2. The challenges of the Slovenian sickness insurance programme
Slovenia’s sickness insurance programme faces very large take-up, with a high and fast increasing share of long-term absences and a continuous increase in the share of mental health cases. Long-term absence is particularly widespread among older workers who increasingly use the sickness insurance as a pathway to retirement. These developments are explained largely by the very characteristics of the sickness insurance programme: the lack of a time limit for very generous benefits, and the lack of any return-to-work attempts and services. New analysis presented in this chapter shows that early intervention is critical and that the likelihood of a return to work false sharply with the duration of sickness absence: those absent for over a year hardly ever return to the labour market. Without significant reform, these problems will get out of hand.
As in most EU and OECD countries, workers falling ill can receive benefits from sickness insurance, following a period of 20 working days of employer-provided sick pay (30 days until end of 2021), which covers part of their earnings lost due to sickness. A particularity of the Slovenian sickness insurance system is that there is no maximum benefit payment duration, turning it into a system also insuring the risks of long-term sickness and, to an extent, disability. This section covers the main characteristics of the programme, the assessment of sickness, and discusses how the features of the programme may be causing a long-term sickness issue.
2.1.1. Benefit provision, funding and organisation
In Slovenia, sickness insurance coverage is a right acquired for those insured under the compulsory health insurance scheme. As such, only persons who are employed, or self-employed, and paying contributions to the compulsory health insurance scheme are covered by public sickness insurance. The system does not allow for voluntary contributions. Unemployed workers and non-standard forms of work not declared as paid employment or self-employment are not covered by sickness insurance. The Public Finance Balancing Act (Zakon za uravnoteženje javnih finance, ZUJF), which entered into force in June 2012, abolished the right of unemployment benefit recipients who were on the sick leave for more than 30 working days to receive sickness benefits.
Sickness insurance is funded through contributions to the compulsory health insurance scheme. The Health Insurance Institute (ZZZS) is the institution responsible for collecting and allocating the public funds for the compulsory health insurance scheme, and thus responsible for determining the rights to sickness insurance and paying the benefits. Health insurance contributions are shared between employers (6.56% of the gross wage plus an additional 0.53% to cover the risks of occupational injuries and diseases) and employees (6.36% of the gross wage). Self-employed workers pay 13.45% from the insurance base determined in accordance with the regulations on pension and disability insurance.
Funding also comes from employers directly, who are responsible for financing the first 30 working days of every sickness absence spell (20 working days as of 2022), for a maximum period of 120 days per year. From the 31st day, the sickness benefit is covered by the ZZZS through compulsory health insurance contributions. Exceptions to this are absence spells due to occupational injuries and diseases, donations, care of a close family member (up to seven days, and 15 days for a child), doctor-mandated quarantine and workers/volunteers for a public service, whose sickness benefits are covered by the ZZZS from the first day.
While the ZZZS manages and allocates sickness insurance funds, the key regulatory role rests with the Ministry of Health, which has consequences for the capacity to reform the sickness insurance scheme. The Ministry of Health is the owner of all public hospitals and national public health institutions, and their key manager and investor (Albreht et al., 2016[1]). While the ZZZS autonomously adopts the financial plans and policies that regulate the rights and benefits of the insured, the Parliament and Ministry of Health have retained a key role in determining the scope of benefits, the financial plan and the confirmation of the elected general manager of the ZZZS (Albreht et al., 2016[1]). This role implies that any reform of the sickness insurance system and the ZZZS needs to be presented to Parliament for approval by the Ministry of Health. This structure creates a political layer to any reform of the scheme.
2.1.2. Coverage and generosity
In Slovenia, the sickness insurance programme covers any insured worker experiencing an illness or injury acquired at work or outside of work. There are no differences in eligibility requirements, sickness assessment procedures, or return-to-work provisions for workers with occupational and general illnesses and injuries.
The reason for sickness absence only influences the generosity of the benefit. Table 2.1 shows how the replacement rate varies depending on the cause of sickness absence. The replacement rate is applied on the base earnings, calculated as the average earnings in the year before the sickness absence. Workers with occupational illnesses and injuries benefit from a statutory replacement rate of 100%, while the statutory replacement rate for general illnesses is 90%, and that of outside work injuries is 80%. Since 2012, the Fiscal Balance Act (ZUJF) imposes a 10 percentage point cut in the statutory replacement rate for sickness absences shorter than 90 days.1 Employers may supplement the statutory benefit through collective agreements, although this is not particularly common.
The Slovenian sickness insurance programme is very generous by international standards (see Figure 2.1). Some countries like Austria, Denmark, Finland and Norway offer a replacement rate of (almost) 100% for a considerable period, e.g. for several months (depending on tenure) in Austria or even a full year in Norway. In the majority of EU and OECD countries, however, sickness benefits are (significantly) less generous than in Slovenia. The sickness insurance in Slovenia is especially generous when compared to the country’s generosity of disability benefits and other social insurance programmes. On average, people on sickness insurance receive over EUR 1 100 per month (gross), while the average disability pension in Slovenia is only around EUR 500 (also gross, more details in tables below).
The programme allows workers on sickness absence to receive a partial sickness benefit on top of partial earnings if they can do part-time work. Partial benefits are positively seen by ZZZS practitioners, as they present the only opportunity for sickness beneficiaries to be in contact with their employer (see more in the following subsection). Partial benefits are calculated proportionally to the hours worked, without any additional incentive to work.
Notes: Mandatory paid sick leave replacement rates. Figures include sickness insurance benefits and employer sick pay, when applicable. Source: European Commission’s Mutual Information System on Social Protection (MISSOC), United States Social Security Administration’s Social Security Throughout the World (SSTW).
2.1.3. Return to active working life
Employers do not have the responsibility to facilitate and promote the reintegration of their employees during sickness absence. However, employers cannot terminate employment contracts, other than for economic reasons or for incompetence of the employee. In case of loss of employment, the effective date of contract termination cannot be before the first day of return to work of the employee, and no later than six months after the end of the notice period.
While on sickness leave, the insured cannot access vocational rehabilitation, under the rationale that the period of sickness absence should be used to treat the ailment medically. Only upon completion of the medical treatment and medical rehabilitation, people on sickness absence can request a disability assessment and be granted the right to vocational rehabilitation within the disability insurance programme. Exceptions are claimants eligible for vocational rehabilitation from URI-Soča (see Chapter 5).
Unlike in most EU and OECD countries, sickness absences do not have a maximum duration in Slovenia. The personal general practitioner (GP), the ZZZS practitioner and the medical committee are responsible for endorsing the extension of the sickness certificate, and determining the extent to which the insured person is unable to work. The extension of sickness absence, and corresponding sickness benefits, has no maximum limit. In most countries, the maximum duration of sickness benefits ranges from one to two years (see Figure 2.2). Only in Sweden, like in Slovenia, there is no maximum benefit duration.
Note: For Germany and France, the maximum duration of benefits is calculated over a period of three years. For Slovenia and Sweden, there is no maximum duration of benefits.
Source: MISSOC (2020), https://www.missoc.org/.
2.2.1. The process of receiving sickness benefits
Upon falling ill, insured workers need to obtain a sickness certificate approved and signed by their GP. This certificate, which includes the reason for the absence and the expected duration of the leave, is sent to the employer, who is responsible for paying sickness benefits for the first 20 working days of illness (previously 30 days). During this period, in case of doubt, employers can ask for a re-examination of the sickness status.
When the employer-paid sickness period terminates, the sickness certificate is shared with the ZZZS, whose appointed physicians evaluate the reason and duration of the sickness absence (Albreht et al., 2016[1]). If the absence is approved, the ZZZS takes over the payment of the sickness benefit. If the absence is not approved, the insured worker can appeal the decision of the ZZZS physician in front of the ZZZS health commission or the court. GPs can support insured workers in the appeal process, and so can employers.
Extensions of sickness absences have to be agreed between the insured worker and the GP, and presented to the ZZZS before approval. These extensions can be requested indefinitely, to the extent that the GP considers the medical treatment incomplete. This can be a lengthy process, and result in much delayed decisions, putting employers and insured workers in a difficult position: if the ZZZS retroactively rejects a sickness leave extension, and workers were absent from work before the decision, employers have the right to dismiss the worker despite being on sickness leave.2
Since February 2020, the process of receiving sickness benefits has been digitalised, which promises to reduce inefficiencies in the process. The so-called eBOL system, a system of electronic sickness certificates imbedded within the ZZZS digital platform, is the first attempt to digitalise the process of sickness benefit entitlement. For GPs, this system implies that they can digitally sign the electronic sickness certificate, which is automatically shared with the ZZZS. It also allows GPs to access the history of sickness absences of an insured worker, acting as a potential deterrent to “doctor-shopping” for sickness absences. For the ZZZS, the eBOL is imbedded in their digital platform, which also contains data on working hours and salaries, making the calculation of benefits faster and more precise.
2.2.2. Assessment of sickness, a medical perspective
The assessment of sickness benefit eligibility performed first by a GP and then a ZZZS physician, takes a medical perspective. In Slovenia, like in many other EU and OECD countries, sickness benefits are granted exclusively on a medical basis, without taking into account work capacity and occupational factors. Even more, sickness absences are granted (and extended) until medical treatment is completed. This close link with a medical definition of a sickness absence precludes from activating workers that have some work capacity, albeit possibly in other jobs. The United Kingdom’s fit note (which replaced the previous sick note) and the Swedish rehabilitation chain present two recent examples of how work capacity aspects and considerations of job change can be included into the legislation (see below).
If ZZZS physicians prescribe long-term sickness absences to an applicant, the employer has to provide a job description. This is the only exception in which the assessment goes beyond the medical perspective to take into account some occupational factors. De facto, this barely affects assessment, to the extent that medical treatment must, nevertheless, be completed during the period of sickness absence. This is particularly problematic for chronic health conditions and mental health issues, for which treatment may never be completed.
2.2.3. General Practitioners, the key actors in the assessment of sickness
In the current system, GPs are the gatekeepers for sickness insurance without much supervision. For spells shorter than 30 days (now 20 days), a single GP’s opinion forms the basis to grant a sickness absence to a worker. Beyond the first 30 (20) days of a spell, GPs are responsible for i) helping claimants create a dossier for the ZZZS, ii) appealing the ZZZS decisions to the health committee or court, and iii) requesting extension of the sickness absence. Anecdotal evidence suggests that appeals and extension requests to the ZZZS health commission are often a success, which shows the lack of control over GPs assessments.
Lack of training, the close link with the sick worker, and shortages of medical personnel lead to unequal treatment and the risk of overly long absence durations. First, GPs do not systematically receive training in insurance medicine, for which their assessments to grant sickness benefits may not always be adequate. There is a risk of focusing on the medical aspect at the detriment of the occupational aspect of sickness absences, resulting in overly long absence durations. Second, GPs responsible for granting the right to sickness benefits are often family doctors, closely linked to the claimant. This may also result in an excessive leniency to grant benefits, and a too low activation of sickness beneficiaries. Lastly, there are regional differences in GP shortages and capacity in Slovenia, which, compounded with economic and labour market differences across regions, results in considerable differences in sickness absence take-up across regions (see Figure 2.4).
GPs role goes beyond the assessment of sickness absence, to the application to disability benefits. Treating GPs are also responsible to ensure the transfer from sickness to disability benefits, by helping claimants collect the necessary information for evaluating the claimants’ case. After one year of sickness absence, GPs are expected to present the claimants’ case for assessment at the Pension and Disability Institute of Slovenia (ZPIZ).
2.2.4. ZZZS physicians have restricted power
There is limited guidance for ZZZS physicians’ in their sickness assessments. ZZZS physicians rely on ICD-10-based guidelines that remained unchanged since 2003. Physicians are kept up-to-date by participating in lectures covering the diagnostic and treatment for several medical fields, but not on the consequences on (in)capacity to work following a sickness.
ZZZS physicians do not have the competence to qualify an injury or disease as occupational. Occupational diseases and injuries are assessed by a specialist in occupational medicine, transport and sports, or can be qualified after a ZPIZ decision. In this case, the benefit is calculated retroactively.
ZZZS physicians’ role in the sickness assessment is bounded by GPs opinions, but there are some checks-and-balances mechanisms. For instance, if ZZZS physicians have doubts on the independence of a GP’s opinion on a case, they invite the insured person to present his/her case at a personal hearing.
ZZZS physicians, and more generally, the ZZZS, has a limited role in ensuring the transition from sickness to disability insurance. ZZZS physicians may help in the process by providing their opinion on the case, but this is neither standardised, nor frequent. The request to apply to disability insurance must come from the employee or the GP, and while the ZZZS should be incentivising applications to disability insurance after one year on sickness leave, they have a limited role in making it happen.
2.2.5. Checks on sickness assessment
The ZZZS performs checks on the sickness assessment, but these represent a very low share of the claims, and are rarely binding. Data from 2019 suggest that the ZZZS performed almost 4 000 checks on sickness claims, or about 1% of all sickness absences financed by the institute (ZZZS, 2020[2]). The ZZZS sends staff to verify that claimants are abiding with GPs prescribed medical treatment. This is not done systematically, however. In addition, the right to sickness benefit cannot legally be withdrawn, for which the penalties for fraudulent sickness claims is limited to requesting a re-assessment.
The ZZZS partly relies on anonymous reports of fraudulent sickness behaviour. Through their website, anonymous users can flag fraudulent behaviour to the ZZZS, information that can be taken into consideration for further checks.
When the ZZZS assesses the validity of a claim, a relatively high share of claims are identified as violations, with considerable regional variation. On average, 6.4% of the checks the ZZZS performs represent a violation of the conditions for receiving sickness absence, as explained above, mainly due to patients not following GPs’ prescriptions. Table 2.2 shows that there is substantial regional variation. Western regions, like Koper, Kranj and Nova Gorica, have a share of identified violations below 4%. Eastern regions, for instance like Maribor, have a much higher proportion of violations. These differences are not linked to a lower or higher check-up rate in Eastern regions compared to Western regions, but rather coincide with differences in economic and labour market conditions.
Sickness insurance take up has been steadily increasing since the Global Financial Crisis. Figure 2.3 plots the yearly average number of sickness spells per employee (left axis), and shows that the average spells per employee went from 0.74 in 2008 to 1.09 in 2019. Therefore, every employee in Slovenia has, on average, more than one sickness spell per year.
Note: Sickness spells per employee are all sickness incidences in a given year divided by the number of employed persons in that year.
Source: OECD calculations based on data from the National Institute of Public Health (NIJZ) www.nijz.si/en and Eurostat LFS employment data (accessed 3 February 2021).
Incidence of sickness absences has increased for all age groups in the last five years, but particularly for younger workers. As shown in Figure 2.3, the average number of sickness spells for workers aged 15 to 24 has doubled in only five years, compared to a 30% increase overall. For the younger age bracket, while the number of employed has declined over time, the number of sickness cases has remained steady.
There are large regional differences in the incidence of sickness absences, corresponding with an ageing of the population. Sickness receipt rates vary substantially across the Slovenian territory, with some regions having an average of less than one sickness spell per employee per year, while others have an average of 1.3 to 1.7 (Figure 2.4). There is not a clear correlation behind these regional differences in sickness insurance take up and regional economic conditions, neither positively nor negatively.
Source: OECD calculations based on data from the National Institute of Public Health (NIJZ) www.nijz.si/en and the Statistical Office of the Republic of Slovenia (SURS) www.stat.si/statweb/en.
There is a gradual shift over time in the composition of sickness spells, as long-term sickness spells become increasingly common. Figure 2.5 decomposes sickness spells of 45 days3 or longer by duration of the spell, and shows that over time, the proportion of sickness spells of two years and more is increasing: from 6% of all ZZZS-financed sickness spells in 2014, to 14% in 2019. Sickness spells of six months or longer represent half of the spells longer than 45 days, particularly in more recent years.
Source: OECD calculations based on National Institute of Public Health (NIJZ) data, www.nijz.si/en.
Women and older workers have a particularly high long-term sickness insurance benefit take up. Figure 2.3 showed that the average number of sickness spells per employee increased for all age groups in recent years. These figures, however, do not take into account the duration of the sickness spell, which, when under 30 working days (now 20 days), is paid by the employer. Figure 2.6 shows that, when focusing on sickness spells covered by health insurance and accounting for repeat spells, the incidence of sickness insurance receipt increases very significantly with age. Workers from age groups 50 to 64 have the highest shares of sickness beneficiaries, and represent over 50% of the total number of beneficiaries. Among those aged 55 to 59, 1.4% of employees are on sickness insurance for over 30 days, at least twice a year. The figure also shows that women are on long-term sickness leave about twice often as men. Experts suggest that this can be partly explained by the gender-specific pattern in the causes of sickness absence: women often take long-term sickness leave during pregnancy, in addition to having different diagnoses that require longer-term sickness leave, for example mental illnesses.
Note: Included beneficiaries are those with at least two sickness spells longer than 30 working days (e.g. covered by sickness insurance).
Source: OECD calculations based on National Institute of Public Health (NIJZ) sickness data (www.nijz.si/en) and Statistical Office of the Republic of Slovenia (SURS) employment data (www.stat.si/).
Reported sickness spells possibly underestimate the issue of long-term sickness due to the dynamic nature of sickness absences. Statistics reporting sickness spells, like unemployment spells, relate to the period of continued sickness leave. Persons on sickness leave repeatedly over the year may not appear in the statistics as long-term cases, even if the cumulated absence over the year would clearly identify them as de facto long-term absentees. In the Slovenian case, where only employed persons are eligible for sickness insurance, and where the dismissal of persons on sickness leave is difficult, the issue of repeated sickness periods can be particularly important. As Figure 2.7 shows, on average Slovenian workers with longer-term absence spells (i.e. any spell of 45 days and longer) experience over two sickness spells per year, compared to 1.2 such longer spells in the Netherlands and 1.4 in Belgium. Presenting these statistics differently, in 2019 there were over 6 700 persons in Slovenia who had at least two sickness spells covered by the ZZZS within the year.
Source: Slovenia’s National Institute of Public Health (NIJZ) www.nijz.si/en, Statistics Netherlands (CBS) www.cbs.nl/en-gb and SDworx www.sdworx.be/nl-be for Belgium.
Sickness absences due to mental health issues are steeply increasing over time, while absences for other key causes remain constant. Figure 2.8 shows the proportion of sickness claimants for one year or longer over the total number of sickness claimants for those qualifying through mental health diseases, cardiovascular diseases (CVD), diabetes, and work injuries. The proportion of long-term claims among all sickness claims with a mental health condition has doubled since 2012, from 1.3% to 2.6% in 2017. For the other health conditions reported, these shares have remained constant.4 Figure 2.8 illustrates the need to pay particular attention to mental health issues as a cause of sickness absence, which increasingly lead to very long sickness spells and may be affected in different ways by the rules and regulations of the sickness and disability insurance programmes. This is even more important than the data suggest because it is well known that, due to widespread stigma and discrimination, mental health conditions tend to be vastly underreported as a reason for sickness absence in all OECD countries (OECD, 2015[3]).
Source: OECD calculations using linked administrative data from National Institute of Public Health (NIJZ).
Data on average sickness payments are not readily available. Statistics on average sickness payments by the ZZZS are not regularly published and were not available for this report. Replacement rates suggest that benefits are possibly high, at least from an international point of view and when compared to other forms of social support (described in the following sections). This is, however, an imperfect measure of the (relative) generosity of the system. Another approach is to use the yearly expenditures on benefit payments for sickness absences by the ZZZS, the number of days of sickness covered by the public institute, and estimate a daily benefit payment. According to the (ZZZS, 2020[2]), in 2019 sickness benefit expenditures amounted to EUR 381 553 163, and the institute covered 6 730 227 days of absence. This results in an average gross benefit of EUR 56.7 per day, or EUR 680 per month. For comparison, the 2020 minimum wage was set at EUR 940 per month (gross), and gross average monthly earnings in 2018 were EUR 1 719 (EUROSTAT, n.d.[4]).
Sickness payments of spells funded by the insurance have been increasing over the past years. A last approach to approximate sickness payments is to use average wage in the year before sick leave and the replacement rates to calculate the (theoretical) sickness payment. Figure 2.9 reports the daily payment from using this approach, from 2006 to 2016. The figure shows a steady increase from 2006 to 2014 in the average payments made to sickness beneficiaries with spells 30 days or longer. The generosity of sickness insurance appears to have increased over time, in line with the increasing trend in average wages. Since other social supports are determined at the monthly level, it is useful to transform the daily sickness payments into monthly payments.5 Data show that average sickness payments increase with spell duration. This is in line with a higher replacement rate for spell durations of 90 days or longer. It could also be the result of the fact that older workers, who usually have higher earnings, are over-represented among long-term sickness beneficiaries. The increase in average sick pay for longer spell duration, however, could also suggest a selection bias, as those who benefit most from sickness insurance appear to stay on the programme for a longer period.
Note: Sickness pay (gross) per day calculated from pre-sickness wages for sickness beneficiaries with spells longer than 30 working days. These data include only claimants on sickness leave due to the four diagnosis available in the data: mental health, cardiovascular disease, diabetes or work injury.
Source: OECD calculations using linked administrative data from National Institute of Public Health (NIJZ).
As longer-term sickness absences increase, sickness benefit expenditures are expanding. Descriptive statistics of the sickness insurance programme show that longer-term sickness is increasing. There is a shift in the composition of sickness absences, with sicknesses of one year and longer and even two years and longer, increasing and occupying a larger share of all absences. This represents a massive cost for the ZZZS, which is fully responsible for covering the cash benefits to workers on sick leave beyond 30 working days of sickness. In fact, sickness benefit expenditures have increased by 55% from 2015 to 2019, and per-day benefit expenditures by 13% (ZZZS, 2020[2]; ZZZS, 2016[5]).
The assessment of sickness may be partly responsible for causing the long-term sickness trend. Assessment for sickness insurance entitlements hinges on medical files entirely, and does not take into account the work capacity of the claimants. In addition, sickness can only end upon completion of medical treatment while on sickness leave. This may lead to long-term sicknesses through two mechanisms:
Completion of medical treatment or medical rehabilitation may be difficult to assess when it comes to sicknesses that do not have a medical treatment with a clear end (such as some mental health conditions), resulting in overly long sickness absences. Descriptive statistics of the programme suggest that this may indeed be the case, as claimants with mental health conditions have a high, and increasing, share of sicknesses longer than one year.
The long waiting times in the health care system in Slovenia make the condition of completion of medical treatment and medical rehabilitation inherently difficult. ZZZS experts estimate that almost one-third of long-term sicknesses could be due to the lack of timely health care access. The Slovenian health care system experiences further strain currently from the COVID-19 pandemic, which threatens to further extend the duration of sickness absence.
The features of the sickness insurance system could also be partly responsible for the increase in long-term sickness. The lack of maximum duration of publicly funded sickness benefits, the generosity of the benefits (especially compared to disability benefit), and the limited checks on sickness absences could play a key role in discouraging sickness claimants from returning to work (or applying for disability benefit – see below). At the same time, the limited financial costs of sickness born by employers, and the fact that they cannot remain in contact with their employees on sickness leave, prevents them from facilitating the return to work of their employees.
Who are long-term sickness beneficiaries? Older workers, slightly lower-paid, but otherwise with similar characteristics than other claimants. Table 2.4 shows that the individual characteristics of the typical long-term sickness claimant are quite similar to those of short-term sickness claimants in terms of educational attainment, employment characteristics (type of contract, shift work, working hours), and intensity of health shock. Otherwise, short-term sickness claimants are most often women working as professionals (i.e. health, teaching, science, legal and IT professionals) while claimants with longer durations are most often men working as craft or related trade workers. Wage seems to correlate negatively with the length of sickness absence, and age positively.6
What are the consequences of long-term sickness spells? A context of lack of activation and employment protection. To answer what happens to employees returning from sickness leave, it is important to keep in mind two factors (discussed in some detail in later parts of this report). First, long-term sickness spells may lead to a depreciation of the working capacity, particularly accentuated by the fact that there are no processes or measures seeking to maintain the working capacity of sickness insurance claimants. Second, employers cannot easily fire employees while on sick leave. Many will thus wait for their employees to return to work before firing them without any logistical hassle. In view of these factors, it is likely that employees staying longer on sick leave will have a harder time going back to work.
Understanding the consequences of (long-term) sickness absences requires linking administrative records. Data on the exit routes from the programme are not readily available to the ZZZS, which limits the institution’s ability to evaluating the labour market implications of sickness insurance. NIJZ data record the termination date of a sickness spell but contain no information on where to the insured person transfers. Employment data, collected by the ZZZS and managed by both the Statistical Office (SURS) and the Employment Service of Slovenia (ESS), are not linked to sickness insurance records. Neither are disability insurance data, collected through the Pension and Disability Insurance Institute (ZPIZ). There is a legal precedent for linking these three sources of data for research purposes, but again, current data protection regulations prevent this from being a regular operation. Improved data availability is the key for building the evidence needed to inform the direction of reform. This report uses administrative data from multiple registers, owned by different authorities, to study the exit routes from sickness insurance.
Most employees remain in employment with the same employer after returning from sickness absence. Figure 2.10 follows sickness insurance claimants who were absent from work for 1<6 months (Panel A), 6<12 months (Panel B) and 12 months or longer (Panel C), for up to one year after the termination of their sickness claim. Regardless of the length of absence, month zero marks the exit from sickness insurance, where all claimants are formally still in employment, which is a condition to receiving sickness insurance in the first place. At month one, i.e. one month after the termination of their sickness claim, there is a substantial drop in the probability of being in employment with the same employer. At the end of the observation period, i.e. one year after ending a sickness insurance claim, over 50% of employees have remained employed with their current employer. This is a large share for international standards, possibly because only employed workers are entitled to sickness benefits.
Persons terminating a long-term sickness insurance claim have a much lower probability of remaining in employment. Figure 2.10 shows that the largest difference in exit pathways across different claim durations arise during the first month after ending a claim. For intermediate and long-term claims (6<12 months, and 12 months or longer) the initial drop in the probability to stay with the same employer is large compared to that of claimants on sickness leave for 1<6 months. After the first month, the trends over time in the exit from employment are quite similar across all claim durations. This is consistent with employers postponing the discharge of their employees on sick leave to the moment they return to work.
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 calculations using administrative data from National Institute of Public Health (NIJZ), Employment Services Slovenia (ESS), Pension and Disability Insurance Institute Slovenia (ZPIZ).
Transitions from sickness to disability insurance are rare after short absence spells but quite frequent for those who have been absent from work for at least a year. Figure 2.10 shows that one year after the end of a long-term sickness absence, about 17% have retired via a disability and another 6% via an old-age pension; together almost one in four. Among sickness insurance claims of intermediate duration (6-12 months), this share is about 10% and among shorter absences less than 5%.
Long-term sickness claimants do not compensate their more frequent job losses by finding new jobs. Table 2.5 shows the results from a model that estimates the probability of taking a given exit pathway by duration of sickness claim, and confirm the impressions derived from Figure 2.10. The probability to remain employed with the same employers decreases notably with the duration of a sickness insurance claim, as described above. Long-term sickness claimants are not able to compensate their job losses by finding new employment, which Table 2.5 shows to be non-statistically significant.
There is a substantial probability to return to sickness leave, particularly for intermediate durations of sickness absence. People with sickness claim durations of 6-12 months have the highest probability to return to sickness leave within a year (and typically already within three months). Long-term sickness claimants have repeat sickness absences less often, partly because many of them have exited the labour market already and partly because they may not be entitled to benefits if they lost their job and did not find a new one.
Older workers, low-skilled workers and those with low wages or a temporary contract are less able to keep their employment position after long periods of sick leave. Table 2.6 shows the differential effects of claim duration on the probability of employment (either in the same job or in a new job) by individual characteristics of claimants. Poor employment and poor skills largely decrease the probability to remain employed for long-term claimants to an extent that is greater than the severity of the health shock.
References
[1] Albreht, T. et al. (2016), Health Systems in Transition: Slovenia (Vol. 18 No. 3 2016).
[4] EUROSTAT (n.d.), Earnings - Labour Market (incl. LFS), 2021, https://ec.europa.eu/eurostat/web/labour-market/earnings (accessed on 4 November 2021).
[3] OECD (2015), Fit Mind, Fit Job: From Evidence to Practice in Mental Health and Work, Mental Health and Work, OECD Publishing, Paris, https://dx.doi.org/10.1787/9789264228283-en.
[2] ZZZS (2020), Podatki o obveznem zdravstvenem zavarovanju.
[5] ZZZS (2016), Podatki o obveznem zdravstvenem zavarovanju.
Notes
← 1. For COVID-19 related sickness absences, this regulation is lifted temporarily.
← 2. See: https://www.rtvslo.si/dostopno/zzzs-je-skoraj-petino-odlocb-o-dolgotrajni-bolniski-izdal-z-zamudo/511546.
← 3. Figure 2.5 is based on data shared by the NIJZ, where sickness spells were pooled in pre-determined duration bins. Focusing on sickness spells of 45 calendar days or longer, which corresponds to the sickness spells entirely financed by the ZZZS through sickness insurance (30 working days).
← 4. Figure 2.8 is constructed using linked administrative record of sickness and employment records, which contain only four types of qualifying causes: mental health diseases, cardiovascular diseases (CVD), diabetes, and work injuries. These causes are not representative of the stock of sickness claims as they together represent less than 10% of all spells, although probably considerably more for long-term spells. Results from the administrative records, although illustrative, should thus not be generalised to the Slovenian sickness insurance as a whole.
← 5. While this is useful, it adds some noise into the approximations, as sick payments are calculated using the number of working days but the administrative data records calendar days of sick leave. To overcome this issue, one can estimate the number of working days using the information of working days and calendar days in a certain month: (number of calendar days absent in a month / number of days in a month) x number of working days.
← 6. Table 2.4 presents only the median characteristics, but the share of 25-49 year-olds decreases with duration on sickness leave, while the share of 50-64 year-olds increases.