Kingdom of the Netherlands—Netherlands Selected Issues
Author: International Monetary Fund
Publication Date:
08 Aug 2003
eISBN:
9781451829457
Language:
English
Keywords:
ISCR; CR; wage; distribution data; wage dispersion; growth differential; industry wage differentiation; cost competition; Wage bargaining; Wage adjustments; Labor unions
Download PDF (656.4 KB)

The Selected Issues paper of the Netherlands provides an overview of the Dutch pension system. It examines the strengths and weaknesses of the existing bargaining model, compares the Dutch wage bargaining in different periods, examines the past macroeconomic performance of the Netherlands, addresses the issues of wage dispersion and differentiation, and discusses future prospects and challenges for the bargaining model. It also estimates the potential growth in the Netherlands, using statistical detrending methods.

Abstract

The Selected Issues paper of the Netherlands provides an overview of the Dutch pension system. It examines the strengths and weaknesses of the existing bargaining model, compares the Dutch wage bargaining in different periods, examines the past macroeconomic performance of the Netherlands, addresses the issues of wage dispersion and differentiation, and discusses future prospects and challenges for the bargaining model. It also estimates the potential growth in the Netherlands, using statistical detrending methods.

II. Wage Bargaining in the Netherlands18

A. Introduction

43. The Dutch wage formation system is being put to the test once again. During previous economic downturns in the early 1980s and 1990s, the collective bargaining model reacted vigorously, if slowly, through economy-wide multi-year wage moderation. In the late 1990s, rapidly rising wages undermined external competitiveness and the subsequent economic slowdown increased unemployment. The question arises whether the history of appropriate wage responses to these challenges will be repeated.

44. The current juncture also seems a suitable moment for a broader reassessment of the existing bargaining framework. In the past decade, wage bargaining has become somewhat more decentralized and differentiated, but it is not clear to what extent this has improved the functioning of the system. Also, as political differences have widened and economic growth has come to a halt, tensions have mounted. Although the positioning of the bargaining parties should, of course, be evaluated in the context of the negotiation process, the tone of the discussion has sharpened remarkably. In light of a diminishing and graying trade union membership, the employers organization suggested last year that it would favor firms bargaining directly with their works councils, rather than negotiating with the trade unions. Trade unions, for their part, have threatened not to cooperate in renewed multi-year wage moderation out of dissatisfaction with employer attitudes and government fiscal policies. Finally, the minister for social affairs recently reminded social partners that, if they fail to act responsibly, the government has the power to intervene in the wage formation process and could impose a wage cap on the private sector (a measure that has not been applied since the 1970s).

45. This chapter examines the strengths and weaknesses of the existing bargaining model. Section B first provides a brief overview of Dutch wage bargaining in the post WWII period; section C describes the structure of the current wage bargaining framework; section D examines the past macro-economic performance of the Netherlands in light of wage flexibility; and section E addresses the issues of wage dispersion and differentiation; section F concludes by discussing future prospects and challenges for the bargaining model.

B. A Brief History of Wage Bargaining in the Netherlands

46. During the “rebuilding” years after the second World War, wage formation was highly centralized and subject to officially imposed wage guidelines, which were determined by the government after consultation with the social partners. Recognizing the importance of external competitiveness for the small and open Dutch economy, wage moderation was broadly accepted as leading principle. The policy was highly successful in fostering competitiveness and generating growth and employment.

47. The success of the post-war guided wage moderation, however, planted the seeds of its eventual failure. By the early 1960s, as a result of sustained moderation, wages had become very low in comparison with those in neighboring countries. In addition, the labor market had become very tight, with the unemployment rate falling to less than one percent. Under these circumstances, compliance with the formal wage norm proved increasingly difficult to monitor as employers resorted to “black wages” and wage drift increased (Visser and Hemerijck, 1997). As a result, by the mid-sixties, the guided wage system had effectively collapsed, and was replaced by a “free” collective bargaining system.

48. The 1970s saw a lack of consensus, polarized labor relations, and steep wage increases fueled by widespread indexation of wages to prices. Although ritual attempts at central coordination were made every year, social partners generally failed to reach agreement and, in the course of the decade, the government had to intervene several times by imposing wage caps to limit excessive wage increases.

49. By the end of the decade, the economy was characterized by high wage increases, an expanded social security system, low corporate profitability, increasing unemployment, and high public expenditure. In 1982, a new government took office and embarked on drastic measures to address the unemployment and fiscal problems. Against this somber background, and under the treat of government intervention, wage coordination was finally restored in the “Wassenaar” agreement.

50. The 1982 Wassenaar agreement between the employers and the trade unions entailed a return to multi-year wage moderation, based on the shared conviction that a restoration of competitiveness and corporate profitability was a precondition for economic recovery and employment growth. Although this basic principle echoed the policies of the 1950s, the Wassenaar agreement brought two important innovations. First, it was essentially a form of self regulation by the social partners, intended to pre-empt the need for government intervention. It therefore strengthened the role of the social partners in the wage coordination process. The government’s main role was to formally extend wage agreements and support wage moderation through the lowering of the collective burden of taxes and social security contributions, thus softening the effects of wage moderation on net household income. As a second major innovation, the central agreement contained only a broad recommendation with regard to wage moderation: the ultimate wage negotiations were left to the sectoral level. The framework thus combined national coordination with flexibility to meet specific sectoral needs. Initially, this flexibility was regarded by some as a major weakness of the agreement, but it turned out to work well, as the central recommendations proved to have a large impact on the sectoral negotiations.

51. The bargaining framework and the resolve to moderation of wages that were brought about in 1982 broadly remained in effect throughout the 1980s and, after a reconfirmation in the 1993 “A New Course” agreement, also in the 1990s.

C. Structure of the Bargaining Framework

52. The current bargaining framework is broadly the heritage of the Wassenaar agreement. Bargaining predominantly takes place at the industry level, although wage agreements at the company level have been gaining popularity. Nevertheless, there tends to be a considerable degree of coordination at the national level, which sometimes takes the form of formal pacts between employers an union confederations. Traditionally, such pacts do not pin down actual numbers for wage growth, but rather formulate a broad recommendation towards wage moderation, to be interpreted at the sector level. The coordination not only concerns wages but also other policy items that affect workers and social security beneficiaries, including government policies. For example, past negotiations have included such items as working time reduction, training and education, savings tax breaks and the reform of the national disability insurance scheme. Sometimes the national agreements have a multi-year focus (e.g. the 1982 agreement), but not always. Most recently, in October 2002, the social partners and the government agreed on a pact that encompassed only 2003, as broader agreement proved beyond reach. Unusually, this agreement set an explicit ceiling for wage increases (at 2.5 percent).

53. Industrial relations in the Netherlands are characterized by a relatively high degree of organization among employers (covering firms accounting for nearly 80 percent of workers). Trade union membership, on the other hand, is low (less than 25 percent of workers) and declining (Figure II-1). In addition, union membership is graying as the inflow of new members is small. The public sector is substantially more unionized than the market sector. Dutch unions tend to represent not only workers, but also the unemployed and recipients of social security benefits (accounting for about one fifth of total membership). Despite low union density, collective wage agreements apply to over 80 percent of workers, mainly as a result of wide employer participation in the collective bargaining. High employer participation in central bargaining is fostered by the contract extension policies of the government: wage agreements reached by the sectoral representatives of employers and trade unions are usually made binding on non-organized employers and employees in the same industry. This practice provides employers with a strong incentive to engage in central bargaining to ensure their interests are reflected. In the event, administrative extension of collective agreements actually applies to only 9% of workers, mostly in small firms with fewer than 10 employees.

Figure II-1.
Figure II-1.

Union Membership

(Percent of employees)

Citation: IMF Staff Country Reports 2003, 240; 10.5089/9781451829457.002.A002

Source: CBS

54. The basic structure of the Dutch model has similarities to those elsewhere in Europe. The degree of employer and employee organization differs considerably among European countries, with trade union coverage being relatively high in the Nordic countries, and relatively low in France, Germany, the Netherlands and Spain. Nonetheless, the coverage of collective wage agreements is similar, typically above 70% (Table II-1). In addition, European countries generally take the industry level as the primary bargaining locus, and combine this with some form of inter-industry coordination. The degree of coordination varies, however, with Germany and Austria showing particularly high coordination. A notable outlier in Europe is the United Kingdom, where bargaining mostly takes place on the company level, and where coordination is largely absent. In this respect, the United Kingdom more closely resembles the decentralized bargaining structures of the United States and Canada.

Table II-1.

Collective Bargaining Characteristics

(In percent, 1994)

article image
Source: Boeri et al. (2001) and OECD (1997)

D. Wage Flexibility and Macroeconomic Performance

55. Like those in many other European countries, the Dutch wage bargaining system is usually regarded as being roughly in the undesirable middle of the so called Calmfors-Driffill curve.19 Based on a closed economy framework, Calmfors and Driffill (CD, 1988) postulated a “hump shaped” relationship between the bargaining structure and the real wage level. When wage bargaining is very decentralized (at the company level), so the argument goes, wage levels will be contained through the pressure of product market competition. At the other end of the spectrum, fully centralized bargaining will also keep wages in check, because in this case the adverse macro economic effects of wage increases (e.g. the effects on inflation and unemployment) will be internalized in the bargaining process. Bargaining at the intermediate, industry level, however, is not disciplined by market competition - as competitors will typically be included in the same wage settlement—nor does it internalize the macro effects. Consequently, bargaining at the industry level is expected to result in an undesirable equilibrium of high wages and high unemployment. Despite its intuitive appeal, however, empirical cross-county evidence for the CD hypothesis has never been very robust (Flanagan, 1999). Moreover, in recent years, favorable evidence seems to be disappearing altogether. Using a relatively up-to-date data set, including the first half of the 1990s, the OECD (1997) had difficulty finding any relationship between bargaining structure and economic performance.

56. Indeed, the predictions of the CD hypothesis are not borne out by the Dutch experience over the past two decades. The excessive real labor cost growth during the 1970s, was corrected during the 1980s in a remarkable manner (Figure II-2). Under the regime brought about by the Wassenaar agreement, real contractual wages even declined in some years and were roughly flat over the full decade.20 By 1990, most of the earlier imbalance in wages seems to have been corrected, competitiveness largely restored, and the labor income share of GDP brought down to about 81 percent (from a peak of over 95 percent in the first years 1980s). The unemployment rate also declined substantially, to about 7 percent. During the 1990s the real labor costs - wage costs including incidental pay components, labor related taxes, and employer social security contributions - roughly mirrored productivity growth.

Figure II-2.
Figure II-2.

Productivity vs Real Wage Growth

(1970-100)

Citation: IMF Staff Country Reports 2003, 240; 10.5089/9781451829457.002.A002

Source: CPB

57. Apart from the goodwill of the social partners to live up to the Wassenaar agreement, other factors may have contributed to wage moderation. Since the Netherlands is a very open economy, bargaining at the industry level does not eliminate the disciplinary effect of product market competition. Indeed, external competitiveness is generally recognized as a key variable in the Dutch wage bargaining process. Also, an earlier IMF study concluded, on the basis of wage equation estimates made by the IMF and by the Netherlands Ministry of Economic Affairs, that wage moderation in the 1980s seems to have been triggered by underlying fundamentals (productivity growth, changes in taxes, and unemployment), rather than an exogenous shift in labor union attitudes (Watson et al., 1999). This conclusion suggests that the bargaining structure may not be decisive for economic outcomes.

58. In light of the substantial wage adjustment during the 1980s, the Netherlands performs relatively well in cross-country studies on wage flexibility, which is usually defined as the responsiveness of wages to changes in the unemployment rate. Recent wage equation estimates by the Nederlandsche Bank show that, over the period 1975–2001, changes in the unemployment rate had a larger impact on the wage rate in the Netherlands than in France, Germany, or the United States (Peeters and Den Reijer, 2002). This result broadly confirms the earlier findings of Layard, Nickell and Jackman (1991), which used a sample that did not include the 1990s (Table II-2).

Table II-2.

Elasticity of Wages With Respect to Unemployment

article image

59. On the other hand, the Dutch wage formation process has often been criticized for a slow speed of wage adjustment. For example, in the early 1980s unemployment had been rising for three years before the social partners finally reached their formal agreement on wage moderation. Although wage growth had already slowed somewhat before the agreement, sizable imbalances had build up between wages and productivity growth. A rising labor income share and increasing unit labor costs in the early 1990s, as well as during the current downturn, again suggest a somewhat sluggish reaction in wages when the economy switches from boom to bust (Figure II-3).

Figure II-3.
Figure II-3.

Labor Income Share and Unit Labor Cost 1970-2002

Citation: IMF Staff Country Reports 2003, 240; 10.5089/9781451829457.002.A002

Source: CPB

60. Some lags in the reaction of wages may be inevitable, as perceptions of productivity growth and underlying inflationary pressures change only gradually. Nonetheless, the structure of the Dutch wage formation does not seem to minimize these delays. In particular, performance related remuneration elements (such as profit-sharing, bonuses, and stock-options), that would allow wages to reflect changes in economic growth more directly and automatically, play a minor role in the Netherlands. Although individual remuneration elements have gained importance, less than 4 percent of the total wage bill is currently directly related to company performance. The only fiscal arrangement intended to stimulate profit sharing (the “winstdelingsregeling”) was abolished in early 2003. Sluggishness may also be partly related to the fact that wage contracts are often negotiated for two years at the time. Shorter contract periods would increase flexibility, although this gain would have to be weighted against the increased negotiating cost of more frequent bargaining.

E. Wage Dispersion and Differentiation

61. Empirically, centralized or coordinated wage bargaining systems are associated with relatively little wage dispersion (Flanagan, 1999). The Netherlands is no exception. Table II-3 shows three measures of wage dispersion. The first two, from a study by Teulings and Hartog (1998), give the variance of overall individual wages (vertical dispersion) and of wages by industry (horizontal dispersion), based on data from the early 1990s. The third measure is the ratio of the ninetieth percentile of the distribution of gross earnings to the tenth percentile, based on the latest available OECD data for each country. It must be noted that these data are rather rough and unsatisfactory measures of wage dispersion. However, invariably, the decentralized bargaining systems of the Anglo-Saxon countries display the highest dispersion, while the Nordic countries have the lowest. The Netherlands holds a position in the middle.

Table II-3.

Wage Dispersion

article image
Source: Teulings and Hartog (1998); OECD

62. This relatively low wage dispersion may reflect general Dutch attitudes towards what is considered a “fair” distribution of income. In the Netherlands, the view is widely held that cultural factors limit the scope for income differences. In this context, both trade unions and politicians have long criticized the increasingly high remuneration of top managers, which is regarded as undermining the favorable climate for wage moderation and industrial peace. Such preferences aside, however, low wage dispersion may signal a problem, as a lack of possibilities for wage differentiation may tend to distort price signals and could result in misallocation of labor. In particular, there seems to be a risk that labor shortages in specific industries or professions lead to higher wages across the board, thus unduly raising the general wage level and hampering adjustment in the tighter sectors. The resulting inefficiencies can have a negative effect on economic performance. Earlier IMF research has shown that low wage differentiation is associated with lower long term output growth (Thomas, 2002).

63. Contributing to the relatively narrow wage structures in many European countries, including the Netherlands, is the policy of formal extension of wage agreements. Central to the Dutch wage formation framework, since 1937, is the so called “algemeenverbindendverklaring” (AVV), by which wage agreements between employers and unions are more or less automatically extended to the unorganized firms and workers in the same industry. The AVV is a cornerstone of the collective bargaining structure, as it provides both employers and employees with clear incentives to organize and participate in the negotiations at the industry level. The AVV grants labor unions a more central role than might be justified on the basis of their membership alone; although it may also be a cause of low membership, as workers can free ride, benefiting from union representation without having to pay membership fees.

64. The AVV has clear disadvantages. In particular, it establishes minimum or uniform standards for wage developments by industry, thus limiting intra-industry wage differentiation and wage cost competition. Also, in practice, contracts are extended almost regardless of possible undesirable macro-effects. Therefore, some have criticized the AW and suggested its scope be narrowed. For instance, the government could apply more discretion in granting AVV extending only those terms of agreements that are regarded as beneficial to the economy at large. In this case, agreements that entail unduly large wage increases or contain elements that run counter to (the spirit of) government policies - e.g. the topping-up of disability benefits - would not be extended.

65. But the practice also has certain advantages. In particular, it promotes industrial peace - the amount of days lost in strikes is very low in the Netherlands (Figure II-4). The AVV also enables industries to include non wage items (such as training programs and promotional activities on behalf of the industry) in their bargaining, as contract extension guarantees that those agreeing to such terms do not face a cost penalty relative to others in the sector. Since the bargaining concerns a broader package of items, the emphasis on wage increases may be reduced, perhaps moderating wages, if not necessarily overall labor cost. The AVV thus enjoys broad support from the main players in Dutch wage bargaining. In its previous “advisory document on social economic policy”, the SER, a main advisory body to the government (and including the social partners), concluded that there was no reason for a fundamental reassessment of the AVV.

Figure II-4.
Figure II-4.

Days Lost to Strikes: Average 1993-2001

(per 1000 workers)

Citation: IMF Staff Country Reports 2003, 240; 10.5089/9781451829457.002.A002

Source: Eurostat

66. In any event, collective wage agreements have increasingly allowed for more differentiation between and within industries, as well as between employees. Since 1975, the number of different collective industry wage agreements has increased from 180 to 220 and the number of company wage agreements from 450 to 800. In recent years, an increasing number of industry wage agreements contains dispensation clauses that allow for some divergence from the central contract at the individual company level. Also, the role of the works councils was strengthened, facilitating more bargaining at the company level. Finally, individual remuneration items have risen, although they remain relatively small.

67. It is not clear, however, that these developments have actually resulted in more wage dispersion. In a study comparing inter-industry differentials in the period 1973–1982 with those in 1983–1995, the Netherlands Bureau for Economic Policy Analysis (CPB) found that, over the long run, wage growth differentials have been remarkably stable (Van der Wiel, 1998). In fact, the overall differential was even slightly smaller in 1983–1995 than earlier. With regard to individual wage differences, another CPB study shows that dispersion increased in the 1980s, but then broadly stabilized in the period 1993–1998 (Stegeman and Waaijers, 2001). This study also shows that increases in individual dispersion had been mainly due to developments in the tails of the income distribution, whereas the middle ground - which is typically covered by collective agreements - had been broadly stable. As regards recent developments, the OECD gross earnings distribution data have shown a slight increase of dispersion from 2.8 in 1995 to 2.9 in 1999. All in all, the available evidence points to only some marginal increases in differentiation.

F. Concluding Remarks

68. Looking forward, the main question in the short term seems to be whether employers, trade unions and government will once again agree on multi-year wage moderation. The pressure to deliver such a pact is huge in light of the current economic slump, weakened competitiveness, rapidly rising unemployment, and the problems with regard to the funding of occupational pensions. The broad recognition of the issues by all the parties concerned, the marked sense of shared responsibility, and the past response to similar circumstances, all bode well. Last year’s agreement to moderate wage growth in 2003 already signaled a continued tendency towards this time-tested remedy.

69. Admittedly, the hard-won partial agreement and the ongoing discord afterwards also revealed that differences have widened, and that further alignment may prove harder than before. In part, this likely reflects the difficult economic situation and the associated painful policy choices to be made. An additional underlying issue at play is the decline of trade union membership, which may be forcing the unions to build a stronger profile towards their (potential) constituency, resulting in tougher bargaining positions. However, too tough a stance also carries a risk of being sidelined, as that would increase the incentives for employers to seek for alternative bargaining partners. The position of the trade unions thus seems difficult at the moment, and it is as yet unclear where it will settle.

70. The wage formation system also faces medium term challenges that go beyond wage moderation. In the past two decades, the existing framework of industry bargaining with substantial central coordination has usefully facilitated the large adjustment that was needed to redress the imbalances of the early 1980s. Moreover, it did so in a climate of remarkable industrial peace. In those respects, wage coordination has served the Netherlands well. However, the framework also has its weaknesses. When economic circumstances change, wages have tended to be sluggish to react. In addition, the relatively high level of aggregation of bargaining probably hampers differentiation and wage dispersion. Now that the structural labor market surplus of the 1980s and early 1990s has been absorbed and the Dutch economy seems to have landed in a fundamentally more benign equilibrium, there would appear to be scope to address such weaknesses.

References

  • Boeri, T., A. Brugiavini and L. Calmfors (eds) 2001, “The Role of Trade Unions in the Twenty-First Century” Oxford: Oxford University Press.

    • Search Google Scholar
    • Export Citation

  • Calmfors, L. and J. Driffill, 1988, “Bargaining Structure, Corporation and Macroeconomic Performance,” Economic Policy, April, pp. 14–61

    • Search Google Scholar
    • Export Citation

  • Flanagan, R.J., 1999, “Macroeconomic Performance and Collective Bargaining: an International Perspective,” Journal of Economic Literature, September, pp. 1150–1175.

    • Search Google Scholar
    • Export Citation

  • Layard, R., S. Nickell and R. Jackman, 1991, “Unemployment” Oxford: Oxford University Press.

  • Lever, M.H.C. and W.A. Marquering, 1996, “Union Coverage and Sectoral Wages: Evidence from the Netherlands” Empirical Economics, 21, pp. 483–499.

    • Search Google Scholar
    • Export Citation

  • OECD (1997), Employment Outlook, July.

  • Peeters, H.M.M. and A.H.J, den Reijer, 2002, “On wage formation, wage development and unemployment flexibility: a comparison between European countries and the United States” De Nederlandsche Bank Research Memorandum WO No. 712.

    • Search Google Scholar
    • Export Citation

  • Stegeman, H. and R. Waaijers, 2001, “Wage Inequality in the Netherlands” in CPB Report No. 01/1.

  • Teulings, C. and J. Hartog, 1998, “Corporatism or Competition?” Cambridge: Cambridge University Press.

  • Thomas, A., 2002, “The Costs and Benefits of Various Wage Bargaining Structures: An Empirical Exploration” IMF Working Paper 02/71.

  • Visser, J. and A. Hemerijck, 1997, “A Dutch Miracle” Amsterdam: Amsterdam University Press.

  • Watson, C.M., B.B. Bakker, J.K. Martijn and I. Halikias, 1999, “The Netherlands: Transforming a Market Economy” IMF occasional paper.

    • Search Google Scholar
    • Export Citation

  • Wiel, H. van der, 1998, “Inter-industry wage differentials in the Netherlands” in CPB Report No. 98/4.

18

Prepared by David Hofman.

19

See Boeri et al. (2001), Teulings and Hartog (1998), and OECD (1997) for useful overviews of centralization rankings.

20

Consistent with the trade unions’ support for the policy of moderation in the 1980s, Lever and Marquering (1996) find for the period 1971–1990, that the upward effect of union coverage on wages is relatively small in the Netherlands.

Same Series

Other IMF Content

Other Publishers

Cover IMF Staff Country Reports
Kingdom of the Netherlands—Netherlands Selected Issues
Author: International Monetary Fund
Volume/Issue:
Volume 2003: Issue 240
Publisher:
International Monetary Fund
ISBN:
9781451829457
ISSN:
1934-7685
Pages:
44
DOI:
https://doi.org/10.5089/9781451829457.002

Union Membership

(Percent of employees)

Productivity vs Real Wage Growth

(1970-100)

Labor Income Share and Unit Labor Cost 1970-2002

Days Lost to Strikes: Average 1993-2001

(per 1000 workers)