Sweden: Reducing working time

The past decades have seen much change in the regulation of working time. While legislation has reduced maximum statutory working time, collective bargaining in recent years has sought to introduce further reductions, coupled with time banking systems and increased flexibility.European Industrial Relations Review looks at recent developments in working time regulation and examines how working time reduction can be achieved and how it links in to other key social policy areas in Sweden, such as decreasing sickness absence.

Background

Working time has been a constant focus for debate, legislation and collective bargaining in Sweden over the past few decades. During the first 20 years of the post-Second World War period, maximum working hours were reduced by legislation, bringing the working week down from 48 to 40 hours. The maximum statutory working day is eight hours.

The average collectively agreed working week tends to be below this figure, however - according to the Mediation Authority (Medlingsinstitutet), the average collectively agreed working week in 2003 was 38.8 hours; according to the European Labour Force Survey, average hours worked per week by full-time employees were 39.9 in 2002 (39.6 for women and 40.1 for men).

The legislated reduction in the normal working week was accompanied by an increase in statutory annual holiday entitlement, from one to four, and eventually five, weeks. Many collective agreements award annual days of paid annual holiday, ranging between five and 10 additional days.

The past few years have seen a negotiated reduction in working time at sectoral level, with many industries offering their employees a choice between reduced working time, pay increases and contributions into pension funds.

Most recently, the debate on working time has become linked to wider concerns about the changing demographic situation of Swedish society - an ageing population and a falling birthrate - persistently high levels of unemployment and sickness absence.

Below, we examine demographic and labour market developments and look at working time issues such as part-time work, overtime and flexibility by means of time banking systems, in addition to related issues such as parental leave.

The demographic situation

The composition of Swedish society is set to change considerably in the medium term. Over the next 10 years, 1.1 million Swedes will reach the retirement age of 65, increasing the number of people over that age by 20%. However, due to a falling birthrate, the size of the workforce is not expected to increase. The popular interpretation of this is that a smaller base of workers will need to support an increasing number of retired people born in the 1940s.

However, it is not necessarily the case that public finances will be overburdened immediately - the cohort born in the 1940s has saved into public pension funds and can be expected to be healthy and live into their 80s, not needing any support. Further, they will spend their private savings, which will be taxed and therefore improve public finances. The group of people who have reached the age when they can be expected to need local government old age care are those born in the 1920s, which is a small cohort. Therefore, it is unlikely that public finances will suffer during the coming 20 years.

The labour market

Sweden is suffering from considerable labour market problems. The unemployment rate is 6.1%, according to Eurostat figures for January 2005, or 5.5%, according to national figures from Statistics Sweden. However, an estimated further 3% of the workforce is employed in labour market measures, bringing the actual total up to at least 8%. Although an increasing number of workers are entering retirement, they are not being replaced by unemployed people.

This practice can often lead to an increase in productivity in the manufacturing industry. However, in some sectors, such as the local government education and care sectors, replacement labour is vital to replace the large number of teachers, day care staff, nurses and nursing aids who are now nearing retirement. For example, in the Värmland county, in the mid-western part of Sweden, one-third of nursing aids, 25% of doctors and 20% of registered nurses will retire before 2013. The situation is similar in many other regions - with the exception of Stockholm, due to plentiful labour supply from surrounding areas.

The Organisation for Economic Cooperation and Development (OECD) examines the Swedish demographic and labour market situation in its report on ageing and employment policies¹. It paints a pessimistic picture, stating that by 2030 almost one in four people in Sweden will be over the age of 65. It believes that the only way to avoid pressure on public expenditure, acute labour shortages and slower economic growth is to reverse the long-term trend towards early retirement. It does note, however, that action has already been taken to reform Sweden's public pension system to encourage people to continue in work beyond the age of 61.

Decline in overtime working

Swedish law allows a maximum of 50 hours of overtime to be worked each month, provided that there is a special need for the work to be carried out. The annual statutory limit on overtime is 200 hours. Additional overtime can be authorised by the occupational safety and health administration (Arbetsmiljöverket). Compensation for overtime worked in terms of cash or additional rest is regulated in collective agreements.

Despite robust economic growth in recent years, the volume of overtime worked has declined. During the first quarter of 2004, the volume of overtime was the lowest for 19 years, according to a survey of 200,000 employees in 250 workplaces, carried out by the Confederation of Swedish Enterprise (SN).

In the manufacturing industry, the percentage of paid overtime hours in relation to total hours worked was 5% in 1995, dropping to 3% in 2003. Further, between 1985 and 2003, overtime worked in the transport and communications sector fell from 6% to 4% of total working time. In the service sectors - retail trade, hotels and restaurants - the amount of overtime worked fell from 2.5% of total working time in 1999 to 2.3% in 2003.

One explanation for this may be the more flexible local scheduling of working time agreements that make it possible for management and workplace unions to agree to work less one week and more during the following week to meet fluctuations in demand. Further, working time banks allow for longer-term working time flexibility (see below).

Time banking systems

The operation of working time banking systems is provided for by many industry branch agreements in Sweden. The concept of saving working time in a time bank has a long history, taking root at the metalworkers' convention held in 1993. This convention set a target of negotiating a working time reduction of 100 paid hours a year. At the time, the union wanted to focus on gaining more leisure time, as it was felt that pay was at acceptable levels. A time banking system was already in existence, allowing those who so wished to deposit pay for overtime worked.

The original metalworkers' vision was for a pure working time reduction. However, over the years, the scheme has developed, offering more choice to workers. The paper sector agreement, covering employees that mostly work under three- or four-shift systems, was the first agreement to introduce alternatives to a simple working time reduction. It allowed workers to choose between shorter working time, cash or payments into individual pension funds. An analysis of options chosen by paper industry workers nationally in 2000 found that the most popular option was working time reduction, among workers of all age groups excepting those under 30 (who chose the cash option). For more details, see EIRR 324 p.23.

Other industry sectors swiftly followed suit and most now offer workers this choice between working time reduction, cash or pension contributions. Under the engineering industry agreement concluded in March 2004 (EIRR 365 p.29), 84 minutes a week for a dayshift worker may be saved into a time savings bank, rising to 94 minutes a week from 1 April 2006. Workers on a two-shift system may save slightly more per week (202 minutes). Over a year, this amounts to 84 minutes (or 94 minutes) multiplied by 47 (the number of working weeks in the year). Once 100 hours are saved, an employee is asked how they would like to spend the hours. If there is no local agreement, time in excess of 100 hours at the end of the year will be compensated by means of pension insurance premia.

No overall evaluation of local practices has been made, although there is some data available for individual enterprises. For example, it shows that 85% of miners in the metal ore mining industry prefer to take time off. A common choice in the engineering industry appears to be to take the occasional day off (eg if a public holiday falls in the middle of a working week), sometimes enabling a worker to take a week off without losing pay or holiday entitlement.

The time banking system is administered at the workplace level and seemingly without conflicts. No case related to time banking has as yet been referred to national level for central negotiations.

In general, employers do not want to extend the system further. However, commentators note that this is because they do not want to see any further reductions in working time, rather than a sign that the system is not working.

This industry initiative has not been followed by unions in the public and private service sectors, as they believe that there is a need to focus on pay rather than working time. In fact, in local government, many workers would like to extend their working time rather than reduce it, as a significant proportion are working on a part-time basis involuntarily.

Reducing sickness absence

Persistent high levels of sickness absence have been a constant preoccupation for the Swedish government. Sickness benefit is relatively generous, at 80% of previous pay, for an unlimited period. Although the government has put into place a range of reforms to sickness benefit in recent years (EIRR 355 p.14), the level of benefit is currently 80%, after being reduced in the early 1990s. The level of sickness benefit is expected to be a major issue in the 2006 elections, with the political opposition believing that it should be reduced to 65% of previous pay.

In the mid-1990s, sickness absence, and in particular long-term sickness absence (one year or more), rose sharply, almost doubling to 324,000 people by the end of 2003. The government has set itself a target of reducing the 2002 level of paid sick days by 50% by 2008, and the minister of working life is hopeful that this target can be reached. It would appear that he has some grounds for optimism: in August 2004, sickness benefits were paid to 263,700 people, 36,000 fewer than in August 2003. At the same time, the number of people receiving long-term sickness benefit reduced by 15,000, to 118,000. It is thought that there are two reasons for this: an increase in early rehabilitation of people on sickness benefit; and an increase in the number of long-term sick people moving into early retirement, which is not such a positive development.

One way that the government has been trying to rehabilitate people on long-term sick leave is to encourage people reporting sick to return to work on a part-time rather than a full-time basis, while still collecting reduced sickness benefits. The idea behind this is not to sever the links between the employee and their workplace - many employees, once they have taken sick leave, never return to work. This approach has been praised by many, although new research carried out by the Swedish National Institute of Public Health² finds that part-time sickness absence can actually prolong periods of sickness instead of making them shorter. This is because there appears to be a risk that this can be used by individuals to shorten working time, subsidised by the health insurance system.

New strategies and experiments

A range of temporary working time reduction experiments have been held over the years to see if shorter working time can help to reduce sickness absence without compromising the flow of production. The idea was that, if successful, the experiments could be made into general policy. One example involved home help workers in the iron ore mining city of Kiruna. The working week for these employees, who visit elderly and handicapped people in their homes, was reduced to 30 hours. This allowed them more leisure time, and time for sport and exercise and had the effect of reducing sickness absence. It also allowed the recruitment of additional home helpers from the ranks of unemployed people. The experiment was popular among the participants - both workers and clients. However, it was finally abandoned by the local government, as it did not benefit from the reduced health and unemployment insurance costs but was obliged to pay increased wage costs.

Experience shows that most temporary working time reduction projects are well received in the early stages, but most have finally tuned out to be too costly and have been discontinued.

Nevertheless, the minister of working life allocated a further SKr100 million in November 2004 to a large research project involving shorter working time, as a way of reducing sickness absence. The project will be run by the Swedish Institute for Working Life.

Part-time work

The role of part-time work appears to be changing, with commentators observing that an increasing number of part-timers would rather be working full time, resulting essentially in a working time reduction paid for by the worker. Whereas part-time work used to be carried out predominantly by women, an increasing number of students and men over the age of 55 now appear to be working on a part-time basis. Accordingly, in 1990, 40% of women were working part time - a figure that dropped to 34% by 2002. Similarly, in the same year, women held 83% of all part-time jobs. By 2002, this figure had dropped to 77%. This trend has led commentators to surmise that part-time work is becoming less of a choice and is functioning more as a type of part-time unemployment for those who cannot find full-time jobs. In the case of students, increased living and housing costs are leading many to work alongside their studies.

Nevertheless, women still make up the largest proportion of part-time employees, working predominantly in the public care sector, in commerce and in transport. Thus, part-time working - which can also be considered as part-time unemployment for those wishing to work full time - remains primarily an issue for women.

In 1999, the Delta study³, a government commission, analysed part-time working, temporary working and unemployment benefits. This commission's report led the minister of labour to order various labour market bodies, in cooperation with the European Social Fund and the Swedish equality ombudsman, Jämo, to create measures to reduce part-time work by 50% between 2002 and 2004.

In the local government sector, a major effort was made between 1999 and 2000, as a result of the Delta study, to reduce the number of employees who had to apply for partial unemployment benefits to obtain a living income. By 2003, the number of women employed part time in the care sector had been decreased by 50% and the number of temporary workers had been reduced by 26%.

Reducing working time through parental leave

Many parents are able to reduce their working time by spreading out their parental leave entitlement over the period until their child reaches eight, as permitted by law. Parental leave entitlement in Sweden is generous by EU standards. First introduced in 1974, parental leave entitlement has been steadily increased and is now 16 months, paid at a rate of 80% of former salary up to an earnings ceiling of SKr25,000 a month (this will be increased to SKR33,000 a month from 1 January 2006), plus a further two months as a flat-rate benefit. Parents can also take a maximum of 120 days of leave a year to look after a sick child under the age of 12, paid at the same rate as parental leave.

Critics of the government have opposed the move to raise the earnings ceiling, warning that parents may misuse the provision allowing them to take paid leave to care for a sick child. However, the Swedish health insurance administration is confident that this will not happen.

Sabbatical leave

Another way of reducing working time, albeit on a temporary basis, is by using sabbatical leave. The right to take sabbatical leave was introduced for all employees on an experimental basis in 2002, after the Swedish environment party (MP) made this a condition for supporting the Social Democratic minority government.

This programme allows employees to take a sabbatical, paid by the Swedish labour market service (AMS) at 85% of unemployment benefit (which is 80% of former pay up to a ceiling of SKr20,000 a month). The intention was that the person taking sabbatical leave would be replaced by an unemployed person, preferably somebody who has been unemployed on a long-term basis. The programme was introduced on an experimental basis in 12 communities around the country.

Evaluations of the project show that the replacements have generally not been long-term unemployed people but those who already have a job and who are using the programme to find a better job. Women in the public sector were the main beneficiaries of the programme.

Nevertheless, the scheme is to be made permanent across the whole country, allowing 12,000 people to take sabbatical leave. However, the AMS will oversee the implementation of the programme, ensuring that those on sabbatical leave will be replaced by long-term unemployed people, those with disabilities or new immigrants.

Future prospects

Since the early 1970s, a series of national commissions on working time have been set up to look at statutory working hours. Political initiatives in the Swedish parliament from different parties and individual parliamentary members are by tradition referred to national commissions and this is seen by some commentators as a deliberate way of taking the momentum out of certain initiatives. In the case of working time, there has been no government action based on any of the commission reports.

Nevertheless, trade unions have since kept up pressure for further working time reductions, with the trade union confederation LO pushing for a long-term target of 30 hours a week. However, in recent years, these visions have begun to seem increasingly utopian - the LO convention held in 2004 (EIRR 368 p.14) stated that "the demand for a six-hour day remains out of reach for the foreseeable future".

On the other hand, employers are increasingly demanding longer working hours, often without financial compensation, and politicians point to the demographic situation, which they believe will result in those employed working longer hours. Therefore, trade unions have in recent years been concentrating their efforts on defending the maximum 40-hour week in Sweden, in the context of pressure to increase working time in Germany and in France.

Overall, commentators believe that it is unlikely that there will be any more formal legislated working hour reductions. In parallel, the campaign by industry and some politicians to increase working time by means of later retirement, longer shifts or reduced annual holiday is thought unlikely to succeed.

Conversely, it is expected that any changes to working time in the coming years are likely to be the result of workplace bargaining or employer unilateral decisions, where unions do not exist or do not have the necessary bargaining power.

¹ "Ageing and Employment Policies/Vieillissement et politiques de l'emploi - Sweden". Organisation for Economic Cooperation and Development (OECD). 23 May 2003, ISBN: 9264199969. Price: €24/$31/£15/¥2900.

² "The high sickness absence - facts and consequences". Swedish National Institute of Public Health (Folkhälsoinstitutet).

³ The Delta study, SOU:1999:27.

Key working time statistics
Statutory maximum working week	40 hours
Average collectively agreed working week	38.8 hours (2003)
Usual working week for full-time employees (all)	39.9 hours (2002)
Usual working week for full-time employees (men)	40.1 hours (2002)
Usual working week for full-time employees (women)	39.6 hours (2002)
Statutory maximum working day	8 hours
Statutory annual leave entitlement	25 days
Collectively agreed annual leave entitlement	Between 5 and 10 extra days
Overtime working (% of overtime hours worked in relation to overall hours)	3% (in manufacturing, 2003 - down from 5% in 1995)

Sources: Swedish Mediation Authority; European Labour Force Survey; EIRO.