This chapter highlights that stronger disability and work capacity assessments alone cannot substantially improve labour market outcomes for people with disability. It underscores the critical importance of early identification and early intervention, both for workers experiencing long-term sickness and for unemployed individuals with unrecognised or fluctuating health conditions. The chapter proposes a structured return‑to‑work process, enhanced employer responsibilities, and more systematic involvement of the Croatian Employment Service to prevent labour market detachment. It further examines the role of professional rehabilitation, noting its comprehensive design but limited reach, and identifies barriers such as late referral, low motivation driven by financial insecurity, and capacity constraints. The chapter concludes that strengthening early intervention processes, reinforcing professional rehabilitation, strengthening data systems, and embedding rigorous monitoring and evaluation are essential for building an effective, evidence‑based disability and employment support system.
Disability, Work and Inclusion in Croatia
Abstract
Ensuring better labour market outcomes for people with disability in Croatia requires going well beyond improving disability assessment. While developing robust assessments is essential, in the current system employment support during sick leave and professional rehabilitation come far too late for most people. Early identification and intervention, and stronger co‑ordination across institutions – particularly during sickness spells and unemployment – are crucial to prevent long-term labour market detachment.
Supporting workers early during long-term sickness is critical. Many people who develop a health condition enter the disability system via sickness benefits. However, Croatia’s sickness system allows individuals to remain on sick leave for an unlimited time and up to 12 months before a work capacity assessment becomes mandatory, despite clear evidence that early intervention drastically improves the likelihood of returning to work. Delays reduce employability and heighten the risk of permanent labour market exit and a transition to disability benefits. Introducing a time limit to sick leave and creating structured return‑to‑work processes, involving doctors, employers, and employment services, starting no later than three months into a sick leave would help people return to work earlier.
Employers must play a more active role. Although employers bear wage costs during the first 42 days of sick leave, their formal responsibilities in managing return‑to‑work are limited. Strengthening employer duties – such as mandatory return‑to‑work planning, earlier engagement with occupational health information, co‑operative assessment of workplace adaptations, and extending employer‑paid periods in cases of non‑co‑operation – would improve outcomes and incentivise earlier intervention.
Early intervention for unemployed people with health conditions is equally important. Many unemployed people have unrecognised health issues, particularly mental health conditions, placing them at high risk of long-term exclusion. The Croatian Employment Service is well placed to identify these individuals early, yet referral rates to professional rehabilitation remain low. Strengthening screening and possibly making participation obligatory for some groups – such as young people – would help prevent permanent labour market detachment.
Professional rehabilitation is comprehensive but underused and accessed too late. Despite various good practices at the professional rehabilitation centres, participation in rehabilitation remains low because referral often occurs only after long sickness spells or unsuccessful disability benefit claims. Allowing access earlier, including in parallel with medical rehabilitation, delinking it from having permanent or below 50% work capacity and raising the age limit above 55 would ensure that it reaches the people who need it most. Greatly increasing the centres’ capacity will be crucial going forward.
Finally, better data and systematic policy and programme evaluation are urgently needed. Building a strong evidence base, improving data sharing, and embedding evaluation in the design of future reforms will be key to ensuring that Croatia’s disability and employment support systems become more effective, timely, and sustainable.
Whilst developing robust disability and work capacity assessments is an important step towards the labour market integration of people with disability, better assessment alone will not improve the labour market outcomes of people with disability very much. As in many other OECD countries, also in Croatia comprehensive assessment and professional rehabilitation, if indicated, are generally only offered to people at a time when they have given up any labour market aspirations. Across OECD countries, early identification of problems and early intervention has shown to be the single most important element of an effective disability policy. How individuals with disability or health conditions can be supported through any sickness or unemployment phase and what services they receive to support their return to the labour market or their first-time labour market entry, are therefore crucial aspects for improved labour market integration of people with disability.
People who acquire a disability during adult life and have a job typically transition into the disability system through a country’s sickness system. This is also the case in Croatia. Supporting these people during the sickness phase already is crucial to ensure a fast return to their job and, if necessary, a fast referral to work capacity assessment and professional rehabilitation, with the aim to prevent them from potentially exiting the labour market. Early intervention for employed people facing more substantial, recurring or chronic health conditions means developing an effective sickness management approach and process which supports their return to work from sick leave as fast as possible, fully or gradually, ideally long before they (consider to) claim a disability payment.
The sickness benefit system in Croatia has unused potential to support early intervention. GPs have an obligation to refer workers to work capacity assessment after no later than 12 months. The law allows them to refer patients earlier if they think work capacity is permanently reduced, but this rarely happens in practice. Remaining on sickness benefit and, therefore, out of work for as long as 12 months significantly reduces the likelihood of a sustainable return to the labour market. This has been observed in several OECD countries, including for example Slovenia, which has a sickness benefit system similar to that in Croatia: in Slovenia, transitions from sickness to disability insurance are rare after short sickness absence spells but quite frequent for those who have been absent from work for at least a year. At the same time, the likelihood of returning to employment, either to the same or to a new employer, significantly decreases with the time spent on sick leave (Figure 5.1). This shows that intervention would be critical early in the sickness phase to ensure access to return-to-work services and to increase the likelihood of returning to work.
Exit pathways from sickness insurance in Slovenia by duration of sick leave, average of years 2013-2017
Note: At month 0, sickness claims are terminated, and all employees are still employed (employment is a pre‑condition for benefit receipt). Pathways follow ex-claimants for 12 months after termination of their claim. Include claims from mental health, cardiovascular disease, work injuries and diabetes (about 10% of all cases); they are indicative of the pathways but not representative.
Source: OECD (2022), Disability, Work and Inclusion in Slovenia: Towards Early Intervention for Sick Workers, OECD Publishing, Paris, https://doi.org/10.1787/50e655b3-en, Figure 2.10.
Croatia should strengthen early intervention and consider introducing a more elaborate return-to-work process to support sick workers return to the labour market as soon as possible. Such a process should include three steps with the involvement of the employee, the employer and different key stakeholders:
1. Assessing workers on sick leave (three months after the sick leave begins): The first step is to control the sick leaves and assess whether workers could return to work, potentially under certain (adapted) conditions, or if it is necessary for the worker to fully remain on sick leave after three months. This step should be carried out by occupational doctors or doctors who are trained in return-to-work matters and work capacity but also have an understanding of the worker’s health issues, so they are able to consider the health condition and the worker’s job and make a decision about the possibility of returning to work, to either their current job or any other job.
2. Supporting the return to work: Once the doctor has assessed whether the worker can return to work, the required support measures are twofold, depending on the job the person can return to.
a. Returning to the same job: If the assessing doctor suggests that the worker can return to their current job, the employee and their employer should start a discussion to see what support is required in the workplace. Such support could entail phased return to work or workplace adjustments. The employer may also seek support from the professional rehabilitation centres in the assessment of the necessary workplace adjustments.
b. Returning to a new job: If the assessing doctor suggests that the worker is able to work but unable to do their current job, the employee and the employer should decide together based on the information provided by the doctor, or even by requesting the direct involvement of the doctor if the information is not sufficient, whether there is another suitable job within the company for the worker. If it becomes clear that the worker cannot return to the same employer, the employment service (HZZ) should get involved and support the worker in finding a new job with a new employer through the various employment services the HZZ provides. During this process, the worker may still be on sick leave and therefore not traditionally under the responsibility of the HZZ. However, even in the current system it is possible to register as a jobseeker with the HZZ for someone who is on sick leave, although this option is not often used yet. Supporting the worker to find new employment at this stage already would be an investment by the HZZ as workers unable to return to work during or after sick leave will likely end up in the unemployment register later anyway and thus become the responsibility of the HZZ. Earlier HZZ involvement will make returning to the labour market more likely and faster, thereby reducing the costs for the HZZ.
3. Promoting professional rehabilitation (no later than one year after the sick leave begins but preferably and, in most cases, much earlier): If the above measures failed, the worker should be enrolled in professional rehabilitation where they could receive extensive training and support in return to work tailored to their needs by an interdisciplinary team of experts. This support should include an assessment of work capacity and needs and opportunities for training as it currently does. In addition, professional rehabilitation should be available not only to workers on sick leave but also to unemployed persons under the responsibility of the HZZ and persons already receiving disability benefits from the Croatian Pension Institute (HZMO) or having been assessed by ZOSI with a work capacity reduction.
The role of GPs should be limited in the assessment of sick leave and the application for disability benefits. The current system gives GPs a key role as gatekeepers for sickness insurance. They also play a key role in preventing transitions to disability insurance, by suggesting that medical treatment is not completed or promoting the application to the disability programme. This comes at a risk of focussing on the medical aspect at the detriment of the occupational aspect of sickness and disability. That is because GPs often lack the training and workplace knowledge to assess the potential to work or work partially. GPs are also generally too close to the claimant, or patient, to be impartial. In a new system, GPs should be supervised closely or even freed altogether from their current gatekeeper role.
If GPs maintain their current role, it is necessary to provide insurance medicine training to them, both in the initial medical curriculum and by providing corresponding mandatory training for all GPs with a practice. It is also important to increase the number of GPs in Croatia and to compensate them better for their assessment role. This would empower GPs to take assessment decisions on a sounder basis and speed up processing times but would not resolve the issue of close connections with their patients.
Introducing functioning or work capacity aspects in the assessment of sickness would also be important. Croatia could borrow some elements of the UK’s Fit Note in which GPs are not only required to assess whether the sick worker is able to work but to suggest basic changes to the work environment or job role, or other steps to help employees return to work earlier. The Fit Note could be shared with employers as the basis for a jointly created return-to-work plan, excluding any medical information and including the doctor’s assessment on the fitness for work.
Whilst Croatia’s sick leave system already provides some financial incentives for a return to work, many workers remain on relatively low sickness benefit payments for a long time. Croatia provides sickness payments for an unlimited period of time for workers who have made social security contributions for at least nine consecutive months (Boundless, n.d.[1]). It is very unusual across OECD countries to provide paid sick leave with no time limit; only New Zealand, Portugal, Slovenia, Sweden and Türkiye do so. In most OECD countries, the maximum duration for paid sick leave is between six and 12 months. At the same time, Croatia provides strong incentives for workers to not stay on sick leave for too long by reducing the amount of sickness payments over time. During the first 42 days of sick leave, which is covered by the employer, workers receive 70% of their salary although they may receive a higher amount depending on collective or employment agreements, and if the sickness is due to a work accident or the Croatian War of Independence, workers receive 100% of their salary. From the 43rd day onward, sickness payment is covered by the HZZO and cannot exceed EUR 995.45 per month for full-time working hours. This is a reduction for most workers considering that the average monthly wage in Croatia in 2025 was EUR 1 498, but still a considerable incentive to stay on sick leave. After 18 months of uninterrupted sick leave for the same diagnosis, the payment is further reduced by half (Boundless, n.d.[1]), thereby providing another incentive for workers to move off sick leave and back into work. However, this additional incentive will likely come too late for most workers, who will more likely aim to move onto other benefits at this point, especially disability benefit, rather than seeking to return to the labour market. The Croatian system is somewhat unique in combining sickness payments for an unlimited period with a very low payment level as systems in most other OECD countries provide sickness benefit for a time‑limited period but at a higher level. Despite the incentives for a return to work that the Croatian system provides, however, the share of people on sick leave is comparable to the EU average (Figure 5.2). This strongly suggests that early intervention plays a minor role in the Croatian system today.
It would be helpful to introduce a maximum sickness payment duration of, say, one year for uninterrupted sick leave based on the same health condition. A maximum payment period for sickness benefits of one year would align the Croatian system with similar programmes in other OECD countries and provide a major push to the operability of the sickness and disability insurance system. The system could differentiate between first and repeat absences, e.g. with a maximum payment period of one year in the past three years. There could also be extensions to the maximum period of payment, e.g. up to 18 months of sickness in total, granted to persons successfully engaged in a professional rehabilitation process. The aim of introducing a time limit is not to penalise people who need long-term medical care, instead a limit would help ensure that everyone receives the necessary support they need. In the current system, too many people stay on sick leave for too long even if they could benefit from support through other channels such as professional rehabilitation. In some other OECD countries people who need longer-term healthcare and therefore have a temporary work capacity reduction due to their health condition or treatment, such as cancer patients, can receive a temporary disability pension once the sickness benefit limit is reached.
While the Croatian system allows for unlimited sick leave, the HZZO regularly monitors workers on sick leave to prevent system abuse. In general, cases are selected randomly for control, but employers can also request a control. People on sick leave regularly are more likely to be selected for a general control. As part of these controls, physicians of the HZZO review medical documentation to evaluate whether there is an indication for temporary incapacity; if they find no such indication, they can overrule the GP and terminate the sick leave. These controls are an effective way to monitor whether workers rightfully use sick leave. However, the HZZO could take a more active role in the return-to-work process and the recommendation of professional rehabilitation for individuals on sick leave. The HZZO could use the controls to promote early intervention, including by stimulating return-to-work measures by the employer but also by allowing physicians to recommend changing temporary incapacity to permanent incapacity if they see fit, thereby enabling earlier professional rehabilitation, or by referring workers to professional rehabilitation directly. This would require legislative changes and raising awareness about the potential of professional rehabilitation and the corresponding processes among HZZO physicians carrying out the controls. To promote return to work, the HZZO could also develop a system through which people who could benefit from early professional rehabilitation are identified not only randomly but systematically, based on their individual sick-leave history and the causes of their sick leaves.
Hours lost due to sickness absences as a percentage of weekly usual hours worked, 2023
Note: Estimates adjusted for an underestimation of 50% of sickness absences reported in labour force surveys compared with those from Health Surveys and other sources. The blue line is the unweighted average of the 26 countries shown.
Source: OECD calculations based on the European Labour Force Survey.
Finally, financial incentives for employers are as important for early intervention as the benefit system and the resulting incentives are for workers. The Croatian system offers some potential in this regard as employers are faced with a statutory period of continued wage payment of 40‑two days for every sickness spell, after which the HZZO becomes responsible for sickness payments. The length of wage payment periods varies considerably across OECD countries, with Croatia finding itself well in the middle. Some countries, like the Nordic countries, provide continued wage payment by the employer for around two weeks of sick leave only. Other countries, like Austria, Germany, Italy and Switzerland, provide continued wage payment for several months and the Netherlands even for two years, thereby creating very significant financial incentives for employers to prevent sick leaves. Although the obligation for wage payment offers incentives for employers to promote early intervention, in practice employers in Croatia do not seem to be sufficiently involved. By law, they cannot dismiss workers who are on sick leave or undergoing work capacity assessment, and if someone is assessed to have a reduction or a partial loss of work capacity, they must make reasonable accommodations for their worker’s job or offer a suitable job if possible. In case they need to make workplace accommodations, they can also request financial support from a dedicated fund, which is yet again a good incentive for the employer. However, anecdotal evidence (as hard evidence is unavailable) suggests that following a work capacity assessment most workers end up leaving their job upon coming to an agreement with their employer, especially in the case of smaller businesses.
Therefore, employers must have greater incentives to strengthen the return-to-work process. To give employers the right incentives to focus on effective return-to-work strategies during the first three months, the Croatian Government could consider the possibility to extend the employer-paid sickness period from 42 days to three months. HZZO payments would in this case start from the fourth month of sickness only. Employers should remain involved even after the third month, however, if the return-to-work approach considers the return to the previous job or company. To make sure that this co‑operation happens, the employer-paid period could be extended even further than this for employers not providing the information necessary to assess their sick employee’s professional rehabilitation needs and options and not co‑operating in the professional rehabilitation process.
Employers’ responsibilities to support their employee to return to work should also be expanded. Employers and employees should engage in regular meetings and draw up a return-to-work plan in a mandatory manner, as is common in several OECD countries. This can be difficult at present, as employers may not be able to obtain any work capacity information of their employees on sick leave. By changing sickness assessment to include non-medical information relevant to the return-to-work process, e.g. through a Fit Note akin to the approach in the United Kingdom as mentioned above, employer and employee could articulate their needs and return-to-work efforts around it. This would also allow employers to co‑ordinate with employees and provide the right support before the three‑month mark.
Early intervention is also crucial for people who are unemployed as some of them have never held a stable job whereas others may have ended up in unemployment because of lacking (early) intervention during a period of long-term sick leave. This is an especially vulnerable group that often suffers from unrecognised disability and fluctuating health conditions, oftentimes mental health conditions, which make finding and keeping employment difficult. Unemployed people with health conditions are often in a more difficult situation than employed people with health conditions as they do not receive the same benefits and support. It is easier for them to fall through the cracks of the system as they do not fall under the same monitoring as employed people on sick leave, e.g. the HZZO does not monitor them whilst they are ill, and their GP is not obliged to refer them to work capacity assessment if they have been ill for as long as 12 months. It is crucial to develop a system that can identify these people early on and provide them access to the right services to ensure their timely return to, or entry into, employment.
5.2.1. The key role of the Croatian Employment Service
People with health conditions and potentially unrecognised disability not benefitting from sickness insurance will not be able to benefit from early intervention measures provided during paid sick leave. For this group, other institutions play a more important role. Similarly to many other OECD countries, like Austria, Canada, the Netherlands and Norway, recipients of social assistance need to register with the Croatian Employment Service. Through this regulation, the HZZ is well-placed to identify jobseekers with health barriers to employment early and to provide targeted support to people, who may end up receiving disability benefit if not activated quickly. The HZZ also plays an important role in supporting people who already have a recognised disability status and potentially a work capacity reduction.
The HZZ works actively to support unemployed people with health conditions and disability who can participate in the same measures as every other (long-term) unemployed person, and they also screen for people who do not have an official disability status through their discussion with each client. If they find someone might have a disability, they can directly refer the person to one of the professional rehabilitation centres to undergo a disability assessment and participate in other professional rehabilitation services (see more information below). However, the number of people referred to professional rehabilitation by the HZZ remains very low, partly due to the capacity constraints of the regional Centres for Professional Rehabilitation. In line with good practices of public employment services across OECD countries, the HZZ provides various services to people with disability including participation in active labour market programmes and in vocational guidance activities, vouchers to reskill and upskill, extended employment support with their employers for 24 months instead of 12 months for people without disability, and referral to the services offered by professional rehabilitation centres. The HZZ also creates an individualised job search plan for every client based on the client’s motivation and interests, skills and education, work experience, disability (if applicable) and the state of the labour market. If someone was assessed to have a reduced work capacity by ZOSI, they also take into consideration the finding and opinion and ZOSI’s suggestions about what job the person may or may not be able to do.
Going forward, the HZZ must strengthen its efforts to support jobseekers with health problems. Health issues are often an underlying cause of unemployment, especially in the case of long-term unemployed people among whom mental health conditions are widespread. Whilst the HZZ offers various services to support the return to work of unemployed persons, including those with health problems, they must carefully select who they can support themselves and for whom a referral to the CPRs would seem to be indicated. Jobseekers with health barriers to employment that could benefit from professional rehabilitation should be referred to the CPRs as early as possible. The CPRs would be responsible for ensuring the professional rehabilitation of jobseekers with health barriers to employment, but their early identification would remain a responsibility of the HZZ. HZZ caseworkers should be quick at referring jobseekers to the CPRs upon suspicion of health issues that impede their entry in the labour market.
It is also crucial to ensure that the HZZ reaches the people who could benefit from their services the most. To this end, any barriers that prevent persons on benefits received because of temporarily or permanently reduced work capacity to register with the HZZ should be removed. For example, in the current system when a decision of incapacity to work is made within the pension insurance system, the person is eliminated from HZZ records. Whilst people with a complete loss of work capacity are not and should not be the priority target group for the HZZ, they should still be able to use the services of the HZZ without any barriers if they wish. In addition, people who are assessed to have partial remaining work capacity can but are not obliged to register with the HZZ. Similarly to other OECD countries, this results in low participation rates for this group. The HZZ could reach out proactively to reach all unemployed people with disability, including those receiving benefits on the grounds of reduced work capacity. Croatia may even want to consider making participation in active labour market programmes obligatory for certain groups on such benefits, such as young persons and individuals who enter disability benefits or acquire a disability but have relevant remaining work capacity.
5.2.2. Initiatives to support finding employment for people with disability
Croatia has a quota system in place to encourage the employment of people with disability. Companies with at least 20 workers must meet a quota of employees with disability equal to at least 3% of their workforce. Employers can also fulfil this obligation through replacement activities, for example providing scholarships to students with disability or entering a business co‑operation with a self-employed person with disability or with a sheltered or integrative workshop. If the employer does not fulfil the quota in any of these ways, they must pay a levy worth 20% of the minimum wage. While disability employment quotas are a policy instrument used by many OECD countries with the aim to raise awareness among employers, they tend to have a negligible impact on the level of employment inclusion of people with disability. To be effective, the levy for non-fulfilment of the quota would have to be a multiple of the current level. The biggest gain from such a quota therefore generally is the additional funds available through the employer levies, as this funding is usually earmarked for the provision of employment interventions for people with disability.
Croatia has several integrative and protective workshops (ZIRs) to support the employment of people with disability. Integrative workshops are established for the employment of people who cannot be employed in the open labour market, and they must have at least 40% of people with disability in their workforce. Protective workshops are established for the employment of people with lower work capacity who cannot be employed in the open labour market or in integrative workshops. Protective workshops must have at least 51% of people with disability in their workforce. ZOSI regularly organises training for the employees of ZIRs to learn about supporting the employment of people with disability in their workshops. Again, integrative and protective workshops play a certain role for people with more severe disability for whom an inclusion into the regular labour market is difficult. However, the continuation of sheltered employment in whatever form contradicts the intentions of the CRPD – which requests signatories to abolish in the longer run any form of special education and special employment – and will not have any real impact on the low level of labour market inclusion of people with disability in Croatia.
ZOSI and the HZZ, along with other partner institutions, also oversee two initiatives to support people with disability in finding employment. One initiative is a pilot project developing and establishing a personal assistance service for people with disability in the workplace. The aim of the project is to provide such a service at the workplace for people who have been determined, by one of the Professional Rehabilitation centres, to need personal assistance at work, including identifying work tasks and activities in which the employee needs the help of another person. However, in 2023, personal assistance services at the workplace were organised for only five individuals. Another initiative is a work training programme for people with disability in sheltered workplaces within ZIRs. The aim of the project is to implement appropriate and adapted training programmes in sheltered workplaces in ZIRs, aimed at increasing employability and facilitating a transition into the open labour market. Participants were offered a two‑year training to perform work in a specific sheltered workplace and additional courses that were aimed at the development and adoption of social competences to aid users’ participation in the labour market and in everyday life. Between 2019 and 2023, a total of 55 people with disability have been included in the project, of which 26 have been employed afterwards (22 of them under special conditions, and only four in the open labour market).
If someone, whether employed or unemployed, has already reached the point that their work capacity has declined, professional rehabilitation can be an effective way to evaluate how their remaining work capacity could be best used and help them reach their full potential through training or retraining, as necessary. This can be a crucial step towards returning to the labour market and even more so if a person with disability is entering the labour market for the first time.
5.3.1. Croatia has a strong and comprehensive professional rehabilitation system
Professional rehabilitation in Croatia takes place in four regional Professional Rehabilitation Centres in Zagreb, Rijeka, Split and Osijek which were established in 2015. The Centres function based on principles of individuality, interdisciplinarity, confidentiality and ensuring active participation and involvement of individuals. Each centre provides the same 12 services, some of which are aimed at people with disability, and some can be requested by the employers of people with disability. The Centres receive clients from four sources: 1) ZOSI can refer individuals undergoing their work capacity assessment, 2) HZZ can refer unemployed people with (potential) disability, 3) employers can refer their employees with disability, and 4) persons seeking professional rehabilitation can apply directly.
Clients referred from ZOSI have been found to have a reduction or partial loss of work capacity through their work capacity assessment for the HZMO. Clients referred from the HZZ are often long-term unemployed or have an unstable work history. The individuals sent by ZOSI and the HZZ first undergo a comprehensive ICF-based assessment, lasting several days, of their current work capacity, knowledge, skills and professional interests during which an interdisciplinary team of assessors evaluates whether the person can be rehabilitated and, if so, which services the person should receive. The multidisciplinary assessor team is composed of a social worker, a rehabilitator, a psychologist, an educational expert, an occupational therapist, and a medical doctor. Specialists can also be involved if needed, for example a psychiatrist in the case of a client with a mental health condition. The medical doctor determines the health status, the psychologist carries out a clinical interview assessing the personality and professional interest, the social worker assesses the client’s social situation, and the rehabilitator gathers information about the person’s formal and non-formal education, past work experience, and any difficulties they may experience at work. As part of the assessment process the client is tested for motivation, mobility and functioning ability as well as basic knowledge of Croatian language, maths, literacy, and computer usage which helps the assessors understand what further training may be necessary. Based on the results of these assessments, following an interdisciplinary discussion, the client is recommended an individualised professional rehabilitation plan and can participate in the Centres’ services as needed, including e.g. training at different skills levels or on-the‑job training.
When an employer refers an employee to a rehabilitation centre, they can take part in two services, a work performance assessment and a work environment assessment. The latter includes an assessment of architectural, organisational, and reasonable adaptation that may be needed to ensure that the employee is able to carry out their job. These assessments are also important because employers may be able to claim financial aid from ZOSI to provide an employee with disability with the necessary workplace accommodation. Larger companies can often more easily adapt the workplace or find another suitable job within their company to retain their employees with disability. Demand for workplace accommodation in general has been growing in recent years, although the reason behind this is unclear.
When an individual applies to a rehabilitation centre directly, they can participate in the so-called Service 1.1. – Rehabilitation assessment in relation to work (short assessment). Participating in this service and receiving a positive outcome grants the person the right to become registered in the National register of people with disability and/or the Register of employed persons with disability (see Box 2.1 for more information on the different disability registers available in Croatia). This is especially important for individuals not able to obtain these rights through any other system, for example, this pathway is often used by people with mental health conditions.
All services of the rehabilitation centres are free of charge and anyone between age 15 and age 55, who finished medical rehabilitation and completed (at least) primary education can participate in professional rehabilitation. Only for some services, the person also needs to be in the Register of people with disability. Over the past five years, more than half of the clients of the Zagreb rehabilitation centre were referred from their employers, 15% were referred from HZZ, 10% from ZOSI and 20% applied directly for professional rehabilitation services (Table 5.1). While these shares fluctuate over time, the share of people referred from ZOSI seems to have decreased over time, at the expense of a slight increase in the share of people referred from HZZ and a larger increase in the share of direct applications.
5.3.2. Professional rehabilitation outcomes could be further improved
Although the professional rehabilitation system is potentially comprehensive and strong, and the number of participants has been increasing somewhat in recent years (Table 5.1), participation remains very low compared to the number of people potentially able to benefit from professional rehabilitation. The system and the procedures around it could be improved in many ways to facilitate and encourage access to professional rehabilitation services.
Number of people in professional rehabilitation services and distribution by type of referral, 2020-2024
Source: OECD calculations based on data from the Institute for Disability Assessment, Professional Rehabilitation and Employment of Persons with Disabilities.
Professional rehabilitation may not reach the people who could benefit the most
As shown in Chapter 2, Figure 2.11, around half of the applications for a work capacity assessment from the HZMO to receive a disability pension, which requires (i) an assessment of less than 50% of work capacity compared to a fully healthy person and (ii) the reduction in work capacity to be permanent, are rejected. Experience from other OECD countries shows that people whose claim for disability pension is rejected are a highly vulnerable group that rarely returns to the labour market, yet most countries, including Croatia, do not provide any support to this group. It would be important to help prevent these people from dropping out of the labour market and falling into long-term unemployment instead, thereby possibly ending up in professional rehabilitation eventually anyway but only much later, maybe even several years later, when their health condition and work motivation has deteriorated, and professional rehabilitation has become more difficult and less likely to succeed. This is especially important for people who get rejected because they are found to have a work capacity reduction of less than 50% whose deterioration of work capacity due to worsening health could have been prevented through professional rehabilitation at an earlier stage. Professional rehabilitation should be delinked from being assessed as having less than 50% work capacity and the requirement for permanence should be abolished. Access to professional rehabilitation should require an assessment for rehabilitation needs. This would allow sickness insurance claimants to access rehabilitation earlier, if needed, and to return to employment without being labelled as having a disability.
Furthermore, the apparent difference in Croatia between a predominantly medically driven approach to work capacity assessment and a holistic approach to professional rehabilitation could imply that people in need of professional rehabilitation might never be identified. People with mental health conditions may not get referred as they do not meet the necessary conditions of severity or permanence, even though they could benefit from professional rehabilitation. Harmonising the approaches by introducing a more holistic approach to the ZOSI work capacity assessment as well, could ensure that professional rehabilitation reaches the right group of people at the right time.
Finally, the current age limit for enrolling in professional rehabilitation is 55 years, which is ten years below the statutory retirement age of 65 years, preventing many people from potentially working for many more years. As populations are ageing, the need to keep people in employment longer is increasing across all OECD countries. The employment rate in Croatia for people aged 55‑64 was one of the lowest across the European Union in 2023, at 51.6% compared to the EU average of 63.9% (Eurostat, 2025[2]). This suggests that Croatia could also benefit from policies aimed at increasing the employment rates of older workers. By raising the age limit for professional rehabilitation, Croatia could gain important contributions to the labour market. Job and career changes tend to be more difficult for workers at older ages, therefore, undergoing professional rehabilitation and reskilling would be key to ensure that older workers with health conditions or disability have the skills to successfully return to or stay in the labour market. Ensuring that older workers with chronic health conditions are not omitted from professional rehabilitation will become even more important over time as population ageing will result in a continued increase in the number of people with disability, driven by the strong age‑disability gradient.
Professional rehabilitation often comes too late
Most clients of professional rehabilitation centres have been unemployed or on sick leave for a very long time before they are referred to this service. The low take‑up of professional rehabilitation, the use of which is entirely voluntary, also likely reflects a limited interest in employment at this stage, when people are more likely to seek to transition to a permanent disability pension. However, even if someone chooses to enrol in professional rehabilitation, it is less likely to lead to a successful and sustainable return to work if someone has been out of the labour force for a long period, as described above. The obligation to wait with professional rehabilitation until after the completion of all treatment and medical rehabilitation adds to late intervention and further prolongs the process. Loosening this requirement and allowing for simultaneous, intertwined medical and professional rehabilitation could be beneficial for several health conditions, especially mental health as well as many other chronic health conditions.
Financial stability could improve motivation and consequently rehabilitation enrolment rates
As active participation of clients is a key part of successful professional rehabilitation, a lack of motivation can be an important barrier to enrolling in rehabilitation. Especially in the case of clients referred from ZOSI, many individuals are not motivated to undergo professional rehabilitation as they perceive it as a long process with uncertain outcomes and financial instability and would rather choose the financial stability and security of a disability pension. This implies that considerable potential work capacity may be lost. Providing financial incentives may help to increase motivation. This is already in place to some extent as the current salary compensation provided during professional rehabilitation is equal to the minimum wage well above the level of the minimum and average disability pensions. However, making financial supports more permeable and flexible is critical, to allow people with disability to transition into the labour market while receiving benefits and to return to benefits in case employment integration fails. Finally, entitlements to disability benefits could also be made conditional on participation in training or apprenticeships, in line with the individual’s remaining work capacity, to increase participation rates in professional rehabilitation.
Raising awareness about the benefits for which individuals would still be eligible even if they started working could also be important for motivation. Since the recent legislative changes were introduced in the field of social welfare, most HZSR benefits are not means-tested anymore therefore they would not be lost even if someone started working and earning an income. In addition, individuals aged 58 and over who were assessed to have reduced work capacity, or a partial or complete loss of work capacity, and who have undergone professional rehabilitation and have registered with the Croatian Employment Service but remained unemployed for a period of five years, are eligible for a temporary disability pension. However, given the strict eligibility requirements, it is important to note that only a very small number of individuals qualify for this type of pension; there were only four beneficiaries in early 2025.
Training could be expanded to target a larger variety of skills
Global megatrends, including the green and digital transitions and population ageing, are shifting skills needs and causing structural changes to labour markets across the OECD. Tackling these challenges requires reskilling but also upskilling to ensure that workers have the right skills to find jobs. However, in Croatia today training as part of professional rehabilitation is only possible at the same or a lower level of initial education someone has attained. Harnessing the full potential of people with disability will often necessitate the possibility of training also at a higher level of education. This would have a double positive effect as a higher level of education reduces the risk for people with disability to drop out of the labour force and increases their chances of finding a job (OECD, 2022[3]). Workers with lower education levels are more likely to work in physically demanding jobs in which it is harder to make reasonable accommodations for people with disability. Workers with higher education levels are also more likely to work in office jobs where workplace adaptation is more feasible and in jobs that are more amenable to teleworking.
Increasing the capacity of the professional rehabilitation market will be crucial
The capacity of the four regional professional rehabilitation centres in Croatia will have to be increased very significantly to strengthen the return-to-work process. Today, the number of people enrolled at the professional rehabilitation centres is very low, partly explained by the voluntary nature of professional rehabilitation, but also due to capacity constraints. Going forward, it will be important to increase the capacity of the Centres very substantially to be able to accommodate a larger inflow of clients especially if further changes to strengthen early intervention more generally are implemented. Increasing the capacity for professional rehabilitation to the necessary extent will present an organisational and financial challenge.
Finally, there is also one other critical issue for Croatia which is a significant barrier to better policy design and the introduction of more efficient and more effective interventions and approaches: the absence of a strong evidence base. Many OECD countries are gradually moving towards more frequent evaluation of the effectiveness of social and labour market programmes and policies. Croatia has yet to go down this route. Monitoring and evaluating of any new reform to assess its impact warrants investments into data collection as it requires information about applicants and claimants of the system. At the moment, Croatia faces substantial data gaps, which make it impossible to conduct impact evaluations on the sickness and disability insurance programmes.
Therefore, the collection of robust data and indicators will be crucial. This report includes some of the data that is available but also shows the limits of the current evidence base. The MIA application in which all information on disability and work capacity assessments is being stored is a very promising starting point. Moving forward, this database could be used to regularly produce a set of robust indicators which could then be linked with data from other stakeholders, such as the pensions insurance institute, the health insurance institute, the institute of social work and the public employment service.
The evaluation approach should be planned ahead and integrated it in the design phase of the new system. Evaluations planned in advance offer more options for evaluation methods than when the programme or policy has been rolled out, allowing to lower the cost and minimise institutional disruptions. In addition, planning ahead allows to prepare the necessary data collection and data sharing systems, and to ensure good baseline data are available in order to establish and test whether the evaluation design generated appropriate comparison groups.
Findings from the evaluation of the new system must be disseminated in a form that decision makers can easily access and use. The evaluation team needs to distil a manageable set of key messages summarising the most policy-relevant results and recommendations, and to communicate these messages consistently across audiences. The sequencing of dissemination activities may also be critical. For example, prior to public dissemination, an internal round of presentations and consultations should be conducted with programme staff and managers to avoid that premature results hurt the new system’s reputation.
References
[1] Boundless (n.d.), Employee Leave Entitlement in Croatia, https://boundlesshq.com/guides/croatia/leave/ (accessed on 21 March 2025).
[2] Eurostat (2025), Employment rate of older workers, age group 55-64, https://ec.europa.eu/eurostat/databrowser/view/tesem050/default/table (accessed on 21 March 2025).
[3] OECD (2022), Disability, Work and Inclusion: Mainstreaming in All Policies and Practices, OECD Publishing, Paris, https://doi.org/10.1787/1eaa5e9c-en.