Sorting people in and out: The plasticity of the categories of employability, work capacity and disability as technologies of government
Introduction: Employability and disability as floating signifiers
The trend in contemporary Western labour markets is one towards enhanced emphasis on competition, mobility, flexibility, and continuous learning. Increasingly, people are expected to assume individual responsibility for the development of their professional portfolios, their capacities, and for their career trajectories. In current labour market policy, there is an emphasis on the self-responsibilization of the individual as a recipe for achieving a greater degree of dynamism in labour markets (e.g. Clarke, 2005).
In this context, employability has gained salience as a policy category (e.g. Garsten and Jacobsson, 2004). Despite its wide-ranging usage, employability remains a contested concept in terms of its use in both theory and policy (e.g. McQuaid and Lindsay, 2005; Peck and Theodore, 2000). Some have seen the discourse on employability as a form of lofty ‘planetspeak discourse’ with no structural roots or social location, free-floating (Fejes, 2010: 90) and open to be translated in various ways in different local contexts (Garsten and Jacobsson, 2004). Nonetheless, the ‘employable individual’ is an influential normative category in current labour market discourses, and it is ‘drenched’ with assumptions about what it takes to be attractive in changeable labour markets (e.g. Garsten and Jacobsson, 2004; Fejes, 2010). Professional skills and work experience are no longer considered to be enough; ‘softer’ social skills, flexibility, and adaptability, are also required, as well as the capacity to market and to sell oneself. Self-help discourses are evoked (Rimke, 2000; Sharone, 2007); and job coaches, career counsellors (Fogde, 2009) and other ‘experts on subjectivity’ (Rose, 1989) are engaged in seeking to transform unemployed subjects into ‘entrepreneurial’ and ‘self-reliant’ selves (Thedvall, 2004). The individual is expected to actively assume responsibility for her own employability by being prepared to work on her ‘self’ in order to improve her attractiveness to the labour market – often without knowing exactly what is expected of her. Ultimately, employability turns out to be a ‘floating signifier’ in Lévi-Strauss’ sense of the term (1950), i.e. a term in itself void of meaning, thus apt to carry diverse definitions. As a floating signifier, employability is open to imply a variety of demands and expectations placed on individuals and, as Cremin (2010) suggests, always bears the risk of the person never being ‘employable enough’.
This begs the crucial question: What happens to people who cannot meet all those expectations? As we will show, for some, the path to employment and employability goes through a process of detecting and coding of disability at the Public Employment Service (PES). By coding unemployed individuals as ‘disabled’, the PES can devise special assistance for them, such as subsidized or sheltered employment or other types of special resources for a period. Disability coding can thus be a means to access public resources and to enhance one’s employability. Being coded as disabled is thus one way of ‘becoming employable enough’ (cf. Cremin, 2010).
The aim of this article, then, is to analyse the processes of evaluating work capacity for marginally employable people as part of the Employability Rehabilitation Programme, with special focus on examination, judgement, and functional impediment codification. By looking more closely into the actual classification procedures as technologies of government (Rose, 1999), we aim to show how the administrative labels work to mediate and distribute employability. The labels that are ‘offered’ and opened up for those who undergo rehabilitation are positioned as voluntary. Yet, it is by accepting and agreeing on being coded as disabled that one may receive the benefits and resources available. While the codes may enhance the employability of the individual, he or she is ‘formatted’ in a process in which the desirable criteria for employability are set according to a specific organizational intervention programme. Employability thus emerges as a result of a template for what is considered acceptable and desirable by the organization, i.e. the PES.
As our analysis shows, employability is constructed and defined by classificatory procedures that involve a number of interlinked actors, organizational typologies, and routines. Whilst the intention of the caseworkers involved is to make disabled people employable through the operations of interventions tailored to the needs of the individual, the primary implication of these is that they contribute to the functioning of a bureaucratic process of classification and governance. We show how the classification practices work as ‘technologies of government’ (Rose, 1999), serving the organizational need to make individuals ‘readable’ and ‘process-able’. Following Scott (1998) and developed further later on, we conceptualize the bureaucratic assessment practices as techniques of ‘legibility’ in that they allow for an organizational ‘reading’ and processing of individuals.
Moreover, the article shows how employability, work capacity, and disability tend to become floating signifiers in this evaluation process. The categories are themselves plastic, relationally defined, and interdependent. To be non-employable becomes a disability and conversely, to be disabled can make one employable. Moreover, the categories demand versatile individuals who can adjust to and ‘qualify’ for their distinctions and demands.
In the following, we will first locate our study in the larger policy context before describing the study conducted in more detail. We then outline our theoretical perspective, which will guide us in the empirical sections that follow. The concluding section summarizes the main findings and discusses some implications.
Targeted policy: The identification of employability gaps and dis-ability
Social citizenship, while defined in legal regulations, is implemented by street-level bureaucrats and caseworkers as part of their daily routines for client encounters; it is in the ‘interface zone’ between the individuals and the state (cf. Martin, 1997) that social citizenship is ultimately defined. It is also here that clients’ subject positions are assigned and subjectivities are negotiated and shaped (e.g. Korteweg, 2003), and it is therefore a place where the calibration of normalization, normativity, moralization, and occasionally the compulsion that constitutes policy implementation takes place. This interface shifts according to ideological changes, organizational transformations and policy adjustments. With the present policy focus on employability, the interface between state and individual has shifted in the direction of individual responsibility and activation. The discourse on employability is individualizing, and in part de-politicizing, in the sense that explanations to unemployment are sought in the supply of labour, more precisely in the qualities and characters of individuals, rather than in the supply of jobs as determined by macro-economic policy (Garsten and Jacobsson, 2004; Sharone, 2007).
The individualist emphasis goes hand in hand with the shift in administrative techniques towards advanced liberal forms of government (e.g. Miller and Rose, 2008), which involves not only a new diagram of control exercised by authorities and institutions, but a constant and never-ending modulation of conduct (Rose, 1999). It is exercised, for example, through the practices of continuous training, lifelong learning, perpetual assessment, continual incitement to improve oneself, constant monitoring of health, and never-ending risk management. Correspondingly, there is a growing market for catering to the needs of ‘responsibilized’ individuals, and to individualized labour market provisions. This may involve coaching, CV-writing, presentation skills, and even dress code management (e.g. Fogde, 2009). A growing cadre of specialists, such as coaches, rehab experts, educational experts, medical experts, personal trainers, therapists, and so on, supply their services to individuals and help out in their fashioning of selves along with current labour market ideals (e.g. Rimke, 2000).
The governmentality perspective, inspired by Foucault (e.g. 1994) and used in increasingly many studies of the encounter between welfare bureaucracies and clients (e.g. Dean, 1995; Østergaard Møller and Stone, 2012), alerts our attention to the fact that these subtle forms of government readily turn into self-government. Indeed, following Foucault, it is precisely the subjectivity of the individual that is the ultimate object of intervention. Hence, the individual is simultaneously an object for intervention and an active subject, and the intervention often aims at infusing agency in the individual. Nevertheless, it is a pre-given, and scripted – thus institutionally shaped – form of agency (e.g. Korteweg, 2003; Mik-Meyer, 2006). The ‘technologies of self’, i.e. the devices used to encourage the individual to evaluate, monitor and improve the self that are mobilized in particular ways to shape how individuals perceive and conduct themselves (Foucault, 1994), are also normalizing in their effects. Practices such as individual action plans for the unemployed, career advice, self-evaluation sheets or other types of evaluation instruments help in establishing normalcy standards. Evaluations do not just objectively measure existing qualities – they also signal what qualities are desirable, and thus help in shaping the subjectivity of the individual being evaluated.
It is against this backdrop, we suggest, that we should understand the increasing number of people coded as disabled in a country like Sweden. In many countries there is an emergent division between active citizens (capable of, and expected to, managing their own risks) and targeted populations (disadvantaged groups, or people ‘at risk’) who require various types of interventions in the management of risk (Dean, 2008: 167; Caswell et al., 2010). Also in Sweden, recent developments have accentuated the increasing dualization of labour market policy; a categorization of the unemployed into ‘normal job-seekers’, which can be offered job counselling or self-service activities, and ‘disadvantaged’ groups, which need special measures (Peralta Prieto, 2006).
This dualization is reinforced by a dramatic reduction of investment in active labour market policy spending (Bengtsson and Berglund, 2012), whereby Sweden in the year 2010 was below the average of all OECD countries. While Sweden has a long tradition of being an active welfare state in the sense of offering the unemployed ‘active’ measures such as placement support, employment training, retraining opportunities, and mobility support instead of just ‘passive’ measures in the form of financial subsidies (cash assistance), the cornerstone of Swedish labour market policy, the Work Strategy, has over the last decades been translated into a sharpening of the qualifications for receiving unemployment benefits and social insurance. Activation in the form of increased conditionality and re-commodification has replaced investments in up-skilling and training (Bengtsson and Berglund, 2012). Labour market policy programmes are reduced in favour of employment on the regular labour market and job-matching measures have been emphasized in the national labour market’s policy directives. The main assignment of the PES is now to match jobseekers against job openings in the labour market. The more costly training measures have been replaced with the less expensive job counselling services. The main support measures of the PES are now targeted to jobseekers who are no longer covered by unemployment insurance or those judged to have a special need, such as those deemed disabled.
At the same time, the political assignment of the PES in Sweden to rehabilitate and activate jobseekers with disabilities has been more clearly articulated. To counteract social exclusion, resources have been placed on offering special jobs and traineeships to prevent and rehabilitate illness and disability. The aim is to provide the same opportunities to people with disabilities as to regular jobseekers to partake in the labour market (Prop. 2007/08:1). In other words, while the tool-box of measures available for ‘ordinary’ unemployed has shrunk considerably, a disability coding opens up for a wide spectrum of measures and support for people deemed in special need (see also Holmqvist, 2009). It is therefore an attractive option both for caseworkers and for individuals with a long history of unemployment or presently finding themselves far from the regular labour market.
In Sweden, subsidized employment is by far the most common assistance given to people classified as disabled. A market for disabled individuals is created by ‘manipulating’ the pricing of productivity. While this in itself is not a novelty, the emphasis on ability, competence and marketability of oneself in contemporary labour markets means that requirements for employability are raised to higher levels than ever before. As a result, an increasing number of people risk ending up as less employable or even ‘disabled’ (see also Holmqvist, 2009). This is reflected in the number of people who are coded and registered as disabled at the PES Service in Sweden, which has increased dramatically in recent decades. In 1992, 10 per cent of all registered unemployed at PES Sweden were coded as disabled while in 2011 the corresponding number was 25.3 per cent. In real numbers more than 171,000 people were coded as disabled in an average month in 2011. The ‘psychological disability’ and ‘learning disability’ codes have seen the most dramatic rise in numbers, having increased about seven times between 1992 and 2011 (statistics from the statistics department at PES, Sweden, 2012).
The number of people coded as disabled has also increased during periods of lower unemployment. Even if the increase reflects changes in the requirements in contemporary work life, with fewer unqualified jobs and high productivity demands in all sectors, it also, we suggest, reflects changes in work life standards of normalcy. The PES staff in our study state that today they classify people as ‘disabled’ who would not have qualified as such 20 years ago.
The floating signification of employability described above thus has implications also for the understanding of work capacity and disability. At the PES, the individual’s work capacity is evaluated in relation to the labour market. To be dependent on a wheelchair is, for example, would not be considered a functional impairment if you were to work as a switchboard operator. However, and in contrast to medical evaluations of work capacity, it is not enough to test work capacity in relation to a fictional labour market. The PES has to try to find an actual existing job for the person in question. This renders the concept of employability inevitably significant, since an individual may be found to have work capacity, yet still not be employable, simply because no employer would consider hiring her.
The study builds on 22 interviews conducted in 2007 and 2008, 15 of which are with staff at PES Rehab (Public Employment Service Rehabilitation or AF Rehabilitering) in two Swedish counties, and seven of which are with local employment officers and staff at the PES central administration. PES Rehab is a consultative function of the Public Employment Service organization, where work life psychologists, occupational therapists, and social welfare supervisors work. These specialists investigate the work capabilities of the individual on assignment by administrators at the PES. Since our research rests on interviews with staff, we are not in a position to provide a clear picture of what disability coding implies for the individual client in question and what it means for his or her subjectivity (see however Holmqvist, 2009; Holmqvist et al., 2012). What we will show is how staff at PES Rehab work to classify disability, to investigate work capacity, and thereby judge the employability of people, and the process of negotiation that this entails. Following Hacking (2007), we are interested in the five elements at work in the ‘sorting out of people’: the classification into kinds of people, the people in question, the institutions that are engaged in the work, the knowledge on which the classifications are made, and the experts who generate or legitimate the knowledge.
The codes used to classify the disabled are also interesting in their own right. Categories shape not only policies but also individuals, a process Hacking (1986) calls ‘making up people’. From this perspective, classifications are normative in aspiring to shape the conduct of individuals in a certain direction. They also contribute to shaping the self-understanding and subjectivity of those being classified. They are thus performative in that they contribute to constitute and frame further actions as well as expectations (Mik-Meyer, 2006). The methods used are practices that tend to enact realities as well as describe them (cf. Law, 2009).
In relation to related previous work on disability coding (see e.g. Holmqvist, 2009; Prior and Barnes, 2011; and Østergaard Møller and Stone, 2012), our study articulates the organizational technologies of government, involving the use of organizational typologies and procedures to make individual characteristics clearly legible. More specifically, we show how the categories used in the processing of people, i.e. categories used to convey information about employability, disability and capability, are themselves floating signifiers, defined relationally and plastically, within the context of a bureaucratic apparatus. This bureaucratic context furthermore works to de-politicize processes that are in effect politically driven and charged.
In the following we will spell out a critical theoretical perspective for analysing the disability coding as organizational techniques of visibility and legibility.
Techniques of legibility and the making up of people
We all classify on a daily basis. We divide things and people into ‘kinds’ as a way of making sense of the world. As pointed out by Mary Douglas (1966: xvii), rational behaviour involves classification, and the activity of classifying is a human universal. Yet as Bowker and Leigh Star (2000) pointed out, large-scale infrastructures are in special need of classification tools. Public organizations, such as welfare state bodies, need to classify individuals in order to be able to direct assistance to them. Organizations cannot treat individuals as unique creatures; they need to classify them according to their organizational schemes in line with their own organizational logic. In this process, the individual is transformed into a subject of the organization in question, and complexity is reduced. The individual is made into a manageable case, which may in a next step require a ‘re-subjectification’ along the lines of the template of the organizational intervention programme in question (cf. Bergström and Knights, 2006). Thence organizational classifications create ‘institutional identities’ (Mik-Meyer, 2006).
The disability codes are part of the assemblage that makes up the technology of government, we contend. In Rose’s view, ‘technologies of government are those technologies imbued with aspirations for the shaping of conduct in the hope of producing certain desired effects and averting certain undesired effects’ (Rose, 1999: 52). It is an assemblage of forms of practical knowledge, with modes of perception, practices of calculation, types of authority, forms of judgement, human capacities, devices, and so forth. In this context, routines for making legible are pivotal for the ability to govern (cf. Scott, 1998). Thus, we conceive of the bureaucratic assessment practices as techniques of legibility in that they allow for a ‘reading’ of individuals. Through the usage of routines and typologies aimed to enhance the legibility of individual strengths and weaknesses, the bureaucratic apparatus enhances its governing functions. The processing of individuals, we argue, relies on techniques for making strengths and weaknesses legible and thus actionable. The range of available diplomas, certifications and other signs of achieved learning goals, are other examples of skills and competences made legible. Legibility is crucial for governance, in that it allows for the follow-up of actions, for verification, control, and for sanctioning, or reward. By making legible, i.e. ‘readable’, for example the functional impairment of a person, the individual may be funnelled to the corresponding labour market intervention programme. The results of this intervention may then be followed up, evaluated, and compared. Legibility thus makes the individual ‘process-able’. To make someone ‘legible’ is by no means a neutral process, but a process predicated on organizational priorities and political aspirations. It is thus a partial and selective process, in which a particular and discerning organizational gaze is operative. The public administration observes certain traits and features and re-constructs them in its organizational categories, which has led Michailakis (2003: 209) to conclude that ‘one is not born a disabled person, one is observed to be one’.
Thus, we understand disability codes as legibility techniques that serve an organizational need, in particular in meeting the New Public Management logic of management by objectives and auditing of performance. The range of available diplomas, certifications and other signs of achieved learning goals, are examples of making legible an individual’s skills and competences, as are the evaluations and tests performed by the PES in its assessment of work capacity. Note here that the organization, that is the PES, needs to ‘read’ individuals, but that the environment also needs to ‘read’ the organization and its performance.
NPM principles as implemented in welfare bureaucracies have huge implications for the client-related work of street-level bureaucrats. They tend to constrain the caseworkers’ discretion, as the emphasis on standardization typically reduces the scope for individualized interventions (e.g. Østergaard Møller and Stone, 2012). Studies from various countries report signs of de-professionalization in the wake of NPM reforms, where professional service delivery is increasingly replaced by bureaucratic programme administration (see research overview in van der Berkel and van der Aa, 2012). One consequence of NPM practices, then, is that ‘mechanical objectivity’ (Porter, 1995), for instance through standardized assessments, tends to replace trust in professional expertise. The work of caseworkers is increasingly to ‘read’ the clients according to standardized templates, which has led Caswell, Marston and Larsen (2010: 400) to speak of ‘screen-level bureaucrats’ replacing ‘street-level bureaucrats’. Moreover, how clients are categorized is sometimes determined by the kind of information that the computer-based classification systems can handle, which in turn is often determined by legibility concerns. Thus, the discretion that traditionally characterized street-level bureaucracy, and which enabled some flexibility in matching solutions to clients, has been considerably reduced with the current administrative practices. Nevertheless, one way of widening the range of support measures available and to match a solution to an individual’s needs is to classify the client as disabled; the price attached is the disability code – and thus the label as disabled.
Classifications form the base for organizations to act upon individuals but their influence exceeds the particular policy measures they enable. Organizational categories, like all social categories, are not just practical devices by which to classify and sort people out (Bowker and Leigh Star, 2000; see also Diedrich et al., 2011); they also have consequences for those classified – practical consequences, affecting individuals’ subjectivity and self-understanding. The power of the diagnosis is essentially the power of shaping subjectivity by providing the ‘map’ and clues by which to understand one’s situation. In this sense, classifications and classes, categories and people, emerge hand in hand, a process of ‘making up people’ by which people come to fit their categories (Hacking, 1986). Once categories are available an increasing number of people come to fit them. Ultimately, categories and social orders are co-produced. Social and structural changes create new categories of people, which may then be reflected in national statistics (for instance statistics over deviant behaviour, illnesses, or disabilities). In Hacking’s terms, new slots are created in which to fit and enumerate people. Yet, counting is no mere report of developments. It creates new ways for people to be (Hacking, 1986: 223). Categories rely on some degree of participation of the parties involved. Moreover, categories have a tendency to become frozen, barring alternative conceptions of subjectivity. Once established, the memory of all the negotiations and contingency is lost. Hence, they are powerful devices for making people up.
As has been noted by Willmott and Knights (1989), our subjectivities are particularly open to the influence of organizational classifications and practices since they offer a way to gain a sense of self-worth and individual significance, but also recognition of a valued social identity in a larger community. We tend to seek security in those social identities that are both available and valued to us. However, precisely because our subjectivity is open, and indeed encouraged to be open, we also become vulnerable to the workings of this form of power.
In the following, we will look into the practice of disability assessment and coding. Consonant with Hacking’s work (1986; 2007), we are especially interested in the ways in which the social, medical, and biological sciences, mediated by organizational routines, create new knowledge and new social classifications. This new knowledge and concomitant new classification schemes render certain qualities and capabilities legible and transparent, thus providing a certain ‘governmental gaze’.
Sorting people out: Classification of disability at the Public Employment Service
The PES has a special mission to support jobseekers who face difficulties in the labour market. On top, the PES has a sectorial responsibility in national politics for the disabled, aiming to cater for the needs of the functionally impaired (AF, 2008a). For the PES, the impairment appears in relation to a specific job and a specific work environment, i.e. it is labour market relative. The PES specific programmes for functionally impaired aim to compensate for the reduction of work capacity and, in this way, increase their employability.
Officials at the PES refer clients who face difficulties in the labour market to further counselling or investigation of work capacity. This is done by a specialist unit at the PES (PES Rehab), where the main experts are work life psychologists, occupational therapists, and social welfare supervisors. The different professional categories undertake ‘work life psychology investigations’, ‘activity-based investigations’, and ‘work-related social investigations’ respectively. The aim is to investigate clients’ work capability prerequisites. The three main categories of functional impairment are: physical, psychological, and socio-medical impairments (for more detail, see the next section). The investigation then provides the ground for the further treatment of the case at hand.
Classification serves practical, procedural purposes. The coding of disability and its registration in the internal PES database forms the basis for record keeping and thereby serves an administrative purpose. The records are intended to gain an overview of the number of people in need of support and of changes over time. The classification is considered necessary in order to inform the distribution of resources and to channel assistance to the right targets. The record thereby functions as a signal to policy-makers. The classificatory schemes or templates employed by the PES professionals serve to make legible what is, in reality, a complex assemblage of capabilities and ‘weaknesses’.
The classificatory schemes we are interested in here are thus not just descriptions of the ‘actual’ work capacity, but more so, they ‘make up’ these capacities and subjectivities through their ability to provide the categories with official regulatory force. They thereby assist in the distribution of employability. Legibility thus works in tandem with and facilitates comparison, categorization, and intervention. Nevertheless, the techniques at work in detecting, measuring, and governing work capacity or employability, are not just devices serving a practical purpose – they also assist in establishing normalcy standards.
Our interviewees at the PES stress that medical diagnoses are made by medical doctors – the PES staff merely place the administrative codes onto a person. The code is considered a planning instrument, and ‘does not have too much to do with the individual,’ as one interviewee put it (Coordinator, PES central organization). As phrased by two other informants:
To know how much money is spent on these groups you have to have a system. And we manage this by having everyone with a functional impairment accept that we provide him or her with a code in the electronic system. This means that we can count how many people with a functional disorder there are in this country. (Social Service supervisor)
The reason for administering functional impairment codes in the system of the PES is mainly that we have to make sure the more exclusive interventions are directed to the right people. This is why we do these assessments and set the functional impairment codes. (Work life psychologist)
The explicit organizational ambition and expectation is to code all clients because, as one informant expressed it, ‘if you would only put a code on every second case, that would be wrong. However, if you feel that this would go straight to hell you don’t have to’. The formal instruction tells that if a person is considered to have a disability, he or she should be coded, in order to gain access to the measures even if, in exceptional cases, this may not be done. Accordingly, the PES staff have an incentive to use the disability code and the client to accept his or her code. The client must approve of the code and this can be a sensitive matter. At times, the code may be hard to accept:
What may be problematic is that the code is about the obstacle while we try to focus on the possibilities. So there may be a little clash there […] It may affect the motivation negatively […]. However, in most cases it works out well anyway. Especially those who work with this a lot, they learn how to talk about it. You try to explain the purpose of it, which is to get access to those resources that exist. (Coordinator, PES central organization)
Accordingly, clients may be persuaded to accept the code. As also described by Holmqvist (2009), the disability code grows out of a process of negotiation between the PES staff and the client. Holmqvist found that most clients did not want to be coded as disabled but accepted it in order to increase chances of getting employment. Our interviewees say that only a few clients resist the code. However, it can be difficult for a client to do so if a number of tests have shown a clear-cut result, and if the code is presented as an offer and a possibility of getting more ‘exclusive’ help.
Our interviews show that there are divergent opinions within the PES about the practices of coding disability, such as:
Let a thousand flowers bloom! Let people be the way they are. Why should we be stigmatizing? Then there is this other fraction, to which I belong, which has as often as possible made the judgement of disability. I realize that this is the only way for this person to actually get access to this money and get this assistance. (Work life psychologist)
The codes may be seen as enabling in that they make possible more exclusive forms of assistance to the individual:
I see it more as a possibility coding. You give this person who might have been a criminal a possibility to return to a job. (Social welfare supervisor)
Nevertheless, while the codes can be enabling in some respects, they can be disabling in others. Holmqvist concludes, in a study of sheltered employment in Sweden, that the longer the employees stay at Samhall (a state company and the biggest provider of sheltered employment in Sweden), the more disabled they become in the sense of acquiring a ‘disabling self’. Even if the initial idea was that the sheltered employment would be temporary, in practice very few managed to get a job in the regular labour market (Holmqvist, 2009). Thus, even if the aim of the measure is rehabilitation, it might have the reverse effect. While the disability codes are simply administrative tools for the PES, for the individual who is labelled ‘disabled’ the code might significantly alter her self-perception; she might even assume the identity of being disabled. Yet, the process of evaluation may also be helpful for the individual, who may indeed gain the support she needs to get a job. What we wish to point to here is the dual nature of codes and classifications: enabling and empowering in some respects, disabling and disempowering in others. While opening practical possibilities, the administrative codes at the same time streamline and enforce standards of normalcy. Moreover, as we will see, a disability is not just a reality to be discovered but an identity to be negotiated.
Objective tests and ‘subjectivity work’
The task of PES Rehab is to investigate the work capacity of the individual. This investigation is based on conversations, self-assessments, tests, and on probation at a workplace. The work psychologists’ tests focus on the clients’ interests and aptitudes, intelligence, logical and spatial capability, word and reading ability, numerical skills, and the like. The occupational therapist tests, among other things, physical mobility, pain, understanding of instructions, process skills (such as organizational skills and problem-solving abilities). The specialist decides exactly how the balancing of tests and conversations plays out. Most of our informants emphasize the conversation as the most important instrument, but tests are still used to a large extent. Sometimes, a specially assigned test leader may undertake tests during a full day.
While these tests may contribute to providing a clearer picture of the jobseeker, her strengths and weaknesses, technologies such as self-assessments and tests may imply a subtle form of control. Not only do they discover objectively existing characteristics, but they also create, to a certain extent, these very characteristics. By making certain characteristics visible and legible, the tests provide a picture that is then objectified. A personality test may, for example, appear as a neutral instrument, but in fact formats the person in the vocabulary of the test (Benson, 2008: 275). The technologies contribute to constituting the qualities that are made visible (Benson, 2008: 276). Thus, they contribute to an objectification or reification of the qualities and characteristics presented by the test. For an individual that is presented, or confronted, with such a picture, questioning or objecting to it may be difficult, or at least counter-productive.
The assessment of the often invisible ‘shortcomings’ is associated with a degree of uncertainty. One way to reduce uncertainty and to objectify knowledge is through standardization and quality assurance. The PES staff we interviewed were keen to call attention to the fact that the methods are quality assured. The central organization of the PES has also standardized manuals for the assessment of work capability and functional impairment in order to achieve a more uniform usage within the PES (e.g. AF, 2008a, 2008b). These manuals can also be seen a instruments for ‘mechanical objectivity’ (cf. Porter, 1995).
Hence, a functional impairment or disability grows out of a process in which certainty is created step by step, and whereby a preliminary fact is eventually established. Moreover, in contrast to the clinical psychologists, psychologists at PES (as well as the other specialists) usually have less time at their disposal and on which to base their assessment; the conversations have to be problem- and possibility-oriented. The tests and the conversations are part of the ‘subjectivity work’ that the clients undergo, and through which a problem is articulated and a certain self-understanding and self-image are fashioned. Working on oneself (Foucault, 1994) requires the capacity to continuously evaluate and correct oneself to enhance one’s attraction in the labour market. Career counselling, for example, is currently largely about learning how to present oneself and how to market oneself (Fogde, 2009). Thus, the ‘subjectivity work’ aims at creating a selling and agentic self. At the Rehabilitation Programme, we suggest, subjectivity work is about learning about and accepting one’s shortcomings and strengths:
Our jobseekers cannot be remoulded in that way. They have their functional impairments and they need support in that. (Social welfare supervisor)
The investigation of a person’s work capacity aims to frame the problem in an objective manner. The implications of this process are in no way neutral, however, but have consequences that infringe on the subjectivity of the person. The outcome of an investigation often involves a paradoxical combination of liberation and limitation. Identifying and objectifying the problem may be liberating for the individual client – who may achieve some self-understanding as to his or her problem in getting a job – but at the same time limiting for the future, in the sense of acquiring a disabled self. Accepting a functional impairment often demands a ‘destabilization’ of one’s identity (cf. Benson, 2008: 276), a re-thinking of one’s qualities and capabilities. It also involves a process of ‘qualification’ into the role as functionally impaired.
It is basically about building people mentally so that they feel more secure in their own role and their way of being. (Work life psychologist)
In a situation where there are ever-increasing demands on people to be enterprising, active, attractive etc, and in relation to which many people experience that they are ‘never employable enough’ (cf. Cremin, 2010), a subject position as ‘unemployable’ and disabled may paradoxically even be experienced as providing more ontological security – especially as it is supported by objective test results. For a client with a long record of unemployment and of being turned down by employers on the open job market, this identity may even be quite attractive.
Socio-medical disability – a market-related invention?
If employability is a floating signifier, so is disability. At the PES, functional impairment is defined in relation to the labour market. This also means that when market requirements increase, more people become disabled, i.e. the number of disabled, to some extent, follows economic cycles. In times of recession the PES, by government decision, gains more possibilities to provide special support through a larger repertoire of intervention programmes, which means that there is both more supply of, and demand for, subsidized or sheltered employment in times of economic down-turn. In good economic times, there are fewer intervention measures available. However, there are always possibilities for special assistance for those people classified as functionally impaired. This provides an incentive for PES staff to use the disability codes to help the client.
For the physical disability codes, a diagnosis by medical experts is required. Often medical certificates are also used for the psychological disability code and the ‘specific’ or ‘general learning disability’; however, these codes can also be based on the assessments of the PES Rehab staff. As mentioned, these are the codes that have proliferated during the past decades. They are arguably more open for interpretation than the physical disabilities. Of special interest to us, however, is the socio-medical disability code, where the contemporary labour market’s requirements become even more obvious. Interestingly enough, this code lacks international equivalence. It is not listed in the WHO’s ICF (International Classification of Functioning, Disability and Handicap). In our interpretation, the socio-medical disability code becomes a residual category for people who are not considered to have a clear psychological or physical disability but for various reasons fail to live up to the expectations of social skills, flexibility or adaptability. ‘This is something we have created ourselves’, as one interviewee puts it. He continues:
It has been formulated by experts in the National Labour Market Board without penetrating at depth […] The reason why the Board chose to see it socio-medically as a disability was in order to be able to use earmarked money to help those individuals to strengthen their opportunities in the labour market. (Social welfare supervisor)
Examples of ‘socially problematic circumstances’ that can be coded into a socio-medical disability are criminal background and previous or on-going drug or alcohol abuse (AF, 2008a). However, other circumstances appear as well in our interviews, such as obesity, exaggerated piercing, or insufficient personal care or hygiene. We interpret this to be taken as evidence of a lack of self-reflexive capacity to work on one’s self and to present a self that is attractive in the eyes of an employer. The fact that the socio-medical disability can be stretched quite far is illustrated in the following quotation:
Altogether there is reason to put a socio-medical code if you have not been out on the labour market for many, many years. You lack an anchorage in the labour market. And that makes one use the socio-medical code. (Social welfare supervisor)
Compared to a physical disability, where a statement from a medical doctor is required, the socio-medical code leaves greater space for the staff at PES to make the judgement themselves, i.e. for the social welfare supervisor and her work-related social investigation.
Our informants at the PES stress that social skills have become more important for employers in recent decades, and they are required in order to be considered employable:
You want the person to function in a social context, which means he or she can’t be too much out of the ordinary. One has to have social empathy, be able to communicate, take instructions and transform them into action. (Social welfare supervisor)
The socio-medical disability coding reflects the view that social skills are increasingly important in today’s work life, even if there are differing views on the need for social skills within PES:
Of course there are those who say that ‘it is exaggerated and ridiculous with those demands for social skills and to be outgoing and be able to speak for oneself. We should not accept this and [we should not] consider those who have a problem with this as having a disability’. While others see it the way I do: It is a disability. It does not work. You can’t find your way on the labour market if you do not dare to look people in the eye. If you are so nervous that you can’t behave properly in a job interview, then you won’t make it the ordinary way. (Work life psychologist)
Who you are as a person is more important than formal qualifications, according to the following interviewee at PES who deals with employer contacts:
Who you are. How you behave. How interested you are. Your social competence, in other words. That is what is important today. Many employers say that if you lack some knowledge, that can be dealt with – as long as you are a good person.
The demands for social competence – and the response of PES by way of a socio-medical functional impairment – reflects a shift in the perception of normalcy in work life, with tougher demands and less space for odd personalities and behaviour out of the ordinary: ‘You cannot afford someone who is not productive, who is slow and tardy and somewhat quaint’ (Work life psychologist). Our informants at PES agree on the fact that it is difficult to be an odd personality these days, even if they are critical of this development:
There is no question about it. Today we judge someone as being functionally impaired that we would not have considered being so 20 years ago. You don’t have to be very clumsy not to be passable on the labour market. (Work life psychologist)
Another work life psychologist says: ‘I have experienced that we are narrowing down what is perceived as normal’. This may pertain to looks, weight, or behaviour: ‘That you are just too much’. She is critical of the notion of employability:
I would think it’s very dangerous if we as a public authority decide that people are not employable. Then we would contribute to a negative development. If we disregard people who are not mainstream, what is left? Well, we may get a small group of yes-men (…) I think everyone merits a chance to be a part [of society].
In her experience, the demands for substitutability lead to people ‘being fitted into a mould, of some sort’:
Since demands in the labour market have tightened so much, more and more people cannot pass the bar – the bar of demands. And then more and more people are being coded as functionally impaired.
In today’s labour market context, the plasticity of available categories for ‘sorting people out’, i.e. defining who is employable and not, manifests itself in the occurrence of an overlap between lacking employability and having a regular functional impairment.
Employability, work capacity and disability as slippery categories
Since the PES staff strive to find an actual job for the client, their evaluations of work capacity and employability, as well as disability, need to consider the actual labour market situation. Ultimately, the employers decide on who is employable. However, work capacity and disability are also defined in relation to the actual labour market. All three are thus relational concepts, and as such fluctuate in their precise meaning. In the process of coding disability, defining work capacity, and thus judging employability, the distinctions tend to become fuzzy. As pointed out by Bowker and Leigh Star (2000), the sheer density of classification systems also means that they are likely to ‘meet up’ in various ways. As we will show, employability, work capacity and disability are liable to melt together in the local practice of the PES, at times creating less transparency and legibility instead of more.
Work capacity is both context- and task-relative; it is defined in relation to a specific workplace and a specific job. The supply of jobs and the demand of employers become decisive not just for employability but also for work capacity:
We can’t say anything decisive about work capacity since the market lowers its demands in times of unemployment and raises them in times of prosperity. (Work life psychologist)
As Brülde contends: ‘In actual fact, the changes in the labour market give rise to a weakened job-related work capacity for many, without their competences or their state of health being weakened at all’ (Brülde, 2008: 218, our translation). As a consequence more people risk ending up as unemployable and disabled.
If you link work capacity to employability you have to consider what the labour market looks like. What jobs are there? What do the employers want? That’s our focus. Work capacity depends on what jobs there are. In what job are you employable? (Social welfare supervisor)
As indicated by the citations above, work capacity and employability tend to become mixed and distinctions between them blurred.
Even if work capacity and employability tend to coalesce, our informants in the PES Rehab staff maintain that work capacity resides mainly in the individual, whilst employability is ultimately defined by the employer:
This concept is owned by the employers to a large degree. To be employable during a boom may not be the same thing as being employable in times of recession. This is partly dependent on the particular needs for workforce that employers are facing. It’s obvious that they lower the bar at times, and sometimes they raise it. The individual may also change in her employability but it is just as much the employers and the labour market that change. (Social welfare supervisor)
Some of the interviewed PES Rehab staff criticize the normalcy expectation inherent in the concept, whilst others contend that the concept of employability is important since it puts the supply of employers who are looking to hire in focus. The interviewees all say that the demands from the employers are more articulate and more influential today. This is viewed not so much a result of changed attitudes among the employers as a result of labour market changes and enhanced competition in the economy. For the individual, this means that expectations are raised:
The individual has to adjust to a large extent. She may have to accommodate to higher demands, maybe to shorter contracts. You are not supposed to place too high demands because then you may not be employable. The employers want flexible people who may adjust easily. (Social welfare supervisor)
This social welfare supervisor is sceptical towards the employability talk:
See, I think it’s dangerous. Because this term will become ever more difficult to define – what’s the content? And it will glide. It will depend on supply and demand.
Thus, both work capacity and employability are relative concepts, according to our informants, with employability being the more plastic one. A person’s employability is determined entirely by the demand for workforce and hence by market forces. But as their task is to find a job for the client, the local PES staff cannot disregard employability. As stated by one interviewee: ‘I may think that you have work capacity but no one wants you’. Another interviewee says:
I may have the physical resources as well as the psychological ones. But am I employable? How do I act? What do I look like? What background do I have? How do I live? What is my motivation? In the end it’s all about employability. (Work life psychologist)
In practice, a person who, after assessment, is judged to have work capacity might still be unemployable. Staff at PES Rehab thus have to work with both the concept of work capability and that of employability.
The higher demands in work life and the more strictly defined boundaries of normalcy have also created grey zones, with people whose personal characteristics influence their degree of employability without being clearly seen as functional impairments (see also Holmqvist et al., 2012). Some informants perceive it as easier to deal with those who have clear, visible impairments:
If we can see with our eyes that this person has a functional impairment, then we can adjust and accommodate. But we who work on this, the specialists, we see that it is employability that’s really the problem today. (Manager, PES Rehabilitation)
That is to say, there are increasingly many who are not considered to have a visible functional impairment, but who are still not employable:
Then we are dealing with something other [than a functional impairment]. You may not be sufficiently active, you may not be sufficiently attractive, and you may not be sufficiently communicative.
Oftentimes, it is to do with the ways of the individual:
That you are somehow difficult, that you are at the margins of what is considered normal. (...) Or else, most things can be fixed. (Occupational therapist)
The socio-medical disability coding as well as the codes of learning disability and psychological disability are ways for the organization to deal with the grey zone cases. But it also implies that the boundary between lacking employability and having a functional impairment is floating, and for some individuals with reduced employability the way to employment goes through a disability code. The classificatory system here functions as a mediator and distributor of employability. By the same token, it also serves to normalize characteristics by sorting them into established categories.
Conclusion: The plasticity of the categories of employability and disability as technologies of government
We have shown that a functional impairment is something that is developed in a negotiation process, where a wide range of individual shortcomings in relation to the job market are being subject to observation and interpretation, made legible, and classified into a documented functional impairment. The disability coding practices are to be understood primarily as technologies of government (Rose, 1999), assemblages of practical knowledge and procedures imbued with aspirations to shape the conduct of people in certain ways. Access to special assistance and resources is conditioned by such a disability codification. Getting a disability codification, in turn, is dependent on ‘strengths’ and ‘weaknesses’ being made legible and hence actionable. The enhancement of legibility is also a way for the organization to be transparent to its environment (cf. Scott, 1998), an organizational requirement reinforced by new public management practices.
Disability thus develops through the mediation of a classificatory logic, where a disability code is the path to special assistance and employment. The PES Rehab staff have to handle and assist clients judged to have work capacity while employers find them unemployable. To understand these dynamics, it is not enough to study the internal work of PES (cf. Holmqvist, 2009). We need to locate the problem in a wider social and organizational context, in relation to the particular characteristics of labour market policy that are articulated in present labour market discourses. More precisely, we need to pay attention to a set of combining factors: changes in work life and in standards of normalcy, but also changes in labour market policy with a reduction of active labour market policy spending, limiting the supply of measures available to unemployed in combination with NPM practices. As Mary Douglas (1966) reminds us, classificatory systems are never absolute, but relative, and cannot be understood out of context. Employability and disability coding becomes a way in which the PES organization can respond to fluctuations in the labour market and maintain a degree of control of its workings at the same time. In concordance with Hacking’s (2007) emphasis on the institutional context of ‘making up people’, we maintain that the administrative procedures of coding, the organizational techniques of legibility and transparency, work to link the organization, i.e. the PES, flexibly with labour market policy goals. A broadening of view also reveals the flexibility of codification practices, the plasticity of the categories available, and the interrelations between policy measures and market fluctuations. The categories of employability, work capacity and disability are shaped by these contextual and contingent factors, but they also have shaping effects on those subject to them.
We have moreover shown how the administrative procedures reinforce standards of normalcy. Categories assist in providing slots for what is inside and outside the box, so to speak, i.e. considered ‘normal’. Characteristics of the individual that fall outside the box may be combined to make up a ‘disability’ coding. To lack employability thus becomes a disability. The disability coding, in turn, opens up for targeted interventions and programmes that may eventually lead to employment. The bureaucratic categories are tightly interlinked and independent. Following Lévi-Strauss’ (1950) notion of the floating signifier, categories that may appear stable and definitive, may in themselves be void of meaning, and thus apt to carry a diversity of definitions. Work capacity, functional impairment, and employability may produce the fiction of categorical stability, but are open to be shaped by administrative, organizational, and political priorities.
The social and policy implications of these coding procedures are potentially significant. For the client the consequences of these taken-for-granted organizational practices may be wide-ranging. The self-perception of one’s impairment is developed in dialogue with ‘experts of subjectivity’ (Rose, 1989) and appears, after the assessment process, as an increasingly indisputable ‘fact’. The procedures may have empowering as well as disempowering effects. The moulding of subjectivity that the process of work capacity assessment implies is a subtle form of control in which the subject herself is, so to speak, invited to participate. It may have a mobilizing and empowering effect, in that it may serve to support and articulate qualities and strengths in the individual, or assist in creating an awareness of shortcomings that may be addressed. On the other hand, it may as well exert a highly conservative effect, an effect that serves those whose interests and priorities are invested in these administrative interventions (see also Knights and Willmott, 1989: 542).
The form of power that is being exercised by the PES, however benevolent (hence more powerful), works by inducing individuals to subject themselves to the classificatory systems offered to them. The road ahead, to a job or to a supportive intervention, lies in accommodating, accepting, and working with the codes and their administrators. In the view of Foucault, subjectivation is an economical form of power, since it is a technique of the ‘social’ and of the ‘self’ which produces a self-disciplining subjectivity (Foucault, 1982; Knights and Willmott, 1989: 550). Modern forms of power and subjectivation, such as those evinced at PES Rehab, work by forcing individuals back onto themselves so that they become ‘tied to (their) own identity by a conscience or self-knowledge’ (Foucault, 1982: 212).
At a larger scale, the classificatory procedures may as well have implications for the interface between the individual and the bureaucratic organization, i.e. the PES, and the ways that policy implementation operates. The procedures of classification come to exert a certain control over individuals by sorting them into an administrative grid of social relations, with differential access to social welfare resources. And it does so by inviting individuals to agree to collaborate in placing a code onto themselves and in improving their work capacity and employability. The organizational and administrative powers at work articulate the individualizing effects of such new regimes of power, as they push individuals back onto themselves as the primary source for productive work capacity and for employability. As qualities and properties are made visible, legible, and thus highlighted as single elements, individuals are separated off from one another and made governable, as well as more acutely responsible for themselves.
To conclude, an analysis of contemporary labour market policies, and their specific technologies and procedures, must pay attention to the practices and procedures at the interface of individual and state. It must be attentive to the implications for subjectivity, empowerment, or disempowerment, of the usage of large-scale classificatory schemes. Our subjectivities are open and vulnerable to these technologies precisely because they appear to recognize and confirm our individual sense of identity as well as confer a sense of linkage to larger social structures. They are, in Hacking’s (2007) view, a moving target. In our view, theoretical inquiry should be directed towards exploring and exposing the contradictions inherent in labour market policy interventions and their sets of technologies, not least pertaining to the simultaneous inducement of empowerment and disempowerment. The contemporary currency of employability points to the double implications that technologies of legibility may have for the individual. It speaks as well to the plasticity with which bureaucratic practices align themselves with policy goals and political ideals.
 Apart from wage subsidy, there are the following supportive measures available: support for technical aids, support for a personal assistant, special support for starting up a business, special introduction and follow-up support by way of a personal coach at the prospect of employment, and different forms of alternative employment such as Development Employment, Security Employment, employment at Samhall, i.e. a government-owned company providing development opportunities for people with disabilities through sheltered employment, and Sheltered Public Employment. Some of the support measures are temporary.
 We gratefully acknowledge the assistance of Ida Seing, who did the interviews, and all the interviewees who generously shared their knowledge, experiences and dilemmas.
 Scott (1998) argued that the state, in its attempts to simplify the classic functions of state control, strives to arrange the population in ways that simplify interventions. These state simplifications, he argues, function rather like ‘abridged maps’; they neither successfully represent the activity of the society they depict, nor do they intend to do so. Rather, they represent only the slice that interests the official or the organization division in question. Hence, they make legible those dimensions that are relevant for the functioning of state power, while leaving others illegible, or invisible. More importantly, state simplifications are not just maps, but when allied with state power, they enable much of the reality they depict to be remade (ibid.: 3).