Introduction

In Norway, schooling frames the life situation of children and youth, representing an expanding institutionalization of their lives. Each year, approximately 95% of the pupils leaving lower secondary in the spring term start upper secondary school in the autumn term. Approximately 10% of this age cohort has special educational needs (SEN) (Myklebust 2012). Following the principles of inclusive education in the national curriculum plan, the aim of education is to qualify all pupils for independent adult living, i.e., having work and their own home. However, the increased abstractness of school content (Hegna et al. 2012: 24) combined with the higher age at which students complete their initial schooling postpones students’ entry into the more challenging labour market. The number of young people who continue their education has increased considerably, while access to manual work has decreased. There are fewer jobs for unskilled individuals, leading to a higher risk of marginalization (Ellingsæter 2009: 154–155). Experiences in the school context seem to be an important pathway for understanding transitions from school to adult life.

Those in need of support with job strategies are young people out of work due to a lack of formal competences, psycho-social difficulties (PSDs) and social problems – or a combination of these factors (Dyrstad et al. 2014: 231). Living independently of the parental home involves taking responsibility for one’s own life, both economically and socially, without receiving regular parental guidance (Mulder 2009: 203). Leaving the parental home is an important marker of independent adult life. Successful transitions to adult life are normally characterized by growing independence. However, former SEN pupils clearly face a risk of being dependent. Modern society emphasizes individualization (Giddens 1991). Although this argument is rather weak empirically, it would lead us to expect a non-linear and varied transitional pattern of life course for vulnerable youth.

This article addresses the school experiences and the risk of dependence in adult life of former upper secondary pupils with non-obvious PSDs and learning difficulties (LDs) in Norway. In this article, we analyse and discuss the factors that push these pupils’ life course towards dependency by analysing prospective data, covering the upper secondary school years (1995–1999) and 2007, from a rather large number of youth (Myklebust et al. 2016).

Previous research

The theme of this article indicates that we are limited to reviewing studies on transitions to adult life with an emphasis on the employment and living situations of vulnerable youth. ‘Vulnerable youth’ refers to groups of people in a potentially vulnerable situation where difficulties can arise or be exacerbated and reinforced, leading to special education measures related to one of these difficulties (Osgood et al. 2005). What is known so far?

The review of relevant research is based on searching Norwegian (e.g., Idunn, Bibsys) and international (e.g., ERIC, JSTOR) databases by applying search words such as transitions to adulthood, vulnerable youth, learning difficulties, psycho-social difficulties and special educational needs. Dissertations and relevant books by leading researchers in the field are searched as well. Knowing the ‘shoulders’ on which a study is positioned is necessary.

Research on the situation among vulnerable youth experience may focus on a variety of aspects, such as transitions to adulthood among young people with chronic diseases (Sattoe 2015), transitions from child services to adult services (Beresford 2004) or the situation among individuals with intellectual disabilities (Kittelsaa 2011).

Olin and Ringby Jansson (2009) note an imbalance emphasizing that for youth categorized as having LDs and PSDs, individual-level problems hinder integration into work and everyday life more than external structural factors. Youth from socially disadvantaged backgrounds have further reduced chances of social inclusion (Riddell 2009: 87). Individuals with mental health difficulties or intellectual disabilities have low employment rates (World Health Organization [WHO] 2011: 237–239). Vulnerable youth experience poor outcomes in education, employment and family formation, marking problems in the transition to adulthood (Osgood, Foster & Courtney 2010: 216), and those with LDs are judged to be particularly vulnerable (Riddell 2009: 87). Both gender and type of difficulty are important in explaining participation in the labour market among former students with special needs (Rabren, Dunn & Chamber 2002: 25). Swedish (Tøssebro & Wik 2015: 19) and Norwegian (Brage & Thune 2015) research reviews have concluded that people with cognitive difficulties and PSDs have a very low employment rate. For marginalized youth, training opportunities lead away from the labour market and often involve repeated cycles of retraining (Riddell, Baron & Wilson 2001); these youth often face a position of dependency even early after their school years, either in the parental home or in shared group homes with other disabled people. This pattern is confirmed by the National Longitudinal Transition Study-2 (NLTS2) in the US (Sanford et al. 2011: 41). Young people with PSDs and LDs are vulnerable, and those with PSDs face the highest risk of dependence, although there is a noticeable variation within the groups (Levine & Wagner 2005: 228).

The reviewed studies indicate that standardization and a risk of dependency among vulnerable youth are collective indicators. However, definitions of different difficulties vary, as do the contexts of transitions in different countries. The review results are therefore not readily transferable to the Norwegian context. The present study will focus on a group of young persons who are not automatically considered disabled persons. They were categorized with difficulties in school (LDs and PSDs) but often lack a formal diagnosis.

The methodological pattern of the reviewed studies is varied as well. Longitudinal studies are valuable, providing an opportunity to compare and understand development over time. Cross-sectional studies are without explanatory direction. Qualitative intensive and prospective longitudinal studies of life histories covering a certain time span (Henderson et al. 2007) as well as larger extensive, longitudinal survey studies conducted over many years (Bynner 2011; Wadsworth 2011) are identified. Longitudinal studies of learning disabilities focus almost exclusively on youth and their earliest transitions, neglecting the later stages of life (Gerber 2012: 42). Most studies concentrate on one transition at a time (Shanahan 2000), for example, from school to work. Transitions are most often studied in isolation (Bynner 2012: 252). Quantitative cross-sectional studies dominate in Norway. Further literature searches identified two relevant Norwegian longitudinal studies (Helgeland 2007; Tøssebro & Wendelborg 2014) in addition to the studies with which the present research is associated (Myklebust et al. 2016). However, Norwegian studies based on prospective longitudinal data about work and independent living among individuals with PSDs or LDs compared with students with other categories of difficulties were not found. We know little about how these ‘invisible’ difficulties influence transitions to adult life. The present article explores this research theme.

Theoretical perspective

Life course transitions is the theoretical grounding of the present study. The life course is the sum of life trajectories and the transitions to different life areas formed by interacting structures and actors. From the life course perspective, people are placed in time and space, emphasizing the interaction between historical and biographical time – for example, how the abstract content of The Norwegian Reform 1994 affects SEN pupils’ transitions to adult life.

The aim and research theme of the present article points to the need for a duality of theory: 1. The lifespan psychological perspective, analysing the effects of individual characteristics (Boyd & Bee 2015; Elder, Johnson & Crosnoe 2003), e.g., variations of LDs, PSDs and other individual characteristics and how they interact over time, and 2. The life course sociological perspective, studying changing structural conditions over time, framing agency (Elder & Giele 2009). This duality is also closely related to the theory of deviance – the medical model focusing on individual and psychological characteristics, the social model emphasizing contextual factors and the balance between these two models in the relational model (Barnes, Mercer & Shakespeare 1999; Priestley 2003) – as in the present study. Life course theory complements the theory of deviance by analysing the norms embedded in transitional patterns – when is, e.g., upper secondary school normally expected to be completed? A stern norm (used by Statistics Norway as well) requires pupils to complete upper secondary school within five years. After that time, students are considered ‘drop outs’ or deviants. This approach has negative effects, particularly for those in vocational studies (Vogt 2017) – and SEN pupils most often follow vocational branches of study.

A life course is shaped by interactions with other persons; lives are socially interwoven. Social integration through linked lives of actors in family and school (Elder & Giele 2009) can be studied by analysing the social organization of special educational measures. Human agency is the capability of making weighted choices for future situations (Elder & Shanahan 1997) and is also related to individual characteristics, such as LDs and PSDs. Combined with human agency, the concept of linked lives contributes to the understanding of adult life adaptations, such as employment status and independent residence. Life course theory also encourages studies of how time, place and changing contexts (home, school, work and spare time arenas) and the transitions (George 1993) between them influence life course through new expectations and roles in adaptive situations. The transitions can be analysed by their location in time, that is, the timing of the transitions. The institutionalization of childhood through the school system communicates norms about what is to be completed at given time points – being ‘off-time’ or ‘on-time’ (Neugarten 1996: 102), indicating that society has a concept of age-appropriate behaviour and associated age-linked sanctions and options. This reflects a normative concept of time (Vogt 2017), connecting timing to concepts of deviation and normality. Life course trajectories analysed as timing processes also have turning points, producing a change in direction (Wheaton & Gotlib 1997: 1). Quitting school puts a young person off-time in a new direction. Whether they occur early, late or not at all, off-time transitions are considered unfortunate for later opportunities.

However, chances and misfortunes are unpredictable elements that can cause new chains of events and make changes more complex; ultimately, these elements have a crucial influence on the life course. A strong emphasis on human agency would portray the life course as more rationally controlled than it is (Myklebust 2007: 211). In longitudinal studies, unintended consequences must be studied, such as whether and how special educational measures with good intentions have the opposite effect intended. Reflecting on this very complex reality, we think of explanations in terms of risk and probability rather than in strictly causal terms.

Material and methods

In many cases, longitudinal studies cover small time spans for rather small samples (Gerber 2012). The data for the present study are selected from the 20-year longitudinal research project (Myklebust et al. 2016) and are critically judged on research ethic and approved by the Norwegian Data Protection Authority. The study is funded by the Norwegian Research Council, the Ministry of Education, Research and Church Affairs and Volda University College.

The participants were recruited from six Norwegian counties. All 760 participants received their education on special terms in upper secondary school. From spring 1996 to spring 1999, data were obtained via questionnaires about the learning situations and 13 different indicators of difficulties. This data collection was repeated twice every school year for each student based on expert judgments by the municipal office of educational and psychological counselling (PPT) or the school’s own expert on special education. A common scale (Skårbrevik 1996) was used to report the indicators of difficulties. The school provided information on the special education measures for each student, such as class type, the presence or absence of a teaching assistant, aspects of the learning environment and educational progression in upper secondary school.

Later, the former SEN pupils themselves answered telephone interviews or recorded their answers on a digital or paper version of the interview according to their preference. This extensive registration provided information about 760 persons, constituting the initial main sample of the present study. Later, several rounds of telephone interviews took place in spring 2002 (494 participants aged 23–24 years), amounting to 77% of those who could be contacted. Unfortunate circumstances at some of the schools, related to the storage of the lists of participant names and case numbers, resulted in the loss of contact information for 118 persons. In addition, two participants had died. Five years later, seven more had died. In 2007, 373 participants answered the request for information, amounting to 59% of the 633 participants available at that time. This sample was further reduced by 16 participants (moderate to severe LDs). The analysis sample of the present study is N = 357. Gerber (2012) found that the median size of 23 longitudinal studies reviewed is only 44 individuals. The analytic sample of the present study is, in comparison, rather large, and the time span from the late teens to nearly the thirties is rather long; both aspects are uncommon in disability research in Norway.

In longitudinal studies, respondent attrition is often a considerable problem. However, we met it by controlling for possible biases when comparing the central variables of the main sample and the analysis sample. In the present study, Myklebust et al. (2016) shows that possible effects of self-selection and sample bias are acceptable when samples from 1996 and 2007 are compared. Minor changes in the distribution of students along the central variables can be observed; the gender distribution is approximately the same (38% girls, 62% boys), and the proportion of students in vocational education programs is also quite similar, 76% and 78%, respectively. Many of the students have combinations of different difficulties. The variation in the 13 difficulties between the original main sample and the 2007 analysis sample is generally small; for example, the presence of PSDs is 34% and 32%, and the presence of LDs is 38% and 36%, respectively (Myklebust et al. 2016).

Variables and operationalization

Based on the research review and theory, a model that shows the possible longitudinal relationships between the dependent variables and independent variables is presented in Figure 1. The dependent variables – as indicators of dependence or independence are the presence or absence of full-time employment and dependent living situation.

Figure 1 

Assumed relationships among individual and structural factors, participation in full-time employment and type of living situation at age 28 to 29 years.

Individual traits in the model refers to the complexity of individual and psychological characteristics – i.e., the range of difficulties and functional levels. Gender is a relevant independent variable due to the different work experiences of women and men. The covariates under Conditions and processes in situations of interactions are contextual school factors referring to educational measures – for example, special classes and having a teaching assistant – and the status of school progression at the age of 20–21 years.

Dichotomized dependent variables (scored as either 0 or 1) are analysed by logistic regression, which is usually employed in attempts to explain whether something will occur or not, such as quitting school or not, having full-time employment or not, or more generally experiencing any phenomenon that can be expressed as an event/non-event. Having full-time employment in Norway means working 37.5 hours a week (reference category, 0). Not having full-time work means, e.g., being on sick leave, having occasional work, being a student, being unemployed or receiving social benefits. Independent living situation means living alone, together with a partner, with one’s own children or with a friend (reference category, 0). Dependent living situation means living with parents, in a residence hall, or in a residential collective with others who receive help.

All the independent variables in the model refer to a point in time before the measurement of the dependent variables, making reverse causation impossible. Generally, in dichotomizing the independent variables, the value of 1 (for example, being in a special class) expresses what is expected to have an explanatory effect on the dependent variable compared with the reference category 0, not in special class, which is expected to have no such effect. This analysis is based on answers from individuals with LDs or PSDs. The informants (N = 357) are categorized into four subgroups:

Neither LDs nor PSDs (N = 155)

LDs (N = 78)

PSDs (N = 44)

Both LDs and PSDs (N = 80)

Participants in the reference category 0 have neither general LDs nor PSDs; however, they are categorized as SEN pupils because of other difficulties. Of the 760 persons in the main sample, approximately one-third had PSDs. Considerable variations exist within the LD and PSD groups; the majority (nearly 80%) had LDs or PSDs of the mildest degree. The term general learning difficulty is defined as a cognitive difficulty linked to learning and intellectual functioning. The categories used here are identical to the Skårbrevik (1996: appendix 8.5, 3) classification in a national study of special educational practices. Those identified as having general LDs are graded according to the WHO’s categorization of intelligence.

Psycho-social Difficulties refers to difficulties in adaptation, problem behaviour, and social and emotional problems (Befring & Tangen 2004: 253) within the normal range of intellectual functioning. The category is widely used in educational practice in Norway, even without a more precise definition (Barne- og likestillingsdepartementet 2009). The following gradation of difficulties applies to this group (Skårbrevik 1996): less serious PSDs, interaction problems; serious PSDs, requiring treatment; very serious PSD, for example, psychosis.

Functional level is an additive index that does not include general LDs or PSDs; it is calculated from the four grade classifications of the 11 remaining difficulties. As such, functional level reflects the complexity of the full picture of difficulty – the extension of difficulties and the sum level of functionality degree. By including the variable into the analysis, it should be possible to control the results and avoid spurious relationships. However, the experts filling in the form might apply different criteria for disabilities in some cases. To overcome this problem, additional information was included in the registration form on each category, thus increasing data reliability. Such information includes signs of problems with which students struggle rather than indisputable facts; thus, these signs are useful indicators of the unified functional difficulties of the students. We chose to organize the variable values as terciles. The first and second terciles are reference categories (0) for the third tercile. The third tercile includes individuals who are believed to function at the lowest level in this range. A Cronbach’s alpha of 0.634 is considered acceptable for an index of 11 indicators.

The first variable from the school context of inclusive education is having a teaching assistant giving extra help or not (reference category, 0). The second variable, special class or not, reflects whether, in the first year of school, the student received all of his/her teaching within the ordinary class (reference category, 0) or in special classes (small groups of four or eight). The third variable, educational results, refers to whether the student, at the age of 20 to 21 years, has delayed school progression or quit, which is categorized as off-time, or has attained a formal competence/apprenticeship (reference category, 0).

Analysis

A main concern in logistic regression analyses is the gradually decreasing sample size, risking flaws in the analysis because of too many covariates in the model. We therefore followed Hosmer and Lemeshow (2000) focusing on the size of the smaller response category of the dependent variable rather than on the overall sample size; the size guideline is 10 (p + 1), where p is the number of covariates in the general model (cf. Figure 1). Having six covariates in the model, therefore, is acceptable. A logistic regression (SPSS Statistics version 23) analyses the simultaneous and relative explanatory power that independent variables have on the individual dependent variable. Ethical considerations have been carefully undertaken according to national criteria for social science research, e.g., informed consent, protection of informants’ anonymity, the possibility for the informant to withdraw from the research at any point of time and securely storing of data at institutional level. The procedures are approved by the Norwegian Data Protection Authority (DPA).

Results

The bi-variate analysis is the pairwise evaluation of the effect of an independent variable on the dependent variable: Having a teaching assistant, being in a special class, and being off-time in the school progression are clearly pairwise associated with the lack of full-time employment among those with PSDs. Women are largely without full-time work. Having PSDs or LDs is associated with a dependent living situation, as is functional level, having a teaching assistant and being in a special class. The multivariate analysis by logistic regression shows the simultaneous relative explanatory power of the type of difficulties, the functional level, the special educational measures and gender, analysing indicators of independence. The following effects, significant at the 0.05 level or lower, are observed:

  • Lack of full-time employment and independent living situation.
  • Individuals with difficulties, including PSDs, have a far greater risk of lacking full-time employment (expressed by the odds ratio) than students on special terms without such difficulties. Those with PSDs have a 2.5–3.5 times higher risk of lacking full-time employment in their late twenties, controlling for the other covariates in the model.
  • School-related factors still have a marked impact nearly 10 years after upper secondary school. Having a teaching assistant more than doubles the risk of lacking full-time employment. Being off-time in school progression also increases this risk.
  • The importance of gender is significant. When controlling for the effects of the other covariates, women in their late twenties face nearly seven times the risk of being without full-time employment compared with men of this age.
  • Individuals with PSDs have a much greater risk of being in a dependent living situation than individuals with other difficulties when controlling for the effects of the other covariates.
  • The risk is significant and even higher for those with both LDs and PSDs. The lowest functional level has a significant and rather large effect on the risk of being in a dependent living situation.
  • Gender has little effect on being in a dependent living situation.

Table 1

Multivariate analysis. How six independent variables affect adaptation to adult life in two domains of life among former students on special terms aged 28–29 years, odd ratios.

Risk of not having full-time employment
N = 357
Risk of dependent living situation
N = 357

Types of difficulties

0. No PSD or LD
1. LD but no PSD 0.7 1.3
2. PSD but no LD 3.4 ** 2.6 *
3. PSD and LD 2.4 * 2.9 *
Functional level

0. Functional level, first and second tercile
1. Functional level, third tercile 0.9 2.4 *
Teaching assistant or not

0. No teaching assistant
1. Teaching assistant 2.3 * 1.3
Type of class

0. Regular class
1. Special class 1.4 1.1
Educational results at the age of 20–21

0. Not delayed or did not drop out of school
1. Delayed or dropped out of school 1.7 * 1.0
Gender

0. Male
1. Female 6.8 *** 0.5 *
Nagelkerke R2 0.314 0.124

Note: +p < 0.10; *p < 0.05; **p < 0.01; ***p < 0.001. 0 is the reference category for each variable. The Nagelkerke R2 coefficient indicates that the model fit is reasonably good, especially in terms of full-time employment.

Discussion

The data are not from a large probability sample. This mid-range study is based on a theoretical sample (e.g., SEN pupils from all upper secondary schools in a selected number of counties [based on research-economic reasons] located north, south, east and west in a country with changing contextual conditions due to upper secondary school reform), restricting the generalizability of the results. The analyses of attrition show rather high response rates over time and only a minor sample bias. However, certain generalizations can be made when the findings are compared with those of other studies. In addition, an outcome cannot be determined from either the individual traits or structural factors alone because it is the result of complex interactions, and chance and coincidence may also have an effect. Studies cited from other countries may have contextual characteristics that are different from those of the Norwegian context, although such studies may still be considered relevant for illustration purposes. Vulnerable youth with PSDs who display normal cognitive functioning clearly face a far more challenging situation in adapting to adult life compared with other subcategories of youth. The vulnerability of youth with PSDs seems to be rather special. What can explain this pattern of actor-structure conceptual duality in light of the life course theoretical perspective of the present study?

Lack of full-time employment – trends of inclusion or exclusion?

The risk of a difficult adjustment to adulthood is far greater for individuals with PSDs, alone or in combination with general LDs, than for those with no such problems, controlling for the effect of other covariates. The relative explanatory power of individual characteristics – PSDs or LDs – when judging the risk of not having full-time employment is strong. Having PSDs seems to pose a special challenge, and this result is supported by other studies as well: Newman, Wagner, Knokey et al.’s (2011: 53) study on unemployment and (Riddell 2009: 87) a study on the transition to higher education. However, individual characteristics through the mechanism of linked lives can elicit and expose vulnerable youth to structural non-inclusive events when facing the labour market.

School factors, teaching assistant and educational results, have a significant relative effect 10 years after the start of school. The participants in this study grew up at a time and place where youth normally attend upper secondary school. Furthermore, Norway is estimated to be the OECD country with the lowest percentage of unskilled work (Hammer & Hyggen 2013: 16–19). Vulnerable youth, therefore, face stronger competition for employment, indicating how historical time meets biographical time in a way that affects an individual’s future.

The opportunities for an individual with PSDs to succeed will therefore depend on the expertise of the facilitators in school and later in working life. Shifrer (2013: 475) shows that both teachers and parents are more likely to focus on disabilities and hold lower educational expectations for labelled adolescents than for similarly achieving and behaving adolescents who are not labelled. This is an indication of the negative effects of being labelled with LDs or PSDs. In this way, structural mechanisms are sources of social expectations and linked lives affecting a life course trajectory. A lack of knowledge and competence in interactive support systems and the handling of the density of interrelated transitions (work and own home) might pose barriers for vulnerable youth, building unintended dependence.

The use of a teaching assistant as a structural factor increases the risk of lacking full-time employment almost 10 years after upper secondary school. Having a teaching assistant creates an informal flag that indicates to others that a person is different – possibly eliciting reduced expectations. In this way, school practices have negative effects on the balance between social integration and segregation. Ideologically and rhetorically, inclusion is emphasized, but its practice in Norwegian schools and communities is not consistent with this ideal (Kvalsund 2004: 174). Structural factors, such as well-intended special educational measures, tend to have unintended negative effects.

However, the structural mechanism of social expectations associated with transitions can also be related to deviations in timing: Those who were off-time in upper secondary school have an increased risk of lacking full-time employment nearly 10 years after completing upper secondary school – effects of time distant factors rather than proximal factors. Lack of education generally presents challenges in the transition to work (Simson 2014: 56). Newman, Wagner, and Knokey et al. (2011) found that former SEN students who had delays have an enhanced risk of not having full-time employment in their late twenties.

The structural aspect of gender has effects as well: Women are far more likely than men to lack full-time employment at the end of their twenties. Gender differences influence young people’s experiences in domains such as employment and family formation (Levine & Wagner 2005: 210). Women with disabilities generally have a higher risk of unemployment than men with disabilities (Borg 2008: 96). This tendency is observed for females in this study as well.

In an independent living situation – or not?

The structural covariates of the classroom/school context have no significant effects on the risk of being in a dependent living situation. Being out of work and the possibility of having one’s own home can influence each other. Exclusion from employment has a series of ‘knock-on’ effects (Riddel, Baron & Wilson 2001: 71). Being out of work and lacking money hinder young people from obtaining their own home, thus demonstrating how marginalization processes can interact across arenas (Anvik 2011: 5). Being without full-time work can have indirect barrier effects on being in an independent living situation. Indirectly, the contextual factors behind not having full-time work are relevant to whether individuals have their own living place or not. A lack of full-time employment means a lack of income, restricting self-determination and available alternatives.

We have shown that PSDs – even in milder cases – lead to a high risk of being in a dependent life situation. The explanatory factors behind this pattern can reflect the special qualities of the difficulty itself, helping actors’ interpretations and reactions to these difficulties as well as wider contextual aspects. Most children in Norway move out of the parental home during early adulthood. Only a small proportion of young adults aged 25 years still live with their parents (Dommermuth 2009: 1). Relatively few of the former SEN students are in a dependent living situation in their late twenties. However, individuals with PSDs or general LDs are over-represented. Those with the lowest functional level are over-represented in dependent living situations, which may be caused by a need for closer monitoring.

Youth with PSDs are judged as having a reduced capacity of self-determination (Lane et al. 2006: 334), which is linked to the character of the difficulty itself, and, through agency, reduced executive functioning (EF). EF includes functions such as short-term memory, planning, flexibility in facing changes, starting activities and restraining inadequate responses (Stenberg 2007). It is reasonable to expect that reduced EF influences people with PSDs as well. This conclusion implies a potentially increasing gap between external expectations towards invisible difficulties and individual capacity to handle the challenges of social adaptation. Helpers interpreting this gap as a lack of will rather than a lack of capacity probably contribute to the process of marginalization. Wendelborg and Tøssebro (2010: 709) refer to this process as a barrier to developing agency and independence among vulnerable youth with cognitive disabilities. To better understand the special adaptive pattern of the dependent living situation of young adults, previously categorized as upper secondary pupils with PSD, it seems necessary to study their adaptive situation more closely, for example, through retrospective qualitative interview data (Langøy 2017).

The interconnection of being out of work and without an independent living place can also be an indication of the effects of transition density elicited by the individual difficulties and the expectations and interpretations of PSDs through the mechanism of linked lives (family, peers and school). A lack of knowledge and adequate assistance from actors in school and support systems can contribute to cumulative disadvantages – the systemic tendency for inter-individual divergence in a characteristic (e.g., money, health, status) with the passage of time (Dannefer 2003: 327). Social barriers producing exclusion and withdrawal in school and society influence work status, independent living and the life course of the participants in this study.

Conclusion

Statements about the emphasis on the de-standardization and individualization of the life course in modern society would make us expect a non-linear and varied transitional pattern for vulnerable youth. The present study, however, indicates a standardization and risk of meeting collective mechanisms of dependency for vulnerable youth with PSDs or LDs. This study, conducted in Norway, shows that individuals with PSDs appear to face special challenges in adapting to adulthood compared with peers with other disabilities. The discussion reveals complex causes for this outcome, probably owing to the interactions among the nature of the difficulty, structural conditions and how their surroundings influence these individuals. However, it is thought-provoking that special educational measures with constructive intentions have negative effects on adaptation to adult life many years later. The mechanisms of life course transitions contribute to a deeper understanding of this pattern. It is no longer acceptable to legitimize the situation by referring to the negative consequences as unintended. The situation has special educational and socio-political consequences. Further intensive research is needed to better understand these paradoxical mechanisms.