Changes in health-related rehabilitation trajectories following a major Norwegian welfare reform | BMC Public Health

Our study addressed three main empirical questions:

  1. 1.

    What were the most typical health-related rehabilitation trajectories for young Norwegian inhabitants aged 23–27 between 2004–2019?

  2. 2.

    Did trajectories and composition of health-related benefit recipients change over the observation period?

  3. 3.

    In parallel with the welfare reform, do we see improved labour market outcomes in our study population?

Preventing early labour market exclusion is top priority in Norway and internationally, however little is known about the health-related rehabilitation trajectories of high-risk young people who have not yet transitioned into permanent disability pension. An important priority for our research is to help close this knowledge gap.

Our study population is, in general, a vulnerable group with poor labour market outcomes. The majority are early school leavers, following disadvantaged trajectories consisting of welfare dependency, unemployment, minimal educational activity and unstable, low-income work. We identified six broad trajectory types, and while some could be considered less “problematic” than others, almost no-one in the study population ended up self-sufficient through work.

Reclassification of welfare statuses and medicalisation

Over the study period we witnessed considerable increase in the proportion following trajectories dominated by long periods of health-related rehabilitation (Clusters 1, 4), accompanied by a substantial decline in the share following the unemployment O.H trajectory (Cluster 5). Furthermore, the average time spent in health-related rehabilitation increased by nearly twenty months while the average duration of “occupational handicapped unemployment” decreased by nearly seventeen months.

What we could be observing here is an administrative re-categorisation of unemployed people waiting for rehabilitation. After WAA was introduced, many individuals categorised as “Occupational handicapped unemployed” were instead provided with a primarily health-related status. It is possible that the increased disability pension incidence over the observation period can partly be attributed to this reclassification of welfare statuses. Giving people a dominant label of poor health, rather than a status related to absence of work, may have the unfortunate effect of steering people into permanent disability benefits (which are also primarily focused on poor health).

Societal perspective

Policymakers in Europe have expressed growing concerns about the medicalisation of unemployment, as it significantly burdens society [30]. In Norway, the combination of economic transformations, welfare system reforms and healthcare seeking behaviours may have created a paradigm shift towards medicalisation of young people’s labour market struggles [6]. Particularly troubling is the high incidence of young individuals receiving disability pensions, with historical data indicating low success rates for reintegration into the workforce [3]. This situation not only imposes substantial financial obligations on the government regarding social security payments but also deprives society of the valuable contributions these young individuals could have made.

Recent research in this field indicates that the medicalisation of unemployment has become more prevalent [41,42,43]. This trend can be partly attributed to better understanding the unemployed people’s health challenges. However, there is evidence that strict eligibility requirements for non-medical benefits [44] increases the emphasis on illness or disability as justifications for accessing benefits or being exempted from certain obligations [45,46,47] This could be a contributing factor in Norway where unemployment benefits are only available to those who have earned the right through work. The only non-medical benefit option available to individuals without previous work experience is a meagre, means-tested social assistance benefit, considered the last safety net in the Norwegian welfare system.

As welfare reforms worldwide move towards consolidated, one-stop-shop services and eligibility criteria for non-medical benefits become stricter, there is a concerning possibility that a growing number of marginalised young adults could be reclassified as ill and receive health-related benefits. Such a trend could increase permanent welfare dependency and labour market exclusion among this group internationally.

Individual perspective

While increasing health-related welfare dependency impacts society negatively, individuals reliant on health-related welfare may perceive it more favourably. This perception could be due, in part, to the reduced stigma associated with being categorised as sick rather than unemployed. Social legitimacy research on welfare benefits has shown that society perceives sick persons as more deserving than unemployed individuals [48]. Moreover, the medicalisation literature suggests that sickness relieves individuals from social role obligations, which helps justify inactivity and benefit receipt [30].

Furthermore, our study found that individual sequences became less turbulent over time, indicating increased stability. This stability resulted from less shifting between low-income employment and unemployment statuses accompanied simultaneously by longer spells spent in health-related rehabilitation, and increased uptake of disability benefits, which often represent a stable and generally permanent state. While previous research consistently highlights the health benefits of being employed compared to being jobless, evidence also suggests that poor working conditions can deteriorate one’s health. Current trends towards work fragmentation and flexible labour markets [49] have negatively impacted low-skilled young people [50], trapping them in low-paid, insecure work and unemployment cycles [51]. Job insecurity poses a comparable threat to health as unemployment [52], emphasising that societal perspectives of what is beneficial may not always align with the best-case scenario for individuals.

Work participation

Between the first and last cohort we observed a 16.3% increase in the proportion of individuals following the least problematic trajectory, characterised by normal unemployment, unstable income, and sickness absence (Cluster 3). However, regression analyses found no significant association between the trajectory probability and a specific cohort. Over the observation period, the average time spent with any income remained relatively stable, increasing by only one month, while the average time spent in “no work or education” decreased by one-and-a-half months.

At first glance, this is a rather unintuitive finding. Given the large-scale economic transformations and increased proportion of early school leavers that occurred over the observation period, we expected to see decreased work activity overtime. This suggests that WAA caters to different individuals than the previous health-related rehabilitation benefits, which makes sense given that composition of health-related benefit recipients changed across cohorts.

Compositional changes

Parental factors

We discovered a decline in parental disability pension dependency over time, while the average level of parental education improved. These findings could indicate that individuals on WAA came from increasingly less disadvantaged social backgrounds, which may have brought them closer to the labour market than their predecessors.

The decreasing prominence of intergenerational welfare transmission is an intriguing finding, running counter to current international and Norwegian literature on the topic [53, 54] and requires further exploration. However, it is also important to note that the trajectory ending in permanent disability pension (Cluster 6) was significantly associated with having two parents who were disability pensioners. Moreover, the proportion of individuals following this trajectory tripled over the observation period. Interestingly, the trajectory probability was not associated with parents having a below average level of education.

Educational attainment, country background and mental health problems

We also find evidence that, in some aspects, young people on health-related rehabilitation benefits are becoming more disadvantaged overtime. The role of education in determining employment and income prospects for young individuals is crucial [55], for example incomplete upper secondary education increases the risk of long-term exclusion from the labour market [55]. Our study reveals a concerning trend of a higher proportion of early school leavers over time, indicating a growing educational disadvantage across cohorts. Additionally, we found a significant association between high school dropout and following a trajectory leading to a disability pension dependency.

Our study population became more ethnically diverse over time, likely due to the inflow of non-Western immigrants into Norway [56]. This has resulted in an increasing disadvantage based on country background. Individuals from non-European countries tend to have lower educational attainment than native-born individuals and face a higher likelihood of unemployment, especially during challenging labour market conditions [10]. Moreover, they encounter more difficulties in finding new employment opportunities [10].

Existing literature also highlights the significant contribution of mental health conditions to welfare dependency among young adults. Over time, there has been a noticeable rise in self-reported mental health problems among Norwegian adolescents and young adults, accompanied by an increased proportion seeking treatment for mental health issues from primary care services. These findings align with our observations regarding educational status and country background. Young people with mental health conditions are more susceptible to dropping out of education and face substantial obstacles in accessing the labour market [57]. Markussen & Seland, 2012 [58] find that approximately half of early school leavers attribute their dropout to poor mental health.

Individuals with a non-European country background face a double disadvantage concerning education and mental health. Not only do they have a higher risk of leaving school early, but a larger proportion report mental health problems [59]. Furthermore, refugees are more likely to consult their general practitioners about mental disorders than the general population [60]. Several factors contribute to this inequality, including lower socioeconomic status, discrimination based on immigrant backgrounds, language barriers, and exposure to adverse life events [61, 62]. Interestingly however, country background was not a significant risk factor for any trajectory type. Most notably, it did not increase the probability of following the trajectory ending in disability pension.


Our analysis also provides evidence that gender plays a role in young people’s work and welfare trajectories. Men are more likely to follow trajectories characterised by unemployment and some labour market activity while women were more likely to participate in long spells of health-related rehabilitation. Women are overrepresented in all cohorts, although the gender gap remained relatively stable overtime.

Future research

Investigating young health-related rehabilitation trajectories using sequence analysis has provided valuable new insights. The influence of administrative status re-categorisations and our observations regarding the diversification of health-related welfare dependency are intriguing findings that warrant follow-up with causal research.


Using high-quality, register data sets allowed us to bypass the quintessential challenges associated with longitudinal surveys such as low return- and high attrition rates. Furthermore, our data encompasses the entire Norwegian population rather than a representative sample, which makes it possible to study small but important and hard-to-reach groups.


We have identified several limitations. Firstly, our analyses do not include diagnostic information, it would be both interesting and relevant to know if the probability of following a particular trajectory was influenced by one’s diagnosis. Another limitation concerns generalisability, the study context is primarily relevant for countries with comprehensive, generous welfare systems and skills-biased labour markets. In addition, due to the descriptive nature of our study, we cannot determine whether observed associations reflect cause-and-effect relationships.


Leave a Reply

Your email address will not be published. Required fields are marked *