Go to content

5. Preventive and rehabili­tative efforts

Preventive and rehabilitative efforts are central instruments for the PES in the Nordic countries, since physical or mental health issues which (strongly) limit individuals in daily activities constitute a central employment barrier in the Nordic countries. Among individuals outside the labour market in the Nordic countries, 43 pct. experience either physical or mental health issues, highlighting that various rehabilitative efforts cannot be neglected (Højbjerre et al., 2023b).
The chapter is separated into the following sections, which each presents the most recent Nordic literature on the topic:
We present evidence from caseworker interviews in section 5.3 and end the chapter with some concluding remarks in section 5.4. The literature from the systematic review is summarised in Table 5.1 at the end of the chapter.

5.1 Preventive efforts

Description of the instrument

Preventive efforts refer to proactive measures and strategies designed either to prevent individuals from becoming unemployed, i.e., through close ongoing support when an individual has found employment, or to minimise the duration of unemployment, i.e., through early intervention programmes and profiling with the aim to identify those who are at higher risk of long-term unemployment. In this chapter, we will focus only on the latter; the former will be discussed in chapter 8 on support, cross-sectoral and coordinating efforts.

The effectiveness of the instrument

The literature review did not reveal any articles on this topic, which is primarily a result of the fact that we required the articles in the review to 1) uncover a causal effect and 2) relate to one of the traditional target groups (e.g., young people, seniors, immigrants, and individuals with disabilities). Nevertheless, preventive efforts (including profiling) are an interesting topic with a possible potential, given partly by the latest developments in machine learning and AI, partly by the vast amount of personal data collected in the Nordic countries.
Profiling can be used in various settings and with various purposes. We understand profiling as a statistical identification or categorisation of unemployed people who are, for example, at risk of becoming long-term unemployed, and early interventions can be crucial for these individuals (see, e.g., Anderson & Minas (2021)). Experience shows that some unemployed people find work again very quickly on their own, while others need active programmes of varying length and content. If it were possible to determine in advance which people are at risk of ending up in specific unemployed groups, it would be possible to target preventive efforts at an early stage, thereby achieving better resource utilisation in the PES and a more meaningful effort for the unemployed person. 
Profiling is not a new tool in the toolbox of employment instruments in the Nordic countries, and it is already used in practice in various ways inside as well as outside the Nordic countries (Eskelinen, Seier Petersen & Bolvig, 2015; Beskæftigelsesministeriet, 2014; Dahl & Myklathun, 2022; Forslund & Vikström, 2011). 
For example, profiling has been used in Sweden for more than a decade, and the Swedish Employment Service (Arbetsförmedlingen) recently introduced a new, more complex model (a flexible survival model with time-dependent covariates) to assess the support needs of unemployed individuals. In this model, they use 24 covariates, including variables such as age, educational level, country of birth, and medical history, to predict the unemployed individual’s probability of getting a job. This probability determines which effort category the tool recommends (Helgesson et al., 2023). The Swedish experience with profiling for determining support needs among jobseekers is further described in Box 5.1.
In Norway, they have also introduced a profiling tool in the PEC. They use a simpler, rule-based algorithm compared to Sweden’s. The algorithm is based on inputs such as age, education, and work experience. Recently, the Norwegian PES (NAV) investigated whether the profiling tool resulted in more accurate case management for the jobseekers. First of all, they find that there is some discrepancy between the suggestions from the profiling tool and the caseworker’s decision when the profiling tool recommends more than no or little effort. This may be due to the fact that the algorithm has been deliberately designed to over-profile, i.e., suggest too much effort rather than too little (Dahl & Myklathun, 2022). Relatively simple algorithms like the rule-based algorithm in Norway have also been introduced in Iceland (The Minus 12-Project). Further, some job centres in Denmark have also partly used a profiling tool, which, however, has resulted in compliance issues related to the use of data, and the tool is therefore no longer in use (Moreau, 2022; Moreau, 2023).
Box 5.1 Profiling for determining support needs among jobseekers – experiences from Sweden
Profiling in the PES (Arbetsförmedlingen) in Sweden has been in use for over a decade, and a new model to determine the support needs among unemployed individuals was recently introduced.
Inputs and data in the model
The new model employs a flexible survival model with time-dependent covariates utilising various inputs, such as gender, age, education level, specialisation, country of birth, municipality, and disability status. It also includes enrolment time in the PES to enhance accuracy. Data for the model comes from jobseeker information provided during registration and from the PES.
Output of the model
The model predicts the probability of an individual transitioning to employment or education, recommending one of three support levels:
  • Minimal support, utilising digital services.
  • Extra support needed for job search or education.
  • In-depth support to be further investigated by the public employment service.
The model performs well compared to similar models, with up to 80% accuracy in predicting long-term unemployment. The tool's recommendations serve as inputs for caseworkers, who make final assessments based on comprehensive data about the jobseeker.
Potential in profiling
Profiling tools as the one described above should never be used in isolation, but can provide useful inputs to the caseworker. In that way, profiling tools can help focus the caseworker’s limited time on the jobseekers who need support the most. As we will see in section 8.1, the caseworker’s time spent face-to-face with the jobseeker is crucial for positive employment outcomes among the most vulnerable individuals.
Source: Helgesson et al. (2023) & Arbetsförmedlingen (2024).

Key considerations regarding the instrument

The latest advancements in AI and machine learning increase the potential of profiling in the active employment programmes

Advancements in artificial intelligence (AI) and machine learning are revolutionising the way we can profile individuals at risk of long-term unemployment. By leveraging sophisticated algorithms and vast datasets, these technologies can now more accurately identify key indicators and patterns that signal potential risk of long-term unemployment. This enhanced profiling capability allows for the development of targeted interventions and support programmes, aimed at addressing specific needs and improving employability. There is, however, some important considerations to make before designing and implementing such profiling tools. 

Important to involve caseworkers in the recommendation from the profiling tool

To effectively utilise the profiling tool in assisting the unemployed, caseworkers need access to the results of the profiling assessment. This includes the individual’s score and the significance of each variable in determining their categorisation. By understanding the importance of each variable, caseworkers can develop a targeted approach for further support. This insight enables them to identify and address specific barriers relevant to each individual. Without this ‘behind-the-scenes’ access, there is a higher risk that the profiling tool will be seen as an additional burden rather than a valuable resource (Eskelinen, Seier Petersen & Bolvig, 2015; Beskæftigelsesministeriet, 2014). Even though they have implemented a relatively complex profiling tool in Sweden, it is still possible for the caseworkers to get an insight into how the various variables contribute to the recommendation from the model (Helgesson et al., 2023). 

Important to consider how to involve the jobseeker and to be compliant with national legislation

When designing and implementing profiling tools, there are various other considerations to make. For example, it is important to consider how the jobseeker is involved. On the one hand, it is important to make the jobseeker aware that the public employment service uses a profiling tool. On the other hand, it is also important to consider how much you want to involve the jobseeker in the results from the profiling tool. If the unemployed are made aware of the key barriers and/or their assigned profile category, it can lead to unnecessary concern when this information is not followed up with a detailed action plan to address the problems they may face (Eskelinen, Seier Petersen & Bolvig, 2015).
Further, it is also important that the profiling tools comply with national legislation in the Nordic countries regarding which personal data is handled and how it is managed. For example, a recent decision by the Danish Data Protection Agency (Datatilsynet) highlights that municipalities cannot rely on consent as a legal basis for processing personal data when using algorithms to predict long-term unemployment. This ruling emphasises that simply obtaining consent from individuals is not sufficient, as it may not be considered voluntary due to the power imbalance between the unemployed and the authorities controlling their welfare benefits (Moreau, 2022; Moreau, 2023). This decision highlights the importance of correct use of data when implementing various profiling tools.

5.2 Rehabilitative efforts

Description of the instrument

The employment instrument rehabilitative efforts refers to initiatives and tools designed to help individuals re-enter the labour market, primarily after shorter and longer sick leave. It includes efforts which help individuals gradually return to their previous work, but can also include reskilling efforts that help individuals who are not able to continue in their current work find new job opportunities. These efforts typically include a combination of traditional active labour market initiatives (such as labour market training, job placement services, counselling, and support systems) and medical initiatives (such as physical therapy), with the purpose of enhancing employability and ensuring sustainable employment.

The effectiveness of the instrument

In the systematic review, we found relatively comprehensive literature on this topic, primarily concerning individuals with varying degrees of health issues. Some studies focus on early interventions for those with less severe health problems, while others evaluate efforts to help long-term sick-listed individuals return to employment. Most of the literature originates from medical journals, where Randomized Control Trials (RCTs) are more common. This is advantageous in terms of identifying the causal effect of various interventions, although the sample sizes in these studies are typically smaller. One challenge regarding RCTs is the external validity of the results, i.e., whether the findings can be generalised to broader populations or different settings beyond the specific context in which the study was conducted.
Several recent studies have focused on interventions aiming to promote work participation among individuals with health issues. Among the various interventions, vocational rehabilitation services have been associated with favourable outcomes in previous reviews (see, e.g., Suijkerbuijk et al., (2017)). Vocational rehabilitation encompasses goal-directed interventions with the core objective of enabling work participation. These findings are in general supported by the literature from this systematic literature review. For example, research from Finland demonstrates that vocational rehabilitation among recently employed individuals with musculoskeletal and mental-related work disabilities has positive and significant effects on work participation. The main services of the scheme included work try-outs, work counselling, and training, and the effect is estimated to be between 11.8 and 7.2 percentage points 1–3 years after programme participation (Leinonen et al., 2019). Similar positive indications of vocational rehabilitation are found in Iceland (Guðbrandsdóttir & Ingimarsson, 2022).
Similar positive results are found in a Swedish study investigating the return-to-work factor among patients on long-term sick leave due to mental health issues and/or chronic pain. In this study, patients were randomly allocated into either 1) multidisciplinary team management (which included meetings with a psychologist, a physician, an occupational therapist, and a social worker), 2) acceptance and commitment therapy (ACT), or 3) a control group. The findings suggested that the multidisciplinary team management increased the return-to-work factor among the patients, whereas the ACT did not result in an increase (Berglund et al., 2018). These results are confirmed among women with long-term sick leave due to pain syndrome or mental illness (Lytsy et al., 2017).
An interesting question in this context is whether the waiting time before participating in vocational rehabilitation is decisive for later employment outcomes. Norwegian research exploiting a natural field experiment demonstrates no effect of reduced waiting time on subsequent employment. In the short run, however, it increased the fraction of individuals receiving permanent disability benefits, whereas reduced waiting time had no effects on receiving permanent disability in the long run. Hence, the reduced waiting time speeds up the transition process from temporary to permanent disability benefits in the short run (Hauge & Markussen, 2021).
Another interesting point of view is whether gradual return to work (i.e., some combination of part-time work and sickness benefits) can produce positive employment outcomes in the long run. This is investigated in a Danish paper, and the results suggest that a greater emphasis on offering gradual-return-to-work programmes is associated with an increase in regular employment and self-sufficiency among sick-listed workers. These programmes are associated with strong and long-lasting effects, however primarily for workers sick-listed from regular employment and for those with physical disorders (Rehwald et al., 2018). These findings are supported by research not identified in the systematic review. For example, research from Norway concludes that the use of graded (partial) rather than non-graded (full) sickness absence significantly improves the propensity of employment in subsequent years (Markussen et al., 2012). 
Finally, part of the literature also examines the employment effect for sick-listed workers from various types of cognitive training, such as self-management support, as evaluated in Rotger & Rosholm (2020), or compensated cognitive training, as evaluated in Fure et al. (2021). The results from the literature regarding various forms of cognitive training are rather mixed. For example, Danish research evaluates the effect of a control-enhancing intervention, i.e., a self-management support course, on the return-to-work factor among long-term sick-listed workers. The research demonstrates that the intervention had no effect on individuals who were control-optimist prior to taking the course, whereas individuals who were control-pessimist prior to participation in the course experienced positive employment outcomes from the course (Rotger & Rosholm, 2020). Research from Norway also demonstrates positive effects from compensatory cognitive training among individuals who have suffered from traumatic brain injuries. The results suggest that the intervention might help patients with mild-to-moderate TBI who are still sick-listed 8 to 12 weeks after injury to an earlier return to stable employment. It is, however, important to mention that the compensatory cognitive training was combined with vocational training in the form of SE, where participants were supported by an employment specialist in returning to their current jobs by working at their actual competitive workplace, which might also explain some of the positive effects (Fure et al., 2021).
On the other hand, results from an evaluation of treatment with acceptance and commitment therapy do not demonstrate any positive employment effects among women with long-term sick leave due to pain condition or mental illness. The therapy is a form of cognitive behavioural therapy that uses acceptance and mindfulness strategies together with behavioural strategies to increase function and quality of life (Lytsy et al., 2017). Similar results from acceptance and commitment therapy are found in Berglund et al. (2018).
Lastly, another important aspect of rehabilitation includes reskilling efforts to help individuals, who are not able to continue in their current work, finding new job opportunities. We have not identified any literature concerning this issue, but an Icelandic paper investigates the potential in validation of transversal skills, which can be an important part of a reskilling-process. In this paper they use semi-open interviews with participants who had completed a transversal skills validation process. The results revealed, increased awareness of what they are capable of, what they had learned in life and work, and how they could use their skills in the labour market. This made them more hopeful and capable in career planning, increasing self-efficacy (Jónatansdóttir et al., 2023). Similar positive results are found in Hreinsdóttir & Sigurðardóttir (2022).

Key considerations regarding the instrument

Gradual return to the labour market appears to provide long-lasting employment effects among long-term sick-listed workers

There is a vast amount of evidence suggesting that gradual return to work produces favourable employment outcomes compared to full-time sickness absence, also in the longer run (see, e.g., Rehwald et al. (2018)). The effect, however, seems to be present only among workers sick-listed from regular employment and for those with physical health issues (Rehwald, 2018). Gradual return to work not only facilitates a smoother transition back to regular employment but also helps individuals rebuild their confidence and work capacity incrementally. By allowing a gradual return to work, employees can better manage their health conditions, while progressively increasing their working hours, thereby reducing the risk of relapse or further health complications.

Multidisciplinary efforts are in general found to be effective in increasing the labour market participation among sick-listed workers

People on long-term sick leave often have a long-lasting process back to work, where the individuals may be in multiple and recurrent states, i.e., working or receiving different social security benefits, and they may shift between these states over time. Collaboration among health care providers and vocational experts ensures a holistic support system, which ultimately seems to promote faster and more sustainable reintegration into the workforce. For example, multidisciplinary efforts seem to produce positive employment outcomes among sick-listed individuals in Sweden. These efforts include individual meetings with professionals from a multidisciplinary team, including a psychologist, a physician, an occupational therapist, and a social worker. The team then meet without the participant to establish an individualised rehabilitation plan, which is later brought back to the participant by one of the team members. The team meet weekly during the project period to evaluate the situation and synchronise the planned or ongoing activities for each participant (Berglund et al., 2018). However, such interventions are comprehensive, making it important to consider their overall socio-economic impact.

Traditional ALMPs seem to have limited effects on regular employment for sick-listed workers

Opposed to the results in chapter 4, the literature on the return-to-work factor among sick-listed workers does not demonstrate positive employment effects from traditional ALMPs such as vocational guidance, internships, and on-the-job training. Participation in a traditional activation programme promotes unemployment during the first year after enrolment and is clearly not helping sick-listed workers reintegrate into the regular labour market. Similarly, traditional activation programmes are ineffective in reducing the degree of welfare dependency (Rehwald et al., 2018).

5.3 Caseworker interviews

The caseworker interviews bear witness to the great diversity and complexity involved in assisting sick-listed or other citizens intended to benefit from preventive and rehabilitative efforts by the PEC in the Nordic countries. There are important differences between sick-listed citizens who are temporarily absent from a job (that they still have) because of illness and citizens who have lost their job because of illness or accident and have to find another one once their health situation allows renewed attachment to the labour market. Moreover, there are also differences in the types of rehabilitation needed for different types of health problems, whether physical or mental, and whether these problems are well-diagnosed from a medical point of view and their consequences well-elucidated in an employment perspective. It may take time to elucidate exactly what type of job a sick-listed citizen is able to return to, for example if this person has acquired a physical impairment through many years of hard physical work, such as chronic back, neck, or shoulder problems.
While this chapter initially surveyed some literature on preventive efforts and profiling, the caseworker interviews generated no knowledge on that topic, because almost no caseworkers were involved in that type of work. The caseworkers who had knowledge to share on the topics of this chapter rather worked with gradual return-to-work schemes and, especially, rehabilitation.

Gradual return to work

Gradual return to work appears to be a commonly used tool in different Nordic countries in relation to sick-listed persons who need to maintain a relation to their workplace in the aftermath of being affected by illness or an accident. For instance, one Norwegian caseworker being asked about the extent to which they used gradual return to work in her employment services (NAV) explained that they evaluated together with the sick-listed citizen how much this person could work. For example, depending on the evaluation, the person could be 20% sick-listed and work four out of five days a week. In Norway, the right to sickness unemployment benefits extends up to a year depending on how ill the person is. During this period, the employment services (NAV) would hold dialogue meetings with the employer and the sick-listed person to consider what could be done to help this person return to work. They might consider different types of physical aids and appliances or home-office solutions. It is clear from the interviews that such dialogues and work adaptations can be crucial in assisting a sick-listed person in maintaining a job. That being said, the interviews also revealed that cases where the citizen had a job and became ill but was able to gradually return to the job were not the most difficult cases and typically not cases involving a more complex set of barriers, because the citizen still had a job, an employer, and typically also recent work experience, qualifications, employment-relevant social networks, etc. The much more challenging cases arose when a citizen became ill for a long time and lost his job or had only had jobs intermittently for a long period because of physical or mental illness and/or alcohol abuse.

Rehabilitation

Although many types of rehabilitation and rehabilitation schemes were mentioned during the interviews with the Nordic caseworkers, it is difficult to extract very specific knowledge about which types of rehabilitation are most efficient. Again, this is due to the heterogeneity of problems and the diversity of persons that the caseworkers meet, as they may require (very) different types of rehabilitation. One Finnish caseworker gave this account in relation to young people suffering from mild psychological problems:
We have a rehabilitative activity. It’s adjustable: it can be one day or two days a week. It would often be two days a week. We have to collaborate with social services. It’s not [time-] limited. They can opt out if it works out and maybe afterwards go to ‘work try-out’. After that, they will be using the wage subsidy. That would be the ideal palette of tools.”
The quote shows that the right dose and the right sequence of activities are crucial to success (we return to wage subsidies in chapter 7). A Norwegian caseworker explained in relation to seniors who had lost their job after a period as sick-listed that job training was an important part of the rehabilitative efforts and sometimes largely sufficient. Initially, neither the caseworker nor the citizen would know exactly how much effort and work the citizen would be able to sustain during such job training and whether it would be sufficient to help them become ready for the labour market again. He explained:
We still have the problem that they don't know how much they manage in their jobs. It is my experience that it is largely sufficient because it goes much better than they thought when they are allowed to warm up in the jersey.
However, it is also clear that a different type of long-term rehabilitation can be needed for citizens who are sick with severe physical (e.g., neurological) or mental health problems. A Swedish caseworker an occupational therapist, working with persons with mental health problems, e.g., autism, ADHD, ADD, and burn-out, explained that bringing such persons closer to the labour market was often a lengthy process that could easily take up to a year. Many of those persons might also suffer from different types of pain and/or fatigue, and part of her job was to figure out what in their occupation or their personal life situation might cause these symptoms. Testing the person with different types and doses of work tasks would help her establish an ‘occupational balance’, which might eventually lead to the person returning to work at, e.g., 40–50% of ordinary working time. However, it often took a substantial amount of time for the process to be a success.

Time as a crucial factor

Many caseworkers return to a particular point of importance in these processes, namely that they take time – the healing from a medical perspective, the citizen’s reorientation towards a new type of work life (which the caseworkers are involved in bringing about), and the citizen’s retraining for new work tasks. Some caseworkers criticize national employment legislation for not allowing sufficient time for these processes and for making some citizens return to searching for jobs before they are ready to do it. Moreover, the rehabilitation processes may be slowed down or even stalled if the doctors cannot provide a diagnosis of an underlying health problem. It appears that particularly intractable barriers must be overcome if a citizen has a substance or alcohol abuse problem combined with some type of psychiatric illness that the psychiatrists are unable to diagnose properly as long as the citizen is actively abusing drugs or alcohol. A Danish caseworker told a story of such a case: a male craftsman that she met repeatedly over several years because he kept losing his jobs due to this combination of alcoholism and psychiatric illness that could not be properly diagnosed in the public health system. Some employment barriers appear very difficult to overcome, and, unfortunately, our interview material contains no examples of good tools to solve problems where the barriers seem to become a Gordian knot.

5.4 Concluding remarks

In conclusion, this chapter examines the effectiveness of preventive and rehabilitative efforts in the Nordic countries. These efforts are particularly essential, given the high prevalence of health-related limitations among those outside the labour market, with 43 percent experiencing significant health challenges.
On the one hand, the systematic literature review did not reveal any articles related to preventive efforts. However, the discussion highlighted the potential of profiling tools, especially with advancements in AI and machine learning. Profiling can help identify individuals at risk of long-term unemployment, allowing for early and targeted interventions. Various profiling tools with varying degrees of complexity have also been implemented in Sweden, Norway, Iceland, and – to some degree – Denmark, illustrating that it might be fruitful, though these tools should complement caseworker expertise, not replace it. However, none of the interviewed caseworkers had experience with working with profiling tools.
On the other hand, the systematic review contained a lot of literature related to rehabilitative efforts. These efforts are pivotal for individuals returning to the labour market after health-related absences. The systematic review revealed substantial evidence supporting the effectiveness of vocational rehabilitation, multidisciplinary approaches, and gradual return-to-work programmes. These interventions seem particularly beneficial for those with physical and mental health issues, facilitating a smoother and more sustainable reintegration into the workforce. These findings are supported by the Nordic caseworkers. However, they highlighted the complexity and diversity of rehabilitative efforts and emphasised the importance of tailored, individualised support, which often requires time and flexibility.
Table 5.1 Identified literature in the systematic review related to preventive and rehabilitative efforts
Study
Instrument
Target group
Time perspective
Effect
Method
Berglund et al. (2018)
Sweden
Two different vocational rehabilitation programmes: multidisciplinary team management or acceptance and commitment therapy
Persons with mental health issues and/​or chronic pain
1 year after programme participation
Positive employment effect from the multidisciplinary team management
RCT
Fure et al. (2021)
Norway
Compensatory cognitive training and SE
Individuals with traumatic brain injury
1 year after programme participation
Positive
RCT
Hauge & Markussen (2021)
Norway
Reduced waiting time for vocational rehabilitation
People with partial work capacity
Up to 2 years after programme participation
No employment effect but led to earlier transitions into permanent disability benefits
Natural field experiment
Hoff et al. (2021)
Denmark
Integrated mental health care and vocational rehabilitation intervention.
Individuals on sick leave with anxiety and depression.
12 months after programme participation.
The integrated intervention resulted in a higher proportion in work
Logistic regression
Langagergaard et al. (2021)
Denmark
Brief and multidisciplinary intervention (including brief intervention and coaching by a case manager)
Individuals on sick leave due to lower-back pain (LBP), with either strong or weak job relations
1 year after programme participation
Brief intervention resulted in higher RTW rates than multidisciplinary intervention for employees with strong job relations. There were no differences in RTW rates between interventions for employees with weak job relations
RCT
Leinonen et al. (2019)
Finland
Vocational rehabilitation
Individuals with musculoskeletal and mental-related work disability
Up to 3 years after programme participation
Positive
DiD
Lytsy et al. (2017)
Sweden
Multidisciplinary assessments and individual rehabilitation interventions
Women with long-term sick leave due to pain syndrome or mental illness
1 year after programme participation
Positive
RCT
Rehwald et al. (2018)
Denmark
Activating sick-listed workers through a combination of weekly meetings with caseworkers and intensive mandatory RTW activities
Long-term sick-listed individuals
3-year follow-up
Partial sick leave had positive effects on regular employment
Traditional ALMPs appear to have no effect at all
Large-scale RCT, DiD, and IV
Rotger & Rosholm (2020)
Denmark
A self-management support course
Long-term sick-listed individuals
19 months
No overall employment effect; however, positive effect on individuals with negative control-beliefs prior to the programme
RCT, FE, and causal forest
Stapelfeldt et al. (2020)
Denmark
A combined acceptance and commitment therapy and IPS model
Cancer survivors undergoing cancer treatment
The intervention continues for a maximum of 1 year or until RTW
A small insignificant effect across all cancer patients
Current Controlled Trials
Note: DiD is an abbreviation for Differences-in-Differences. RCT is an abbreviation for Randomized Control Trial. IV is an abbreviation for Instrumental Variables. FE is an abbreviation for Fixed Effects.
Nordic Council of Ministers I www.norden.org/publications I pub@norden.org