Incomes and Labor Market Policies in Sweden, 1945–70

In reconciling the policy objectives of price stability, on the one hand, and full employment and economic growth, on the other, Sweden has employed a governmental incomes policy only in the period immediately after World War II. At that time the authorities attempted to achieve price stability through direct control of prices, maintenance of low and stable interest rates, and exhortation to the trade unions to exercise restraint in their wage demands. Since this strategy was not complemented by a demand management policy that was adequately restrictive, it failed to ensure domestic and external stability. Consequently, there was a deliberate shift in the policy mix from direct controls to the active use of monetary and fiscal policies during the 1950s. Increased reliance was placed on labor market policies1 in the late 1950s and beyond, but these have been combined recently with measures to stimulate industrial development in depressed areas. In the housing, agriculture, and public service sectors, the system of price control has been retained.


In reconciling the policy objectives of price stability, on the one hand, and full employment and economic growth, on the other, Sweden has employed a governmental incomes policy only in the period immediately after World War II. At that time the authorities attempted to achieve price stability through direct control of prices, maintenance of low and stable interest rates, and exhortation to the trade unions to exercise restraint in their wage demands. Since this strategy was not complemented by a demand management policy that was adequately restrictive, it failed to ensure domestic and external stability. Consequently, there was a deliberate shift in the policy mix from direct controls to the active use of monetary and fiscal policies during the 1950s. Increased reliance was placed on labor market policies1 in the late 1950s and beyond, but these have been combined recently with measures to stimulate industrial development in depressed areas. In the housing, agriculture, and public service sectors, the system of price control has been retained.

In reconciling the policy objectives of price stability, on the one hand, and full employment and economic growth, on the other, Sweden has employed a governmental incomes policy only in the period immediately after World War II. At that time the authorities attempted to achieve price stability through direct control of prices, maintenance of low and stable interest rates, and exhortation to the trade unions to exercise restraint in their wage demands. Since this strategy was not complemented by a demand management policy that was adequately restrictive, it failed to ensure domestic and external stability. Consequently, there was a deliberate shift in the policy mix from direct controls to the active use of monetary and fiscal policies during the 1950s. Increased reliance was placed on labor market policies1 in the late 1950s and beyond, but these have been combined recently with measures to stimulate industrial development in depressed areas. In the housing, agriculture, and public service sectors, the system of price control has been retained.

I. Swedish Policy Experiments

The adoption and constant modification of this “new” mix of general fiscal-monetary policies and energetic policies for the retraining and redeployment of labor were promoted by the continuity of a government that has been open-minded to experiments in economic and social policy. The policy mix, particularly the adoption of active labor market policies, has to a large extent been inspired by economists in the trade unions.2

Despite the challenge of some academic economists,3 the Swedish Government in the 1950s and 1960s did not attempt to interfere with wage bargaining because of opposition from the labor market organizations. This standpoint has been defended both by the Government and by the labor market organizations, on the grounds that the bargaining process, on the whole, has managed to preserve peaceful labor relations and that union leaders could not guarantee sufficient support among members for a policy of wage restraint in the event of governmental interference. The Government has thus not intervened directly in the bargaining process.

The wage policies of the unions have been guided by the principle of wage solidarity. In particular, this principle has aimed at obtaining uniform average increases in wages throughout industry without regard to low productivity and stagnating sectors, a policy which has been facilitated by highly centralized wage bargaining. Combined with more or less efficient demand management policies and international competition, which limited the rise in international prices during most of the period after World War II, this wage policy of the unions played an important role in closing down inefficient enterprises and production lines and in promoting the desired restructuring of the Swedish economy.

The success of the restructuring process has to no small degree been dependent on the adoption of active labor market policies, which have allowed the skill content and geographic distribution of the labor force to adjust to the requirements of a changing industrial structure. Under these policies, generous grants are made for the training and movement of labor. Originally designed to assist anticyclical monetary and fiscal policy, labor market policies have since become an instrument of structural policy, used independently of cyclical conditions. By aiding the expansion of efficient firms with above-average productivity growth, this instrument has helped to foster economic growth. Considering this and the fact that those in retraining and public relief works are not viewed as unemployed persons in Sweden, labor market policy has become an important instrument in the Government’s efforts to reconcile price stability with rapid growth and full employment.

On the whole, the mix of stabilization policies in Sweden has been more successful in preventing recessions than in combating inflationary tendencies, except for certain periods (particularly 1955–63) when stabilization policies were quite successful. In assessing the success of the Government’s stabilization efforts, it should not be overlooked that economic policy in Sweden has been characterized by a continuous determination to attain full employment and full social welfare. The upward drift of prices following World War II met with little opposition from the public, except in years with notably accelerated price increases. This study examines the debate surrounding the adoption of the “new” policy mix in the 1950s and early 1960s, as well as the institutional framework which has permitted some success in reconciling the objective of price stability with the aims of sustained full employment and economic growth in the absence of a governmental incomes policy.

direct controls and voluntary wage restraint after world war ii

In the five years immediately after World War II (1945–50), there was intensive debate in Sweden about the appropriate set of economic policies to be adopted in the face of an expected world-wide depression, and there was also some experimentation with direct and indirect control over economic activity. Following the recommendations made under several postwar programs for economic policy, the Government in 1945 and 1946 sought to sustain a high level of demand to maintain full employment. It pursued expansionary fiscal and monetary policies, and it abolished or liberalized several wartime controls in 1945 and 1946—rationing, import controls, and building controls. Controls over prices and capital movements were retained, and efforts were made to attain “voluntary” wage restraint by the labor unions. Price policies were designed on the assumption that, during a period of direct price control, prices were determined more by cost than by demand factors. The price policies at that time were aimed at stabilizing or reducing prices in reaction to increased productivity of enterprises and by increasing the supply of goods from abroad at cheaper prices. Various measures were introduced to this end, including a reduction in interest rates (February 1945), an appreciation of the krona by 17 per cent (July 1946), the abolition of the general sales tax (July 1947), and liberalization of imports. The Government exhorted the labor unions to show restraint in their wage demands in return for reasonable price stability.

Although these policies kept unemployment at a low level (averaging 1.5 to 2 per cent in 1946 and 1947), they did not succeed in holding down prices and wages nor in maintaining external equilibrium. The failure was attributed to substantial excess demand, which developed as a result of the expansive fiscal and monetary policies and the liberalization of direct controls on investment. Thus, in spite of price control, consumer prices rose by 8 per cent in the three years to mid-1948, and despite government pleas for wage restraint, wages in industry (including fringe benefits) rose by over 20 per cent in the two years 1946–47.4 The current account balance shifted fom a surplus of SKr 1.3 billion in 1945 to a deficit of SKr 1.5 billion in 1947 under the influence of excess demand, currency revaluation, and import liberalization.

The Swedish Government reacted to these developments by reintroducing import controls and certain forms of rationing, but monetary and fiscal policy remained rather passive. Building controls were tightened and were applied to discriminate against investment in housing and building in the services sector in favor of the export- and import-competing sectors. In addition, the Government was successful in achieving a wage freeze in 1949 and 1950 by persuading the unions to accept a prolongation of the previous wage agreement in return for a promise to hold down price increases by price control, subsidies, and low interest rates. As shown in Table 1, prices rose very little in 1949 and 1950, and there was only a rather moderate increase in wage rates in industry in these years. The current account balance changed from deficit to surplus, amounting to SKr 460 million in 1949 and SKr 160 million in 1950. However, the success of these policies was of only short duration. In 1951 and 1952, very strong inflationary forces coming from abroad gained the upper hand. With an increase of as much as 16 per cent in consumer prices in 1951, the system of direct controls and passive demand management policies could not ensure price stability—the major precondition for continued wage stability. Dissatisfaction among the unions with this price performance and the uneven wage drift during the wage freeze in 1949 and 1950 resulted in a sharp upsurge in wage rates in industry of some 20 per cent a year in 1951 and 1952.

Table 1.

Sweden: Hourly Wage Costs, Prices, and Profit Margins, 1947–70

(Annual percentage change)

article image
Sources: National Institute of Economic Research, The Swedish Economy, various issues; Monthly Digest of Swedish Statistics; and for 1947–50, Assar Lindbeck, Svensk economisk politik (Stockholm, 1968), p. 200.

For 1947–54, adult industrial workers only, including fringe benefits; for 1955–70, the total economy, excluding fringe benefits.

Ratio between gross profits and receipts from sales.

Even though the upsurge of prices in 1951 was partly due to external factors, such as the Korean boom, the Government became disillusioned with the policy of repressive controls on prices, investments, and imports.5 This disillusionment, stimulated by criticism from academic and labor union economists,6 was not only fostered by the failure of the policies of direct price controls but also by the adverse effects on resource allocation and economic growth created by an inflationary atmosphere and by the direct controls themselves.

transition to demand management and labor market policies in the 1950s and 1960s

Proposals by academic economists

The economic problems encountered in the late 1940s and early 1950s led to several proposals on the mix of policy instruments needed to reconcile price stability and full employment. Most of the proposals did not challenge the principle of noninterference by government in the wage negotiation process, a principle to which both the centralized labor and employer organizations were strongly attached. Except for that put forward by Hansen (see below), the proposals did not advocate a policy mix embracing reinstitution of an explicit incomes policy.

A widely discussed proposal was made by Rehn.7 As a method of reconciling price stability and full employment, the Rehn model advocated a combination of restrictive general demand management policies and active labor market policy, designed to alleviate unemployment in inefficient sectors by shifting labor to more efficient sectors through retraining and compensation for the geographic movement of labor. One important aspect of the Rehn model was the profit squeeze to be brought about by fiscal action to curb demand and by increases in wage costs. This was considered necessary to remove the disturbing provocation for unions to seek large contractual wage increases and to lessen demand-induced wage drift. Another important concept of the Rehn model was the “wage solidarity principle,” under which unions should not seek differential wage increases according to differences in profits or productivity among economic sectors or branches but should demand uniform average wage increases for all members. Rehn considered that the pursuit of this principle—apart from its social attractiveness to the unions—was necessary to avoid the general wage inflation that would result from the attempts of wage earners with lower than average wage increases to restore their lost wage relativity. The shift of labor from the inefficient to the efficient sectors of the economy without the inducement of a wage differential between these sectors would be satisfied through labor market policies.

The main problem with the Rehn proposal stemmed from the possibility of conflict between stabilization and allocation in his model.8 Doubts were also raised as to whether a lower average level of profits would be sufficient to moderate wage increases if unemployment was kept at a low level. The wage solidarity principle was criticized on the grounds that it would make it impossible for efficient firms with temporary low profits to survive. Despite the criticisms, the Rehn model was included in the economic policy program of the Swedish Federation of Trade Unions in 1951 and the report of a government ad hoc committee on stabilization policy in 1961. From the late 1950s, official policies tended to follow Rehn’s proposals, although they could not remove entirely the problems of excess demand for labor and of wage drift.

Hansen’s proposal9 for solving the conflict between stabilization and full employment was not translated into official policy, but his ideas for the conduct of government incomes policy cannot be overlooked. He suggested that the Government should announce in advance of wage negotiations the desirable wage increase on the basis of expected productivity increases. Should the negotiated wage increase exceed the announced rise, the Government would increase direct taxes so that real disposable income would be lower than it would have been if the announced rise had been accepted. If actual wage rises nevertheless exceeded productivity increases in the agreement period (through higher wage drift), Hansen proposed to use a combination of reduced indirect taxes (to eliminate the inflationary effect on the cost side of undesirably large wage rises) and higher direct personal taxes (to lower demand) to the disadvantage of wage earners.

The reorientation of policies

Replacement of direct controls by general demand management policies. Following the Korean boom, there was a gradual shift from direct controls to flexible general demand management policies. Import controls were liberalized between 1952 and 1954, and licensing control of building was discontinued between 1954 and 1959, except for control by the labor market authorities of the timing of building starts designed to counteract the seasonal fluctuations in building activity. Price control was gradually abolished between 1954 and 1956. From 1955 onward, the authorities favored price competition, supported by new antimonopoly legislation and by continuous scrutiny of developments in prices and profit margins in the business sector. However, as already noted, controls of rents and agricultural prices have been retained.10

During the 1950s, the authorities activated both fiscal and monetary policies for anticyclical purposes. There has been considerable experimentation with a wide array of policy instruments in this field, but a detailed analysis of the instruments and their effectiveness would go beyond the purview of this paper. However, some behavioral characteristics of stabilization policies seem to be relevant.

In the field of fiscal policy, reliance has been placed on both tax policy and, since the later 1950s, on anticyclical variations in public investment and housing investment—in addition to variations in investment from the so-called investment fund reserve scheme (see below, The development and structure of labor market policies). An important characteristic of public expenditure policy has been strong expansion during periods with recessionary tendencies (1958–59, 1962–63, and 1966–67), whereas in boom periods the expansionary effect of public spending policy was merely reduced and almost never replaced by a contractionary effect.11 This reflects a greater priority in preventing recessions rather than in checking inflationary tendencies and is consistent with the greater importance attached to full employment than to stabilization of prices.12

Such priorities have been helped by the absence of a conflict between domestic and external policy in the period 1948–67. On the other hand, tax policy has become increasingly restrictive since the 1950s. The total marginal tax rate (including indirect and corporate taxes, as well as social security contributions), with respect to changes in gross national product (GNP), rose from nearly 40 per cent in the mid-1950s to about 50 per cent toward the end of the 1960s—a record level by international standards. This resulted mainly from: (1) a movement by an increasing proportion of taxpayers to higher and more progressive income tax scales; (2) contributions to the General Pension Fund, introduced in 1960 and levied on employers in proportion to their wage bill; (3) some increase in the rate for basic pension contributions of income earners; and (4) increasing reliance on indirect taxation, as is reflected by the reintroduction of the sales tax in 1960 and by increases in indirect tax rates in subsequent boom years. The increase in the tax burden has not, however, prevented substantial deficits in the central government accounts in the five years to 1970.

In contrast to the immediate postwar period, tax and stabilization policies have generally been guided since the 1950s more by considerations of demand management and resource allocation than by considerations of the development of costs and wages. The reliance on direct taxes and the gradual increase in their rates in the 1960s reflected this. Because more pressure could be exerted on inefficient firms and on the growth of private consumption through higher indirect rather than direct taxation, indirect taxes were preferred. There was no reduction in the sales tax or other indirect taxes when price stability was endangered by greater cost rather than by demand pressures in postboom years (for example, in 1962 and 1966). The return to reasonable price stability appears to have been endangered under the considerable price pressure prevailing at the time of wage negotiations (for example, in 1966).

While the policy of increased direct and indirect taxation in the 1960s achieved a stabilizing effect on the economy in the short run, it appears that the policy induced an extra push on wages and (through the cost structure) on prices in the somewhat longer run. This has been facilitated by a free bargaining process that does not inhibit the labor market organizations from retaliating in their wage negotiation proceedings against increased taxation and its effects on prices. Such attempts have been made in the past. The unions of relatively high-paid Swedish academic professionals and civil servants and some union groups of the Central Organization of Salaried Employees (TCO), accounting together for nearly 20 per cent of the labor force in 1969, are known to have bargained with reference to income after taxes.13 On the other hand, the Swedish Federation of Trade Unions (LO), in attempting to improve the income position of its members in relation to other groups, is known to have bargained on the basis of wages before taxes. However, since their wage demands have been influenced by price increases, they have also retaliated against indirect tax increases, insofar as these influence prices.

The development and structure of labor market policies. Since 1958, general stabilization policies have been supplemented to an increasing degree by active labor market policies. This instrument now forms an integral element of overall economic policy as reflected in legislation passed in 1966, when the purposes and application of labor market policies were fully accepted.

The introduction of labor market policy was occasioned by recessionary tendencies in 1958/59 and did not at first constitute a deliberate step toward the application of the Rehn model. In that period, the authorities used selective measures for the first time on a large scale to counteract unemployment in sectors and areas having particularly weak demand. With this step, the authorities abandoned to some extent the conventional alternative of providing a general (that is, undifferentiated) stimulus to demand as a method to counter unemployment.14 A further step toward supplementing general stabilization policy was taken by the authorities when they retained the labor market measures after the recession of 1958/59 and broadened and improved them in line with the Rehn model. The labor market measures facilitated structural adjustment of the economy by promoting geographical and vocational mobility of labor, by helping to procure job opportunities through relief work and regional development programs, and by encouraging the entry of women into the labor force.

It was necessary to broaden the range of labor market instruments during the 1960s for several other reasons. Excess demand for labor in certain sectors and areas during most of the 1960s, particularly the metropolitan areas in southern Sweden, along with considerable unemployment in other areas (especially in northern Sweden where it rose as a result of the rationalization of the forestry industry) were only two of these reasons. Another reason was that the total labor supply grew only moderately as the result of a low rate of natural growth in the labor force (not over 0.5 per cent a year) and of immigration to Sweden. Rapidly growing industrial branches therefore had to rely on labor released from the stagnating and declining sectors of the rest of the economy. The structural element of labor market policy—alone a stabilizing factor in the longer run—gained increasing importance in the 1960s, a fact underlined by the adjustment measures in the field of retraining which were taken on a considerable scale even in periods of boom. Public expenditures under a labor market policy that enhances the efficiency of the economy have been considered preferable from the standpoint of taxpayers and consumers to outlays on subsidies and unemployment relief in the traditional sense.15 The development of labor market policy has also been motivated by welfare purposes, inasmuch as it helped to enhance freedom in the choice of occupation and working place.16

Labor market policy is formulated and executed by a labor market administration working directly with the Government and comprising the National Labor Market Board with 25 regional agencies (the county labor boards) and the local employment offices and agents.17 The National Board consists of five government representatives (including the Director-General), four union representatives, and two representatives from the employers’ organizations. Broadly defined as employment-creating measures, the activities of the labor market authorities are far-reaching.18 They are not only concerned with labor market measures in a narrow sense, such as the provision of relief work and the promotion of labor mobility and retraining programs, but also with the coordination of labor market policy and specific demand management measures, such as the release of investment reserve funds 19 and the acceleration of public investment programs or government purchasing from industry. Total appropriations in the government budget for manpower and regional policies were increased over the years from SKr 1 billion in fiscal 1965/66 to SKr 2 billion in 1969/70, raising the share in total appropriations from 3.5 per cent to 5 per cent, equivalent to 1.5 per cent of GNP. The number of persons affected by employment-creating measures in a broad sense rose strongly during the 1960s, and by November 1970 (a boom year) they amounted to an estimated 78,500 persons, or 2 per cent of the labor force.20 The most important activities of the labor market authorities relate to relief work, retraining, and regional development.

Under the relief work program, employment in public works is offered to persons unable to find employment elsewhere. The program is phased with seasonal and cyclical variations in employment and is also coordinated with public investment in depressed areas. During the 1960s, government expenditure for relief work has shown a fair degree of anticyclical variation (Table 2). The number of persons affected has also varied countercyclically (Table 3), but there has been a rising trend in such “relief” employment as a result of the growing number of older people with a labor market handicap and a worsening structural unemployment problem in northern Sweden.

Table 2.

Sweden: Principal Labor Market Expenditures, 1963–70

(Expenditures in millions of Swedish kronor)

article image
Source: Central Bureau of Statistics, Statistisk Årsbok för Sverige, 1964–71.

As percentage of average of gross national product for the two calendar years, at 1959 prices.

As percentage of total government expenditure before lending.

Included in expenditure on relief work.

Table 3.

Sweden: Effect of Labor Market Measures on Unemployment, 1959–70

article image
Sources: Central Bureau of Statistics, Statistisk Årsbok för Sverige; National Budgets of the Swedish Economy; and Monthly Digest of Swedish Statistics.

Average for February, May, August, and November, according to labor force surveys.

Includes handicapped persons in relief work.

However, the public relief work program has not eliminated labor redundancy in depressed areas and has not greatly contributed to an improvement in the productivity of the economy.21 Partly for these reasons, increased emphasis has been placed on policies of labor adaptation, the most important element being the retraining programs. Government expenditure for retraining and the average number of persons engaged in it have grown steadily during the 1960s (Tables 2 and 3). As could be expected, however, the rate of increase has diminished in boom periods and accelerated in recession periods during most of the 1960s. The retraining programs are mainly designed for the metal, engineering and building industries, but they are also offered to married women in order to facilitate their re-entry into the commercial sector. The average length of retraining courses in 1970 was about six months.

Another element of labor adaptation is promotion of the geographical mobility of labor. Certain measures in this field (travel and removal grants) have had a long tradition in Sweden but have been greatly improved since the latter part of the 1950s, particularly to bring about the transfer of unemployed persons from depressed areas to areas of labor shortage. The types of allowances used to stimulate labor mobility are listed in Table 4. The number of persons benefiting from such allowances has risen at a rate similar to that of the retraining programs.

Table 4.

Sweden: Stimulation of Labor Mobility, 1965–70

article image
Source: National Labor Market Board.

A person may have been granted more than one type of allowance.

The outflow of labor from the less industrialized areas in northern Sweden—aggravated by active labor market policies—led to sharp criticism of labor market policy in that part of the country. In 1962–64 the Government therefore started to place more emphasis on regional development, with measures involving preferential government loans and subsidies to firms willing to invest in northern Sweden.22 In recent years, the investment reserve fund scheme has been used for the same purpose—a feature which has tended to complicate its operation as a tool of demand management.

In spite of the impressive policy efforts to obtain a better regional balance in the labor market, excess demand for labor in industrial areas and a high level of unemployment in northern Sweden persisted throughout the 1960s. Except for the political resistance to the movement of labor from northern Sweden, labor mobility seems to have been hampered until recently by housing shortages in pressure areas.

the wage negotiation process

Wage bargaining

The Swedish Federation of Trade Unions (LO), the most influential union organization,23 had a membership at the end of the 1960s of over 40 per cent of the labor force (Table 5), exceeding by a large margin the combined membership of the other unions, which comprise mainly salaried employees. By 1970, the unions for salaried employees represented one fourth of the labor force because of their rapid unionization since the 1930s. The Swedish Employers’ Confederation (SAF),23 the dominant organization on the employers’ side is the counterpart of LO in the bargaining process, although a considerable number of employees (particularly salaried employees) engaged by members of SAF are not formally covered by the frame agreements between SAF and LO. The powers that SAF can exert and the benefits it can confer on its members—through the right to impose fines and the mutual assistance fund in cases of strikes or lockouts—enable it to present a unified front against the union organizations.

Table 5.

Sweden: Organization of the Labor Force, 1969–70

article image
Sources: Central Bureau of Statistics, Statistisk Årsbok för Sverige; and Organization for Economic Cooperation and Development, Economic Survey of Sweden (Paris, April 1970).

December 31, 1970.

December 31, 1969.

May 1, 1970.

October 1, 1970.

One of the main principles guiding the wage policies of the two major negotiators (LO and SAF) is the absence of government intervention in the wage bargaining process. This precept was enshrined in the 1936 Basic Agreement between LO and SAF, which established machinery for the parties to settle conflicts between them and to protect third parties from harmful effects of labor disputes. The Basic Agreement still underlies wage settlement procedures today. The desire not to provoke government interference has induced LO and SAF to observe an established code of conduct and to reach agreements wherever possible in a harmonious manner.24 As an outgrowth of this behavior, labor conflicts have been fewer and shorter in Sweden than in many other industrial countries since World War II. In addition, the labor market organizations have generally attempted to avoid wage agreements which would have seriously endangered the national interest, especially when they might have threatened the competitiveness of exports. This has not, however, prevented the organizations from occasionally arriving at contractual wage increases in excess of what the authorities considered appropriate for price stability. The Government has made its views known in public statements or economic surveys rather than by direct communication with the labor market negotiators; these have had some impact on the wage policies of LO and SAF.

One important objective of the wage policy of LO is wage solidarity. This principle seeks to eliminate “unjustified” wage differences between job categories and branches of industry by adhering to equal pay for equal work, without regard to the profit situation in the individual industrial enterprises and branches. Although the principle implies the elimination of marginal firms, it has been accepted by SAF on the grounds that it maintains and increases the efficiency of the economy. The implementation of the principle led to a greater need for job evaluation in determining the appropriate wage structure and to the adoption of active labor market policies to bring about a redistribution of labor between the stagnant and growing sectors of the economy. Equal pay for equal work in its more conventional sense between males and females had already occurred on a considerable scale before LO and SAF agreed in 1960 to omit from collective agreements any reference to the sex of wage or salary earners.

The procedures for wage negotiation between LO and SAF are basically simple.25 Through collective bargaining and with due recognition of the present and prospective economic situation, both parties determine the major lines of wage policy at the national level for the recommended period of the wage agreement (one or two years).26

During the period from September of one year to March of the following year, when much of the bargaining for the next agreement takes place, the Wages Conference of the LO meets and reviews the general economic situation. Unofficial consultations take place in September or October between SAF, LO, the economic ministries, and other trade unions. The publication of a semiofficial survey of economic developments, both past and future, by the National Institute for Economic Research (Konjunkturinstitutet) in October or November provides an independent analysis for all parties—government, labor, and employers—of the overall capacity of the economy to absorb wage rises. The analysis is supplemented with others carried out by LO and SAF economists and by discussion in the Parliament (Riksdag) on the economic situation.

After the LO Wages Conference and after members of SAF have directed their executives on respective negotiating positions, negotiation between the bargaining delegations begins in December or January, frequently under the chairmanship of an employer’s representative. The parties take into account a further report of the National Institute for Economic Research and the budget speech of the Minister of Finance (particularly his examination of the requirements for economic stability and his views on the scope for wage increases compatible with price stability).

In the absence of disagreement (see Provisions for settlement of deadlocks, below), LO and SAF arrive at a frame agreement in February or March. The frame agreement does not acquire legal force until the date when its provisions are incorporated into the individual collective agreements, which are negotiated between national trade unions and employers’ associations concurrently with the central negotiations.27 This two-level approach to collective bargaining in Sweden began in 1956. In practice, the national trade unions usually follow the directions laid down in the frame agreement, because: (1) the composition of the negotiating teams at the central level assures that the frame agreement does not contain terms unacceptable to them; (2) the frame agreement provides for “exceptions” applicable to certain industries where wages have lagged behind the general development; and (3) LO has the means to ensure discipline. The same rationale applies to SAF membership. In addition, the LO-SAF frame agreement finds wide acceptance beyond the sphere of LO and SAF as it stipulates that LO and SAF shall follow the same lines (as laid down in their agreement) in negotiations with labor market organizations which are not members of LO and SAF, respectively. Frequently, the frame agreement serves as a model in negotiations between labor market organizations other than LO and SAF. For example, price agreements between the Government and agricultural producers have in the past been made with reference to changes in the contractual industrial wage rates.

Provisions for settlement of deadlocks

In accordance with the policy of government noninterference in wage bargaining, official mediators are not brought in until late in an intractable negotiation to settle a deadlock. Before the mediators are called, the parties themselves may appoint an impartial arbitrator. It is only when he is unable to settle the differences that a district or special mediator28 can be appointed by the National Labor Market Board under provision of law adopted in 1906. If the dispute is serious, the Government may appoint a conciliation commission. Should these procedures fail, the parties can resort to direct action (either strikes or lockouts, subject to conditions laid down by law) or submit to arbitration by the National Labor Market Board (as allowed for by the 1936 Basic Agreement), or the Government can adopt special legislation, if the situation warrants it.

In the 1950s and the first part of the 1960s, the labor market parties 29 generally had less and less recourse to the mediation system. This, together with the fact that industrial unrest has been kept at a low level (Table 6), indicates that the negotiation procedure has been successful in minimizing the friction inherent in the collective bargaining system.

Table 6.

Sweden: Industrial Unrest, 1946–70

article image
Source: Central Bureau of Statistics, Statistisk Årsbok för Sverige.

Annual average of these years.

II. Some Conclusions on the Swedish Policy Mix

Swedish economic policy, characterized since the late 1950s as a mixture of stabilization and structural policy, has to some extent realized the proposals laid down in the Rehn model. Tax policy has been continually tightened in an effort to hold down the growth of domestic demand and to reallocate resources from the private to the public sector. The wage policy of the labor unions has been guided by the concept of solidarity within a system of free, collective wage bargaining. In the first half of the 1960s there was a profit squeeze, the deterioration being most pronounced for enterprises exposed to foreign competition.30 Furthermore, there was considerable acceleration of structural change during the 1960s, measured by the increased rate of shutdowns of firms and production lines. Finally, the active labor market policies since the late 1950s assisted structural adjustment and helped to maintain unemployment (in the traditional sense) at a low level (Table 3).

However, general demand management policies have not been used since the late 1950s to bring about a sufficient reduction in the excess demand for labor. This may be attributed to the higher priority accorded by the authorities to full employment than to price stability. Labor market policies in the form of retraining and resettlement programs appear to have been hampered until recently by the housing shortage in the industrial area in southern Sweden, where labor mobility has been reduced and excess demand for building workers has continued throughout the 1960s.

Excess demand in the overall labor market during the 1960s contributed to an increase in wage rates considerably in excess of the overall growth in productivity. The wage increases originated in high wage drift, which developed in spite of the far-reaching influence of the labor market organizations (Table 7).31 Wage drift has made the wage policy of the labor organizations difficult to implement, for it has worked against attempts by LO to raise relative wages of low-paid workers and against attempts of the unions of salaried employees to keep pace with wage developments of industrial workers. Wage drift occurred on a considerable scale in the building trade during most of the 1960s. In industry it reached sizable proportions in the boom years 1964/65 and 1969/70, when price stability was endangered more than at any other time since the Korean boom, partly because an upsurge of export prices raised the scope for wage increases in the export industries and partly because the expansionary effect on demand of public sector operations, initiated in preceding recession years, was reduced too late and too little. The wage settlements concluded in an atmosphere of rapidly rising prices in the postboom year 1966 provided for considerable rate increases.

Table 7.

Sweden: Contractual Wages and Wage Drift in Building, Industry, and Total Economy, 1955–70

(Annual percentage change)

article image
Sources: National Institute of Economic Research, The Swedish Economy; Monthly Digest of Swedish Statistics; and Sten Borg, “Is There a ‘Critical’ Level of Unemployment?—A Case Study of Sweden,” in On Incomes Policy, Papers and Proceedings from a Conference in Honour of Erik Lundberg (Stockholm, 1969), p. 271.

Excluding fringe benefits. The addition of fringe benefits would add between ¼ and 1 per cent to hourly wage earnings annually, depending on the year in question.

Estimated figure.

The increase in wage rates has exceeded the substantial growth in industrial productivity during most of the 1960s and particularly in 1970. From the early 1960s until 1967, the increase in unit labor costs (including fringe benefits) in industry was greater than that in Sweden’s competitor countries (Table 8). Thereafter, the development of unit costs was slightly more favorable in Sweden than in its major competitor countries. The sharp wage increases in Sweden have probably influenced the development of productivity within the industrial sector. While undoubtedly helping to force inefficient firms out of the market, wage rises may have also stimulated the efforts by other firms, particularly in the export sector, to increase their efficiency.32 However, with the accelerated closing down of import-competing firms and of enterprises producing for the home market during the 1960s and with a liberalized trade regime, the propensity to import has shown an increase—a development which underlines the need for greater cost stability than in the period 1948–67, when there was no conflict between the objectives of external and internal policy.

Table 8.

Sweden: Comparison of Productivity and Costs in Manufacturing With Prices in Sweden and Abroad, 1962–69

(Annual percentage change)

article image
Sources: Organization for Economic Cooperation and Development, Economic Survey of Sweden (Paris, April 1970); and “World Summary of Reported Balance of Payments Statistics, 1963–68” (unpublished, International Monetary Fund, August 21, 1970).

Based on national currency.

Based on U.S. dollar.

Annual average of these years.

In judging the success of the Swedish policy mix, it should not be overlooked that the task of containing the upward pressure on prices in conditions of full employment has been hampered by a combination of factors that are partly of structural origin and outside the control of demand management—for example, the differences in productivity gains between the sectors exposed to international competition and the “sheltered” sectors of the economy, as well as occasionally (as in 1964/65 and 1969/70) a relatively rapid rise in export prices in a system in which unions persist in following the principle of solidarity. The “competitive” sector (accounting for about 30 per cent of total employment in 1967) covers branches with above-average productivity growth, while the sheltered sector comprises branches with below-average productivity growth. The competitive sector has been the wage leader in Sweden,33 and through the union policy of wage solidarity it has transmitted to the sheltered sector increases in wages in excess of the latter’s productivity gains. This situation has resulted in price increases in the sheltered sector far in excess of those in the competitive sector, where prices are determined by international competition. A study 34 by economists in the labor market organizations has revealed that the average rate of price increase in the sheltered sector over the period 1960–67 exceeded its average increase in productivity by 2.5 percentage points a year, compared with an excess of average productivity growth over the average price rise in the competitive sector of 6.5 percentage points (Table 9).

Table 9.

Sweden: Comparison of Sector Product and Employment in 1960 and 1967 With Price and Productivity Movements for 1960–67 in the Sheltered and Competitive Sectors

(In per cent)

article image
Source: Gösta Edgren, Karl-Olof Faxén, and Clas-Erik Odhner, “Wages, Growth and the Distribution of Income,” Swedish Journal of Economics, Vol. 71 (September 1969), p. 143.

Includes industries protected from import competition, public and private services, the building sector, and the public sector.

Includes raw material production, intermediate production for home and export consumption, industries facing import competition, and finished goods industries.

In the competitive sector, the annual average rise of 9.4 per cent in wage costs over the period 1960–67 was somewhat in excess of the “room” for wage increases (that is, its productivity growth plus an increase in its foreign prices). This “excessive” wage increase was probably a reflection of the wage pressure stemming from excess demand conditions in the labor market. Given the uniform increase in wage costs, the differential price rise between the competitive and sheltered sectors over this period was explained by their differences in productivity growth. In its position as a wage leader, the competitive sector played a key role in both wage and price developments, although it accounted for a smaller proportion of output and employment than the sheltered sector. An attempt by demand management policy to contain price increases of structural origin (namely, those originating in the differential growth of productivity between economic sectors and uniform average wage increases) might have achieved its objective by reducing the ability of the sheltered sector to pass on increased costs in prices. However, experience in Sweden has shown (for example, in 1958/59) that, to succeed in price restraint, such an attempt would have led to an unacceptable impact on employment and economic growth.35

On the other hand, wide productivity differences between economic sectors and lack of control over wage drift by the labor market organizations would have made it difficult for the authorities to adopt an effective incomes policy.36 The need for stricter wage discipline, particularly better control of wage drift in boom periods, is strongly felt by the whole labor market. The need to arrive at an enduring improvement of the balance of payments position is only one reason. This is reflected in the 1967 proposal by SAF for an improved method of wage bargaining that would be designed to obtain greater stability in wages and prices.37 This proposal was, however, rejected by LO, mainly on the grounds that it would result in a transfer of the problem of income distribution from negotiation between LO and SAF to “objective” estimates by experts who might be influenced by the Government and thus would deprive the unions of their chance to bargain freely.38

APPENDIX Swedish Labor Market Organizations

In addition to its size and the close association of its leadership with the ruling Social Democratic Party, the influence of the Swedish Federation of Trade Unions (LO) is reflected in its comfortable financial position. In 1963, LO and its affiliates had accumulated funds totaling Skr 702 million, of which Skr 202 million were held in liquid form. Certain large unions, affiliated with LO, also have considerable funds of their own—for example, the metal workers’ union (Skr 7.9 million in 1964) and the building trade union (Skr 2.2 million in 1964).39 LO works through a tripartite structure consisting of: (1) a 13-man executive committee (the Secretariat); (2) a Representative Assembly, which comprises the Secretariat and representatives of over 40 affiliated national unions; and (3) a Congress comprising 300 delegates (including the Secretariat and Assembly). The supreme decision-making body is the Congress, convening once every five years. The Representative Assembly is the supreme authority between each Congress and convenes twice a year, partly in order to consider collective bargaining problems. The Secretariat has executive power in all important matters, namely, intra-union disputes, labor conflicts, and wage policies and negotiations.40

On the other hand, the Swedish Employers’ Confederation (SAF), founded in 1902, represents a large section of the manufacturing, transportation, and building industries and, with respect to wage bargaining, acts as the leader for the independent employers’ associations in the private sector 41 and the employers in the public sector.42 Membership of SAF includes members of employers’ associations in different industries. Individual employers or companies are affiliated with SAF through their employers’ associations, in which they are “partners.” All activities of SAF are carried out through the members rather than through the partners. Members are entitled to a voice and vote in SAF, while partners can vote only through the members. The partners can receive compensation from SAF in the case of approved industrial conflicts.

The structure of SAF is tripartite, comprising: (1) an Annual General Meeting as the highest authority, consisting of representatives of the members who have voting power; (2) a General Council, which is also composed of member representatives and whose main task is to fix annual dues and to serve as a body of reference for major lockout decisions; and (3) a Board which has executive powers in all important matters—namely, lockout decisions and wage negotiations. A partner and member can thus call a lockout only after approval by the Board. The powers of SAF over its members are also reflected in its right to impose fines against members or partners not observing its rules, and in its administration of a substantial mutual assistance fund.43

Cooperation between SAF and other employers’ associations is achieved through a joint consultative committee with employers in the commerce, shipping, and agricultural sectors. In practice, this committee does little more than maintain representation on certain public bodies, such as the Labor Court, or exchange information on common problems.


Mr. Brehmer, Chief of the Northern European Division of the European Department, is a graduate of Kiel University, Germany, and has served as a staff member of the Deutsche Bundesbank. He is the author of Struktur und Funktionsweise des Geldmarktes der Bundesrepublik Deutschland Seit 1948 (Tübingen, 1956 and 1964) and of a number of journal articles.

Mr. Bradford, a graduate of the University of Canterbury, Christchurch, New Zealand, was an economist in the Northern European Division on temporary assignment from the New Zealand Treasury when this article was written. He has now returned to the New Zealand Treasury.


Labor market policies mainly comprise programs to promote the geographical mobility and retraining of labor and to generate employment.


See especially, Rudolf Meidner, “The Dilemma of Wages Policy Under Full Employment,” and Gösta Rehn, “The Problem of Stability: An Analysis and Some Policy Proposals,” in Wages Policy Under Full Employment by Erik Lundberg and others, edited and translated by Ralph Turvey (London, 1952).


See, for example, Bent Hansen, The Economic Theory of Fiscal Policy (London, 1958), pp. 353 ff.


See Assar Lindbeck, “Theories and Problems in Swedish Economic Policy in the Post-War Period,” American Economic Review, Vol. 58 (June 1968), pp. 12–14.


See, for example, Gösta Rehn and Erik Lundberg, “Employment and Welfare: Some Swedish Issues,” in Industrial Relations, Vol. 2 (February 1963), pp. 2–3.


See, for example, Meidner and Rehn, op. cit. In his criticism of direct controls, Lundberg focused particularly on the distortion of investment between sectors subject to control and uncontrolled sectors. See Erik Lundberg, Business Cycles and Economic Policy (London, 1957), Chapters 11 and 12.


See, for example, Meidner and Rehn, op. cit.


Lindbeck, “Theories and Problems in Swedish Economic Policy” (cited in footnote 4), pp. 22 ff., and Lundberg, Business Cycles and Economic Policy (cited in footnote 6), pp. 258–61.


See Hansen, op. cit., pp. 353 ff., and Lindbeck, ibid., pp. 24–26.


Rent control is designed to prevent high profits on housing in the interest of consumers. Rent controls have been retained for social welfare reasons. However, the substantial decline in the relative price for housing (compared to prices of other consumer goods) since the 1930s contributed to considerable excess demand for housing until recently, despite a high rate of residential building. The waiting period for apartments remained very long throughout the postwar period, particularly in the high growth areas of southern Sweden. This has formed a considerable obstacle to the policy of promoting geographical mobility of labor. The excess demand conditions in the building market have also sometimes caused a faster rise in the wage rates of building workers than for other groups, particularly in the 1960s.

With continued price control of agricultural products in the 1950s and 1960s, the Swedish authorities retained an instrument to guarantee reasonable incomes to farmers. As an outgrowth of this policy, agricultural support prices have exceeded world market prices by a considerable margin (on average by 70 per cent at the end of the 1960s). Such a policy has not been without consequences for union wage policy. Although the unions have publicly agreed with the agricultural support policies of the Government, they have frequently demanded wage compensation for price increases caused by higher agricultural support prices (see Lindbeck, “Theories and Problems in Swedish Economic Policy” (cited in footnote 4), p. 73). As the income of farmers has nevertheless remained relatively low, agricultural price support has not hampered the outflow of labor from agriculture to other occupations.


See Assar Lindbeck, Svensk economisk politik (Stockholm, 1968), p. 209 (table).


The margin of public tolerance of unemployment appears to be very narrow in Sweden. For example, the 4 per cent rate of unemployment in the winter of 1958/59 was viewed by the unions as a serious failure of stabilization policy. See Rehn and Lundberg, op. cit., p. 2.


See Erik Lundberg, “Incomes Policy in Sweden: Some Issues,” in On Incomes Policy, Papers and Proceedings from a Conference in Honour of Erik Lundberg (Stockholm, 1969), p. 19.


See Rudolf Meidner, “The Goals of Labor Market Policy,” in ibid., p. 190.


See Rehn and Lundberg, op. cit., p. 6.




See Organization for Economic Cooperation and Development, Labor Market Policy in Sweden, Reviews of Manpower and Social Policies (Paris, 1963), pp. 25 and 52.


For details, see ibid., pp. 64–65.


Sweden operates an investment reserve fund scheme which allows firms to set aside sums for future investment at times designated by the Swedish authorities. In setting aside funds, the firm obtains a tax benefit of advance depreciation and, at the time of use, an additional tax allowance. The release of these funds is coordinated with the labor market authorities.


See Organization for Economic Cooperation and Development, Economic Survey of Sweden (Paris, April 1972), p. 27.


See Rehn and Lundberg, op. cit., p. 7.


See Lindbeck, “Theories and Problems in Swedish Economic Policy” (cited in footnote 4), pp. 31 and 65.


See the Appendix for a more detailed description of the labor market organizations.


Jean Mouly, “Wages Policy in Sweden,” International Labour Review, Vol. 95 (March 1967), pp. 174 and 182. The Basic Agreement was concluded in an atmosphere in which the Government had threatened to legislate in the field of labor relations at a time when labor relations were particularly difficult.


For more technical details, see T. L. Johnston, Collective Bargaining in Sweden (Harvard University Press, 1962), pp. 263–75, and Mouly, op. cit., pp. 168–74.


The 1966 agreement was valid for three years, but the provisions relating to wages concerned only the first two years. Most industrial wage agreements expire on January 30 or March 30 each year and are automatically renewed for another year, unless notification is given three months before expiration date.


Provisions for wage increases under the frame agreement are related to actual earnings. For remuneration by piece rates (the most prevalent in Sweden), the agreements foresee an average percentage increase in earnings subject to a minimum increase expressed in ore per hour, while for workers paid on a time basis the average increase is expressed in terms of ore per hour. The reference to average increases in the frame agreement implies that some wages may be raised by a greater and some by a smaller amount.


There are eight mediators officially appointed for one year, representing each of eight labor districts.


See Johnston, op. cit., p. 272.


Finance Ministry, Svensk ekonomi 1971–1975 med utblick mot 1990:1970 års långtidsutredning (1970), pp. 155–56.


Empirical evidence shows that the rate of increase in wages in Sweden is sensitive to the degree of demand pressure in the labor market, particularly at low unemployment rates. See L. Jacobson and Assar Lindbeck, “Labor Market Conditions, Wages, and Inflation—Swedish Experiences, 1955–67,” Swedish Journal of Economics, Vol. 71 (June 1969), pp. 64–103.


See Lindbeck,”Theories and Problems in Swedish Economic Policy” (cited in footnote 4), p. 76. Lindbeck believes that in this case the adoption of an incomes policy in the form of guidelines is complicated, as the productivity increase cannot be taken for granted.


See Gösta Edgren, Karl-Olof Faxén, and Clas-Erik Odhner, “Wages, Growth and the Distribution of Income,” Swedish Journal of Economics, Vol. 71 (September 1969), p. 149.


Edgren and others, op. cit.


Ibid., p. 157.


Ibid., pp. 158–59.


See “The 1967 Proposal for Wage Stabilization,” in On Incomes Policy (cited in footnote 13), pp. 225–31.


Clas-Erik Odhner, “The Government Agency for the Determination of Wages,” in ibid., pp. 235–42.


See Mouly, op. cit., p. 172.


For a thorough description of the structure and legal status of the labor market organizations, see Johnston, op. cit., pp. 23 ff.


This group is estimated to employ about 400,000 persons. It includes the Commercial Employers’ Association (about 113,000 employees in 1967), the Swedish Shipowners’ Federation, and the Newspaper Employers’ Association. See Mouly, op. cit., p. 173.


The counterpart of central government employees in collective bargaining is the Government Collective Bargaining Board. For local government employees, the central bargaining partners are three federations—the Federation of Municipalities (covering about half of local government employees), the Federation of Rural Councils, and the Federation of County Councils, which provide a common bargaining front for local governments.


The insurance fund of SAF amounted to SKr 277 million in 1964. In addition, each partner is obliged to give a formal guarantee, which can be used to cover the cost of any dispute if the fund is not adequate. Such guarantees totaled SKr 340 million in 1964. See Mouly, op. cit., p. 172.