Limits to vocational inclusion?: Disability and the social democratic conception of labour

The article discusses the difficulties faced by disabled people to enter the ordinary labour market. Despite increasing political emphasis on employment integration, and a gradual integration on most other areas of society, statistics indicate stagnation and even decline in labor market participation among disabled people in the majority of EOCD countries. Several studies also report that many disabled people experience a disinterest from employment authorities in helping them find employment.Drawing on a biographical study from Norway based on life-history interviews with 66 disabled people, the article discusses possible explanations to the meagre results of disability employment policies. The article suggests a number of feasible explanations: Parts of the explanation must be sought in the early institutionalization of waged labor. Defining the ‘ability to work’ was the primary criterion used to define who is disabled and, subsequently, entitled to public support. The article suggest that the increasin...


Introduction
The Scandinavian welfare states, also known as the social democratic type of welfare state regimes (Esping Andersen 1990), are characterized by their comprehensiveness, their universalness and their emphasis on equality and solidarity (Kolberg and Esping Andersen 1992;Hanssen, Sandvin, and Sö der 1996), i.e. by their inclusiveness. These are the welfare states that have been most ambitious in terms of realizing what T.H. Marshall (1965) called social citizenship. The condition for doing so was the ability to create and maintain high employment, which again, was made possible by a strong alliance between government, capital and organized labour (Pierson 2006). Moreover, the Scandinavian countries have also been pioneers with respect to disability rights (Priestley 2003). Although this role may not be as obvious today as it once was, we have witnessed a gradual inclusion of disabled people into virtually all parts of society over the past decades, with one important exception: the labour market. Despite the broad implementation of antidiscrimination legislation and a strong emphasis on disability employment policies, employment among disabled people seems to have stagnated and even fallen compared to other groups in the majority of industrialized countries (OECD 2009). Pearson and Prinz (2005) claim that disability employment policies is the area of social policy that has been the least effective in achieving its goals, and employment is probably the single most important political issue for disabled people today (Shah and Priestley 2011).
The question raised in this article is why the labour market seems to be more difficult to access for disabled people than other parts of society. We start by looking at some of the statistics available on disabled people's labour market participation, and introduce a biographical study carried out some 10 years ago (Sandvin 2003) that illustrates some of the changing experiences of people with mobility and sensory impairments approaching the Norwegian labour market at different stages of the welfare state development. The study is used as a qualitative background and source of possible explanations as to why the ordinary labour market seems so difficult to access for disabled people. Even if the qualitative study is limited to Norway, we believe that some of the experiences revealed may have a more general relevance, at least within a Scandinavian context.

Background
The general pattern exhibited from available international data is that the employment rate among disabled people is substantially lower than for the rest of the population. According to the OECD (2009), the average employment rates for persons with impairments across the OECD is just over half of that for people without such impairments, and there is little evidence to show that the situation is improving. Figures show a standstill or a slight decrease in the employment rate among disabled people, both in Scandinavia and in the majority of industrialized countries (Lein 2007;Bø and Håland 2010;OECD 2009).
With respect to variations among countries, Halvorsen and Hvinden (2009, 26) contend that the 'available figures for the proportion of employed persons with a disability are uncertain' (our translation), which is probably due to both linguistic and operational differences. They show that although huge differences are found among countries with respect to disability employment, these differences are very unstable across different studies. On a national level, however, surveys may be useful in monitoring the development over time, provided that the operational definitions and questions used are kept constant.
In Norway, the most reliable source in this respect is the Labour Force Survey operated by Statistics Norway. According to this survey, 42.3% of the disabled population aged 15Á66 was employed in the 2 nd quarter of 2011, as compared to 73.9% of the population as a whole (Statistics Norway 2011), which is a slight decrease from the three preceding years (45.3%, 43.1% and 43.6%). Among those employed, 47% of the disabled worked part time as compared to 26% overall. The official unemployment rate among disabled people differs only marginally from that of the entire population, though the real unemployment rate is much higher. The same survey detects that 25% of the disabled people not employed wanted paid work, whereas only 17% of those who wanted a job were classified as unemployed according to the criteria on active job seeking and availability to work, with the figures remaining the same during the last three years. In a comparative study that included Norway and the UK, Hansen et al. (2011) find that the level of such 'concealed unemployment' is about the same for the two countries. She believes that many find it too difficult to remain an active job seeker over time, and that being dependent on a stable income makes many settle for a disability pension.
Some studies also indicate that the employment rate among disabled people in Norway has decreased over a longer time span. Hanssen and Bliksvaer (1999) have used the Level of Living surveys carried out at regular intervals by Statistics Norway, in which the same filter question was used for the four surveys included (1980, 1987, 1991 and 1995). What they found is that the workforce participation among disabled people aged 17 to 64 increased from 55% in 1980 to 59% in 1987, and then fell to 47% in 1991 and 48% in 1995. The high levels of workforce participation in the 1980s were also supported by Barth (1987), who found that 61% of disabled people were employed compared to 75% of the non-disabled population. Even if these figures are not comparable to the more recent figures shown above due to changes in the questions' delimiting disability, we believe that the available evidence suggests that there was a marked decrease in work force participation among disabled people from the late 1980s onward, thus forming a level that remained in the ensuing years. One explanation suggested by Hanssen and Bliksvaer (1999) is that the period of relatively high unemployment between the late 1980s and mid-1990s hit disabled people particularly hard, as the unemployment rate among disabled people increased far more than for the non-disabled population, particularly among young people (Hanssen and Bliksvaer 1999, 79). It seems as if this period of relative exclusion turned in to a more permanent pattern, in which many of those who would normally turn, or return, to the labour market instead settled for a disability pension, thereby increasing the 'concealed unemployment' demonstrated by Skog-Hansen above. The number of recipients of disability pensions increased from approximately 140,000 in 1980, to 240,000 in 1995(Hanssen and Bliksvaer 1999) and reached 320,000 in 2005(Statistics Norway 2006.

A generational perspective
The individual accounts referred to in this article stem from a generational study carried out some 12 years ago (1998Á2002). The study is based on life-history interviews with 66 people with different impairments (mobility, visual and hearing impairments) from three different age groups. When the interviews were conducted, the interviewees in the oldest group were in their mid-60s, the middle group were in their mid-40s and the youngest were in their mid-20s. The additional criteria for selection were that the impairments occurred early in life and were severe. The majority of the visually impaired were legally blind and among those with a hearing impairment, half defined themselves as being deaf and were interviewed in sign language, while the other half described themselves as being hard of hearing and used various types of communication devices. The interviewees were quite equally distributed across generations (25 from the youngest generation, 19 from the middle generation, 22 from the oldest generation) and types of impairment (22 had a mobility impairment, 23 had a hearing impairment, 21 had a visual impairment), and came from all over Norway, representing a broad spectrum of communities and socioeconomic backgrounds. All the interviewees contacted us after having seen a request in a magazine or receiving an invitation through one of the disability organizations. This way of recruiting interviewees, which is also called self-selection, has ethical advantages compared to other strategies, first of all by not putting any pressure on people to participate, though it also involves a risk that those who come forward do so with a certain purpose, hence creating a so-called self-selection bias. In this case, people with the respective impairments were invited to talk more generally about their lives, and there is no reason to believe that this caused systematic bias with respect to specific issues.
The interviews were conducted with a somewhat different purpose than what we are concerned with in this article. At that time, we were interested in how modernization and the evolving welfare state affected the orientation and identities of the interviewees (see Sandvin 2003). Confronted with complaints about the slow progress of disability employment policies, we decided to go back to this data to look at how interviewees from different generations with different impairments described their vocational careers, their dreams and their accomplishments with respect to education and labour.
Biographical research, also called life-history methodology, is a broad category of approaches to qualitative studies (Schwandt 2001). One such approach is the study of generations, which can refer to family-based generations, or as with this case, to an aggregate of people born around the same time and with the same social location, therefore being similarly positioned in relation to the social and historical processes taking place in the society where they live (Mannheim 1952). To a large extent for disabled people, their historical location determined even their social position. As Priestley (2001, 244) contends: disabled people's life choices and life chances can be very fixed and limited, by comparison with non-disabled people, within particular social conditions. (. . .) Disabled people have been, in general terms, much less free to reinvent themselves or to manage their own lives.
One important aspect of life-history methodology is to link individual experiences with social history, as generational studies make it possible to compare individual experiences made under different socio-historical contexts. In this case, we are particularly interested in the experiences of people with different impairments approaching the educational system and the labour market under different sociopolitical conditions. The study provides no basis for determining the factors that continue to limit disabled people's access to paid employment, but it can reveal patterns that may exist within or across generations, which may point to possible explanations.

Gradual inclusion
Education is an important component in understanding disabled people's position in relation to the labour market, providing what Leisering and Leibfried (2000) call the foundation for opportunities and direction in life. Education is also the area where we find the clearest and most substantial differences between the three generations in our study.
Regarding compulsory education, the oldest generation is a segregated generation, where special schools represented not only the default choice but also the only option for most pupils with a defined impairment. For some of the blind children, the segregation started even earlier when they were sent to orphanages at the age of three to five. Schools for the deaf were based on the 'oral method' and sign language was prohibited, which is something that seriously hampered the children's theoretical learning (see also Rydberg, Gellerstedt, and Danemark 2009). Several of the deaf interviewees said that they did not understand anything in school before they learned some sign language in the schoolyard, where older classmates could explain what was going on. The right to education was less clear for children with mobility impairments who were supposed to go to ordinary schools, although no strong enforcements were made to ensure they achieved access. Many of the mobility impaired report about long stays in orthopaedic institutions, often for years, where education was often absent or at least far from adequate. One of the male interviewees expressed it in this way: I have had a slightly different upbringing than most people, because I have been away so much in hospital. I have calculated that half of my life until I was 14 years old was spent in hospitals. At the time I started school, there was no teaching in the hospitals, so the first years were bad. The education was far from satisfactory.
It is interesting to note the priority given to correction at this stage, as education and participation were generally subordinated to bodily and behavioural corrections. The question is how this affected these children's opportunities and orientation in later life, with Leisering and Leibfried (2000) stating that one's life course is political, and that politics is extensively concerned with life course policy. The policy contained in the education provided for these children was one of correction and segregation.
As Vedeler and Mossige (2010) describe, the focus of national disability policy in Norway shifted in the late 1960s, from segregation to integration: 'The integration of children with disabilities into mainstream education was a manifestation of this shift' (ibid, 258). Still, even the middle generation represents a segregated generation, at least with respect to education. The politics of integration came too late to have any major impact on their primary education. Special schools were still the main alternative for pupils with sensory impairment, and sign language was still abandoned in schools for the deaf. For the mobility impaired, a firmer enforcement of the right to education implied a stronger mainstreaming, although several of the interviewees still report about long stays in orthopaedic institutions without sufficient teaching. The picture revealed is more that of continuity than of a break with the past, even if accounts indicate that conditions in special schools were less harsh than they used to be.
If the picture to be drawn based on the two oldest generations is one of continuity, the youngest generation definitely represents a break with the past, and is clearly marked by the new integration policies. All of the interviewees (except for some of the deaf) speak about their mainstreamed schooling as being the most natural thing of all. There are stories about teasing and bullying, and about a lack of assistance and the necessary technology, but not in a way that alters the general picture of mainstream schools as the only sensible, or even thinkable alternative. From the early 1980s, sign language was accepted as a first language in schools for the deaf, and was the only legitimate reason for segregated schooling.
The same picture appears when we take a look at secondary education. Getting an education was considered very important by the representatives of the oldest generation, perhaps more important than for nondisabled people in the same generation. A female interviewee from the oldest generation with mobility impairment describes what this meant for her as a disabled woman: My parents saved up money so that the girls in the family could receive an education first. This was a time when those who had gone to secondary school were considered to be a good match, especially the women. The standard was to marry and to become a housewife. But I was very aware that I was not a good match at all. I had to get an education to get by.
Most of the older generation received a crafts education from one of the vocational schools run by their respective organization. The scope of choice at these schools was limited to what was regarded as a suitable occupation for people with certain impairments such as a tailor, basket worker or piano tuner for the blind, carpenter and welder for the deaf and so on. For most disabled people, entering an ordinary higher education was unthinkable since it would require a secondary education, but neither secondary schools nor universities were accessible. Additionally, it would also be rather expensive and thus risky because job opportunities were uncertain.
For the middle generation, the picture revealed is somewhat mixed; narratives indicate that the possibility of entering higher education indeed existed, but very few continued towards a degree (only one teacher and one social worker). One reason for this was that financial arrangements were still inadequate, and very few would take the risk of running up a debt and still not acquiring a job. The dilemma is evident in a statement from one of the hearing impaired from this generation: I had a big problem compared to hearing people when it came to education. (. . .) I would probably need a longer period of time and run up a larger debt than others. And the question was: Would I get a job afterwards? There's no point in putting yourself in debt if you do not get a job.
There was still no state policy to actively support disabled people in entering the generic institutions of higher education, not to mention making these institutions universally accessible. There was always an easier alternative; referring to the vocational centres under the auspices of the disability organizations.
At the time that the youngest generation finished their compulsory education, a reform act from 1994 had given all the youngsters the legal right to enter secondary school, which implied that all secondary schools in fact had to be accessible to disabled people. Disabled pupils were even given economic support during secondary school, and had the opportunity to use more time than their able-bodied peers. Almost all of the interviewees from the youngest generation had finished secondary school, some had started their higher education and some had even obtained a degree. Even so, some chose not to acquire a higher education and quite a few were in the process of trying to secure a job, but all saw the option of taking higher education as a matter of course. One of the female interviewees from the youngest generation, a wheelchair user, had just been accepted to a folk high school in Denmark (institutions for adult education that generally do not grant academic degrees, most commonly found in the Nordic countries). When asked about further educational plans, she responded: Consequently, what we see with respect to education is a clear picture of social change. While the oldest generation, and to a large extent even the middle generation, were excluded from an ordinary education and directed towards a separate path, the educational career of the youngest generation was mainstreamed and followed pretty much the same pattern as the rest of their generation. They used some additional time, but their choices seemed to reflect pretty much the same orientation as their non-disabled mates.
These differences had a clear impact on the further direction of these people's lives, which was most easily depicted for the oldest and youngest generation. To a large degree, the oldest generation remained a segregated generation throughout their life course, not so much by force as by a collective association with the disabled community (Sandvin 2003). This was also inflicted by state policy, or rather by a lack of such policy. The welfare state acted by refusing to act, and the responsibility for securing disabled people's rights to education was primarily left to the disability organizations, for whom the space of action was limited not only by the prevailing discourses of the time but by the active exclusion of disabled people from ordinary schools. For the middle generation, the exclusion was less firm, and inclusion was accepted and even rhetorically encouraged, although little was done to enforce it.
For the youngest generation this picture was completely changed, insofar as the very idea of such institutional segregation appeared strange and horrifying to our youngest interviewees. As stated by Mannheim (1952), the mode of thought was radically changed due to the experience of growing up in an integrated setting, As a result, the collective identity and association with the community of disabled people was substantially weakened due to a more deliberate and active mainstream policy. A young man with cerebral palsy stated this almost in its extreme when he declared that: I've never cared so much about other disabled people to be honest. It may sound a bit strange, but I actually have not. I've had only non-disabled peers and I cannot recall . . . . It may just sound like crap but I have not looked at myself as disabled.
This seemed a bit strange to the interviewer since the man had just said that he had been quite active in disability sports. When confronted with this apparent inconsistency, he replied: 'Yeah, that's true, but they are just like me. The ones that I met have the same background as me. They have never thought of themselves as disabled in that respect.' The use of 'in that respect' refers to the experience of being excluded.

Inclusion discontinued
The changes described above are best perceived as a gradual inclusion of disabled people into mainstream society, particularly with respect to education. The question for us to ask, then, is how this gradual convergence towards an integrated system of generic education is reflected in the further life course of disabled people. As asserted by Leisering and Leibfried (2000), if education provides the foundation for opportunities and a direction in life, we would expect a corresponding and gradual integration into the ordinary labour market, but that is not what we found. Instead, it is our oldest interviewees who have the most extensive experience from the ordinary labour market, with virtually all of them having extensive experience from ordinary work, though not always in occupations consistent with their education. Quite a few were employed by a disability organization or by a business run by one of these organizations such as print shops, libraries or magazines, but with regular contracts and most often after quite a few years in other sectors. The interviewees representing the middle generation had to struggle more to gain admittance to the labour market, especially in relation to having a stable position. They changed work more often, partly due to temporary contacts. One interviewee from the mid generation who had a mobility impairment said: 'I've had many different jobs, but none that were long-term and none that were particularly well paid. I have had several retail jobs and I've worked as a gardener's assistant.' Quite a few turned to the disability organizations for work, often earlier than was the case for the oldest generation, and quite often without a regular contract. Furthermore, many had already left the labour market by the time of the interview.
Obviously, our youngest generation did not have the time to gain extensive job experience by the time of the interviews. Nonetheless, the majority of the 25 from this generation had made their first attempts at obtaining a job, with some striving for this for quite a while. A young man with a visual impairment reported on his attempt to approach the labour market: 'It was very difficult to get a job. I applied again and again, but I was always rejected. At that time I wasn't admitted to any education either, so I just had to sign up for the dole.' Based on these young people's experiences, the opportunities for disabled people to access the regular labour market does not seem to have improved. The problem of finding a job was a recurring theme in most of the interviews and the majority of those who were no longer being educated were either unemployed, an apprentice paid by the employment office or in some form of sheltered workshop, all with some form of income compensation provided by the welfare state. These alternatives were all regarded as unacceptable: '[M]y priority now and for the future, is education and work', as one of the deaf interviewees asserted. No one had yet obtained an ordinary job, and some had already been offered a disability pension, most often in combination with some sort of sheltered work, which they felt was rather offensive. In addition, many reported a feeling of not being taken seriously by the employment office, or a kind of indifference in relation to their strong desire to find employment, which stands in sharp contrast to the official policy. One girl with a mobility impairment said that when she finished secondary school she was recommended for a full disability pension, which she opposed. She might have accepted receiving a half-pension, but she insisted on getting the opportunity to work: 'I think it is totally wrong, because I do not need to go on full disability pension.' The same experience is reported by Vedeler and Mossige (2010).
The findings are congruent with the employment statistics referred to earlier; hence, education does not provide the opportunities and direction for the adult life of disabled people in the same way as for non-disabled people. It is well established from quantitative studies that higher education generally improves the possibilities of disabled people in the labour market (Bliksvaer and Hanssen 2006;Borg 2008), but such studies employ a very broad definition of disability. Qualitative studies, both our own and others, show that people with profound sensory or mobility impairments still face significant obstacles in relation to the labour market even if they have an education (Anvik 2006;Vedeler and Mossige 2010).
Our own study is more than ten years old and is used here to illustrate the experience of approaching the labour market at different historical times. The question is whether there have been changes in recent years that are likely to have altered this picture. The main changes that have occurred in Norway during this period is the introduction of a stronger protection against discrimination in employment, partly inspired by Sweden and partly to ensure the same level of protection as in the EU. This regulation has not yet had any noticeable effect (Halvorsen and Hvinden 2009). The law must itself be considered a response to the lack of integration in the labour market, but the experiences of countries that have had such legislation for years indicate that the effects may be questionable (Doyle 1996). In an interview with USA TODAY (2010) in connection with the 20 th anniversary of the Americans with Disabilities Act (ADA) Carol Glazer, the president of the National Organization on Disability, said that 'While education has improved considerably, joblessness has not. We as a nation must figure this out' (USA TODAY 2010). Michailakis suggests, with reference to Sweden, that the meagre effect of anti-discrimination legislation is because such legislation individualises problems that in fact are systemic.
Hence, despite the convergence taking place in the area of education, disabled people are still deprived of the signifier of adult status provided by gainful employment to a large extent, and thus admittance into a normative life course. The exclusion may start later than it did some decades ago due to mainstreamed education, but irrespective of educational or other personal achievements, disabled people are still largely maintained in subversive positions and deprived from entering what is perceived of as the height of the normative life course; namely that of a productive and independent adulthood (Priestley 2003, 117Á118).
Despite an increased political emphasis on employment integration, what we see in our study is instead a reverse development with respect to the type of support received by our interviewees. The oldest generation was subjected to job coaching, training and retraining, as well as re-education and rehabilitation, with an active orientation towards (re-)entering the labour market. Most of these services, however, were offered by the disability organizations, because the government had not yet developed such strategies. By the time the middle generation reached adulthood, the welfare state had expanded considerably, also with respect to employment policies. Still, very few of our interviewees received what they would call active assistance in order to enter the labour market. They were better catered for through various economic measures, of which some could be difficult to return to if they used their chances on vague or short-term employment contracts.
For the youngest generation, the number and generosity of measures and schemes had increased even further, though first and foremost the measures are related to income security, educational support and independent living. Officially, disabled people are also an important target group for employment policy, but the strategies employed are mainly universal in nature, presumably reflecting a more general conception of mainstreaming. There are very few selective measures aimed at assisting disabled people in entering the regular labour market. More targeted measures, such as employment quota schemes, have been put forward from time to time, but have been turned down thus far. It is an interesting paradox that the selective strategies employed by disability organizations, before any mainstream policies were adopted, seem more effective in providing disabled people access to the regular labour market, than mainstream policies seem to do.
The picture drawn here may seem very pessimistic, perhaps even exaggerated, in the context of the political ambitions stated with respect to disability employment. Yet, the available data reveals that radical reforms in education that aim at Á and to a great degree succeed in Á providing young disabled people with the same foundation for opportunities as everybody else, seem to exert a meagre effect on employment rates. It is surprising that the oldest generation, who were collectively excluded from the ordinary system of education, had greater success in the labour market than the youngest generation seems to have had. It is also surprising that despite the political rhetoric and a move towards a more activating and 'enabling state' (Gilbert 1999), the vast majority of our youngest interviewees report a feeling of disinterest from the employment office and other welfare state representatives in their labour market ambitions. How can this be explained?
Disability and the labour market: Limits to inclusion?
As pointed out by Priestley (2003), issues of employment have long occupied the centre ground of thinking and research on disabled adulthood. For this reason, the answer to the questions raised is not a lack of attention, thereby making the question even more pressing to raise. Even so, there is hardly any single explanation for the features described above, as the reasons are most likely both complex and contextual. We will suggest some possible explanations and discuss their relative importance with respect to a Norwegian context, but we will also point at some deeper structural barriers that could transcend national contexts.
One way of explaining the lack of progress of including disabled people into the labour market is simply that the labour market itself has changed in a way that makes it gradually more difficult to access for disabled people (Askenazy 2004), and it has been suggested that competition has increased, that an emphasis has been placed on performance as much as on qualifications, and that the general pace is faster today than it was some decades ago (Sennett 1998). But during the same period, the labour market has also become more automated and computerized, which should make many jobs easier to manage for people with impairments. Hence, a thorough investigation of the relation between disability and a changing labour market would require a more detailed distinction between types of work, between entry-level work and professional work, between manual work and non-manual work, etc. (Roulstone and Barnes 2005). The argument that the labour market has become too difficult to manage for disabled people also contains a perception of disabled people as someone who are slower or weaker and unable to keep up with the pace, which may be true for some, but which is also a conception that may be part of the problem.
Another explanation is that differently from the two succeeding generations, the oldest generation lacked in their formative years what Lesiering and Leibfried (2000) calls the 'expectation of a safe life span' provided by an old age pension and measures of risk management, therefore providing a stronger motivation to work. For most of our interviewees from the oldest generation, work was a bare necessity in order to live an independent or decent life, while the middle generation, and the youngest generation in particular, were guaranteed a certain income even without working. The necessity of the oldest generation to find work may also explain the determined and targeted support from the disability organizations on this matter as we discussed earlier. This explanation carries some weight, as income compensation removes the necessity to accept whatever jobs are available, but once again the explanation contains some underlying assumptions that are questionable, not least with reference to our data. It assumes that the problem is a question of motivation, although what our young interviewees report is that it is a struggle to find a job, and a fight against (permanent) allowances. We have also referred to the 'concealed unemployment' as shown by Hansen et al. (2011). Hence, even if the incentives were stronger for the oldest generations, this explanation seems far from sufficient.
A third explanation for why the oldest generation succeeded more than the following generations is the fact that the responsibility for both vocational education and employment, was left entirely up to the disability organizations: The educations offered were more targeted, as most organizations had job coaches working actively to help people obtain work, and the larger fabric of organizational services represented a labour market of its own. This may explain why our oldest interviewees managed to get a job, but it is not a sufficient explanation to as to why the youngest generation struggled, and it does not explain the lack of energy and enthusiasm they experienced from labour authorities.
A fourth explanation, which is more directly related to why the gradual inclusion seen with respect to education is not reflected in a corresponding inclusion in the labour market, is that the systems of education and labour are structured differently. Education in Scandinavian countries is principally public, or at least publicly administered, and hence highly responsive to public policy, while labour is primarily regulated through market mechanisms and therefore far more resistant to public interventions (Michailakis 2000). Policies adopted to increase the participation of disabled people in the labour market, be it through anti-discriminatory legislation or by compensatory models (Gooding 1996;Priestley 2003), represent indirect measures, while policies in other areas such as education, health and social services most often contain direct measures. Because disability policies have generally become more progressive, increased expectations and pressure have been placed on the labour market, though not with the same response as with other areas. Instead, as pointed out by Hyde (1998, in Priestley 2003, workfare and market deregulation may exert a negative impact on disabled people's opportunities for work insofar as they are increasingly exposed to the disabling social relations of competitive labour markets. In fact, political control over the labour market is very restricted and even decreasing, thereby meaning that ambitious employment policies are most likely to fail. This suggests that explanations for the meagre results of expansive disability employment initiatives must be sought within the very structures of the labour market itself, as well as in the structured relation between labour market, disability and the welfare state. This requires a closer look at the historical construction of the dependency of disabled people upon the welfare state, as disability is historically linked to the early institutionalization of waged labour. As pointed out by Priestley (2003, 133): 'Employment is (. . .) significant in the construction of disability as a social category, since 'ability to work' is the primary criterion that states use to define who is disabled.' Hence, an important task for welfare states has been to monitor the demarcation line between those capable of working and those exempted from the duty of work (Midré 1990) in order to protect the ethics of work. The essence of work ethics is the will to work, but as stated by Bauman (1998) the paradox, however, is that the blessing of work was introduced in the early stages of industrial production when work was about to become a curse. The welfare state praises the work ethics which it is utterly dependent on, but the regimes of demarcation between worthy and unworthy recipients of public support testify to a deep suspicion about (some) people's willingness to work, thereby making the 'suspicious cases' (those suspected of being unwilling to work) the most important cases to supervise. On the other hand, disabled people are above suspicion and do not represent a threat to the prevailing social order. It is suspected, however, that not all of those falling into a wide disability category should be exempted from the duty of work (OECD 2009), though that is because they question their very status as disabled. As Roulstone and Barnes (2005, 21) point out: 'There is now growing evidence that the target group of disabled adults within flagship policies (. . ..) are those that may be perceived in costbenefit terms to be the most easily returned to the labour market, but who may also be the most resistant to paid work.' This does not apply to those participating in our study since they are 'obviously disabled' and thus per definition exempted from the duty of work. Hence, as seen from the welfare state, there is no pressing need to spend too many resources and energy to assure that disabled people have access to the labour market. Even if such policies are demanded and politically required, they do not represent a pressing issue to effectuate. On the contrary, doing so could possibly hamper the fine line of demarcation between those obliged and those not obliged to work. After all, it is those not exempted from the duty of work who are the primary target of workfare policies, which may help explain not only the meagre results of disability employment initiatives, but also the lack of energy and enthusiasm from employment officers as experienced by the youngest generation. But for disabled people, exclusion from labour market participation may mean being kept in a dependent and subversive role, being deprived of the economic opportunities of waged labour and the valued status of independent adulthood.

Conclusion
The point of departure for this article was the low or even declining participation of disabled people in the ordinary labour market, despite an increasing political emphasis on employment integration. Accounts given by three generations of disabled people growing up in different phases of the welfare state draw a picture of the gradual equalization of opportunities in terms of education. It would therefore be reasonable to expect that this would lead to a corresponding equalization in terms of labour market participation. Nevertheless, what we find is that accessing the labour market is just as hard, if not harder, for the youngest generation as it was for the two older generations.
The question is what makes vocational inclusion so much more difficult to accomplish than other forms of inclusion? Changes in the labour market may serve as part of an explanation for this, as may the improvements in disability benefits. More important is probably the fact that the welfare state does not command the labour market in the same way they do with education. Consequently, to create a labour market more receptive to disabled people would require far stronger regulative measures than the welfare state seems capable or willing to apply. In addition, the individual experiences accounted for above indicate that not even street level bureaucrats seem willing to put too much energy into assisting disabled people in obtaining ordinary work. We have suggested that this hesitation, as well as the political hesitation to put into service more radical measures, at least partly has to do with the institutionalized role of modern welfare states in administering and monitoring the demarcation between those regarded as being capable of work and those exempted from the duty of work and hence qualified for income compensation, a demarcation through which disability itself is defined. Disabled people are publically defined as being exempted from the duty of work. Thus, monitoring those capable of working, particularly those suspected of being unwilling to work, easily becomes more imperative than assisting those who per definition are exempted from such duties. Moreover, the latter might even hamper the former by blurring the criteria on which to act.
If these arguments hold, the political ambition to open the labour market more fully to disabled people seems stuck in structural relations between the labour market, the welfare state and the institutional conception of disability. A restructuring of these relations would probably require a political will for stronger regulation of the labour market, but it will also require a deeper understanding of the way in which disability and dependency are socially produced, in which both the labour market and the welfare state play substantial roles. This again will require that policymakers pay closer attention to disabled people's experiences and understanding. As Roulstone and Barnes (2005, 3) point out, unlike other policy areas: 'disability employment policies have been almost hermetically closed to disabled people's influence.' Nonetheless, there is a risk that such an influence may challenge the prevailing alliance between government, capital and organized labour particularly characteristic of social democratic welfare regimes, in which responsibilities are carefully shared. Access to the labour market is controlled by employers and organised labour, while disability belongs to the welfare state.