Printer friendly page

Yonatan Reshef
School of Business
University of Alberta
Edmonton, Alberta

Based partly on: 
Mark Harcourt Relations Industrielles, 1996, Vol. 51, 1: 177-201

Before reading this document, take a look at Sweden: General Timeline.

Low unemployment is the primary goal of the active labour market policy in Sweden, rather than a secondary effect as it is with lifetime employment in Japan. The importance of this goal in Sweden reflects the long-governing Social Democratic Party's (S.A.P.) desire to satisfy the immediate needs of its Swedish Trade Union Confederation (L.O.) supporters for more jobs and income (Ginsburg, 1983). The S.A.P. government has also used full-employment to stimulate the demand for unskilled labour and thereby reduce income inequality and relief dependency, which together threaten to impose major financial burdens on an already costly welfare state (Ginsburg, 1983). Note, the Social Democratic Party had an unbroken run in office from 1932 to 1976 (apart from a three-month interregnum in 1936), then returned to power in 1982 and lost power again in the 1991 election, only to regain it in the 1994 election and keep it in the 1998 election.

In the fight against unemployment, the active labour market policy is the main weapon in the Swedish government's arsenal. It consists of three sets of interrelated programs which are used to improve the demand for labour, supply of labour, and matching of supply and demand for labour. The public costs of these programs are usually in the range of two to three percent of gross national product per year (Swedish Institute, 1993). The demand-oriented programs primarily consist of wage subsidies, recruitment subsidies, and public relief work, which have been used to provide up to sixty thousand additional jobs per year in occupations, industries, and regions where there are labour surpluses (Swedish Institute, 1993). The supply-oriented programs incorporate vocational training and mobility assistance which are used to increase the availability of qualified workers by more than one hundred thousand people each year in occupations, industries, and regions where there are labour shortages (Swedish Institute, 1993). The matching programs encompass guidance counseling, job placement, aptitude testing, occupational rehabilitation, and job clubs which are used to match more than one million workers per year with jobs that fit individual needs and abilities (Swedish Institute, 1993). These programs are available to all job-seekers as part of an attempt to ensure a universal right to work, even though the active labour market policy caters predominantly to heavy user groups of female, young, or disabled workers who represent the most employment disadvantaged people in the work force (Swedish Institute, 1993).

Low unemployment is difficult to ensure in a dynamic capitalist economy where job opportunities shift as production methods and consumer tastes rapidly change. Nevertheless, the Swedish labour market policy is 'active' in preparing for, and adjusting to, labour market changes. The Labour Market Board (A.M.S.), which oversees the policy, is sufficiently independent of government bureaucratic and financial control to exercise discretion in rapidly dropping obsolete programs and adopting innovative new ones, even when major increases in expenditures are at stake (Rothstein 1985). Central government direction of labour market activities is proscribed by the constitution, except in narrowly defined circumstances and only then through published, official 'letters' (Rehn, 1984). In fact, the A.M.S. is even encouraged, under the 1966 Labour Market Ordinance, to undertake rapid and flexible measures in promoting the optimal functioning of the labour market and in ensuring the elimination of unemployment whenever or wherever it occurs (Rehn, 1984). This far-reaching trust in the Labour Market Administration has stemmed from its corporatist governance structure in which all major industrial relations actors previously participated in an executive capacity. However, the employer and union federations have assumed a more limited role since 1992 through their membership on the A.M.S. Advisory Council (Trehorning, 1993).

The Active Labour Market Policy

The preconditions for the active labour market policy were established by the 1938 Saltsjobaden Agreement between the Swedish Trade Union Confederation, the L.O., and the Swedish Employers' Confederation, the S.A.F. This accord signified a turning point to a tumultuous relationship between labour and capital that had been marred by forty years of severe industrial strife. Employers had previously relied on their superior economic power to dominate unions through the use of mass lockouts. However, the electoral victories of the social democrats (S.A.P.) in 1932, 1934, 1936, and 1938 provided the labour movement with the political power to counter employer actions (Korpi and Shalev, 1979; Korpi, 1983). Employers then realized that they were unlikely to defeat unions through labour market pressure, and so settled on a compromise as a way of preserving the power they still had (Korpi and Shalev, 1979; Korpi 1983). Both parties thus agreed to recognize and cooperate with each other in promoting income growth as an alternative to fighting over income shares (Korpi, 1983). Business declared its nominal political neutrality with respect to the distributive policies of the welfare state in exchange for policies that favored capital accumulation and development in the private sector. The L.O. consented to these arrangements in the belief that greater industrial concentration, following naturally from increased economic development, would facilitate the government's intentions to plan the economy and nationalize private businesses (Korpi, 1983). S.A.P. and L.O. Leaders also regarded business acquiescence as indispensable to the rapid introduction of policies for redistributing income and achieving full-employment (Korpi, 1983).

The active labour market policy began as a modest unemployment reduction program in 1947. It was established on the recommendation of the 1944 Myrdal Commission that employment services, vocational counseling, and incentives for geographical mobility were needed to successfully redeploy labour from wartime to civilian industries (Jangenas, 1985). A Labour Market Administration was established with policies conceived by the Labour Market Board (A.M.S.) and executed by several hundred local employment services (Rothstein, 1985). A full-employment mandate was assigned, but only as a complement to the Keynesian fiscal and monetary measures used to achieve full-employment. However, this mandate greatly expanded in the 1950s when the active labour market policy was given the pre-eminent roles of achieving both full employment and accelerated economic growth in accordance with the spirit of the Saltsjobaden Agreement.

The trigger for change came with the failure of the social democratic (S.A.P.) government's Keynesian fiscal policy to generate low unemployment and inflation in the late 1940s. In 1951, economists Gosta Rehn and Rudolf Meidner of the Swedish Trade Union Confederation (L.O.) responded by devising the 'Swedish model' to simultaneously achieve full-employment, moderate inflation, decreased wage inequality, and increased economic growth. Their comprehensive plan thus promised to fulfill the main aims of the Saltsjobaden Agreement. The socialists were assured full-employment and reduced income inequality, while the capitalists were assured a vibrant private-sector economy with the capacity to fund the socialists' welfare state. However, the S.A.P. government originally rejected this proposal because of the unpopular taxes required to finance it and because their Agrarian Party coalition partner feared that any abandonment of Keynesianism would raise the interests costs of heavily-indebted farmers (Rehn 1984). Instead, the government chose to continue with a "a high level of overall effective demand to ensure full-employment", but "in combination with a vast system of regulations (price, trade, and investment controls) and more or less formalized agreements with the trade unions on anti-inflationary restraint" (Rehn 1984: 5). These arrangements broke down when employers, in their desperate efforts to overcome a chronic labour shortage, offered wage increases in excess of those demanded by the L.O.. At this point, the government realized that continuing its policies would encourage members to desert their unions in order to obtain higher wage increases, thereby damaging the S.A.P.'s primary source of electoral support and financing (Rehn 1984). For this reason, the 'Swedish model' was eventually endorsed in 1955 by both wings of the labour movement, the S.A.P. and the L.O., and implemented in a coordinated effort after 1956.

The two arms of the labour movement divided their responsibilities for achieving the different goals established by the Saltsjobaden Agreement. The S.A.P. government agreed to achieve low inflation and unemployment through its active labour market policy. The L.O. unions agreed to achieve decreased income inequality and increased economic growth through a wage solidarity policy negotiated centrally with the Swedish Employers' Confederation, the S.A.F. The L.O.'s wage solidarity policy provided equal pay for equal work in the same jobs across all industries in addition to fair and reasonable differentials for different jobs within each industry (Filcher 1991). These measures reinforced other welfare state efforts to narrow income disparities. Moreover, the equalization of wages for similar jobs across all sectors raised the labour costs of inefficient firms and lowered those of efficient firms, thereby forcing the former to rationalize or close and permitting the latter to retain more of their profits for reinvestment in new machines, equipment, and facilities (Jangenas 1985). As a result, the policy received strong support from the S.A.F., which was dominated by large, efficient engineering firms (Filcher 1991). However, the accelerated pace of structural change displaced many workers from their jobs in traditional industries, even though it produced more rapid economic growth.

The active labour market policy rectified the unemployment effects of the wage solidarity policy by hastening the transfer of people from traditional to advanced industries (Rehn 1985). It therefore fulfilled a crucial supporting role in making faster economic growth and smaller income differences acceptable to the union members who were forced to relinquish their jobs. Moreover, the active labour market policy obviated the need for wage increases to attract new recruits to shortage occupations and industries, thereby dampening any tendency toward wage and hence price inflation.

Recent Developments

The following is taken from, Nils Elvander, A New Regime For Collective Bargaining and Conflict Resolution. 
Paper presented at the International Industrial Relations Association 6th European Congress, Oslo, June 25-29, 2001.  

Click to download the paper.

In March 1997 the foundation was laid of the most important innovation of the rule system for bargaining and conflict resolution on the Swedish labour market since the Basic Agreement of 1938 between the peak organizations of capital and labour, the Swedish Employers' Confederation (SAF) and the Swedish Trade Union Confederation (LO), better known as the Saltsjobaden Agreement. This was done through the ”Agreement on industrial development and wage formation”, signed by all the employer associations belonging to SAF and all the national unions within the industrial sector. The Industrial Agreement (IA) is unique in its broad extent: it covers by and large the whole competition-exposed sector in the Swedish economy, it bridges old class distinctions on the trade union side between blue collar and white collar workers, and it presents an entirely new model for collective bargaining and conflict resolution.

The new model aims at avoiding industrial action as long as possible. This is done by the introduction of a kind of private mediation institute ("impartial chairmen" appointed by a joint steering committee). The rules are to be used in place of the legal stipulations on mediation. So far, the IA has turned out to be a great success. It has been strengthened in daily work through intensified co-operation between the parties on a number of important issues, such as energy and tax policies. In a very successful way it went through the ordeal of the first industry-branch negotiations according to the new rules around the turn of the year 1997-98. The next, and much more arduous, ordeal of this kind comes with the bargaining round of 2001. Finally, the IA has become a model for the rule systems on other parts of the labour market in two ways. On the one hand, it has to a large extent inspired the government bill on a new and stronger mediation institute which was accepted by parliament in spring 2000. On the other hand, it has recently been followed by similar negotiating agreements within the whole public sector. In these ways, a new Swedish labour market regime is emerging; more than half the Swedish labour market is covered by the new functions of  ”private mediation”, and the new public mediation institute is adapted to this development.

The negotiating agreement’s point of departure is that new industry-wide agreements within the different industrial branches shall be reached before current agreements have expired. This is a great improvement compared with previous protracted bargaining procedures. One month before the day of expiration the parties shall be supported by an impartial chairman, who shall intervene in the negotiations on his own initiative and take measures necessary to bring negotiations to an end in due time. An impartial chairman may order the parties to analyze or define special bargaining issues, put forward his own proposals for resolving a negotiating issue, and postpone industrial action for a maximum of two weeks after the day of notice. Obviously, the powers of an impartial chairman are much more extensive than those of an ordinary mediator. The IA has in fact introduced a kind of strong private mediation institute in the industrial sector: "The tasks and powers of an impartial chairman under the terms of this agreement apply throughout the duration of the negotiations and set aside the provisions of the Mediation Act."

The Early 2000s
New collective bargaining procedures.  On 14 April 2000, trade unions and employers in the various parts of local government - country councils, local authorities and parishes - concluded an agreement on new procedures in future pay bargaining rounds, effective from 1 May 2000. The parties state that the agreement will contribute to preserving the national economic balance in future pay agreements without any party having to resort to strikes or lock-outs.

 The new agreement contains practical regulations on timetables, mediation and industrial action during collective bargaining. Negotiations should start three months before an agreement expires. The mutual aim should be to finish the negotiations before the previous agreement expires. In the event of industrial action, notice shall be given to the other party and to a mediator appointed beforehand by both parties. The mediator's role will include a duty to postpone notified industrial actions for up to 14 days. From 1 June 2000, the Co-Determination Act (medbestämmandelagen) has been amended to create a new Mediation Authority to mediate in labour conflicts and work towards satisfactory pay determination (SE9912110F). By registering agreements on bargaining procedures - such as the new local government agreement - the bargaining parties in a sector may follow their own agreed rules instead of the statutory rules on mediation. It is believed that the parties themselves are best placed to agree how bargaining should be conducted.

Read the full story.

Figures on industrial action in Sweden in 2002, published in January 2003, confirm that labor conflicts are at a low level in historical terms. Changes in legislation, new bargaining procedures and economic factors seem to be the main reasons for the decline in action.

Read the full story.

Unionization trends.  Statistics published in March 2003 indicate differing membership trends for the three main Swedish trade confederations in 2002, continuing the pattern of recent years. The largest confederation, the blue-collar LO, lost 1.4% of its members, while the white-collar TCO grew by 1.2% and the professional SACO by 4.5%.

Read the full story.

In 2003, the membership of Sweden's blue-collar LO trade union confederation fell, continuing the trend of recent years. It lost about 1.4% of its members, with only the Commercial Employees' Union, which organizes in the services sector, registering an increase. SACO, which groups professional workers' unions, grew by 3.5% in 2002, while the membership of the white-collar TCO confederation remained stable.

Read the full report (2004).
Read the full record (2005).

The European Union Working Time Directive implemented in the Swedish Working Time Act.  The Swedish Parliament (Riksdag) approved 16 February 2005 the Social Democrat Governmental proposal of making changes in the Swedish Working Time Act in order to implement the European Union Working Time Directive (93/104/EC) in a 'clearer way' . The new rules contain, in short, that the average weekly working time should be at the most 48 working hours. The night/day rest should be at least 11 continuous hours. Some exceptions will be allowed for public tasks, for example in the police force and the armed forces.

Read the full report.