balance of direct and indirect labour hours and costs, were an integral part of the way Ford management imagined and then understood the nature of the untried mass-production process. Ford’s accounting was not restricted to cars, costs and productivity, however. Ford’s Sociological Department also collected detailed records about individual employees, their family lives and morals, all of which was compiled and analysed centrally. This was, then, a form of proto-governmentality, a new way of imagining the individual and governing the population numerically. In short, Ford had the capacity to combine factory accounting and employee records to construct a new form of sociology of mass-production. By combining these information flows Ford could have, for instance, examined the interaction of nationality, length of service and productivity. Instead, just as this new possibility of a combined form of moral and management accounting was made practical, so Ford violently abandoned its innovative analytical capacities in favour of a regime of direct supervisory control and direct, immediate alterations to the assembly line based on crude data and physical observation. Ford’s abrupt rejection of data as a way of imagining as well as controlling production and people was a rejection of an emerging system of governmentality. Governmentality – that combination of signs, systems and practices – always seems to be fully formed and impregnable. The Ford example suggests that we should be more circumspect and understand systems as governmentality as vulnerable, especially when untried and reversible, although seldom with the violence of Ford in 1921.

Economics remains more or less untouched by Foucaultian scholars. Yet in his 1978–9 lectures, Foucault considered the nature of neoliberalism at some length. For Foucault, Gary Becker’s notion of ‘human capital’ signalled a fundamental break in the way that capital and labour were conceptualized. The radical individualism of neoliberalism regarded labour as a kind of capital and the worker as a new kind of entrepreneur of the self. Classical economics, from Smith through Marx to Keynes, studied the relationship between capital and labour in terms of its societal utility. Neoclassical economics, by contrast, transforms homo economicus into an entrepreneur of the self. That is, the worker/entrepreneur has his own stack of human capital in which to invest and from which to generate a return: wages and salaries become analogous to profits (Vigo de Lima 2010). The firm no longer has even the most abstract responsibility for the long-run employment of staff. At most, the socially responsible firm accepts the much more limited obligation to ensure the employability of its staff. This is the subject of Eric Pezet’s essay on the universalization of the career in France after 1968. Faced with the need to restore industrial order, the French state looked for ways of tying the individual to the enterprise for the long run. This entailed the construction of elaborate internal labour markets in which tasks and seniority systems were designed by a combination of job evaluation and industrial psychology. Scientific management, job evaluation and industrial psychology represented a complex settlement between capital, labour and the state, a settlement that was rooted in its objectivity or at least its political neutrality. Important elements of the new career system owed much to the industrial psychology pioneered by the Lyons truck-maker, Berliet, after the Liberation. Berliet’s discredited management – a family under investigation of collaboration – and a Communist party trying to offset a failing system of workers’ control turned to industrial psychology as a way of preserving hierarchy while increasing employees’ role in operational decision-making. Just as the career was a vital part of Berliet’s reconciliation of management prerogative and employee involvement, so it offered an ‘objective’ route to industrial order after 1968. For the state, making the career universal was the foundation not just for an orderly society but also for mass consumption based on extended credit. However, career management required specific organizational techniques and measurement that could not simply be read from the ‘capital logic of the state’ (Donzelot 1991: 171). The appeal for the unions was clear. The career implied an automatic, neutral mechanism for increasing earnings over the long run. Here Foucault’s concept of governmentality goes beyond political economy in that the state and employers independently by mutually constitute a common, novel object, the career. Classification devices such as job evaluation and the career are vital in ordering industrial populations and shaping the life chances of individuals, but have a logic and temporality that is independent of the strategies of firms, unions or the state. The career is the secularized calling of modern capitalism, a way of focusing the individual on the task, a way of constructing a durable bureaucratic morality (Sayer 1991). Here, of course, Foucault comes very close to Weber (Clegg, Courpasson and Phillips 2006; McKinlay forthcoming). The career requires the organization to establish systems that measure tasks, establish job ladders, produce records that monitor individual performance and progress, and evaluate the trajectory of the employee with the population over the long run. Data cannot be private and ephemeral but must be aggregated, permanent and analysable, which entails a significant and continuing commitment to storage and information-processing technologies. Paradoxically, the availability of data that established populations and norms can also be used by its subjects to monitor that procedural justice and individual liberty are protected.

Foucault’s engagement with philosophical questions was of pivotal importance to his scholarship. Indeed, it is arguably the feature of Foucault’s work that has been most enthusiastically embraced by the critical management studies community. The trade route from Foucault into philosophy has undoubtedly inspired a generation of critical management scholars and, in turn, has led to the production of important scholarship. In their analysis of Foucault’s reading of Descartes, Nick Butler and Stephen Dunne contribute to this body of work. They rehearse the charge sheet, compiled by critical and post-structural scholars, against Descartes. His mind–body dualism has been invoked to explain why contemporary organizations possess phallocentric, misogynistic and patriarchal tendencies. Having stated the case for the prosecution, Butler and Dunne proceed to mount a defence of Descartes, in which they debunk popular representations of his work as little more than a fairly vulgar caricature. Their broader point is that the organization studies rendering of Descartes does not stand up to close scrutiny. Accordingly, they advocate, more generally, that organization theorists should take an altogether more careful and nuanced approach in their engagement with philosophy.

Butler and Dunne sprinkle scepticism on the widely held view, at least in organization studies circles, that a Cartesian dualism is something that needs to be overcome if one is to conduct a credible critique. Instead, they take a counter-position: they insist that in relation to Descartes it is in fact the dualism between reason and madness that has the most far-reaching promise for a critical analysis of organizations. Their wider point is this: Descartes belonged to a broader discursive formation that comprised an assemblage of institutions connected to the production of particular forms of knowledge. It is this assemblage that made Descartes’ work possible. For Butler and Dunne the appeal of Foucault’s project is that it enables the identification of the historical conditions that allowed for the formation of Cartesian thought. The appropriation of philosophy by organization studies has yielded great insights into how and why people organize. Butler and Dunne, following Foucault, call for a far more historical approach to understand philosophy and the context in which that philosophy was made possible.

Conclusion

The danger is that we pay such close attention to periodizing Foucault’s work – and allocating relative values to the early or late Foucault – that we ignore the deep continuities that underpin it. Alternatively, and more usefully, one can think of the changing emphases of Foucault’s research not so much as ‘periods’ in which each surpasses and rejects that which has gone before but as layers of a complex whole. This alternative allows for developments in – or additions to – a corpus of work, which maintains a sense of overall cohesion that is all too easily lost by a ‘periodized’ Foucault.

Identity rather than institutions has become the central concern of organization studies. Where a broadly Weberian concern with measuring organizational structures dominated organization studies for almost four decades after 1945, this has been displaced by a concern for interpersonal processes and identity formation. Individuals are permitted to construct reports on themselves, and are answerable not just to managers but also to themselves in terms of their personal effectiveness and how they are remaking themselves to achieve organizational goals.

Equally, there is little sense of the ways in which systems of governmentality gain coherence, widen or narrow their practice or adapt as the object itself alters. There is little sense of systems of governmentality as rough sketches, arbitrary and tentative; they are always fully formed and perfectly coherent. It is difficult to avoid a nagging doubt that a slippery, teleological functionalism is never too far from the surface: that which becomes is that which had to emerge.

Where Foucault’s influence on organization studies has been sustained and profound, his work has had little or no impact on business history. Yet Foucault’s work poses at least three major challenges for business history. First, his work involves a complete rejection of arguments that rely entirely on economic logic, no matter how subtly qualified by constraints or bounded rationalities. Foucault insists on the independence and the independent efficacy of the social. Second, the main institutional focus of business history remains the firm. That is, the evolution of competitive strategy by the firm and how – structurally – this is operationalized and to what effect. A major thrust of British business history has been the limitations of strategy at firm level and the many ways that this has proved inadequate in the long run. The firm has become the metaphorical hair shirt, the marker of long-run economic decline. By contrast, Foucault asks us to consider the ideas, practices and languages used by managers to think about organization, competition and markets. This move means that the object of business history ceases to be the firm and becomes, for instance, how ‘the worker’, ‘the consumer’ and ‘the organization’ are conceptualized and put into practice. A Foucauldian business history would, then, displace the firm from centre stage. One difficulty with this is that it risks underplaying the privileged legal, economic and social position of ‘the firm’. After all, the firm is the institution in which we are managed, which produces goods and services, and is itself subject to regulation and negotiation. Reconciling the specificity – and historicity – of particular firms with these wider languages of governmentality remains a major challenge for Foucauldians. Third, governmentality has struck its deepest roots in critical accounting. The consensus in critical accounting is that accounting numbers do not simply register, reveal or clarify economic ‘realities’ that would otherwise remain obscure. Rather, accounting procedures construct new objects to be known, problematized and managed. The gains made through this approach have been significant and suggest that business historians should also reject any notion of accounting as a neutral, objective tally of performance.