Effects of supplementary sickness compensation on absence from work

In 2007, the Institute of Labour Market Policy Evaluation conducted a study examining the effects of changes in the Swedish national sickness insurance policy in the municipal sector. The research reveals that the increased income replacement rate, due to the reintroduction of collectively agreed additional compensation in 1998 for periods of sick leave of at least 91 days, increases the rate of sickness absence by 4.7 days.

Changes to sickness insurance system

During the 1990s, the national sickness insurance system underwent frequent changes, most of which led to reductions in the income replacement level during a period of sickness absence. At the beginning of the 1990s, the total income replacement level was 90% of the lost income. A relatively high level of sickness absence and the economic crisis in Sweden motivated subsequent reductions.

From the 1970s until 1991, employees working in sectors covered by collective agreements could receive additional compensation from collective sickness insurances, in addition to the income replacement from the national sickness insurance system.

In 1991, the income replacement levels were lowered and a reduction policy was introduced, preventing additional compensation from collective insurances after the 90th day of sickness absence. Any extra compensation after that day would lead to an equivalent reduction in the benefit paid out by the national sickness insurance. In 1998, the government changed the Swedish national sickness insurance policy to allow for additional compensation from, for example, collective agreements after a worker’s 90th day of absence, without any reduction in the public sickness benefit, up to a maximum of 10% of the standard wage during all days of sickness absence from the 91st to 360th day. The extent of the extra compensation was, as before, negotiated between the social partners.

Study methodology

Against this background, a study (415Kb PDF) published by the Institute of Labour Market Policy Evaluation (Institutet för arbetsmarknadspolitisk utvärdering, IFAU) aims to evaluate the effects of the change in the Swedish national sickness insurance policy on the duration of sickness absence. The study compares the extent of sickness compensation for employees in the municipal sector affected by the government’s policy change introduced in 1998 with the results pertaining to the central government sector which was unaffected by this change. The reference years used in the study are 1997, which was the year before the policy change, and 1999, with the full implementation of the change affecting all municipal sector employees.

IFAU uses a form of econometric calculation with a ‘difference-in-differences’ approach to estimate the effect of the collectively agreed additional sickness compensation. The difference-in-differences approach is a method used to look at relative differences and control the effects of a sequence of interventions – such as time trends – when comparing different groups. The data used in the study derives mainly from the IFAU database and includes all municipal and central government employees who were absent from work due to sickness absence for more than 91 days. The sickness absence of the two groups of employees is compared over two time periods between 1997 and 1999. The number of sickness absence days for which workers receive benefits is calculated for durations of at least 91 days up to a maximum of 360 days – spells of sickness absence longer than 360 days were treated the same as for 360 days’ absence – when comparing the municipal and government sectors.

Results of study

The report shows that, with increased economic incentives, the number of sick days increased, on average, the conditional absence duration by 4.7 days. The average sickness duration for municipal employees in 1999 was 259.8 days when calculating periods of at least 91 days up to a maximum of 360 days of absence. The reintroduction of additional sickness compensation increased the duration of sickness absence by 4.7 days for periods of sickness absence longer than 91 days, which corresponds to a 2.8% increase compared with the previous reference period. However, the additional compensation did not cause any effects prior to the 91st day or after the 360th day in terms of sickness absence. The results also indicate that the changes in the municipal sector are not due to sector-specific time trends.

The theoretical explanation for the increase in sickness absence associated with the collectively agreed additional compensation is due to economic incentives to prolong the duration of sickness absence. Previous research has shown a positive relationship between sick-leave income replacement rates and sickness absence. One category of people who opt for sick leave includes workers who find themselves in a so-called ‘grey’ zone, whereby it is possible, to differing extents, for these workers to attend work or to get a certificate of illness from a doctor. If an individual can choose between remaining absent from or returning to work, and has no preference for either alternative in terms of their current health status, working conditions or income replacement level, then a change in the income replacement level is likely to influence a worker’s decision to attend work.

As a consequence of the increased sick leave duration of 4.7 days, the costs for the public sickness insurance has increased by 3% for sickness absence durations of between 91 and 360 days. The total public cost for municipal workers’ sickness absence of more than 90 days beginning in 1999 increased by SEK 102 million (about €10.9 million as at 29 February 2008). According to the research, in order to cover the increased costs of the additional compensation, insurance premiums should be increased by at least 22%.

Commentary

Drawing on the results of the IFAU study, the government proposed a reform (in Swedish) of Swedish legislation in order to reduce the national sickness compensation equivalent to the gains of any supplementary compensation. The total replacement rate ceiling would be limited to a maximum of 75% of the lost income after 12 months of sick leave. Both additional sickness compensation from private sickness insurances and collective agreements were to be affected by the government’s proposed policy to reduce compensation. However, after strong criticism from the social partners, the government recently withdrew this proposal (SE0802029I).

The study supports the claim that additional sickness compensation increases the duration of sick leave between day 91 and day 360. It concludes that sickness durations for individuals on longer periods of sickness absence beyond 360 days, or on shorter periods of less than 90 days, are not hugely affected by supplementary insurance.

Paul Andersson, Oxford Research