Jump to ContentJump to Main Navigation
Welfare and Work in the Open Economy Volume II: Diverse Responses to Common Challenges in Twelve Countries$

Fritz W. Scharpf and Vivien A. Schmidt

Print publication date: 2000

Print ISBN-13: 9780199240920

Published to Oxford Scholarship Online: November 2003

DOI: 10.1093/0199240922.001.0001

Show Summary Details
Page of

PRINTED FROM OXFORD SCHOLARSHIP ONLINE (www.oxfordscholarship.com). (c) Copyright Oxford University Press, 2019. All Rights Reserved. Under the terms of the licence agreement, an individual user may print out a PDF of a single chapter of a monograph in OSO for personal use (for details see www.oxfordscholarship.com/page/privacy-policy). Subscriber: null; date: 21 May 2019

How Small Countries Negotiate Change Twenty‐Five Years of Policy Adjustment in Austria, the Netherlands, and Belgium

How Small Countries Negotiate Change Twenty‐Five Years of Policy Adjustment in Austria, the Netherlands, and Belgium

Chapter:
(p.175) 5 How Small Countries Negotiate Change Twenty‐Five Years of Policy Adjustment in Austria, the Netherlands, and Belgium
Source:
Welfare and Work in the Open Economy Volume II: Diverse Responses to Common Challenges in Twelve Countries
Author(s):

Anton Hemerijck

Brigitte Unger

Jelle Visser (Contributor Webpage)

Publisher:
Oxford University Press
DOI:10.1093/0199240922.003.0005

Abstract and Keywords

Although Austria, the Netherlands, and Belgium are so seemingly alike in their tightly coupled, consociational, and corporatist democratic structures and in the “Bismarckian” origin of their welfare states, they have had radically different experiences since the 1970s. While the Netherlands, which appeared in the 1970s and early 1980s to be afflicted with a terminal ‘Dutch disease’, has seemingly been cured, Belgium, with a similar initial profile, has been malingering and Austria has managed to avoid the crises from which the others are recovering. Since all three countries have internationally exposed and hence vulnerable economies as well as policymaking structures with plural veto positions, the success or failure of adjustment policies did depend on the ability of actors to adopt action orientations that emphasize common, rather than separate, interests. The Austrian social partners succeeded in maintaining this ‘encompassing’ perspective throughout the period under study; the Dutch had to relearn it after dismal failures; and in Belgium, the increasing salience of linguistic cleavages added to the difficulty of achieving, and acting on, convergent perceptions and interest definitions.

Keywords:   Austria, Belgium, Bismarckian welfare state, consociationalism, corporatism, Dutch disease, language communities, Netherlands, social partners, veto positions

5.1. Introduction: Birds of a Feather?

Austria, Belgium, and the Netherlands seem like ‘birds of a feather’, albeit with different colors (Alber 1998). They are small and open economies, each with a tradition of social partnership, Bismarckian‐type welfare state, and consociational democracy based on political coalitions that bridge cleavages of class, religion, or language. Comparing these ‘most similar cases’ (Przeworski and Teune 1970) can potentially contribute to our understanding of the interaction among economic vulnerability, institutional capacity, policy legacies, and policy responses.

‘Birds of a feather flock together,’ says the Duchess in Alice's Adventures in Wonderland. But in our case, they did not. In response to the economic shocks of the 1970s, Austria, Belgium, and the Netherlands went different ways. Austria managed to keep unemployment low and was able to check inflation, whereas Belgium and the Netherlands accomplished neither. Why? Did policymakers perceive or approach the problems differently? Did they act under different external and internal constraints? Did they pursue different objectives? When the oil shock of 1979 and rising interest rates in the 1980s had to be addressed, imbalances were much more pronounced in Belgium and the Netherlands than in Austria. But crises do trigger learning processes and create political conditions for change. This is most clearly seen in the Dutch case, where reforms in the 1990s were more daring than in Austria. Did the Dutch learn their lesson better, or (p.176) will they be just as vulnerable when another international shock hits? Why was learning of this kind blocked in Belgium, despite highly unsatisfactory labor market outcomes?

Our aim in this chapter is to present description and explanation. In this first section we offer a bird's‐eye view of the similarities and variations among our three countries. In the following three sections we offer a dynamic, fine‐grained account of each country's problems, perceptions, legacies, policies, and outcomes. The chapter ends with an attempt at explaining the diversity in policy responses and an assessment of vulnerabilities and capacities.

5.1.1. Small, Open Economies

Austria, the Netherlands, and Belgium are three small countries bordering on Germany, their main trading partner (see Table 5.1). Belgium and the Netherlands trade heavily with each other, Austria with Switzerland and Italy. As transit economies, with the two largest ports of mainland Europe, Belgium and the Netherlands are two very open European economies, with imports and exports making up 75 percent and more of their gross domestic product. Austria's share is 60 percent, still above average in Europe but not unusual for a small economy. With limited domestic markets, these countries are highly dependent upon access to foreign markets and have always given a great deal of attention to trade liberalization and competitiveness. Not surprisingly, they are advocates of European economic integration. Belgium and the Netherlands were among the six founding nations of the European Economic Community. Austria, a former member of the European Free Trade Association, joined the European Union in 1995. All three qualified for membership in the Economic and Monetary Union in 1999.

5.1.2. Different Industrial Profiles

Small domestic markets encourage open economies to seek competitive advantage through product specialization and economies of scale in export markets. For historical and geographical reasons, our three countries exhibit different patterns of specialization. As continental Europe's oldest industrial nation, Belgium developed a strong position in the production of raw materials, coal and steel, and related industrial products, until decline began in the 1960s. In the past forty years, the economic axis has shifted to light industry and services, with a strong impact on the port of Antwerp and domestic investment by multinational, mainly US, firms. The Netherlands industrialized late and never became a truly industrial (p.177)

Table 5.1 Austria, Belgium, and the Netherlands Compared, 1994

Population (in 1000s)

GDP (in million ECU)

GDP/capita (ECU)

Germany's share of exports (%)

Openness of the economya(%)

Austria

8,030

168,865

21,029

38.0

60.9

Netherlands

15,381

283,920

18,459

26.8

74.6

Belgium

10,116

193,610

19,139

19.4

81.4b

Flanders

5,857

113,740

19,420

Wallonia

3,538

50,538

15,274

Brussels

950

29,332

30,864

(a) Exports plus imports in percentage of GDP. See: OECD (1997) Statistical Compendium (National Accounts) (Paris: OECD, 1997); own calculations.

(b) Including Luxembourg.

Sources: Eurostat, 1997; MPIfG, Adjustment Database, 1999; OECD, Economic Outlook 1999; Belgian Statistical Office.

nation, despite its post‐1945 efforts. The Dutch economy is specialized in transportation and logistics, international finance, business services, agro‐industry, and foreign trade, a tradition that goes back to the seventeenth‐century colonial ventures of the Dutch East‐India Trading Company. The country is home to large indigenous and Anglo‐Dutch multinational firms, like Philips, Unilever, Shell, Heineken, and some of Europe's largest banks. Postwar Austria lagged even further behind, and in 1970, 12.2 percent of its labor force was still employed in agriculture, compared with 3 percent in Belgium and the Netherlands. Today Austria has a larger industrial sector in terms of people employed or value added. Unlike Belgium and the Netherlands, Austria nationalized large parts of industry and the banking sector after 1945. In the 1970s nationalized industry comprised 70 percent of banking, credit, and insurance, 33 percent of industrial manufacturing, 23 percent of food distribution, and 40 percent of housing construction (Kurzer 1993: 38). The private sector used to be dominated by small and medium‐sized firms, and none of the major multinational enterprises had its home base in Austria.

Neither the Netherlands nor Belgium has had a particularly happy or interesting experience with industrial policy. After 1945 both countries undertook large‐scale industrialization programs for the purpose of recovering from war damage (Netherlands) or developing rural areas, especially in Dutch‐speaking Flanders (Belgium). In the 1970s both also became increasingly immersed in hopeless rescue operations for ailing industries. At this time Austria showed fewer signs of strain, and its nationalized (p.178)

                   How Small Countries Negotiate Change Twenty‐Five Years of Policy Adjustment in Austria, the Netherlands, and Belgium

Figure 5.1 Annual Real Growth, 1970–1998

Source: OECD, Economic Outlook (Paris: OECD, various years).

industrial firms maintained high levels of employment. But problems arose in the mid‐1980s, when the nationalized sector was restructured and privatized under both domestic and international pressure.

5.1.3. Growth, Employment, and Unemployment

On a cumulative basis, economic growth hardly differs among the three countries. All three experienced the slowdown in output and productivity growth that has characterized all advanced economies since 1973 (Figure 5.1). Until 1985 Austria did better. From 1986 to 1996 the economy grew at a rate of 21 percent in Belgium, 25 percent in Austria, and 26 percent (the OECD average) in the Netherlands. The number of people employed increased by 4 percent in Belgium, 6 percent in Austria (just above the OECD average), and 18 percent in the Netherlands. Labor force growth was much slower in Austria and Belgium, yet only in the Netherlands did employment growth outpace labor force growth. Hence the Netherlands was the only one of the three countries in which unemployment was lower in 1996 than it had been ten years earlier (Figure 5.2). It should be noted, though, that the Dutch unemployment rate had to come down from very high levels. This predicament was shared with Belgium but had been avoided in Austria, where unemployment never exceeded 4 percent of the labor force until quite recently. 1998 was the first year in two decades during which the Dutch unemployment rate dipped under 4 percent. In 1998 Belgium unemployment fell to 8.8 percent (OECD 1999).

(p.179)

Table 5.2 Output, Employment, Labor Force, and Unemployment, 1970–1996

Countries

Variability of outputa

Employment responsiveness to outputb

Labor force responsiveness to employmentc

Variability of unemployment rates

Austria

1.79

0.47

0.83

0.35

Belgium

1.88

0.45

0.15

1.40

Netherlands

1.55

0.52

0.12

1.02

(a) Standard deviation of the first difference (in log and multiplied by 100) of output.

(b) The estimated coefficient b in the regression. (Employment deviation from trend) = a + b* (output deviation from trend), where the trends have been established using the Hodrick‐Prescott filter, imposing identical smoothing factors for employment and output in each country.

(c) The estimated coefficient b in the regression. (Labor force deviation from trend) = a + b* (employment deviation from trend), where the trends have been established using the Hodrick‐Prescott filter, imposing identical smoothing factors for employment and output in each country.

(d) Standard deviation of the first difference (in log and multiplied by 100) of unemployment.

Source: Pichelmann and Hofer 1999: 34; Belgium: own calculations.

Belgium and the Netherlands experienced very high levels of unemployment while Austria did not (Figure 5.2). How did this relate to the rather small difference in the (accumulated) growth of output? Figure 5.1 shows that the recession years of 1975 and 1993 were greater periods of setback in Belgium than in the Netherlands and that the Dutch fared worse than the Austrians until the mid‐1980s. Yet the variability in Austria's output is not lower (Table 5.2). By means of labor hoarding and labor force withdrawals (foreign workers, women), however, Austria matches variability in output with considerable responsiveness in how the labor force reacts to employment opportunities, so that the overall result is a much lower, and hardly varying, unemployment rate (Pichelmann and Hofer 1999). This adjustment mechanism was not available in the Netherlands and Belgium. Foreign workers had obtained residential rights by 1975, and the Netherlands in particular experienced strong growth in the female labor supply, which was unresponsive to cyclical demand changes (Hartog and Theeuwes 1983). Overall, unemployment rates in Belgium and the Netherlands showed greater variation over the cycle (Table 5.2). Once unemployment is high and long‐term unemployment is allowed to rise, joblessness tends to be persistent (Layard, Nickel, and Jackman 1991; OECD 1994).

How the Netherlands caught up is also shown in the employment/population ratios. Around 1970 the Netherlands, like Belgium, had one of the lowest employment rates in Europe. At the time this was entirely due to the fact that married women were not participating in the workforce. After a further decline lasting through the early 1980s, the Dutch employment ratio leaped from 50 percent in 1990 to 61 percent in 1998, almost level (p.180)

                   How Small Countries Negotiate Change Twenty‐Five Years of Policy Adjustment in Austria, the Netherlands, and Belgium

Figure 5.2 Annual Employment Rates As a Percentage Of the Labor Force, 1970–1997

Source: OECD, Statistical Compendium: Labour Force Statistics (Paris: OECD, 1997); own calculations.

with Austria and five points above the OECD average. From Figure 5.3, based on OECD Economic Outlook data, we learn that over a thirty‐year period the Dutch got back to where they were in 1970, with employment rising after 1984. Austria has been unable to keep up its high level of employment from the 1970s; in the 1990s it experienced a jobs decline. Downward drift and stagnation characterize developments in Belgium, which started from the same position as the Dutch and had almost the same experience until the late 1980s. The difference with the Netherlands is to be found in the 1990s. Behind these trends there are significant shifts in female and male employment, a pattern of withdrawal from the labor market by older males, a unique rise in part‐time employment in the Netherlands, and very different employment levels and trends in domestic private services.

5.1.4. Negotiated Economies

In Small States in World Markets, Katzenstein (1985) classifies our countries under the label of ‘democratic corporatism’. In response to their dependence on exports, the imperative of competitiveness, and their inability to exert control over external events, small and open economies tend to develop tightly‐knit organized networks of consultative bodies at (p.181)

                   How Small Countries Negotiate Change Twenty‐Five Years of Policy Adjustment in Austria, the Netherlands, and Belgium

Figure 5.3 Employment As a Percentage Of the Working Age Population (15–65), 1970–1997

Source: OECD, Statistical Compendium: Economic Outlook (Paris: OECD, 1999); own calculations.

the national level oriented toward advancing economic and social progress and preserving social peace. The Parity Commission and its various subcommittees in Austria, the Central Economic Council and National Labor Council in Belgium, and the Foundation of Labor and Social‐Economic Council in the Netherlands are all postwar inventions. They embody the idea that organized capital and organized labor, together with representatives of the state and central bankers, share responsibility for the welfare of the nation.

In each country the umbrella organizations of employers and unions are involved in consultation, both with each other and the state. Overall,

Table 5.3 Some Characteristics of Labor Relations in the Mid‐1990s

Countries

Union density rate (total and private sector) (%)

Employees in firms joining employers' associationsa(%)

Employees covered by collective agreementsa(%)

Extension of agreements through public law

National minimum wage

Belgium

53/40

80

82

Significant

Statutory

Austria

44/37

96

97

Significant

Agreement

Netherlands

28/20

80

79

Limited

Statutory

(a) Private sector.

Source: Visser 1996.

(p.182) social partnership is more fully developed and stable in Austria (Tálos 1996; Traxler 1997) than in the Netherlands (Hemerijck 1995; Visser and Hemerijck 1997) and Belgium (Van Ruysseveldt and Visser 1996; Vilrokx and Van Leemput 1997). The state is weakest in Belgium, for reasons we will indicate later, while cooperation between unions and employers is most balanced in Austria. Belgian trade unions count among the strongest outside Scandinavia (Ebbinghaus and Visser 1999). Compared to the other two countries, they are more integrated with (and divided among) separate political parties, and they must constantly perform a delicate balancing act between Belgium's two main linguistic communities and regions. From a comparative perspective, Belgium's central organizations of unions and employers are rather weak, with little authority over affiliates and with pronounced regional divisions. Dutch trade unions weakened considerably between the mid‐1970s and the mid‐1980s, whereas Dutch capital and employers' associations remain well organized (Visser 1997). In all three countries collective bargaining is pervasive and primarily organized by economic sector. In each country there is intensive coordination by the umbrella organizations, but in Belgium agreement has been rare. Dutch wage negotiators operate under a ‘shadow of hierarchy’. The opposite trend is found in Belgium. Here the state has taken over the role of wage bargainer and negotiates under a threat of dissent and implementation failure. Table 5.3 captures some features of current labor relations.

5.1.5. Economic Performance

Institutional variation is associated with variation in performance. Austria had much less trouble with inflation in the 1970s than the other two countries. The Dutch got inflation under control more quickly than the Belgians (Figure 5.4). Each of the three countries assigned monetary policy the task of fighting inflation, but Belgium and the Netherlands paid a high price in terms in terms of competitiveness. While the schilling and the gulden were pegged to the Deutschmark in an early stage, Belgium had to accept a major devaluation of the franc in 1982 and waited till 1990 before it tied its hands. The extent of continuity in Austria and policy reversal in Belgium and the Netherlands is well captured by the competitiveness indicator, based on unit labor costs in manufacturing (Figure 5.5).

In the wake of the 1973 oil‐price shock, policymakers in each country tried to combine a hard currency policy with Keynesian demand stimulation. But only in Austria did congenial wage‐setting policies help to avoid the negative outcome of cost‐push inflation, a loss of competitiveness, and higher public debts. Austria (along with Switzerland) was among the few Continental European countries where levels of industrial and societal (p.183)

                   How Small Countries Negotiate Change Twenty‐Five Years of Policy Adjustment in Austria, the Netherlands, and Belgium

Figure 5.4 Annual Percentage Change Of Consumer Prices, 1970–1997

Source: OECD, Economic Outlook (Paris: OECD, various years).

                   How Small Countries Negotiate Change Twenty‐Five Years of Policy Adjustment in Austria, the Netherlands, and Belgium

Figure 5.5 Index Of Unit Labor Costs, In Real Terms, Relative To 1990 (= 100), 1970–1997

Source: OECD, Statistical Compendium: Economic Outlook (Paris: OECD, 1997); own calculations.

(p.184) conflict had remained low in the late 1960s (Crouch and Pizzorno 1978). Lijphart (1968) and Huyse (1970), when publishing their studies on the wonders of consociational democracy in the Netherlands and Belgium, noted that ‘the times were changing’. Steiniger (1975), in his comparative study, found no such changes in Austria.

Austria started the 1970s with a budget surplus and reached a maximum net deficit of just over −5 percent of GDP in 1995. The deficit varied with the business cycle but remained within a −5 percent margin. Belgium, where the net deficit peaked at −12 percent in 1981, shows the most extreme variation and managed only with great difficulty to qualify for the EMU norm of −3 percent in 1998. The Netherlands' public sector deficit peaked in 1982 at −6.6 percent and has declined since, approaching a balanced budget in 1999. The debt–output ratio increased in all three countries until the 1990s. Austria started with a low public debt of 19 percent in 1970, increasing to 69 percent in the 1990s until it recently declined to 64 percent thanks to austerity measures in 1995. Belgium entered the 1970s with a debt–output ratio of 64 percent, rising to a peak of 137 percent in 1993, and declining to 122 percent in 1998 with the help of a special 3 percent EMU surcharge. The Netherlands began with 52 percent, allowed the debt–output ratio to rise to 81 percent in 1993, and then managed a reversal to 71 percent. The debt problem is reflected in debt interest payments, which have been lower in Austria (3.5% of GDP in 1997) than in the Netherlands (4.3%) and Belgium (7.5%). While Austria performs best in terms of the misery index (unemployment plus inflation rate), it runs a slightly negative current account balance. Belgium experienced severe current account deficits between 1977 and 1984. The Netherlands always had strong surpluses, except in 1978 and 1980. The toleration of current account deficits in Austria may reflect the country's overall sound economic position and a monetary policy that is perceived as stable by international creditors.

Austria's remarkable stability also shows up in its functional income distribution. Labor's share in national income has hardly changed during the past four decades. The slow upward drift was reversed in the second half of the 1980s. In the other two countries, labor's share, adjusted for changes in employment structure, rose significantly after the mid‐1960s and decreased in the early 1980s. In Belgium labor's share increased again in the latter half of 1980s and in the 1990s. In the Netherlands a similar upward trend in the late 1980s was again sharply reversed after 1993. Labor's share of Belgian national income is today still above its 1960s average and above the European average, whereas it has fallen under its 1960s level in the Netherlands (Table 5.4). Underneath these trends lie different wage policy experiences.

(p.185)

Table 5.4 Labor's Share in the National Income, 1960–1997

Country

1960–69

1970–79

1980–89

1990–97

Austria

68.5

69.6

71.4

69.9

Belgium

69.6

74.5

73.2

71.5

Netherlands

69.9

74.2

67.6

65.8

EU average

71.3

75.2

73.1

69.9

Source: European Commission, DG II, European Economic Review 4/1998; adjusted for changes in employment structure.

In terms of personal income distribution, Austria is the most unequal of our societies. In 1993 the D5/D1 earnings ratio was 2.00 in Austria, as opposed to 1.43 in Belgium and 1.54 in the Netherlands (OECD 1996: 61–62). In the upper earnings range Austria is the most, and Belgium the least, unequal. Inter‐sectoral earnings differentials in Austria are particularly large on account of inequality between male and female workers (Pichelmann and Hofer 1999). The incidence of low pay, or the percentage of workers who earn less than two‐thirds of the median, is higher in Austria (13.2%) and the Netherlands (11.9%) than in Belgium (7.2%) (OECD 1996: 72). Or should we say that wage compression at the lower end of the labor market is strongest in Belgium? The Belgian national minimum wage tends to be high by international standards. In 1993 medium wages in manufacturing were only 43 percent higher than the minimum wage, which is rather similar to the situation in German manufacturing. In the Netherlands, the distance between medium and minimum wage had risen to 56 percent, in France to 65 percent, and in the UK even to 80 percent (Roorda and Vogels 1997). In Austria minimum wages are contractual, vary by sector, and are much lower in most services than in manufacturing (Dolado et al. 1996). If measured in purchasing power parity, the Belgian minimum wage is estimated to be about 10–20 percent above the Dutch minimum wage, which stood at just over 2,000 guilders, or 1,120 euros in 1999. More Belgian (9%) than Dutch (3%) workers actually earn the minimum wage (Roorda and Vogels 1997).

5.1.6. Continental Welfare States

In the comparative literature on European welfare states, Austria, Belgium and the Netherlands are classified under the regime‐types of Continental, ‘Bismarckian’, conservative, or Christian Democratic welfare states (Esping‐Andersen 1990; Kersbergen 1995). Protection is based on employee insurance against occupational risks, financed by earmarked payroll contributions from employers and workers. Traditionally, trade (p.186) unions and employers' associations are involved in the supervision and administration of social security provision. Programs revolve around income replacement and are targeted at the male breadwinner in order to preserve traditional family patterns. In contrast to the Nordic welfare‐state model, social services are underdeveloped. Instead, the system is highly transfer‐oriented. Active labor market policies are a recent discovery. In terms of resources spent on such policies in 1995, Belgian ranked fifth and the Netherlands sixth among fifteen European countries, after Sweden, Denmark, Finland, and Ireland. If recalculated as percentages of GDP per percentage point of unemployment, Belgium (0.14%) trails the Netherlands (0.21%) but is ahead of Austria (0.09%) (OECD 1997).

A comparison of policy provisions addressing the social risks of unemployment, sickness, disability, poverty, and old age renders a differentiated picture of the three countries' social security profiles. Here we present the state of affairs in the early 1980s (Flora 1986: vol. ii); in Sections 5.2–5.4 we narrate the most significant changes and reforms. Unemployment insurance is compulsory in all three countries and financed by contributions from employers and employees, often with additional state subsidies. In the event of unemployment, dependent workers must register as unemployed and be available for work. Entitlement to benefits requires a minimum record of 26 insurance weeks in the year preceding dismissal. Benefits are payable for 12 to 30 weeks in Austria; 26 weeks in the Netherlands, with an additional, lower but not means‐tested provision for another two years; and without limitation but at a lower benefit level in Belgium. In all three countries insurance benefits are related to the level of wages last earned: the range extends from 58 percent for low‐income workers to 40 percent for high‐income workers in Austria, in the Netherlands the maximum is 80 percent of former earnings (70% after 1987), and in Belgium during the first year of unemployment benefits were fixed at 60 percent, then cut to 40 percent for single persons in the second year, and thereafter to a (non means‐tested) flat rate of half the national minimum wage. Family supplements, up to 80 percent of prior wages for the lowest income groups, are granted on application in Austria. In the Netherlands people who are still unemployed after 2.5 years, have been entitled to means‐tested flat‐rate social assistance, guaranteeing 70 percent of the national minimum wage, with extra provisions in the case of breadwinners. Trade unions in all three countries are involved in the administration and supervision of sectoral insurance funds, but only Belgian unions have a role handling individual cases of unemployed workers. This is generally seen as a powerful incentive for workers to join the union and an explanation for the high unionization rate in Belgium during periods of high unemployment (Rothstein 1992; Ebbinghaus and Visser 1999).

(p.187) All employees in these three welfare states are covered by sickness and disability insurance. In Austria, sickness benefits are payable from the fourth day of illness for a maximum of 78 weeks. Legally guaranteed benefits range from 50 percent to 75 percent of wages last earned, depending on the length of illness, the degree of incapacity to work, and family circumstances. Disability benefits vary from 60 percent of wages last earned for single persons to 100 percent for breadwinners. In Belgium and the Netherlands sickness benefits are payable from the first day up to a maximum of one year, at a legally guaranteed level of compensation of 60 percent in Belgium and 80 percent (now 70%) in the Netherlands, though in most collective agreements this is topped up to 100 percent. After one year Dutch workers qualify for a disability benefit, which is payable until the legal retirement age of 65 at a maximum of 80 percent (70% after 1987) of last‐earned wages. The actual benefit level corresponds to the degree of claimant's incapacity. After one year of sickness leave, Belgian workers may claim a disability benefit equal to 65 percent of their last wages in the case of workers with dependent relatives and to 43.5 percent for single persons (subject to income ceilings). In Austria and Belgium disability coverage is restricted to accidents at the workplace or going to and from work, whereas in the Dutch disability program entitlement is unrelated to the cause of the incapacity.

Social assistance provides the public safety net for citizens who no longer have other entitlements or means of subsistence. Assistance programs are financed by general revenue, administered by municipal social service agencies, and subject to means testing. In the Netherlands, benefits are set at the so‐called social minimum, pegged since 1974 at 70 percent of the national minimum wage for a single adult, 90 percent for a single‐parent household, and 100 percent for a married or cohabitant couple. In 1974 a national ‘right to a subsistence minimum’ was promulgated in Belgium, with maximum social assistance amounting to 28 percent of a single industrial worker's average monthly wages and 40 percent in the case of married couples. In Austria social assistance falls within the jurisdiction of the provinces, which administer their own social assistance provisions, with significant regional variation.

The prior employment record is the main criterion for entitlement to old‐age pensions in Austria and Belgium. Pension schemes are financed on a pay‐as‐you‐go basis. In Austria, benefits can add up to about 80 percent of wages last earned. There is no public basic pension in Austria and Belgium. In Austria older citizens in need must seek support from the family, voluntary organizations, and social assistance. Since 1969 there has been a minimum guaranteed income for the elderly in Belgium, providing a safety net for all senior citizens who are insufficiently covered by (p.188) occupational pension schemes. Benefits levels of this scheme are flat rate, rather low, and means‐tested. The Dutch pension system has three tiers. The first tier is a basic public pension, which guarantees everybody over the age of 65 a (non means‐tested) flat‐rate basic income of 1,400 guilders (660 euros). This part is financed out of general taxation and financed on a pay‐as‐you‐go basis. The second tier provides for obligatory additional pensions, up to 70 percent of last earned wages. In contrast to Austrian and Belgian pensions, this second tier is funded. The third tier consists of personal equity plans and other voluntary private pensions, many of which are tax‐deductible.

In the 1970s and 1980s each of the three countries used different social policy provisions to subsidize exit from the labor market and discourage entry. While early retirement and the repatriation of foreign workers provided for the main exit routes in Austria, in Belgium extended unemployment benefits performed similar functions, and generous and lenient sickness and disability pensions became the main ‘escape route’ in the Netherlands.

The three countries reveal similar trends in terms of total government outlays, resources spent on social expenditure, the share of social transfers, the financial basis of the welfare state, and taxation. In all these areas Austria was a laggard but has meanwhile caught up. The data show that total government outlays have continued to increase in Austria while declining in Belgium since 1993 and in the Netherlands since 1988. Total government outlays were 53.2 percent of GDP in Austria, 55.3 percent in Belgium, and 54.5 percent in the Netherlands in 1995. Social transfers as a percentage of GDP were 22 percent in Austria, 24.3 percent in Belgium, and 25.1 percent in the Netherlands. This share has been growing since the 1980s in Austria, whereas it has been rather stable in Belgium and the Netherlands. Almost three‐fifths is paid out of employers' and workers' social security contributions. In 1995 social security contributions covered about 57 percent of expenditure in Austria, 58.6 percent in Belgium, and 59.5 percent in the Netherlands. The non‐wage share of total labor costs is around the average of all OECD countries.

5.1.7. Segmented Pluralism: Class, Religion, and Language

Austria, Belgium, and the Netherlands have long been considered ‘segmented’ or ‘pillarized’ societies (Huyse 1970; Lorwin 1971; Lijphart 1968). In Austria, the two opposing sides reflected the class‐divide between workers and owners, the latter including the self‐employed, small business, and farmers. Before the war they had behaved like two separate societies or Lager (literally, ‘camps’), each with its own set of parties, (p.189) unions, newspapers, sport associations, and para‐military organizations, which clashed in the 1934 civil war. After 1945, with memories of civil war and Nazi barbarism still fresh, they invested a great deal in mutual cooperation.

The Reformation and Counter‐Reformation had divided the Netherlands into Protestants and Catholics who (alongside Socialists and Liberals) emerged as self‐contained communities or vertical ‘pillars' during the development of mass democracy and worker mobilization, each camp with its own set of parties, associations, unions, mass media, and clubs. Following the school pact of 1917, those at the top of these pillars learned to compromise and cooperate, each leadership safeguarding a large degree of autonomy and control within its own pillar (Daalder 1966; Lijphart 1968). This system witnessed its maximum strength in the 1950s but began to disintegrate with astonishing speed after the mid‐1960s. Religious attendance and voting declined, Catholic or Protestant associations and parties disappeared or merged, the ties between interest groups and parties loosened, and elite control over society and in associations was diminished.

The class‐division in Belgium was cross‐cut by other divisions, between Catholics (on the one hand) and anti‐clerical Liberals and Socialists (on the other) over the role of Church and State in education, and by the language conflict between Dutch‐ and French‐speaking Belgians (Lorwin 1966). The school issue was finally put to rest in 1958, but the language conflict intensified. Within a hundred years of Belgium's founding as an independent state in 1830, hopes for French as its sole, unifying language had vanished. The majority of the population spoke various local variants of Dutch, or Flemish, and many were alienated from a state that had been ruled by foreigners most of the time and long refused the use of Dutch in its laws, courts, and in secondary education. Since French‐speaking Belgians never had to learn Dutch, there was no basis for generalized bilingualism. Eventually Belgium became divided in three linguistic territories: a Dutch one, a French one, and a bilingual one, taking into account the reality of Brussels, a former Dutch‐speaking city that was located in Flanders but had gradually expanded and become Frenchified throughout the nineteenth and twentieth centuries.

The linguistic conflict was always contaminated by differences of class and economic fortune between the regions. Ever since the decline of the manufacturing and trading towns of Flanders, this province and other Dutch‐speaking regions had been the backwaters of the Low Countries, whereas the mines and steel mills of Wallonia became the champions of nineteenth‐century railway capitalism. In the 1960s the economic position of the two regions began to reverse. Mutual distrust in the Belgian state (p.190) channeled the preference of each community towards a federal solution: whereas Flemish nationalists wanted more cultural autonomy, Wallonia's Socialists and union leaders hoped to gain more economic power against what they saw as flawed policies from Brussels' haute finance and American capital. The Brussels conundrum ruled out a simple solution (Van Parijs 1998). The Flemish side defended devolution to language communities, Wallonia demanded regional autonomy, and the French‐speaking majority in Brussels feared Flemish domination since Brussels is the capital of Flanders. The solution eventually found was a double federation, i.e. a devolution into two, and later three (taking into account the small German‐speaking minority within Wallonia) language communities and three regions, yielding five sub‐national parliaments and governments (the Flemish community has merged with the Flanders region). The powers and responsibilities of these entities have been defined in extremely complex negotiations, beginning with fixing the language border in 1963 (which had previously depended on a census in which people were asked what language they spoke). In 1993 Belgium was formally declared a federal state and, henceforth, all parliamentary bodies of the communities and regions have been elected directly. The federation remains responsible for defense, justice, security, social security, labor law, and fiscal and monetary policy. Foreign and European affairs are shared depending on the policy issue at stake.

5.1.8. Consociational Democracies

The political systems of Austria, Belgium, and the Netherlands, along with Switzerland, have been regarded as the clearest examples of consociational democracies (Lipset and Rokkan 1967; Lehmbruch 1967; Lijphart 1968). In such democracies party elites are encouraged to cooperate in a spirit of non‐competitive power sharing. In his seminal theoretical case study of Dutch politics, Lijphart sums up the ideal‐typical consociational democracy as an alternative to the majoritarian Westminster model by citing four institutional characteristics: proportional representation, coalition governments, mutual veto rights, and cleavage autonomy. How well party elites are able to accommodate divergent world views in broad coalition governments depends as much on institutional features as upon the various political actors' preferences, strategic goals, power resources, and control over their respective constituencies. How policy choices affect policy performance cannot, therefore, be answered by merely focusing on institutional characteristics (Scharpf 1997). Only when grand coalitions adhere to common objectives and elites have control over their supporters does consociationalism yield the institutional advantage of quick or consistent (p.191) policy implementation. If, however, coalition partners adopt divergent policy goals or are faced with non‐compliance and internal competition, they tend to engage in counterproductive political interactions, ending in policy immobilism. This was the fate of the divided governments in Dutch politics in 1970s and of the many large but unstable coalitions that ruled Belgium in that decade.

The Austrian variant of Proporzdemokratie is rather unique, inasmuch as its cleavage structure between the two prewar Lager has translated into two large parties, each with a potential majority (SPÖ and ÖVP). From 1945 through 1966 they formed grand coalitions; in 1966 the ÖVP formed a single‐party majority that lasted until 1970, when the SPÖ won the first of several general elections, allowing it to govern on its own until 1983, at which point the Socialists lost their absolute majority. In 1986 the SPÖ returned to a grand coalition with the ÖVP, which continued into the 1990s. The size of the two major parties' combined electoral support has, however, decreased and this decline, together with the strong showing of the Free Democrats (FPÖ), a nationalist, anti‐establishment party on the right, tends to threaten the continuity of Austria's consociationalist‐corporatist polity (Table 5.5).

Table 5.5 Electoral Results of Social Democrats and Christian Democrats, 1975–1999a(%)

Year

Austria (%)

Belgium

Netherlands

Flanders

Wallonia

SPÖ

ÖVP

SP

CVP

PS

PSC

PvdA

CDA

1976–80b

50.4

42.9

12.4

26.1

13.0

10.0

33.8

31.4

1981–85

51.9

41.9

13.5

20.5

13.2

7.4

29.1

32.0

1986–90

43.1

41.3

14.9

19.5

14.6

8.0

32.6

35.0

1991–95

42.8

32.1

12.3

17.0

12.7

7.7

24.0

22.2

1996–99

35.6

22.6

10.1

14.5

10.0

6.3

29.0

18.4

SPÖ: Sozialistische Partei Österreichs; ÖVP: Österreichische Volkspartei; SP: Socialistische Partij; CVP: Christelijke ‘kspartij; PS: Parti Socialiste; PSC: Parti Social Chrétien; PvdA: Partij van de Arbeid; CDA: Christen Democratisch Appèl.

(a) Share of valid votes in percentage of total, in general elections for the national parliaments.

(b) Results are averaged for each interval in case of more than one election.

In the Netherlands coalition governments are the rule. The three main parties are, from left to right, PvdA, CDA (before 1977 divided between a Catholic and two Protestant parties), and VVD, the conservative Liberal Party, which draws its support from the middle classes, white‐collar workers, and the business elite. Between 1945 and 1958 the government was built (p.192) round the alliance between PvdA and the Catholic Party. From 1958 to 1973, with a brief interruption, the three Christian parties allied with the Liberals. From 1973 to 1977, around the time of the first oil shock, the PvdA dominated a center‐left coalition with two small left‐of‐center parties and two of the Christian parties. From 1977 to 1989, with an interruption of only nine months, the government was formed by a CDA–VVD coalition, with the CDA dominating. In 1989 it changed partners and allied with Labor. Since 1994 the government has been run by a unique Labor–Liberal coalition, forcing the Christian Democrats into opposition for the first time since 1917. Electoral support for the traditional backers of the welfare state and Dutch corporatism, the PvdA and CDA, has meanwhile dropped below 50 percent, as more voters shift their preference to the Greens on the left and the Liberals on the right.

The Belgian political landscape is more complicated and has recently undergone major changes. Between 1968 and 1978, the three mainstays of Belgian politics, the Christian Democrats (CVP/PSC), Socialists (PSB/BSP), and Liberals (PVV/PLP), split into regional parties that have had to compete with local rivals. However, the religious‐ideological segments of like‐minded political parties and trade unions, especially the alliances between the CVP and the dominant Christian federation (ACV) in Flanders and between the PS and the socialist federation (FGTB) in Wallonia, have survived. The last all‐Catholic cabinet dates from the 1950s, followed by a Socialist–Liberal coalition that put the school issue to rest in 1958. Over the next forty years, Christian Democrats ruled in alternating coalitions with Socialists (1972–74; 1977–81, and 1988–99) and Liberals (1974–77 and 1982–87). Some of these Cabinets were enlarged by regional parties or have been based on grand coalitions of all the major parties for the purpose of constitutional reform. Support for the Socialists and Christian Democrats has declined and even fell under 50 percent in the most recent general elections of June 1999. Belgium's federal government, and four of its five regional or community governments are now based on a coalition of Liberals, Socialists, and Greens.

5.1.9. Negotiated Change Under Domestic Constraints

In small countries like Austria, Belgium, and the Netherlands a sense of vulnerability is usually engraved in the minds of policymakers. External shocks induce policy actors to play down divisions, though it may take time to understand the nature and implications of such challenges. Policymaking in the three countries under study is critically dependent upon the consent of different coalition parties and on support from social partners. None of the actors is autonomous and free to choose its most‐favored (p.193) response to new conditions or external pressures. This is particularly true for the government, whose role under the rules of corporatism and consociationalism is constrained both from without and from within. By incorporating multiple parties in coalition governments and integrating the social partners into the administrative structures of public policymaking, the state can mobilize more resources and rally support for policy change. But consociationalism and corporatism can also inhibit change, precisely because of the need for extensive compromises and coalition formation. All negotiated systems are vulnerable to so‐called ‘joint‐decision traps’ (Scharpf 1988). Where the state is weak and its powers are ‘hollowed out’, this may create prolonged policy stalemates.

In the sections that follow we observe varying configurations of the relationships between the state and social partners, ranging from a very stable, uncontested, and consensual pattern in Austria, through a narrower, and variable though (in major areas) renewed cooperative style in the Netherlands, to a troubled and conflictual mode in Belgium. We will argue that these configurations matter for the formation and implementation of policies affecting wages, employment, and social security, since outcomes do depend on cooperation, at different levels, among a variety of actors with veto power.

5.2. Austria: The Rewards of Slowness

Austria entered the 1970s with full employment and a sound macroeconomic performance. Its postwar economic policy profile was strongly influenced by the fact that all interested national actors shared the goal of freeing Austria from the surveillance of foreign occupation forces and re‐establishing a sovereign nation. This feeling of ‘sitting in the same boat’ explains the highly consensual style and the involvement of social partners in many policy areas. The cornerstone of Austrian social partnership is the tripartite Parity Commission with its subdivisions for Wages and Prices. It was founded in 1957, one year after the Hungarian uprising, under the threat of an influx of immigrants. The subcommittee for prices controlled about 80 percent of all consumer goods. Though the effectiveness of price regulation was contested and imported goods were beyond its scope, the committee did have a dampening effect on price developments and gave Austrian trade unions a unique position to influence wages and prices, i.e. to negotiate over real rather than nominal wages.

Austria's postwar model of cooperation among organized labor, capital, and the state was embedded in a climate of consensus, based on shared preferences for growth, exports, full employment, stability, social peace, (p.194) and on a deep fear of inflation. From the start, Austria's economic policy model included demand‐side as well as supply‐side elements, such as interest rate subsidies, export subsidies and guarantees, subsidies to tourism, and, above all, tax exemptions. The creation of the Beirat für Wirtschafts‐ und Sozialfragen in 1963, with representatives of the social partners and ministries along with academic experts, reinforced the ‘expert’ quality of Austrian social partnership. This Advisory Board issued policy recommendations on a purportedly scientific basis and became an instrument for orchestrating a policy consensus on the basis of a common understanding of the problems at hand. Implicitly, it was understood that the shared interests were those of Austrian male breadwinners. Negative effects, like the benign neglect of consumer interests and a bias against women and foreigners, were ignored most of the time by the social partners and government officials. Austrian unions also had no explicit goal of narrowing income inequalities. Full employment, they argued, is the most effective incomes policy (Guger 1998), while redistributive policies are best left to the state.

Collective bargaining in Austria essentially takes place at the sectoral level, but is highly coordinated via preparatory talks at the central headquarters of the social partners. The sectoral affiliates of the Austrian Confederation of Trade Unions (ÖGB, Österreichische Gewerkschaftsbund) and the Austrian Economic Chamber (WKÖ, Wirtschaftskammer Österreichs, formerly the BWK) negotiate over the wage level for each industry, but under strict supervision about when and how. Compulsory membership in the Chamber and the unitary structure of the ÖGB produce an extraordinarily degree of discipline and membership (100% in the case of employers, 43% in the case of workers, down from 60% in 1970) (Ebbinghaus and Visser 1999; Traxler 1997). From the outset the peak associations were keen to prevent inflationary pressure emanating from competition among their affiliates. The ÖGB always wanted to be involved in all areas of social and economic policymaking and not just be a ‘wage machine’.

This encompassing view of trade unions was enhanced by the fact that the unions had a strong foundation in Austria's unusually large public sector. After the war large firms were mostly nationalized, since indigenous big businesses were scarce and the Austrian government tried to save them from the Soviet army's confiscation of German property. Nationalized industries in Austria accounted for about 20 percent of manufacturing employment, one‐third of these jobs in the two main nationalized steel firms. The three leading banks and the mining, chemical, and engineering industries were also practically under state control. Consequently, many of Austria's key industries were strongly unionized and in the mire of (p.195) domestic policies, and they offered a useful tool for sheltering employment if need be. In 1970 almost 80 percent of all employees in the public sector—nationalized industries, rail and postal services, and government employees—were unionized. This compared with 54 percent in the private sector (Guger 1998).

Macroeconomic policy during the era of postwar reconstruction emphasized cheap money, a balanced budget, and a strong currency in order to dampen imported inflation. The Austrian central bank (ÖNB, Österreichische National Bank) was partly owned by the social partners and the political parties, with a government share of 50 percent. This set the stage for a consensual and negotiated monetary and exchange rate policy. The ÖNB pegged the schilling first to the US dollar and then to a currency basket of its main trading partners before finally joining the European currency ‘snake’ in 1971. Low interest rates set by the ÖNB encouraged private investment by making loans cheaper. Savings were heavily subsidized in order to dampen consumption and encourage private investment. At the time, stable exchange rates with interest rate differentials between Austria and Germany could still exist, because capital controls prevented excessive and sudden capital movements. In case of inflationary pressures, the ÖNB put up credit ceilings. And as long as there was full employment, the budget was balanced or even in surplus.

5.2.1. Collective Mistakes ‘Jointly Corrected’ In the 1970s

When it was confronted with the first OPEC oil‐price shock, Austria appeared to have had the luck of the innocent. It had everything in place for an economic policy model that matched the requirements of fighting stagflation. The oil shock hit unexpectedly, as it did in other countries. With an expected inflation rate of 10 percent, unions and employers had agreed to a 14 percent wage rise for 1975. When, contrary to the expected growth rate of 3.5 percent, a −2 percent contraction occurred, the social partners jointly ‘corrected’ their mistake (Scharpf 1991: 56). The unions and the government assigned priority to full employment over real wage increases, redistributive objectives, or budgetary restrictions. This even applied to the conservative opposition. The SPÖ government pursued an expansionary budget policy, including public investment programs and indirect subsidies to industry, in order to stabilize demand, while (as a way of curtailing inflationary pressures from abroad) the ÖNB kept interest rates low and tied the schilling to an appreciating Deutschmark. This ‘unusual mix’ (Guger 1998) of policies proved extremely successful in countering the dilemma of stagflation. It could only work if every actor played its part and the social partners were able to control inflationary (p.196) pressures and prevent the erosion of competitiveness through wage moderation. Austria's social partners were able to do just that because of their organizational strengths, consensus, and the all‐encompassing corporatist framework. Keynesian ideas had reached the Austria economic policymaking community only late (Rothschild 1993: 137), but in the 1970s the interaction of budget, monetary, and wage policies provided a good fit with the prescriptions of Keynesian demand management.

This combination of deficit spending, hard currency, and incomes policy was later coined ‘Austro‐Keynesianism’. It essentially revolved around five demand‐ and supply‐side instruments aiming at full employment and growth: deficit spending in order to stimulate demand, an accommodating monetary policy of low interest rates in order to stimulate private investment, labor hoarding in the nationalized industries, a hard currency policy, and an incomes policy. Austro‐Keynesianism differed from traditional demand management in that it was not confined to anti‐cyclical policies, but rather represented a long‐term device to stabilize business expectations with the help of incomes and exchange rate policies, encouraging growth and investment (Tichy 1985). Employment‐stimulating budget measures were supported by cheap loans promoting private investment and allowing labor hoarding in the public sector. Inflation was held down by means of a hard currency and wage restraint. The commitment to fixed exchange rates necessitated that current account deficits resulting from such a policy had to be dealt with separately through export subsidies and taxes on imports.

Inflation came down from its 1974 peak of 9.5 percent. The increases in real wages during the recession of 1975 were compensated by subsidies to firms out of the budget and corrected in the bargaining rounds for the next year. Since business could rely on moderate wage claims in the future, it did not need to increase prices. The hard currency choice helped keep the price of imported goods low. Consequently, Austrian inflation fell rapidly in line with Germany's. A comfortable budgetary position allowed enough room for maneuvering in macroeconomic policy. 1975 was an election year and the SPÖ, aiming at maintaining its absolute majority, gave priority to full employment and was willing to experiment with Keynesian deficit spending in a way that overruled the traditional commitment to a balanced budget. The government allowed the deficit to increase from 1.5 percent of GDP in 1974 to 4.5 percent a year later. Since the budget deficit was largely financed abroad, the domestic money supply increased and crowding out private investment through higher interest rates was avoided.

It has been argued that Austria's exchange rate policy in the mid‐1970s was hard, whereas its monetary and labor market policies were soft (Breuss 1978: 6). The currency policy of the early1970s was quite flexible, and there (p.197) were still visible signs of an autonomous exchange rate policy. Having depreciated by an accumulated −13 percent between 1968 and 1973, the schilling next appreciated 2.9 percent against the dollar from 1974 to 1977, which helped keep the price of imported oil down (Tichy 1985: 500). Within the SPÖ Party, the unions emphasized the importance of stable exchange rates as a bulwark against inflation and the loss of purchasing power. As an alternative to depreciation, massive export subsidies helped ease the burden on industry, and they also help explain why Austrian exports (−3.5%) declined less than in Germany (−6.1%) (Uher 1993: 86f.). Trade unions did not take advantage of labor hoarding in the nationalized industries by pressing real wage increases, demonstrating that full employment was their true priority. Between 1973 and 1983, redundancies in the state sector were half of what they were in the private sector—indicating that the public sector did in fact shelter employment. It was a political decision that turned out to be an advantage once the crisis had passed. Nationalized industries were eager to exploit new market opportunities, whereas comparable German firms were reluctant to hire new workers after they had gone through the experience of costly layoffs. The number of workers employed in Austrian industry fell only slightly in 1975–76 and increased immediately afterwards. An additional source of support for employment came from the service sector. Tourism, retailing, and public sector jobs in municipalities, police departments, and post offices soaked up those industrial workers who had been laid off (Scharpf 1991: 58).

The demand‐side success of Austro‐Keynesianism was supported by fairly effective supply‐side measures. First, like other European countries in the mid‐1970s, Austria implemented a previously negotiated reduction of the working week to 40 hours. Second, the government and the social partners jointly decided to implement a more restrictive quota on foreign laborers. The influx of foreign workers had played an important role in postwar Austria, helping to overcome the labor shortage during the period of reconstruction. Immigrant labor served as a buffer during the 1970s, when policymakers agreed to restrict immigration and encouraged repatriation, although not as extensively as in Switzerland. A limit on work permits resulted in a reduction of 52,000 foreign workers (−20%) between 1973 and 1978, equaling a 2 percent reduction in the labor force (Scharpf 1991: 58). The combined effect of all these measures showed up in low open unemployment and a rising employment/population ratio, despite the slight decline in employment (−56,000) in 1975.

Chancellor Kreisky received a political bonus for his economic management in the elections of 1975 and, again, in 1979. Hence, the SPÖ maintained its absolute majority throughout the crisis‐ridden 1970s. Inflation was contained thanks to trade union wage moderation and a (p.198) strong currency. Full employment was preserved because the SPÖ privileged job creation, labor hoarding in nationalized sectors, and repatriation of foreign workers over the goal of budget consolidation. Unemployment remained below 2 percent in the second half of the 1970s. The price of the hard currency option was a deteriorating external balance, but this seems not to have been a major constraint. The problem was addressed with an increase in the VAT rate on luxury goods and cars, intended to restrict imports, and through special investment and export‐promoting measures in the context of the Association Treaty with the European Community (Lehner 1991: 474). In contrast to later periods, confidence in the ability of national crisis management was high (Kienzl 1993: 71). By virtue of the outstanding macroeconomic performance it achieved—maintaining full employment in the face of the 1970s dilemma of stagflation—the policy‐mix of Austro‐Keynesianism was raised to the status of an effective crisis management model.

5.2.2. The Policy of ‘Diving Through’ In the Early 1980s

The oil shock of 1979, in combination with the financial constraints of 1981's high interest rates and increased international capital mobility, posed a new challenge to Austrian policymakers. This time they managed less well. Their anticipated policy response of ‘diving through’ this second crisis in a manner that was tried and familiar underestimated changes in the international environment. Financing the US current account deficit had absorbed 10 percent of world savings, and the resulting high interest rates spilled over into Europe. The effectiveness of capital controls had eroded, and international financial markets had become less predictable. This became evident in 1979, when the ÖNB set interest rates below the German level.

The negative trade balance, due to the increased cost of oil imports in combination with low interest rates, resulted in a massive withdrawal of capital. The ÖNB lost about one‐third of its reserves. This effectively caused the central bank to give up its active nominal interest policy (Winckler 1980). The schilling depreciated 2.3 percent against the Deutschmark in 1979 (Tichy 1985: 500) and inflation increased from 3.7 percent in 1978 to 6.8 percent in 1981. Since all Austrian actors were afraid of depreciation and its inflationary consequences, the schilling was officially pegged to the Deutschmark in 1980. By tying its hands, the ÖNB sacrificed monetary policy in order to concentrate on exchange rate stability (Hochreiter and Winckler 1996). The dramatic loss of currency reserves had persuaded the central bank to abandon one crucial pillar of Austro‐Keynesianism.

(p.199) The budget deficit came under increased pressure from rising interest rates. In principle, the government applied a formula, which prescribed that the annual net deficit should not exceed −2.5 percent. Fiscal policy did not allow heavy deficit spending. But, as in 1974–75, the recession of 1981–82 was fought with the help of expansionary employment programs, letting the deficit slip to −4 percent in 1983. Employment in the exposed sector was kept relatively constant, due to continued labor hoarding by the nationalized industries. Elderly workers were allowed to take early retirement. If unemployed, they received special allowances. Sending foreign workers home by withdrawing work permits, however, proved more difficult now, since many immigrants had applied for Austrian citizenship after an uninterrupted stay of ten years. Still, banking on its successes from the 1970s, Austria managed to keep unemployment down at 3.5 percent in 1982. Full employment policy was not maintained, however. Using national rather than OECD or ILO statistics (the latter being more restrictive since they measure readiness to accept a job), unemployment rates had more than doubled within four years (from 2% in 1979 to 4.5% in 1983). The employment/population ratio fell from 68 percent in 1979 to 63 percent in 1983, reflecting early retirement and discouragement. Active labor market policies were hardly used. Favorable developments in the 1970s had induced Austrian policymakers to continue without major changes, applying the rule: ‘If it ain't broke, don't fix it’ (March and Olsen 1989).

Restricting the labor supply was a major policy response to the labor market problems of the 1980s. Giving in to union demands for a new round of working time reduction, employers granted two extra holidays in 1985, and a year later the minimum vacation was increased to five weeks, six weeks for workers with twenty‐five years of service (Hauth 1989: 19). Negotiations over working hours were later moved from the national to the sectoral level, a practice adopted for wage bargaining since 1982. By the end of the 1980s the average working week was 38 or 38.5 hours. Eligibility criteria for receiving disability pensions were eased, and special allowances were made available to redundant workers too young for early retirement but deemed too old to find a new job. These arrangements, targeting firms in difficulties, were supported with public money and became very expensive in due course. Between 1975 and 1985 the number of older workers fell dramatically: while 98 percent of men and 47 percent of women between 55–59 participated in the labor force in 1975, these percentages had dropped to 70 percent and 30 percent in 1985. Participation rate for males aged 60–64 were halved from 38 percent to 18 percent; female rates in this age group decreased from 13 percent to 8 percent (Walterskirchen 1997). The SPÖ paid the price for its diminished success (p.200) in managing the 1981–82 recession with disappointing results in the 1983 general elections. The Social Democrats lost their absolute majority and turned to the FPÖ (then still quite liberal) to form a new coalition. In 1985 ‘Austro‐Keynesianism’, already dead for some years, was officially buried by the new SPÖ Chancellor, Sinowatz, when he observed that it had been a policy of ‘diving through’, which could only bring temporary relief and never work in the long run (Seidel 1993: 146). It had helped to retard the consequences of internationalization and prevented a rise in unemployment rates.

5.2.3. ‘Stepping on the Brake’ In the Later 1980s

The new SPÖ–FPÖ coalition gave budget consolidation priority over full employment. Labor hoarding in the nationalized sector was increasingly seen as a burden on the public budget. The crisis of the steel industry became a factor in Austrian politics in the 1980s (more than ten years after Belgium or Germany). Between 1981 and 1985 the two major steel firms, Voest‐Alpine and VEW, had received 25 bn ATS (3.6 bn DM) in subsidies. In 1985 Voest needed another 16 bn ATS while, for the first time, it announced plans for a mass layoff of between 25,000 and 34,500 people to be completed by 1990. At the same time, Voest Intertrading announced enormous losses from speculating on international financial markets. Consequently, the issue of privatization stopped being taboo in Austria's negotiated economy, and labor hoarding in the nationalized industries was gradually abandoned. Finally, Austrian corporatism was willing and ready to open up to the outside world, though at a slow pace, as summed up in a famous quote by ÖGB President Verzetnitsch: ‘The most important person in a bob‐sled is the one who handles the brake.’

The 1980s were marked by changes in policies, politics, and perceptions. Nationalized industries were increasingly perceived as rent seekers. When the right‐wing populist Haider took over the FPÖ, the SPÖ ended its coalition with the ostensibly liberal party. After the elections of 1986, Social and Christian Democrats formed a grand coalition. The new SPÖ Chancellor, Vranitzky, came from the banking sector and was committed to a modernizing course. Both parties set their sights on membership in the European Union. They agreed on an extended program of privatization in 1986–87. Begrudgingly, an internally divided ÖGB went along, its member unions squeezed between the never‐ending claims of the protected sector and a private sector that was less willing to foot the bill. The WKÖ wanted privatization and deregulation, but slowly. ‘Let up on the brakes but don't step on the gas' was the slogan of its general secretary (Teufelsbauer 1986). The privatization program for 1987 left 51 percent of (p.201) the shares of the public sector, including banking, Austrian airlines, and energy firms, in public hands (Feuchtmüller 1987). Privatization proceeded in steps, and the impact was cushioned with social plans elaborated by the social partners. This helped keep unemployment stable at 3–4 percent. Yet the abandonment of labor hoarding and the restructuring of the nationalized industries—the country's most unionized sector and home to the Socialists' most stable electorate—led to a decrease in ÖGB and SPÖ membership. Under threat of losing their jobs, the frustrated workers became the ideal constituency for an increasingly right‐wing FPÖ.

In 1987, later than in most other countries, capital controls were abandoned. Now international evaluation of Austria's economic performance became even more important. A sound budget figures high in international country ratings, which governments cannot afford to ignore. One convenient way to create revenue is to sell off the ‘silverware’ of state‐owned enterprises. Once again, this gave Austria time to postpone substantial social policy reforms. By the end of the decade, when privatization and industrial restructuring were well under way, the government finally succeeded in its consolidation effort. Helped by the expansionary effect of German unification, the budget deficit fell from 4.4 percent of GDP in 1987 to 1.8 percent in 1992. The tax reform of 1988 played a minor role in consolidation; it was mainly seen as a step to adjust Austria to international standards. The employment effects were very modest (one calculation put them at a 0.2 percentage point cut in the unemployment rate). Taking its cue from examples in the USA, Germany, and Sweden, the reform was meant to increase the Austrian tax system's efficiency and transparency by reducing exemptions in exchange for lower rates and a broader tax base (Nowotny 1989). It was a political signal to capital that tax subsidies could no longer be expected (Lehner 1991: 475). The corporation tax rate was reduced from 55 percent to 30 percent, later raised to 34 percent again (comparable to Belgium at 39% and the Netherlands at 35%). Going against the international trend, the progressive element in the income tax system was strengthened (Guger 1998), and the SPÖ and unions continued to favor lenient treatment for small investors and savings accounts, many held by their members. The tax reform of 1993 abolished the inheritance tax and the tax on artisan enterprises. Austria appears to have the lowest property tax in the OECD (Farny 1996: 74). As was discussed in the introduction, overall levels of taxation are slightly above the OECD average, though slightly lower than in Belgium and the Netherlands. The overall tax burden rose in the period up to the early 1980s but has been constant since, hiding a shift in the composition of revenues towards social security charges, now just over one‐third of all (p.202) revenues. Again, at this point there are hardly differences with Belgium and the Netherlands. The Austrian tax reforms of 1988 and 1993 shifted the tax burden from capital to labor. The effective tax rate on labor in Austria (43% in 1995) lies now between Belgium (40%) and the Netherlands (46%) and is not greatly out of line with other European continental welfare states (OECD 1998: 60).

While reducing employment in the nationalized industries, Austria increased public sector employment, going against an international trend that even included Scandinavia. Government employment increased steadily from 8.9 percent of total employment in 1970 to 11.6 percent in 1980, 13.3 percent in 1990, and 14.3 percent in 1994. Between 1987 and 1996, the public sector, including education and health services, contributed toward easing the labor market situation by adding 140,000 jobs, equal to two‐thirds of total employment growth (Walterskirchen 1997: 7). Most of these public service jobs went to Austrians, whereas foreign workers found jobs mainly in construction, retail, restaurants, and manufacturing. In 1995 the federal government imposed an employment stop, and all new jobs now have to come from the regions and municipalities, responsible for education and health. Social Security and related public services expanded employment. By separating the post office and Austrian railways (now independent corporations) from the state budget, the government was able to generate new revenue.

5.2.4. The 1990s: Going Europe

The end of the Cold War and the opening of borders led to a revival of an old Austrian fear—to be left in the isolated corner where its capital lies geographically: East of Prague. Until 1989 Austria's place in the West had been assured by the Iron Curtain. To join the EU became now a matter of utmost urgency. Back in the 1960s Austria had already applied for membership, but the USSR had vetoed this. Its robust economic position, based on a hard currency, low inflation, low unemployment, and close trade links with Europe and Germany, made Austria a sure candidate for EU membership. The budget deficit was the only cause for worry. Entry into the EU in 1995 and the austerity measures required to meet EMU membership criteria became two showpieces for the resurgence of Austrian corporatism in the 1990s.

In the early 1990s Austria's relative growth performance had improved and public finances were in good shape again. Though small by international comparison, Austria's labor market problems had however increased. Austria was strongly affected by the crisis in Yugoslavia. In 1991 an inflow from the traditional immigrant countries Yugoslavia and (p.203) Turkey added 7 percent to Austria's labor force. Unemployment increased steadily, to 4.4 percent according to OECD and ILO statistics, 7.8 percent according to national statistics. This is still modest by European standards of today, but high if judged in the light of Austria's postwar accomplishments. Increased public sector employment and the effects of introducing the second year of a maternity allowance helped keep unemployment down, but at the cost of higher budget deficits. Under the impact of lower growth rates, the deficit rose to 5 percent in 1995, despite consolidation efforts. No alleviation was to be expected from tax policy or the sale of state enterprises. Substantial expenditure cuts became unavoidable. In order to divert blame from this unpleasant policy problem, domestic politicians welcomed the use of Brussels as a most convenient scapegoat.

In the late 1980s the SPÖ and ÖVP had presented their ideas on joining the EU and, after some massaging, managed to convince their respective social partners. This had been more difficult for the SPÖ than for the ÖVP. Initially, the unions received Vranitzky's liberal pro‐European ideas with a great deal of skepticism. The ÖGB was split between two wings, one in favor led by the powerful metalworkers union representing mainly members from export industries, and one against led by the union of whitecollar staff in industry and services, now the ÖGB's largest affiliates. The ÖVP had to appease the opposition from small artisans and farmers. Farmers were induced to vote in favor of EU membership by the promise from industrialists to support liberalization of trade regulations, enabling farmers to earn money outside their traditional tasks. Later business reneged under pressures from tourism and retailing, and ÖVP had to orchestrate a new compromise, less favorable to farmers (Tálos and Kittel 2000). In the end, Austria's social partners ‘went Europe’ and joined the grand coalition in support of EU membership. In 1992 unions and employers negotiated an agreement in which they codified the switch from a closed to an open economy. A subcommittee for international affairs was founded and the long obsolete subcommittee for prices was changed into a subcommittee for competition (Nowotny 1998). The hope of being able to meet future shocks of internationalization under an EU umbrella, with closer ties to Austria's big German neighbor, had outweighed fears about a loss of sovereignty and lower social standards. After they had convinced themselves, the social partners played a leading role in persuading the public, and in the 1994 referendum 66 percent of the Austrian electorate voted in favor of EU membership, leaving only Haider's FPÖ and the Green Party opposed. Austria became a member of the European Union in January 1995, together with Sweden and Finland.

With EU membership Austria had set its sights on joining the EMU as well. It effortlessly fulfilled all the membership criteria, except the 3 (p.204) percent norm for budget deficits and the 60 percent norm for the public debt ratio. A consolidation package of rather unpopular fiscal authority measures would have to be implemented. The social partners had been involved in virtually every major economic policy decision since the late 1950s. However, for the first time in forty years, the government decided after initial negotiations in November 1994 to go ahead with its austerity plan without asking the prior consent of the social partners. Vranitzky was the banker‐technocrat type of politician who had no close ties with the union leadership, and even within his own party he was not fully supported. The ÖVP was split among the different interests of big industry, farmers, small business, and artisans. Perhaps the idea was that some demonstration of political power or public leadership was called for. The trade unions denounced the procedure as intolerable and stepped up resistance. Their substantive criticism was that the package, including cuts in maternity and early retirement allowances, family assistance, unemployment benefits, and civil servants pay, as well as higher contributions for health insurance, schoolbooks, and public school transport, was socially unbalanced. This led to fierce tension between unions and the party. The SPÖ was dependent on union support in parliament, and the austerity measures that were ultimately presented and adopted did not go as far as originally planned.

The unions had demonstrated their veto position but they were no longer gatekeepers. They could not avoid accepting the government's agenda and be drawn into negotiations over stronger austerity measures. But its failed show of political resolve had taught the government the lesson that it could not go against the social partners when making painful decisions. Yet the conditions for cooperation had changed. The government now decided the issues and the social partners worked them out (Tálos, interview, Sept. 1998). The second austerity package of September 1995 was negotiated with the social partners but was also far more severe than the first one. It included expenditure cuts of 34.5 bn ATS (2.5 bn euros), including a cut in family assistance and in birth assistance allowances, a cap on health care expenditure, and stricter regulations on unemployment benefits. Furthermore, the Structural Adjustment Acts of 1995 and 1996 raised the threshold for early retirement, subjected disability pensions to stricter criteria, and phased out the special early retirement schemes. Paradoxically, the social partners had gone much further than the government would ever have dared (Tálos, interview, Sept. 1998). The measures were implemented without protest—a unique feat in Europe.

(p.205) 5.2.5. Social Policy: A Pensioners' State

As discussed in the introduction, Austria's social security has developed along lines typical for Continental welfare states. It relies heavily on gainful employment and joint family coverage. This is reflected in the fact that about 95 percent of benefits are devoted to health, pensions, and unemployment insurance (plus family benefits), and only 5 percent to social assistance. Austria's policy of keeping overt unemployment low through the early retirement is clearly visible in the structure of social expenditures. In 1995, 49 percent of social spending went to old age pensions, which is significantly above the EU‐12 average of 44 percent. On the other hand and for the same reason, expenditure on unemployment (6%) is below the EU average (9%).

The 1995 restrictions on unemployment insurance increased the reference period from 20 to 26 weeks; benefits now end after 20 weeks, though under certain conditions they may be extended to one year. Net replacement rates have been lowered from 58 percent to 56 percent of net wages. When unemployment insurance benefits expire, a means‐tested allowance, up to a maximum of 92 percent of the prior unemployment insurance benefit, is available under more restrictive conditions. The extended benefits introduced in 1990 for older workers or workers in crisis regions have been abandoned again. Some local authorities responsible for social assistance have made payment contingent upon the willingness to work, introduced sanctions in cases of abuse, and reduced benefits for asylum seekers (Alber 1998: 18; Tálos and Wörister 1998: 273–79). In the health system small user charges were introduced in 1989 for hospital care, and in 1996 for ambulant care (Tálos and Wörister 1998: 230).

Austria's main problem, so far unsolved, is the pension system. The 1980s and 1990s transformed the Austrian welfare state into a ‘state of pensioners’ (Esping Andersen 1996: 74). Since 1975 the average retirement age for the dependent labor force has declined by roughly three years. In 1996 it averaged 58 years for men and 56.5 years for women (Pichelmann and Hofer 1999: 9). In 1993, 2.2 times the amount of transfers going to young people were spent on the aged. Only Italy beats Austria in this respect. The 1997 Pension Reform raised the assessment base of pension benefits from fifteen years of highest earnings to eighteen so as to harmonize the private sector scheme with pensions for civil servants (Tálos and Wörister 1998: 262). This change will be gradually phased in through 2003 and will make it harder for people with a short employment record, especially women, to draw a full pension (Alber 1998: 14). Early retirement was made financially less attractive by cutting the replacement rate and terminating allowances for the elderly unemployed who go into ‘pre‐early’ (p.206) retirement. This has resulted in a curtailment of premature pensions, drawn after only thirty years of work, by an estimated 7 percent (Alber 1998: 15). The Constitutional Court has ruled against the unequal pension age for men and women, and women's pensions will now be raised from 60 to 65 years, to be fully phased in by 2024. Under pressure from the European Court of Justice, foreigners must be granted access to emergency aid. This is one of the few exceptions to the restrictive trend in social policy during the 1990s. The others are the introduction of health care benefits for the old and sick, demanded by the SPÖ, and the extension of maternity allowances (to 18 months) and two years of unemployment insurance for employees who become self‐employed, a benefit demanded by the ÖVP and business groups. We agree with Alber (1998: 19) that the Austrian welfare state is still intact and that no major shift towards a new welfare mix strengthening private provisions ever occurred.

5.2.6. The Alpine Model: Success or Failure?

Austria's overall macroeconomic performance is quite good. Unemployment rates have been among the lowest in Europe (4.4% in 1999), though higher than what they used to be. Inflation is low, and the current account improved owing to the international recovery and unusually high receipts from tourism. Income inequality is significant and growing, but not a major issue. Social peace and stability are still as solid as the Alps.

Schettkat (1999) used ILO data and called Austria one of the four European tigers, along with the Netherlands, Denmark, and Ireland, on account of these countries' growing employment/population ratios. However, our OECD Economic Outlook data, shown in Figure 5.3, suggest that Austria was more like a paper tiger. Employment has been on a downward trend since the 1970s. There are not enough new jobs to absorb the increase in the labor supply due to more women (re‐)entering the labor market and immigrants seeking work. Employment ratios for those over 55 are extremely low. Roughly four‐fifths of job creation comes from the public sector, the major pillar of employment policy in the 1980s and the 1990s. Between 1980 and 1996 50,000 new jobs were created in the private sector and 208,000 in the public sector. It is more than likely that this public job machine will not continue to work, owing to additional consolidation efforts. Recently a trend towards more part‐time employment is observed. In 1998 15 percent of all new jobs had less then 10 weekly hours, and these were mainly held by women. But part‐time jobs for women are often of inferior quality, involuntary (Muehlberger 1998), and socially unacceptable to men.

What is left, then, of Austro‐Keynesianism? We have argued that this peculiar but successful combination of policies was abandoned in the early (p.207) 1980s. Fifteen years later, upon joining the EMU, Austria delegated monetary and exchange rate policy to the European Central Bank. The nationalized industries have progressively been sold off and privatization has been practically completed. Industrial policy is busy applying EU standards. In order to fulfill the criteria for the Growth and Stability Pact of the Amsterdam Treaty, additional budget cuts have been proposed. These have to come out of the expenditure side, since the near future is unlikely to bring any new increase in receipts. The tax reform for the year 2000 is forecast to result in a reduction of 30 bn ATS in government receipts. Trade unions wanted lower tax rates as compensation for real income losses over the last fifteen years, and the remainder of the lost receipts derives from tax deductions of families demanded by the ÖVP. The original reform goal, as announced in the National Action Plan for employment submitted to the European Union, was to improve Austria's competitive position by reducing taxes and duties on labor costs, especially through a reduction of payroll taxes in low‐wage sectors. But this was not realized. The employment effects will therefore be modest, and budgetary concerns paramount.

Wage moderation is the only instrument remaining from the traditional Austro‐Keynesian toolkit. Throughout the turmoil of the late 1980s and early 1990s, the Austrian wage bargaining system remained a beacon of stability. Bargaining practices remained firmly entrenched at the sectoral level, with the metalworkers assuming wage leadership (Traxler 1997). But what if unemployment keeps creeping up? Will disillusioned trade unions demand higher wages for insiders? Will passive labor market policies be abandoned? Access to the EU Social Cohesion Fund has increased the attractiveness of active policies. Supervised by a tripartite board, the Austrian public employment service is managed by the social partners. The revised National Action Plan (1999) foresees a drop in the unemployment rate from 4.4 percent in 1999 to 3.5 percent by 2003 and the creation of 100,000 additional jobs. The emphasis is on reducing youth unemployment. There are doubts as to whether these measures will be enough to make up for expected job losses in textiles and clothing (−23,000 jobs), food manufacturing (−5,000 jobs), and building materials (−15,000 jobs). Employment in services picked up, but only slowly. In 1998 30,000 additional jobs, many of them part‐time, were created, mainly in the business‐related service sector and in retail. Working time flexibility appears to be rather limited; compared with Belgium and the Netherlands, fewer Austrians work in shifts, at night, or during weekends. After considerable pressure from the government and heavy resistance from trade unions against amending the working time law in 1997, the power to regulate these matters has been devolved to sectoral level collective agreement.

(p.208) In short, the Alpine model is rather old‐fashioned. Despite the rhetoric of modernization, it is an example of ‘defensive corporatism’. It continues to rely on compromise by the social partners reached by unanimous decision between comprehensively organized peak associations based on compulsory membership for employers and a union monopoly on representing workers. Economically, the model has not escaped the ‘welfare without work’ logic of Continental welfare states. Exit from work has been the main strategy to keep overt unemployment low. Public employment has been the most important instrument for confronting the shortage of jobs. Job creation in private services has been lackluster. Politically, the two big coalition parties have succeeded in retaining their exclusive hold on political power, but the challenge of Haider's right‐wing populism has become stronger and now poses a major obstacle to social partnership. In 2000 the new coalition between ÖVP and FPÖ meant an end to forty years of SPÖ government. Most social changes have come from abroad. Equal rights and affirmative action for women were pushed onto the Austrian agenda by the European Union and became law in 1995 after an impressive showing in a referendum. A European Court decision required that Austria stop excluding foreigners from receiving social assistance. The Austrian paradox seems to be that, even with a high level of female labor market participation (higher, in any case, than in Belgium and the Netherlands), this society is still very male oriented.

The Alpine model has its drawbacks, but is it a failure? Is it a dinosaur, incapable of adjustment to the needs of post‐modern society (Crepaz 1995)? We prefer to speak of a midlife crisis for a 42‐year‐old social partnership, based on fears of faltering potency, a perception that the best years may be over, and sudden engagement to a younger bride in Brussels (Unger 1996). If there are signs of modernization, they essentially come from without—Europe—rather than from within. But a midlife crisis eventually passes, even if the best years are over. Union density has fallen by one‐third during these twenty‐five years, and compulsory membership in the Chambers has been challenged, but the ÖGB and WKÖ are still formidable powers. Recent efforts of the new coalition to exclude labor from policy reform negotiations and to split the social partners might change this drastically. In 1997, 60 percent of Austrians (in 1983 it was 69%) agreed that social partnership was an advantage for Austria, whereas only a small minority of 12 percent expressed the opposite view (Tálos and Kittel 2000). Referenda among members of the Economic Chamber and the Chamber of Labor produced a large turnout and considerable majorities in favor of continuing compulsory membership.

Adjustment towards a flexible deregulated labor market progresses faster than the public debate likes to admit. Many plant‐level agreements (p.209) have flexible working time schemes, hardship clauses are used, and troubled firms are exempted from sectoral collective agreements. In the retailing and building industry, in tourism and business services, there is greater regulatory flexibility than before (BMAGS 1999). This is part of Austria's social partnership culture, which accepts internal flexibility, especially when demanded by works councils, which are interested above all in their firms' survival. Typically, they defend internal flexibility and deviance from the rules while simultaneously sticking up for strong external constraints as fixed in collective agreements and law. In the process, Austrian social partnership has witnessed some devolution towards sectoral negotiations and agreements. This implies some loss of central coordination capacities. Associational self‐regulation is primarily confined to wage bargaining and vocational training. In short, the social partners no longer decide which issue should be on the public policy agenda. Rather they must deal with issues presented by the government, which itself is increasingly borrowing from the EU agenda. As the story of the consolidation package has so clearly demonstrated, the social partners are, however, still perfectly able to defend their institutional veto powers.

Of the three countries, Austria was both last and least exposed to the challenges of international trade. In combination with the country's extremely slow pattern of adjustment, this may have helped Austria avoid extreme fluctuations in output and employment. Being a laggard with roughly ten years of retarded development has its advantages. In Nadolny's novel The Discovery of Slowness, the captain prevents his ship from being lost in the driving ice precisely because his innate slowness prevents him from reacting at once. Austria entered the 1990s with lower unemployment rates and fewer fiscal problems compared to most other welfare states. Incremental or defensive corporatism was able to absorb external shocks and avoid internal upheaval by successfully shifting some problems ahead and solving only others that were most urgent. The transformation of the welfare state has only just started. But Austrian corporatism's still strong institutional capacity, demonstrated by a successful accession to the European Union and the austerity measures of 1995, means that implementing difficult changes can happen quickly, once the actors have decided on a course of adjustment.

5.3. The Netherlands: Reinventing Itself

A distinctive feature of Dutch postwar economic policy was its stringent wage policy. Between 1945 and 1963, longer than any other Western democracy, the Netherlands ran a statutory wage policy (Windmuller (p.210) 1969). Collective agreements needed prior approval from a Board of State Mediators, which was bound by wage guidelines issued by the Minister of Social Affairs. These annual guidelines were subject to central negotiations and intense consultation with the central organizations of unions and employers, which kept affiliates under strict hierarchical control. Statutory wage policy disintegrated in the 1960s under the pressure of tight labor markets and considerable wage drift.

After a failed experiment with self‐regulation and intermittent state intervention, the new Wage Act of 1970 handed the responsibility for wage setting back to unions and employers. The government retained the power to order a temporary wage stop or impose a ceiling on wages if the economic situation did in its view warrant such a step. The unions opposed the law and temporarily left the main advisory body, the Social Economic Council (SER), in protest. Experience had told the Dutch unions that government controls placed them in the impossible position of having to defend low wages while profitable firms repeatedly broke the rules under pressure and paid higher wages. Without control over the works councils or some other kind of presence within individual firms, government controls tended to ‘hollow out’ unions' bargaining role and make members disaffected (Visser 1995). Distrusting the works councils and the government, the trade unions defended central wage negotiations, at a distance from the firm but also without government involvement. Dutch employers likewise ruled out decentralization, since they feared that company bargaining would only help a radicalizing union movement gain entry into the firm. Small and medium‐sized firms lobbied for the continuation of government controls. Large and multinational firms might easily have done without government interference, which they increasingly resented, but they did not break ranks with other employers.

By the late 1960s Dutch wage policy, the distinctive element of postwar economic growth and industrialization, was in shambles. Unions and employers were increasingly divided, even within their own ranks, over how to conduct a responsible wage policy and over the role of the government. They still maintained (and would continue to maintain until 1982) that wages should be determined centrally, with annual guidelines fixed by their peak associations represented in the Foundation of Labor. This disarray resulted in only one agreement, in 1972, but that accord was overshadowed by a leadership crisis in the Dutch Federation of Trade Unions and a strike in the metal industry over extra compensation for low‐paid workers.

Comparing the Netherlands with Austria on the eve of the first oil shock, we find a completely different institutional setting. While both countries had a legacy of centrally coordinated wage bargaining, this had broken down in the Netherlands, and there was still no agreement on an alternative. (p.211) Institutional failure was aggravated by the rise in social conflict and breakdown of elite control that accompanied ‘depillarization’. This proved all the more serious because postwar industrialization, helped by low wages and rich natural gas finds, had relied upon labor‐ and energy‐intensive export products. By the 1970s the Netherlands had become a high wage economy as a result of large wage hikes from the 1960s—a formalization, in effect, of what some large firms were already paying out as ‘black wages' (Visser and Hemerijck 1997). It proved impossible to sustain industries like coal mining, textiles and clothing, and shipyards, and we observe that these industries' decline began earlier (before the currency appreciation of the 1970s) and was more intense than in Austria or Belgium (earlier, in fact, than in every European country except Britain). This would have been less of a problem had the Netherlands managed to match its industries' decline with growth in services, private or public. But services did not step into the breach, and the entire 1970s were marked by a decline in manufacturing employment and a stagnation of private services.

5.3.1. The Dutch Disease

When the Bretton‐Wood system collapsed, the Netherlands joined the ‘snake’. There seems to have been agreement on this move among policy actors, reflecting a tradition of defending a strong currency and of placing financial before industrial interests. This choice meant that competitiveness would decline if Dutch unions proved unable to enforce wage restraint while the currency was appreciating. This is exactly what happened, especially when Denmark, Italy, France, and the UK defected from the ‘snake’. Breaking the wage‐price spiral became the key issue of the 1970s. Since labor costs in manufacturing were rising faster than in competing countries, all relevant policy actors agreed on the desirability of some form of incomes policy to combat cost‐push inflation, but there was no consensus. On the eve of the first oil crisis, a left‐of‐center government led by Den Uyl (PvdA) had taken office. Impressed by the recession but expecting the crisis to be short‐lived, he opted for a Keynesian strategy of fiscal stimulation. The stage was set for a corporatist policy package, along the lines of the Austrian example, exchanging fiscal reflation for voluntary wage restraint. But the radicalized Dutch unions wanted more reforms in economic policy than this government, with weak support from its Christian Democratic coalition partners in Parliament, could offer. In 1974 the government imposed a wage and income freeze, but its effect was undone as soon as the ban on higher wages was lifted.

Thanks to its gas exports, the Netherlands did not experience balance of payments problems after the oil‐price hike (the country itself was (p.212) targeted by OPEC's embargo). Hence, an expansionary course seemed feasible and appropriate. There were various sizeable employment programs, but the main rise in government spending came from expanding welfare programs. This was partly the result of more people becoming unemployed or reaching the limits of their insurance and partly a concession to the egalitarian and solidaristic preferences of the unions. In 1974 the minimum wage and, indirectly, public sector salaries and social security benefits were ‘linked’ to private sector wage increases. Youth minimum wages increased sharply and were also indexed. Most collective agreements contained automatic price escalators. The result was that pay increases in the private sector automatically translated into higher public sector outlays. The escalators made the government hostage to the outcome of negotiations in the private sector and an interested party in wage bargaining that unavoidably became tripartite (Hemerijck 1995).

From 1960 to 1980 the public sector's claim on national income doubled. Between 1961 and 1971 the share of government outlays increased from 30 percent to 46 percent of GDP, compared to an OECD average jump from 30 percent to 37 percent (OECD 1974). The share rose further to 60 percent by the end of the decade. This expansion was almost exclusively caused by sharply higher transfer payments to households and, in later years, rising interest payments (Kam 1996: 263). The rapid growth of the public sector was only partly matched by higher taxes and social security charges. Price indexation via semi‐annual automatic cost‐of‐living adjustments in wages and benefits remained the main sticking point. An employer offensive to revise the system in 1977 was badly prepared. After a brief but massive strike, the unions claimed victory. As a result, indexation became the litmus test for solidarity between high‐ and low‐paid workers, and between those with and without jobs. It was hard to convince union members that they should give up this instrument at a time of high inflation, when indexation was most needed. Unlike their Austrian counterparts, Dutch unions were in the dark about pricing decisions for common consumer goods. The rational expectation game of achieving lower prices and wages by anticipating lower wages and prices could not be played. The penalty paid by the unions for their defense of indexation was government intervention and an ever smaller space for negotiated nominal wage increases (Visser 1990).

Indexation limits the room for maneuver in case of external shocks. International comparative data show that the short‐term and long‐term responsiveness of wage rates to changes in unemployment was much smaller in the Netherlands (and Belgium) than in Austria. Thus a much larger rise in unemployment was needed to induce wage moderation. Indexation tends to increase real wage stickiness. Indexation ‘tends to (p.213) make inflation less painful and socially disruptive, but faster and more difficult to brake’ (Braun 1976: 235). Since firms in the sheltered sector and government services do not have to fear a loss in market share, they can compensate higher wages by raising prices with impunity. Higher prices in the sheltered sector, much like rising non‐wage labor costs, end up stimulating wage increases throughout the economy, undermining the competitiveness of the exposed sector. Indexation is associated with a higher overall wage level, since the ‘power of less favored groups to insure that their wages rise more closely in line with other wages will probably not be offset by the reduction in the power of the most favored groups to secure above average wage increases’ (Braun 1976: 234). Dutch unions tried to negotiate lower increases for higher‐paid employees, but this backfired when senior white‐collar staff founded their own unions in the 1970s.

Natural gas turned out to be a mixed blessing, contributing ultimately to the infamous ‘Dutch disease’. The embargo aside, OPEC's oil‐price hike should not have caused major problems for a country that had its own energy resources. Export revenues and import savings from natural gas resulted in a current account surplus, which put upward pressure on the guilder. The Dutch Central Bank (DNB) pursued its hard‐currency policy with the utmost of zeal when it realized that the government was losing control over wage setting and public finances. The guilder's effective exchange rate increased by 30 percent between 1973 and 1977. Profits, real investment, net exports (excluding natural gas), and private sector employment all came under pressure. The inconsistencies in the Dutch policy mix grew to self‐defeating proportions (OECD 1980). In hindsight, one might have thought policy actors would have been aware that the combination of an appreciating currency, fiscal reflation, and wage (cost) growth was bound to worsen the stagflation dilemma. In 1976 some prominent Labor Party economists called for a return to the statutory incomes policy and stronger public sector expansion. But the unions rejected the proposal without so much as a thought. Employers did not want any additional state influence and stepped up their attack against what they called ‘anti‐business' policies. In an unprecedented open letter addressed to the prime minister in 1976, CEO's from the nine largest multinational corporations demanded a change of policy. A massive outflow of capital, from 2.3 bn to 7.5 bn guilders between 1972 and 1982 (Kurzer 1993), was widely seen as a vote of no confidence at the time.

In 1976 Finance Minister Duisenberg put on the brakes. Public expenditure was not to rise by more than 1 percent of net national income per annum (Zanden and Griffiths 1989). By moderating public sector growth the government hoped to make room for private sector investment. Den (p.214) Uyl resigned the next year over the issue of land ownership, then won the largest victory for his party in any general election, but failed to form a new governing coalition. Van Agt, the leader of the newly formed Christian Democratic Party (CDA), brokered a coalition with the conservative Liberals (VVD). Like its predecessor, this government tried to talk the trade unions into restraint, but it had even less to offer. Unable to bring public finances under control, the new government felt obliged (against its own intentions) to intervene in wage setting in 1979, 1980, and 1981. The government's wavering position was epitomized by internal conflict between two of the government's CDA ministers, for Social Affairs and Finance (Toirkens 1988; Hemerijck 1995; Snels 1997). Albeda, responsible for Social Affairs, stressed the importance of cooperating with the social partners in order to fight rising unemployment. Cuts in government expenditures and wage restraint were necessary but had to be made acceptable to the unions through compensating policies. Albeda rightly perceived a change in opinion within the trade union movement after 1978. Worried by the continued decline in manufacturing jobs and the collapse of shipbuilding, the major industrial union made a U‐turn and defended wage moderation in combination with working time reduction. The union won support from the Federation of Dutch Trade Unions (FNV), and late in 1979 there was even a draft central agreement at the Labor Foundation. But FNV leader Kok withdrew his signature at the last minute when affiliates opposed to the agreement threatened with revolt (Nobelen 1983). Following this failed accord, so carefully arranged by Albeda, his Finance Minister Andriessen wanted an extended wage freeze for two years. Since even VVD regarded this demand as outlandish, Andriessen decided to resign. The commitment to concertation had triumphed over the objective of fiscal restraint, but not for long. When the second oil crisis hit, the Social Affairs Minister had no choice but to impose another wage freeze.

5.3.2. The Crossroads at Wassenaar

The restrictive macroeconomic policies of the USA, UK, and Germany in the wake of the second oil shock made more problems for the Dutch. Higher interest rates exacerbated firms' financial liabilities, squeezed profits, and caused a spate of bankruptcies and factory closures. Unemployment soared at a rate of 10,000 to 15,000 per month, to a record 800,000 in 1984 (14% according to national statistics, later revised to 12% in accordance with ILO and OECD standards). Against the background of strong labor force growth, forecasts were extremely gloomy. Within five years trade unions lost 17 percent of their members, and of the remaining (p.215) membership nearly one‐quarter was out of work, on social benefits, or in retirement. The public deficit peaked at 6.6 percent in 1982.

Slowly but surely the relevant policy actors in the Dutch political economy came to understand the economic and social implications of the perverse spillover effects linking high real wages and indexation to low profits, unemployment, and fiscal crisis. In 1980 the Scientific Council for Government Policy issued a devastating report on the state of Dutch manufacturing and industrial policy, with a biting rebuke of the ‘waiting for corporatism’ attitude (WRR 1980). Alarmed by its conclusion, Van Agt installed an ad hoc commission, chaired by the president‐director of Royal Dutch Shell and members selected on personal merits (from the main industrial trade union, among others). Not wasting any time, this commission recommended lowering wage costs to boost profitability, introducing greater flexibility and decentralization into wage negotiations, boosting wage differentials, and disconnecting public sector wages from the private sector (Dellen 1984).

The general elections of 1981 had resulted in a patched‐up coalition between the Christian and Social Democrats, but the new administration immediately fell back into a deadlock and lasted only nine months. New elections in 1982 brought an austerity coalition of the CDA and VVD to power, thanks to gains for the Liberals. This ‘no‐nonsense’ coalition, led by Lubbers (CDA), committed itself to a three‐track strategy of (1) economic recovery through improved business profitability, lower wage costs, industrial restructuring, and less regulation; (2) reorganization of public finances and fiscal consolidation; and (3) cost‐neutral work‐sharing in order to alleviate the unemployment problem. Unveiling its plans on 22 November 1982, the new cabinet declared that ‘it is there to govern’ with or without the consent of the social partners. Two days later the leaders of the main union and employers' confederation, Kok and Van Veen, announced that they had struck a deal. They had started preparing this deal in the Labor Foundation over the summer, but now the social partners were in a hurry to prevent another intervention. The government had already announced that it would suspend indexation in 1983, and its Minister of Social Affairs, De Koning, bluffed that he might have to impose mandatory work‐sharing if voluntary measures were not forthcoming.

With unemployment running into the double digits, 14 percent of the working age population receiving disability or early retirement benefits, and 6 percent on sickness leave, the trade union movement was in no position to fight over wages. Employers saw a chance to rid themselves from recurrent state intervention (van Bottenberg 1995). The Wassenaar agreement was only a recommendation, but one that carried authority. The (p.216) opposition to the agreement—from the food and services union, through Philips, to the metal employers' federation—was defeated and isolated. To help investment and employment, negotiators in sectors and firms were advised to forsake price indexation and use the savings for a cost‐neutral reduction of working hours. With this agreement Dutch unions acknowledged that higher profits were required for the higher level of investment essential to job growth (Visser and Hemerijck 1997). Like the government, employers recognized that working‐time reduction was a small price to be paid for this concession. The response to the Wassenaar agreement was swift. Although negotiations over shorter hours proved cumbersome, in less than a year two‐thirds of all collective agreements were renewed, mostly for two years, during which the payment of price compensation was suspended and a 5 percent reduction of working hours took place. By 1985 cost‐of‐living clauses had virtually disappeared; average real wages fell by 9 percent in real terms (Visser 1990). The adjusted share of income from dependent employment, which had risen from an average 69.9 percent in the 1960s to 74.2 percent in the 1970s, fell to 67.6 percent in the course of a few years.

Assured of restraint, the government had its hands free to regain control over public sector finance. Between 1983 and 1986, a 3 percent decline in the public sector deficit was achieved through a small drop in employment (−1%) and salaries (−3%). In 1983 salaries, minimum wages, and related benefits were frozen, and for 1984 the government cut public sector wages by 3 percent. This pushed the public sector unions to the barricades, but after a strike of three weeks they found themselves isolated. Formally, the government had ended the mechanism that had assured public employees' salary rises in line with private sector wages since 1962. Henceforth, unions and employers could no longer shift the costs of stagflation onto the government, but were forced instead to internalize the external effects of wage inflation and unemployment (CPB 1997: 165). Private sector unions could no longer be held hostage by public sector unions (and benefit recipients). Years later, the unions would accept another ‘binding’ principle, tying their wage demands to increased labor force participation (see below).

Social Affairs Minister De Koning took care of the human face of ‘no‐nonsense’ government. Special measures were taken for the poor and families living on just one income at the social minimum. As the most influential CDA power broker after Lubbers, De Koning had convinced the prime minister, against the advice of the DNB and the Finance Ministry, not to follow the Germans in the EMS currency realignment of 1983. The guilder was devalued by 2 percent against the Deutschmark—a decision that was soon regretted when financial markets demanded an (p.217) extra risk premium. In March 1983 Dutch monetary authorities announced that henceforth the gulden would be pegged to the Deutschmark. There would no longer be an independent exchange rate policy.

Helped by the international economic upswing, investments and jobs benefited from the recovery of profits in the second half of the decade. The reduction of working hours had an additional work‐sharing effect, but the largest boost to jobs came from the rapid expansion of part‐time jobs, mainly for married women and young people (Visser 1999). The initial growth in part‐time jobs during the early and mid‐1980s did not result from policy changes, but rather from behavioral changes by women and employers. In the early 1980s, employers presented part‐time work as the flexible and individual alternative to the campaign for a 36‐hour work week employers were bent on resisting. Dutch mothers once used to withdraw from the labor market upon childbirth, but this habit was rapidly changing under the impact of better education, higher female wages, smaller families, and changing social norms. Faced with high risks of unemployment for themselves or for their husbands and partners, withdrawal from the labor market was becoming more costly and risky. More women decided to stay employed and ask for reduced working hours instead of quitting their jobs. In the absence of childcare facilities, part‐time work became the dominant ‘coping’ strategy for working mothers. Employers in sectors and firms with overcapacity saw part‐time work as a welcome form of voluntary layoffs with little cost to themselves. In health, education, and social services, but also in central and local government, part‐time employment was the least painful means of adjusting to budget restrictions. Hiring young people was often limited to 32‐hour entry jobs.

Lubbers's austerity policy paid off politically. In 1986 the center‐right coalition was re‐elected with gains for his CDA party. Two governments of the same persuasion meant a clear break with the immobilism of the 1970s. Inflation declined to near zero in the mid‐1980s and has disappeared from the country's worries, until recently. The strict exchange rate policy exerted discipline on wage developments, while wage moderation, in turn, enabled the DNB to stick credibly to its non‐inflationary policy. Low inflation allowed unions to forget about indexation. The new mix of macroeconomic policy and wage setting also changed the institutional relations among unions, employers, and the state (Visser 1998). The new pattern became a central dialogue about a wide range of policy issues combined with sectoral wage bargaining, based on the primacy of industrial self‐regulation. The role of the central organizations was confined to redirecting sectoral contracting towards tacit, economywide wage restraint and public‐regarding behavior (Toren 1996). These new principles have been recognized politically. The 1970 Wage Act was revised in 1986, and (p.218) the possibility of intervention was narrowed to situations of extreme economic distress.

Since 1982 there has been no political intervention in wage setting. This is not to say the state has lost all leverage over Dutch wage setting. There is no unlimited Tarifautonomie, not even under the revised Wage Act (which has not been tested yet). Moreover, a 1937 law authorizes the Social Affairs Minister to declare collective bargaining agreements legally binding for all workers and employers in a certain branch of industry. This provision was routinely applied till the early 1990s but has since been used as a bargaining chip to obtain certain ‘public‐regarding’ objectives from wage negotiators, especially lower entry wages just above the statutory minimum wage for people with low skills, little experience, or a history of unemployment. The government, no longer dependent on the social partners for its own finances, is now in a stronger, more independent position from which it has been able to drive unions and employers closer together.

5.3.3. Trapped in a Spiral of ‘Welfare Without Work’

The second Lubbers administration promised to reduce the deficit even further, stop increasing taxes and social premiums, and lower unemployment to under 500,000 people. The new policy challenge was to curtail the costs of social security and reduce the number of welfare recipients, which had continued to rise throughout the 1980s. Cost reduction did take place, mainly through a 1983 freeze (in nominal terms) on the minimum wage and all related benefits, followed by a 3 percent cut in 1984, and the delinking of benefits from wages. The gap between employed workers' average earnings and the minimum wage increased by 12 percent between 1983 and 1989. Since the minimum wage is the basis for calculating state‐guaranteed minimum levels for all benefits, social assistance, and basic old age pensions, the freeze reduced the state's liabilities. Youth minimum wages were reduced even more drastically (Salverda 1997). At the same time, fewer adult workers remained at the minimum—so that the proportion of adult workers earning the minimum wage went down from 10 percent to 3 percent between 1983 and 1995.

Under the growing competitive pressures of the 1980s, firms in the high‐wage Dutch economy could only survive if they were able to increase labor productivity. Many embarked on a strategy of channeling less productive and more expensive (mostly older) workers out of the labor market into unemployment or onto disability. Here there was a clear moral hazard, since the insurance costs were borne by all firms and (in case funds should run out) by the taxpayer. Rising insurance premiums resulted in higher social security contributions, while the wedge between total (wage and (p.219) non‐wage) labor costs and take‐home pay increased. The result was a vicious cycle of job destruction, layoffs caused by firms seeking to match rising costs through increased productivity, leading to higher costs, and so forth. The high average and marginal wedge created clear disincentives for workers to seek a job and threatened the continuation of wage restraint.

It became increasingly obvious that firms, with the complicity of social security administrators, were reducing slack while externalizing costs, leading to an ever‐growing volume of claimants. The so‐called ‘system reforms’ in social security, implemented in 1987, had tightened eligibility and lowered guaranteed benefits from 80 percent to 70 percent of last‐earned wages in case of unemployment, disability, and sickness, but the reforms had not altered the bipartite governance structure. After the reforms, Lubbers had promised ‘peace on the social security front’, but it was not to be. The contest over the rise in disability claimants could no longer be avoided and came to a head in 1991. When it came, Lubbers was fortunate enough to have traded his Liberal coalition partners for the Social Democrats, after a break in the old coalition (over environmental policy) and new elections in 1989.

Already in 1986, workers in the age group 55–64 who received a disability benefit outnumbered those with a job (Aarts and de Jong 1992). These authors list four features that may explain the Dutch disability crisis (Aarts and de Jong 1996). First, disability risks were defined as social rather than occupational, and the scheme did not make a distinction between different causes of disability. Second, until it was corrected in 1987 entitlement was based on an assessment of a worker's particular incapacity to find a job similar to his former job. Third, administration for the sickness and disability schemes was kept separate; workers could first spend a year on sickness leave, virtually unsupervised, and than move on to disability. Fourth, under collective agreements, most firms topped up sickness benefits to 100 percent of previous earnings and supplemented disability benefits to 100 percent for a year or longer. Workers drawing full‐time disability pensions were not required to be available for hire on the labor market.

Legislation regulating dismissals under the 1945 Extraordinary Decree, originally meant to address postwar chaos but still on the statute books four decades later, was another factor. The Decree had instituted preventive controls and required employers who wanted to end an employment contract to ask permission from the regional labor office director. Procedures could drag on for months, with uncertain outcomes, and no permission could be legally granted when workers were sick or pregnant. The route through sickness and disability became the path of least resistance for many employers. Finally, administration and supervision of the (p.220) system was in the hands of the social partners through so‐called Industry Boards. These Boards were responsible for the medical service that assessed the degree of disability of claimants. As it happened, there was an extraordinary lack of transparency in these assessments. All these factors conspired to bring about a seemingly unstoppable rise in the number of claimants. The Disability Program was expected to support no more than 200,000 people when it was inaugurated in 1967, but was coping with almost 900,000 beneficiaries by 1990. Estimates at the time suggested that between 30 percent and 50 percent of all recipients really were full‐time or part‐time unemployed. The scheme also was a harsh welfare trap: once officially recognized as (partially) disabled, workers acquired a permanent labor market handicap, often combined with social isolation.

5.3.4. The Political Risks of Welfare Reform

The third Lubbers administration (1989–94), now including a PvdA led by ex‐union leader and Wassenaar negotiator Kok as Finance Minister, began in an optimistic mood. Union decline had stopped and unions were regaining confidence, demonstrated by a series of strikes in industry, transport, education, and health services. For the first time in a decade, real wages increased (Visser 1998). Economic recovery allowed for a partial restoration of the suspended ‘index linking’ that tied the minimum wage and benefits to wage developments. This had been one of PvdA's demands in its coalition negotiations with the CDA. Top civil servants at the Ministry of Social Affairs designed a new system of linking, which became operative in 1992. Index linking of minimum wages and benefits would henceforth depend on the volume of benefit claims in relation to employment measured by the so‐called I/A (inactive/active) ratio. If this ratio exceeded a given reference level, fixed at its 1990 value (82.8), the government would be entitled to suspend linking, fully or partially. In 1990, 1991, and 1992 linking was fully applied, but in 1993, 1994, and the first half of 1995 linking was suspended, since the I/A ratio had exceeded its reference value. In recent years the I/A ratio has rapidly fallen and linking has been applied every year since 1995. The new conditional linking mechanism gave the unions, whose (aging) members care much about linking of benefits to wages, a direct incentive to take account of the employment effects of wage bargaining.

Soon after the formation of the Lubbers–Kok cabinet in 1989, an unusual tripartite Common Policy Framework promising growth, employment, tax relief, and wage moderation was concluded between the government and the social partners. Employers signed an agreement with the unions to create 60,000 additional jobs for ethnic minorities, among (p.221) whom the unemployment rate is three times the national average. This agreement was not implemented, however, and local employers and unions hardly seemed aware of it (Heertum‐Lemmen and Wilthagen 1996). In 1994 the agreement was renewed, but employers made sure that no precise target was mentioned. Early in 1991 members of the central employer federation withdrew their signature from the tripartite Policy Framework and stepped up their campaign for tax relief, lower non‐wage labor costs, and a cap on government spending. To drive home their point, employers decided to stay away from the customary spring meeting of the Labor Foundation with core Cabinet ministers. They preferred one‐to‐one exchanges with the unions or with the government, rather than tricky tripartism.

At around this time German unification boosted economic growth, but soon high interest rates dampened economic activity. Although the downturn of 1992–93 was not as bad as in other countries, the decline in unemployment was immediately halted and reversed. The government was preoccupied with the lack of a quick response from wage negotiators and prepared to intervene on the basis of the 1986 law. The pressure worked, and early in 1993 the Labor Foundation recommended a two‐month ‘breathing space’ in order to absorb the new facts of the international recession. Later in the year, after lengthy negotiations, a new agreement called the ‘New Course’ was reached, a true follow‐up of Wassenaar, with more organized decentralization and negotiated flexibility (Visser 1998). Employers gave up their blanket resistance against shorter working hours. Both parties agreed to work together to improve the unfavorable employment/population ratio and recommended flexible working time patterns and part‐time work as a solution. There was a separate agreement to abolish residual differences in rights and fringe benefits between full‐time and part‐time workers. This was backed up in 1994 by legislation forbidding all discrimination on the basis of working hours. In the course of the third Lubbers administration, the exchange logic behind organized wage restraint changed character. Increasingly, wage moderation was matched by lower taxes for workers and lower social contributions for employers, made possible by improved public finances and a broader tax base. This was supported by the tax reform of 1990, which integrated taxes and social security charges, lowered the overall and highest rate, while limiting deductions. Since the national social security contributions were integrated, the first band (37.3%) is mainly (32.3%) a social security charge, while the second (50%) and third (60%) bands are exclusively income taxes. (Payment for each of the three kinds of employee insurance—disability, sickness, and unemployment—is collected collaterally, based on a fifty‐fifty division between employees and employers.) By absorbing social (p.222) security into the lowest band and eliminating deductibility for social insurance contributions (the single largest item before 1990), the tax base was broadened by almost two‐thirds. The motive behind the 1990 tax reform, adopted in order to bring about a cost‐neutral rationalization of the Dutch tax system, was largely administrative. However, since the reform was sugared with a tax break for most households, consumer spending surged by 4 percent in real terms, while private savings also picked up (Kam 1996). As a consequence, the tax reform had unintended but highly favorable employment effects.

Lubbers I and II had exhausted the ‘price’ policy of bringing social expenditures under control by freezing and lowering benefits. The PvdA opposed any further measures of this kind. The emphasis therefore shifted to a so‐called ‘volume’ policy aimed at reducing the number of social security recipients. Unavoidably, the political crisis of the Dutch welfare state came to revolve around disability pensions. The number of people receiving disability benefits seemed to approach the magic number of one million in a working age population of seven million. In a well‐played cri de coeur, in September 1990, Lubbers shocked viewers of the TV evening news by saying that the country was ‘sick’ and required ‘tough medication’. The next summer, after long agony, the government decided to restructure the sickness and disability programs. In response, the unions organized the largest postwar protest march in The Hague. This episode had far‐reaching political consequences. The PvdA went through a very deep crisis and Kok, its leader, nearly resigned. Notwithstanding popular resistance, the reforms were enacted. They included a reduction of replacement rates for all workers under 50, including those who were already disabled (and could not take extra insurance). After some time, benefits would decrease to 70 percent of the statutory minimum wage plus an additional age‐related allowance. Anyone under the age of 50 in the system would furthermore be subjected to a new medical examination based on stricter rules. In 1993 re‐examination affected 43,000 people under the age of 35, of whom 30 percent lost their benefits and 18 percent saw their benefits reduced. (Under pressure from Parliament, the regime has recently softened again.) The legal requirement for partially disabled workers to accept alternative employment was tightened. Under the new sickness law of 1994, the first six weeks in which a worker on sick leave continues to receive wages were charged directly to the employer (later this was extended to the first year). In response to lower legal disability benefits, the unions negotiated extra‐legal supplementary benefits in their collective agreements for 1995 and 1996, at the expense of wage increases. As a result, the costs of sickness and disability have been internalized by the bargaining parties to a greater extent.

(p.223) The 1994 election was a vote of popular discontent. The Lubbers–Kok coalition was effectively voted out of power, losing 32 of its 103 seats in Parliament, just four short of a majority. This amounted to a political earthquake (Koole 1995): no single Dutch coalition had ever lost as much support in one election. The CDA's extraordinary decline (from 54 seats to 34 seats) was caused by a leadership crisis and the untimely publication of a party document suggesting that future retrenchment might include a freeze on the basic Dutch pension scheme. Social democratic voters punished the PvdA for its part in the attack on hard‐won social rights. Ironically, however, the PvdA became the largest party after losing 12 of its 49 seats. A restored PvdA–CDA coalition was now only possible with the help of the progressive Liberals (Democrats 66). This party convinced historical enemies, the PvdA and VVD, to form the first coalition government since 1917 without a confessional party. The new ‘purple’ coalition did not slow down the reform effort, but the PvdA had a bottom line condition for its cooperation: the level and duration of social benefits would not be tampered with. Kok had conceded that this position might have to be reviewed after two years if developments should turn out unfavorably regarding public sector deficits, inactivity on the labor market, and employment. From this very defensive position, the party was entirely committed to the ‘job, jobs, and more jobs’ approach as the only way out. This explains its support for reforms aiming at improvements in efficiency by introducing financial incentives through partial re‐privatization of social risks and managed liberalization of social policy administration.

The common thread running through many legislative changes in social security system is placing greater financial responsibility on employers and employees. This applies to privatization of sickness benefits, differentiation in invalidity benefit premiums, and lengthening the number of reference days required before workers receive a benefit under the new Unemployment Benefit Act. Under the new Disability Act of 1997, employers can ‘opt out’ and take full responsibility for the first five years of occupational disability, while legally obliged to pay at least 70 percent of the disabled employee's most recent earnings. If they remain in the public system, they have to bear the risk of rising costs if the incidence of disability in the industry rises. This reform allows differentiation among employers' contributions and encourages firms to invest in health and safety at work. Employers can re‐insure risks, but may have to pay higher premiums if rates are higher than average.

The welfare reforms of the 1990s were helped, politically, by the results of some alarming inquiries into the causes of the social security crisis. A Parliamentary Inquiry in 1993 revealed what was already common (p.224) knowledge, namely that the social partners had made ‘very liberal use’ of social security for purposes of industrial restructuring. The inquiry established that the Industrial Boards had no intrinsic interest in getting people off welfare and that the social partners had all along known that the system was being abused. The government limited the self‐governing autonomy of the social partners and set up an independent Supervisory Board. A separate National Institute for Social Insurance was set up with responsibility for contracting out social security's administration to privatized delivery agencies. Trade unions and employers' associations have retained an advisory status only. A final string of measures concentrated on introducing or intensifying ‘activation’ obligations for the long‐term unemployed, fortified by penalties since 1996. For young people, entitlement to benefits was replaced by the entitlement to a job under the Youth Work Guarantee scheme. Single parents with children older than 5 (no longer 12, as before) must be prepared to take a part‐time job. The standard for ‘appropriate’ work was broadened, and benefit recipients are now expected to accept jobs lower than their former job in terms of earnings. Municipal social services are required to draw up so‐called reintegration plans and training programs, in collaboration with regional employment services. In line with the Luxembourg employment guidelines, the government is spending an extra 500 m guilders for a so‐called ‘integral approach’ in which every unemployed worker will be offered a job or an education within the first year of unemployment.

Unlike Austrians and Belgians, Dutch citizens are entitled to a flat basic public pension when they reach the age of 65. In the face of a rapidly greying population, which is expected to reach a peak round 2020–30, nearly all parties have agreed to put aside extra tax revenue from the current successful growth spurt for a special fund to meet future obligations in this universal tier of the pension system. With respect to the obligatory, occupational‐based and earnings‐related, second tier of Dutch pensions, unions and employers are quietly negotiating a change from a final to a middle wage system, and the replacement of collective early retirement schemes (financed out of the wage sum and adding up to 8% in some industries) with a funded time‐saving system. Finally, with regard to the private non‐compulsory tier, there is a move towards restricting the tax deductibility of private equity plans because of how they narrow the tax base. Owing to its three‐tiered make‐up, the Dutch pension system appears better adapted to flexible, interrupted, or part‐time working careers than most continental European systems (Alber 1998).

(p.225) 5.3.5. From Fighting Unemployment to Increasing Participation

In the late 1980s, policymakers became aware that the low level of labor market participation was the Achilles' heel of the extensive but passive Dutch system of social protection. In 1990 the Netherlands' Scientific Council for Government Policy proposed breaking with the past and advocated a policy of maximizing the rate of labor market participation as the single most important labor market policy goal of any sustainable welfare state (WRR 1990). Gradually the message, though not the policy recommendation to lower the statutory minimum wage, was adopted by the government. In 1993 the social partners followed suit by embracing higher levels of participation in the New Course agreement (Visser and Hemerijck 1997).

The new policy priority made its imprint on all kinds of policy initiatives, beginning with using the I/A ratio as the basis for linking benefits and wages. There was a new focus on active labor market policies, an underdeveloped area in postwar policymaking. The public employment service was reorganized. It had long been run in a highly bureaucratic fashion from the Ministry, held a monopoly on job placement, was avoided by employers, and was swamped by unemployed people with no chances on the open labor market. It was estimated that one‐third of those registered as unemployed were not available for work or were already working. Around 1980, the idea of involving the social partners was placed on the policy agenda. Minister De Koning believed that a tripartite organization would condemn the social partners and the government to mutual cooperation in the fight against unemployment. After lengthy preparations, the 1991 reform brought the service under tripartite control, financed by the government but run independently. There was also a strong element of devolution of responsibility to regional employment boards. A critical evaluation after four years led to another overhaul in 1996 and provided a policy window for issue linkage between labor market and welfare policy. As a result of strengthening ‘activating’ measures in unemployment insurance and social assistance, municipalities and social insurance organizations have allocated a budget to buy placement and training activities from the employment service. Beginning in 2000 the employment service is expected to compete with private providers like the temporary work agencies. Cooperation between social security organizations, the municipal assistance offices, and the employment service is further enhanced by the 1998 Mobilization of Unemployed Persons Act. As new Centers for Work and Income, they are expected to work together in one location. New government plans are bringing all the various kinds of insurance into a single public service, whereas job placement services will (p.226) be privatized, possibly within minimum requirements established by collective agreement. Unions, supported by employers, oppose these plans, which would further reduce the social partners' involvement, but the Lib–Lab coalition appears determined to have its way.

Beginning in the early 1990s, the Lubbers–Kok administration and the two purple coalitions government had started to redress the imbalance between active and passive policies. Under the Youth Guarantee Plan and with the so‐called ‘Melkert‐jobs’ (named after the PvdA Minister of Social Affairs and Employment), permanent jobs in the public sector for unemployed youth and the long‐term unemployed were created. These jobs are usually for 32 hours a week at a maximum of 120 percent of the hourly minimum wage. Additional job programs based on these schemes have been expanded and now absorb around 1.5 percent of total employment. The majority of these programs are carried out by the municipalities. With the introduction of the Jobseekers Integration Act, local authorities have been granted more resources to target projects for the long‐term unemployed. In addition to job creation in the public sector, the government has introduced several kinds of employment subsidy schemes based on reducing the social security contributions paid by employers. Subsidy schemes cover as many as one million workers. The first initiatives were directed at the long‐term unemployed, women, ethnic minorities, and low‐skilled groups. Because of a disappointing take‐up, more recent schemes cover all jobs with an hourly wage of up to 115 percent of the minimum wage. The majority of these jobs go to new entrants. Employment subsidies can add up to 25 percent of the annual wage. First evaluations of these schemes suggest that they accelerate employment creation and the reintegration of unemployed workers (SZW 1998; NEI 1999).

Labor market flexibility is an integral part of the new Dutch labor market policies. Legislation removing constraints on shop hours, business licenses, temporary job agencies, working time, and on dismissals consolidate this development. Following the ‘New Course’ agreement of 1993, unions and employers in decentralized bargaining rounds have concentrated on exchanging shorter working hours, more leave arrangements, and income stability throughout the year for the annualization of working hours, more opportunities for working evenings or on Saturday, and lower overtime rates (Tijdens 1998). The new Working Time Act of 1996, which allowed negotiated contracts to deviate from legally defined maximum working hours or minimum rest periods, supports this trend. In 1993 the Labor Foundation recommended that employers grant workers' requests to work part‐time unless there are compelling business reasons for rejection. A right to part‐time work is currently under parliamentary review. Temporary work agencies provided employment for over 200,000 person (p.227) years in 1997. Almost half of the workers in a flexible job are under the age of 25; among adult workers the share of tenured contracts has been stable at 90 percent during the past ten years. In 1995 unions and employers signed the first collective agreement for temp workers, introducing a right of continued employment and pension insurance after four consecutive contracts or 24 months of service. It prepared the ground for the central agreement on ‘Flexibility and Security’ of 1996, which, in turn, paved the way for an overhaul in 1999 of Dutch law on protection against dismissals. This ‘flexicurity’ law (Wilthagen 1998) is a compromise, not just between employers and employees, but also within the unions between workers with and without stable jobs. A relaxation of statutory dismissal protection for regular employment contracts (more opportunities for negotiated termination of employment) is exchanged for an improvement in the rights of temporary workers and the introduction of a ‘presumption of an employment relation’ in the case of freelance work. Temporary employment agencies no longer need a license, but the law assumes their responsibility as employer.

The incremental individualization of the tax system since 1984, improved opportunities for switching from full‐time to part‐time jobs, and the removal of all remaining elements of discrimination based on working hours have all contributed to a ‘normalization’ of part‐time employment (Visser 1999). The differences in hourly wages between full‐time and part‐time workers is 7 percent in the private sector, but much smaller in the public sector where almost half of all part‐timers work. Dutch social security legislation is especially friendly to part‐time employment, since the main principle for coverage (for health insurance too) is the employment contract, irrespective of hours worked (SZW 1995). The current government has introduced a Framework Bill on Employment and Care into Parliament, which is intended to harmonize different forms of leave and create a framework to save time or money for leave. Employers are lobbying hard to defeat this bill; their alternative is a proposal to have the unions sign a Framework Agreement containing greater sectoral flexibility. A tax reform, coined the ‘Tax Plan of the 21st Century’, was proposed in September 1999. It includes lower tax rates on labor and capital, financed by the broadening and ‘greening’ of the tax base. Capital asset taxes will be replaced by a fixed capital revenue tax. VAT will be raised, with an exemption for labor‐intensive services, for which the VAT rate will be lowered. The Central Planning Bureau expects the new tax system to boost job growth considerably so long as there is wage restraint. But unions deny any relationship and have already twice stepped up their demands for 2000, clearly in response to pressure from a tight labor market. Help on the tax side is still needed to get more low‐skilled workers employed. But (p.228) one might question whether it would not be wiser, under the current circumstances of a balanced budget and full employment, to increase public investment in education, technology, infrastructure, innovation, and stress prevention, rather than heaping more money on house owners and consumers.

5.3.6. The Record: Wage Restraint, Flexibility, Part‐Time Jobs, Services, and Welfare Reform

Since 1982 a new mix of macroeconomic policy and wage setting emerged that generated a virtuous cycle of price stability, fiscal consolidation, restored profitability, and strong (part‐time) job creation in private services. Subsequently supported by welfare and labor market reforms, this virtuous cycle has had important positive feedback effects on female employment and domestic demand, leading in turn to a slow but solid decline in labor market inactivity without having to sacrifice basic social security. According to the Central Planning Bureau, wage moderation has been the single most important weapon in the Dutch adjustment strategy. CPB (1997) estimates that two‐thirds of job growth between 1983 to 1996 should be attributed to wage moderation. Unit wage costs in 1996 were at the same level as they were in 1981. Over the same period they increased by 40 percent in Germany and 15 percent on average in the EU. The return to wage moderation contributed to job‐intensive growth in three ways. First, by restoring profitability, it created a necessary condition for investment and job growth. Second, by lowering the external exchange rate, it supported export growth and employment in the exposed sectors of the economy. Third, wage moderation kept more people on the payroll and, in combination with lower taxes, had a favorable effect on employment in domestic services. As a corollary of the shift to services, growth in labor productivity per hour, although already very high by European and American standards, was smaller than in other countries.

Sectoral employment trends in the Netherlands went in almost opposite directions before and after 1984. In the 1970s the subsidized sector (health, education) and government employment witnessed the strongest growth, whereas private sector employment stagnated. Employment levels in industry fell steeply and employment growth in private services (trade and transport, financial and personnel services, hotels and restaurants) was faint. After 1984, job decline in industry and agriculture stopped. Employment growth is now strongest in private services, 3 percent per year, four times as much as in the public and subsidized sector. The public sector absorbs only 13 percent of total employment, which is low in comparison with most European countries. The reason is that health services (p.229) and social welfare in the Netherlands are often organized by formerly pillarized voluntary associations with support from public subsidies. Together with these semi‐public activities, the public absorbs 25 percent of total employment.

Next to wage restraint, labor time reduction and higher labor time flexibility have played a key role in job growth. Over time policy preferences shifted from across‐the‐board reductions in the working week towards the enhancement of part‐time work and annualization of working time. Two‐thirds of the jobs created since 1982 have been part‐time jobs. The surge in part‐time employment and the shift to services coincided with the rapid increase in labor force participation of women, from 29 percent to 60 percent between 1971 and 1996, the strongest rise in any OECD country. While the labor force increased by one‐fifth between 1970 and 1997, the female labor force more than doubled (SCP 1998).

By taking wage moderation, the proliferation of part‐time work, the shift to services, and increased female participation together, the following picture emerges (Hartog 1999). Between 1987 and 1996, value added in the total economy grew by 31 percent. However, employment volume (in hours) declined by 7 percent in the exposed sector and increased by 33 percent in the sheltered sector. Because private consumption is more service‐intensive, and services are more employment‐intensive, job growth really took off in the service sector. Private consumption has made a major contribution to economic growth and was boosted by a ‘feel good’ factor related to the labor market's steady improvement and rising housing prices (with most double income families owning their house). Strong domestic growth has allowed the Netherlands to be relatively unaffected by the recession in Germany. The positive interaction effects among wage moderation, part‐time work, the shift to services, and increasing female participation have been supported by social security reform and innovations in labor market policy since the early 1990s. Transfers to households in GDP terms came down from 31 percent to 26 percent, and government expenditure's share of GDP declined from 62 percent to 52 percent from 1983 to 1998. Relative to the size of the labor force, there are fewer people on disability and sickness benefits. Today's discussions center on the wisdom of further privatization and on incentives for employers to invest in preventing sickness and reintegrating inactive workers. The reform of the employment service, the strengthening of activation requirements in social security, additional job programs in the public sector, wage subsidies to encourage employers hiring low‐skilled workers, tax reduction for those in work, negotiated flexibility in working hours, statutes and pensions, and policies to enhance the growth of part‐time work—are all components of the Dutch job miracle. However remarkable that miracle is, it should be (p.230) emphasized that inactivity, especially in the form of disability, is still high and that the present success story is no more than a game of catch‐up from a truly dismal situation in the recent past.

5.4. Belgium's sur place

The international monetary crisis, the oil‐price shock, and the international recession of the 1970s had a powerfully disturbing influence on the Belgian economy. During the recession unemployment doubled from 100,000 in mid‐1974 to 228,000 two years later. By the end of the decade Belgium had the highest unemployment rate and the fastest increase of unemployment in Europe with the exception of Spain. Employment contracted by almost 6 percent between 1974 and 1983. In the private sector (not counting subsidized employment programs and the self‐employed) the decline was 15 percent. In manufacturing employment declined by 29.8 percent and, taking into account the fact that average hours were reduced by 11 percent, we find that total labor input in manufacturing fell by 37.5 percent, a sharper contraction than in any other European country in this period.

Sneessens and Drèze (1986), from whom these figures are taken, present a broader picture of the Belgian economy using time series data from 1962 to 1983 showing national income shares for labor, unemployment, inflation, balance of payments, and budget deficits. The striking feature of these data is the sharp break after 1973, indicating a deep transformation and severe imbalance in the Belgian economy. During the 1960s the Belgian economy had performed as well as its neighbors. In the early 1970s inflation and wage growth was above the trend in Germany or the Netherlands, unemployment was slightly higher, the overall employment/ population rate was among the lowest in the OECD (as in the Netherlands), and the debt–output ratio was near the 60 percent limit that would become the EMU norm thirty years later. Belgium had found it difficult to adjust to the loss of its African colony in 1960. Unlike the Netherlands, there was no compensation from gas resources. Thus, Belgium had less leeway when confronted with the new realities of the 1970s.

Comparing the period 1968–73 with 1973–80, we find that GDP growth was halved, export growth fell from 11 percent to 3 percent per annum, and inflation rose to 8%, whereas unemployment increased from 2.4 percent in 1973 to 7.9 percent in 1980. Public sector debt, fueled by costly settlements of Belgium's linguistic conflicts and by steeply rising deficits in the social security funds, reached alarming proportions by the end of the (p.231) decade and would necessitate perennial austerity in the next. In 1981 the public deficit stood at −12.7 percent of GDP, which was the record in the OECD. Debt servicing alone had risen to almost 20 percent of government revenue. Deficit spending, which had helped to boost domestic demand in excess of supply in the 1970s, had as its counterpart an external deficit equivalent to −12 percent of GDP on a cumulative basis in 1981. This had to be financed by increased public borrowing, crowding out private investment. At the end of 1981 net external liabilities were larger than the foreign exchange reserves of the Belgian National Bank (BNB). In that year the economy contracted by more than a full percentage point, and unemployment, already high by the end of the 1970s, rose to 500,000 (10.2%), or 700,000 (16%) if people in various early retirement schemes are included. In sum, when the second oil crisis hit Belgium, its economic imbalances ‘were more marked than in the majority of member countries and were tending to become self‐perpetuating’ (OECD 1983: 2).

5.4.1. Contradictory Responses: Hard Currency and Keynesianism

When the Bretton‐Woods system collapsed, Belgium joined the ‘snake’. It stayed in till the very end, despite the defection of France and other participants of the first hour. It did so for broadly the same reason as the Netherlands: with wage negotiators unwilling or unable to control cost‐push inflation, a hard currency policy was Belgium's main protection against imported inflation. As explained in the case of the Netherlands, this came at the tremendous cost of lost competitiveness in foreign markets. Jobs in the exposed sector were disappearing at an alarming rate. The index of unit labor costs in manufacturing, relative to the OECD average, shows that by 1975 Belgium had slipped to one of the worst positions. The so‐called ‘wage gap’, or difference in wage costs between Belgium and its five main trading partners (Germany, France, the Netherlands, Italy, and the UK), widened from 100 to 113 index points (1970 = 100) between 1972 and 1975, and then to 129 points by 1979 (Goubert and Heylen 1999).

Exchange rate stability had long been a priority in Belgian monetary policy and was not contested by the unions. The experience of the 1949 devaluation had taught them that, in a small economy where most goods are imported, devaluation hurts workers' purchasing power (Kurzer 1993). Difficulties in obtaining voluntary wage restraint posed a dilemma for policymakers, and all governments after 1974 were caught between the brake of a hard currency policy and the engine of a reflationary course for which there were insufficient resources (Jones 1995). The immediate government response to the oil crisis and the recession of 1974–75 had been to rescue troubled industries with subsidies, job creation in the public (p.232) sector, and demand‐stimulating measures. Another response consisted of various measures aimed at reducing the labor supply: a recruitment ban on foreign workers, which was largely ignored; early retirement of older workers; and an additional year of education for young people. Policy did not change much in 1974 when the Christian Democrats shifted coalition partners, from left to right, or when they welcomed back the Social Democrats in 1977. All governments till 1982 were wedded to a Keynesian approach.

If civil servants for both central and local government are counted together with employees in the nationalized industries, public employment increased by 35.5 percent between 1970 and 1984; there was a spurt of 123,000 jobs in the direct aftermath of the 1974–75 recession and another expansion in 1978–79. The share of public jobs in total employment rose from 23 percent in 1970 to 32 percent in 1982 (OECD 1986). It was not enough. The recession accelerated the decline in the very industries that had been the mainstays of employment: steel, metal processing, chemicals, cars, and transport equipment. Until 1975 heavy industry had absorbed redundant workers from labor‐intensive industries such as textiles, clothing, and leather, as well as coal mining, but this was no longer the case. Overall, the employment ratio in manufacturing fell by more than four percentage points between 1973 and 1980, from 18.6 percent to14.3 percent. This contraction was twice as strong as in the Netherlands, not to mention Austria where industrial labor was hoarded.

The BNB estimated that total aid to business in the form of subsidies, capital transfers, loans, and government equity investment, averaged 5.5 percent of GDP per year in the second half of the 1970s, peaking at 8.9 percent in 1982. This equaled 13 percent of government expenditure at the time and was unsustainable. In later years it would fall to around 7 percent (OECD 1986). The largest items in the budget were subsidies to traditional sectors and large firms, which compensated for the hard currency policy but hampered innovation at the same time as they slowed down job losses. In the 1980s attempts were made to target government aid to start‐up firms and innovative projects in new technology. This coincided with the devolution of industrial policy to the regions, as part of the institutional reforms of 1980. However, the so‐called ‘national industries' (steel, coal, ceramics, and glass, mainly located in Wallonia, and textiles and shipbuilding, mainly in Flanders) remained under the responsibility of the central government.

Before 1973 Belgium's main difficulties had been related to its outdated industrial structure and the shifting economic balance between the two regions. As the first industrial country on the Continent, the country had inherited a large manufacturing base, concentrated in Wallonia. The (p.233) decline in the demand for steel and coal began to hurt Wallonia as early as 1960. Postwar industrialization programs and accommodating tax treaties had helped to attract foreign (mainly American) investment (oil and chemical industry near the coast and the port of Antwerp, light metal engineering and transport industry around Ghent) and solve the age‐old problem of underemployment in rural Flanders. With Belgian and foreign capital shifting their weight towards expanding Flanders, Wallonia fell behind. Unemployment had always been much higher in Flanders, but in 1970 unemployment in Wallonia (5.1%) was almost double the rate in Flanders (2.8%). By 1980 unemployment rates in both regions had slipped into double digits. However, unemployment has persisted in Wallonia, where rates are currently three times higher than in Flanders.

5.4.2. Divided Employers and Strong Unions

In the 1970s wage negotiators failed to take account of the country's deteriorating external position (Goubert and Heylen 1999), and Belgium ended the decade with the highest manufacturing wage costs in the European Community (De Grauwe 1994: 67). The unions continued to negotiate real wage gains in spite of the sharp rise in unemployment. This may be thought to reflect their continued strength (Hancké 1993; Ebbinghaus and Visser 1999) and is consistent with our hypothesis that Belgian trade union leaders continued to believe that wage restraint would not help but only depress demand and therefore aggravate the crisis.

With corporate profits under pressure, employers screamed for wage restraint and lower taxes every year since 1975. But they were internally divided and unable to attract the unions to a deal. In 1974 they signed what would turn out to be the last central agreement for a long time. Building on the 1944 Draft Agreement on Social Security and the 1954 Productivity Agreement, the unions had committed themselves to the maintenance of social peace and the promotion of productivity in exchange for giving workers a share in the proceeds of economic progress. In this spirit seven central agreements had been reached between 1960 and 1975. These had expanded workers' rights and benefits, including old age pensions (1971), paid holidays (1960, 1963, 1966, 1973, and 1975), shorter working hours (1969, 1971, and 1973), national minimum wage (1975), equal pay (1975), as well as union representation rights in firms (1971). In this ‘Golden Age of Planning and Growth’ (Dancet 1988: 217), parliaments had signed into law and governments had executed what the social partners wanted.

Until 1974 wage setting was entirely a matter for unions and employers, with no role for the government. Actual wage bargaining took place at the (p.234) sectoral level and has been characterized as rather uncoordinated (Van Ruysseveldt and Visser 1996; Vilrokx and Van Leemput 1997). In comparison with Austria and the Netherlands, the role of the central organizations is limited. Internally divided among sectors, regions, and ideological currents, the central organizations carry little authority over their member organizations. This has proven to be a problem, especially in the Socialist union federation (FGTB), but also in the Belgian Federation of Employers (FBE), which faces a strong contender in a Flemish organization (VEV). The central organizations do meet in direct consultation within the National Labor Council (CNT), where they have had the right since 1968 to negotiate binding agreements. At the sectoral level, collective agreements are negotiated in some 130 parity commissions that meet under an independent chairperson and are assigned the task of negotiating (binding) agreements, settling and preventing disputes, and advising the government. With the major exception of the chemical industry, company bargaining is secondary and mainly concerned with non‐wage issues.

In 1975 the central employers' organization pulled the brake and recommended a small wage increase, but according to FBE chairman Puellinx his advice went unheeded and most sectors negotiated a real wage increase of 5 percent or more (interview, SWAV 1995: 5). Similar attempts in later years failed, and until 1982 gross wages and wage costs outpaced productivity growth by a wide margin, with the result that unit labor costs increased (see Table 5.6 for a comparison with the Netherlands). The result of uncoordinated wage bargaining was, as in a prisoners' dilemma, a higher cost for society in the form of rising unemployment and social security expenditures. The bargaining parties maintain an interest in demanding social protection (unions) and subsidies (firms), the costs of which are shouldered by the community at large. Unavoidably, this development led to government involvement in wage bargaining and a change from bipartism to tripartism (Slomp 1983). Until 1981, however, all attempts at encouraging unions and employers to accept restraint failed. Each side had a sticking point: employers refused to discuss working time reduction, unions ruled out real wage restraint, and governments could not afford to give up the strong currency (De Swert 1989). But the government was not in a position to stand aside. It needed wage restraint, if only to limit the rising claims and costs of social security and the growing deficit in public spending. After 1981 the government took an authoritarian line, and from 1982 to 1986 autonomous wage setting through negotiations between unions and employers was suspended.

The central organizations not only lacked control over their members, they also disagreed over policies. As in many other European countries (p.235) after 1968, Belgium witnessed a period of worker militancy and radicalization of union demands (Molitor 1978) and Belgian employers felt under attack ideologically (Moden and Sloover 1980). With the slowdown of economic growth, it became less easy to find compromise solutions. In 1976 Fabrimetal, FBE's powerful affiliate in the steel and metal industry, alarmed by the rapid deterioration of its members' competitive position, demanded the abolishment of the indexation system. Established in the immediate postwar years, this system guaranteed that wages were fully adjusted to price increases, making wages inflation‐proof. The technique was for each point rise in the consumer price index to be translated, after two months, into a similar rise in wages. Coverage was nearly complete and in 1975, the system was extended to the national minimum wage and social benefits. Fabrimetal's attack received insufficient backing in other sectors, and employers were forced to retreat under pressure from strikes. Indexation became a ‘holy cow’ and remained a sticking point in Belgian industrial relations over the next several decades. The arguments for or against indexation and the deleterious effects on bargaining discussed in the preceding section on the Netherlands apply with equal force to Belgium. Unlike Dutch (and later Italian) union leaders, however, Belgian union leaders remained steadfast in their defense. Their behavior contradicts the hypothesis that union leaders in a continuous low inflation environment lose interest in an automatic cost‐of‐living adjustment wage, since it constricts their role and autonomy in collective bargaining (Braun 1976). The unions countered employers' pressure for wage restraint with demands for working‐time reduction. In 1976 Belgian unions, beginning in the troubled steel industry, were the first in Europe to raise the demand

Table 5.6 Average Annual Growth of Real (Gross) Wages and Productivity (%)

Real (gross) wage growtha

Productivity growthb

Belgium

Netherlands

Belgium

Netherlands

1971–82

4.14

2.72

2.80

2.28

1983–85

−0.75

−0.91

1.35

2.53

1985–89

0.70

0.70

2.02

0.79

1990–95

2.20

0.31

1.73

0.72

(a) Annual change in wage, in percentage, averaged per period.

(b) Annual change in productivity, in percentage, averaged per period.

Note: Real (gross) wage is calculated as the sum of all (nominal) compensations to wage and salary earners divided by the total number of wage and salary earners employed, deflated by the consumer price index. Labor productivity is calculated per person (rather than per hour, explaining the low figure after 1985 in the Netherlands, due to part‐time employment).

Source: Conseil Central Économique, report on the Belgian economy in 1997.

(p.236) for shorter working hours as a work‐sharing measure, but negotiations proved fruitless.

The first of many state interventions in wage bargaining came at the end of 1976 when Belgian's outgoing center‐right government imposed a ‘special levy’ of 50 percent on all contractual pay increases above those linked to indexation in order to finance a special fund for older unemployed workers who were allowed to stay outside the labor market on a transitory pension till legal retirement age. This decision provoked a national strike, jointly organized by the socialist and Christian unions, the first‐ever strike of the latter against the government (including their own ministers). The incoming center‐left government boosted public sector employment and tried to win unions and private sector employers over to a package deal based on wage restraint, working time reduction, and job targets equivalent to 2 percent of the workforce. This attempt at political exchange failed. Unions wanted more assurances that there would be additional jobs, whereas employers rejected the targets as ‘bad therapy’ and interference with management. When the government dropped the targets and offered a voluntary scheme for subsidized working time reduction, employers in some sectors, mainly those with overcapacity, did accept negotiations over working time reduction. The moderating effect on wages was small. The unions were disappointed that the job targets were dropped, and the subsidies lowered the employers' resistance to wage demands.

5.4.3. The 1980s: International Pressure and Recovery

In the early 1980s, Belgium earned ‘the dubious distinction of being the first and only Member State to receive an official recommendation of the European Commission to address its growing public deficits and rapidly growing wages’ (Kurzer 1997: 118). At the European Summit of 1980, Belgium was ‘advised in strong terms’ to modify its ‘prejudicial system of wage indexation if it wants to stay in the EMS' (Dancet 1988: 211). Impressed, Prime Minister Martens proposed a temporary suspension of indexation, apparently without consulting his socialist coalition partners, whose ministers resigned. The next government, again with the socialists, asked special powers of Parliament, and only 48 hours before a wage stop would have taken effect, unions and employers agreed to seek voluntary restraint, packaged with some working time reduction for 1981–82. Unlike the Dutch agreement of Wassenaar, this agreement was not the dawn of a new era in industrial relations. It was too obvious a dictate. Intellectually, union leaders had not come around to the idea that profitability and competitiveness had to rise in order to prevent further job (p.237) losses, as had Dutch union leaders by this time. Moreover, in addition to Belgian negotiators' low mutual trust, there was no stable government that might have protected a compromise between the social partners against defection or abuse. Indeed, within weeks the government fell (over the federalization issue), provoking the fourth crisis in two years. In an unusual speech King Baudouin reminded politicians that the time had come ‘to put our differences aside’ and ‘give priority to survival’, as one ‘would do if we were at war . . . war for the preservation of our economy’ (Deweerdt and Smits 1982: 262). But the next government did no more than solve another piece in the federalization puzzle and postpone what Belgian employers now called ‘the self‐evident decision’, by which they meant the suspension of wage indexation. Instead, the government bought time and offered employers a flat‐rate reduction in social security contributions for manual workers in struggling manufacturing firms (Maribel, Law of 29 June 1981).

The general elections of 1981 resulted in losses for the Christian Democrats in Flanders and gains for the Liberals. Forming his fifth cabinet, CVP leader Martens chose the Liberals as his coalition partner. Despite its narrow majority in Parliament, this coalition would stay in office the full four years and continue unchanged after the elections of 1985. Martens V more or less ruled by decree, seeking special enabling powers from Parliament in an annual vote of confidence. This silenced not only the Socialist opposition but also the prime minister's critics inside his own CVP, especially from the party's Christian union wing. This ‘less democracy for a better economy’ approach (Smits 1983) was defended by Martens as ‘an unavoidable step in the recovery of our country’ (Bastian 1994: 92). Not only Parliament but also the collective wage negotiators were placed in ‘preventive custody’ (Vilrokx and Van Leemput 1997: 336–37).

The recovery began with a large devaluation of the franc. Early in 1982, after negotiations within the IMF and Ecofin, an 8.5 percent devaluation was obtained, the largest since the start of the EMS (Gros and Thygesen 1994: 76). The liberals had wanted more, the unions (like the Bundesbank) less (Kurzer 1993). Devaluation could only restore the competitiveness of Belgian firms if combined with foolproof wage restraint. Hence, a standstill on wages (till May 1982) and a suspension of indexation for the entire year were crucial. Martens and his aides had secretly negotiated the recovery plan with the veteran leader of the Christian workers' association, who used his influence to assure cooperation from the Christian unions. They consented on condition that suspension of indexation would be temporary and special income protection be given to low‐paid workers and large families. The Socialist unions were left in the dark, but their strikes changed (p.238) nothing about government policy. During the next four years, until the end of 1986, the government decreed that three index jumps, each amounting to a price increase of 2 percent, would not be translated into wages and that any wage increase above inflation was prohibited. In combination with the devaluation, these draconian measures did indeed lead to a sharp turnaround in competitiveness. Real wages fell between 1982 and 1986, and all productivity gains were captured by firms, leading to a sharp decline in labor's share in enterprise income. At the same time, however, the government raised the employers' contributions for social security in order to meet its second objective of fiscal consolidation. Hence, non‐wage labor costs decreased much less than wages, and the wedge between labor costs and take‐home pay rose sharply. Since the government had agreed to exempt minimum wages and heads of (large) families from its measures, or to offer them compensation, the distance between minimum and average wages narrowed, and the overall wage structure compressed.

There are similarities, in background and content, between the Belgium recovery plan of 1982 and the ‘no nonsense’ approach that started later the same year in the Netherlands. Both were responses to a very severe crisis, implemented by center‐right governments with a strong electorate mandate for the conservative Liberals. In both countries, as in Germany or in Britain two years earlier, the Social Democrats had exhausted themselves as partners in government by their failure to tame or convince the unions (as well as by their antagonism toward permitting a new generation of nuclear weapons on Dutch and Belgian soil). In both countries the recovery plans aimed at restoring private sector profitability, consolidating the budget, and using job sharing to slow down (if not reverse) the upward trend in unemployment. The overall package was deflationary in both countries and initially increased unemployment. It implied a shift of incomes from households to enterprises, showing up as a fall in labor's share of net national income. The main difference, however, was that recovery in the Netherlands was consolidated by a major agreement between the unions and the employers. As was argued in the preceding section, this allowed the Dutch government to concentrate on fiscal consolidation and uncouple public sector pay from private sector wage negotiations. In Belgium there was no such agreement, leaving the government in a more vulnerable position, with many risks of implementation failure and political horse‐trading.

The Belgian government's choices were constrained by the fact that it started the decade with a huge public debt and social security costs that had risen spectacularly. There was no money to lower taxes on employers or workers or boost consumer demand through tax reform. While the cost crisis in social security was probably more severe, the austerity measures (p.239) were rather piecemeal and drawn out over a longer period than in the Netherlands, possibly because opposition from the more powerful Belgian unions and within the Christian Democratic Party was stronger. Social security for workers in Belgium is contributions‐based, but premiums in 1983 covered only 62 percent of total expenditures, as against 84 percent in 1970 (Peeters 1989: 202). Many funds had fallen into debt. The deficit in social security (all programs combined) had increased nearly tenfold, from 26 bn Belgian francs in 1970 to 254 bn francs in 1980, with the rise in unemployment benefits as the largest contributing factor. The number of unemployed receiving full compensation from the insurance funds had trebled between 1974 and 1980, while expenditure on unemployment rose from 0.7 percent of GNP in 1970 to 5 percent in 1985. In 1980, as part of an initial austerity package, the beneficiaries receiving unemployment insurance entitlements had been differentiated among principle breadwinners with dependent families, single households, and persons earning a second household income. For the latter two categories, benefits were lowered and phased out earlier. In later reforms, an extended ‘waiting time’ of 150 days has been applied to school‐leavers before they may claim a reduced benefit. After 1982 benefits no longer followed wage increases (a principle only introduced in 1975), but breadwinners and large families received compensation.

Another cost explosion concerned the various early retirement schemes. Faced with the explosive growth in youth unemployment, which had risen to 25 percent in 1980, and with the structural crisis in coal mining, steel, ceramics, textiles, and shipbuilding, Belgian governments believed they had no choice but to encourage the ‘pre‐pensioning’ of workers. It was, moreover, the only policy response on which there was agreement between the social partners (Bastian 1994). The appeal of this approach shows up in the sharp fall of the employment ratio for older males between age 60 and 65, from just over 60 percent in 1970 to 32 percent in 1981. By 1983, the first year for which we have internationally comparable data, the employment/population ratio in this age group was lower in Belgium than in all other countries except Luxembourg. Older workers received an early pension until legal retirement age in case of unemployment, and public subsidies were made available for early retirement in case a young worker was hired as a replacement. The costs were initially fully borne by public funds, like the unemployment insurance fund, but after 1982 the government introduced special levies.

Work sharing was the third plank of the 1982 recovery plan. Under the so‐called 5 + 3 + 3 operation, initiated in 1982, unions and employers were encouraged to negotiate a 5 percent reduction in working hours in exchange for a 3 percent wage sacrifice and a 3 percent increase in employment or, (p.240) failing that, an obligatory employer contribution to the National Employment Fund. Attempts to negotiate a central agreement along these lines came to naught. Employers, once assured of government‐imposed wage restraint, needed to make no further concessions. Unions felt that they had been robbed of an instrument—wage pressure—to wrest concessions from employers on labor time reduction and job creation. Yet in sectors and firms with overcapacity, employers and unions did reach agreement for about 1.3 of the two million employees in the private sector (Béguin 1985: 37). With an average working time reduction of 1.4 percent and a net employment effect of 52,100 jobs (Werner and König 1987: 62), about half of the government's target was reached. The Federal Minister of Labor defended state intervention on the grounds that he had no choice but to ‘oblige the labor market organizations to negotiate amidst a web of constraints, which both sides had refused in previous years’ (Hansenne 1985: 56). This approach was also applied in the so‐called ‘Hansenne working‐time experiments’, in which the state offered to suspend legal norms on maximum daily or weekly hours and overtime rules if firms could show positive effects on investment and employment and sign an agreement with the unions and the Ministry. Fifty‐five plant agreements of this kind were signed between 1983 and 1986, in total affecting 26,000 employees (Denys, Hedebouw, and Lambertl 1985), nearly all in large manufacturing firms, including Siemens, Phillips, Samsonite, and General Motors, all of which were interested in increased flexibility and located in Flanders. The Socialist unions in Wallonia would have nothing of it (Bastian 1994: 101; Colpaert 1987: 41).

5.4.4. Safeguarding EMU Participation

In 1990 the monetary authorities, anxious to quell the occasional speculative frenzies against the franc, promised to preserve parity with the Deutschmark in the event of a realignment, while the BNB gained the same degree of independence as the German and Dutch central banks. Having signed the Maastricht Treaty on Economic and Monetary Union, Belgium set itself the objective of belonging to the first group of EMU members. It had much to gain from borrowing the Deutschmark's reputation (Dyson 1994: 207) and, as a transit economy, it had much to lose should it be excluded. But on at least two major indicators among the EMU convergence criteria—the public sector deficit and the debt–output ratio—Belgium was way off target.

With the entrance of the Socialists into Martens's seventh cabinet (1988–91), fiscal control had been relaxed and the links among price indexation, wages, and benefits restored. From 1987 to 1993 wage bargaining (p.241) was again in the hands of unions and employers, albeit under a ‘shadow of hierarchy’, as defined by the 1989 Law on Safeguarding Competitiveness of Enterprises. This law authorized the government to intervene, ex post, if wages in Belgium had risen faster than the average trend among its five major trading partners (Michel 1994). Twice a year, the Conseil Central Économique (CCÉ), with bipartite representation by unions and employers joined by economic experts each side gets to appoint, issues a report on the state of the economy. If it sees fit, the CCÉ may recommend wage guidelines and other measures. The balance sheet on this period of limited ‘free’ wage bargaining (1987–93) is mixed. The central organizations were able to conclude (biennial) central agreements, but ‘devoid of content and largely dictated by government’ (Vilrokx and Van Leemput 1997: 341). The fact of the matter is that these agreements did not uncouple, or differentiate, collective bargaining in different sectors, as was by now the case in the Netherlands (and would keep happening there even more intensely in the 1990s), and as had always been the case (under strict supervision) in Austria. In Belgium, instead, wages in high productivity sectors like manufacturing kept setting the norm for wage increases in low productivity sectors like domestic services. By their very nature, statutory interventions are unable to differentiate by sector, because of asymmetrical information and monitoring problems. (The Dutch had already tried this, and failed miserably, in the early 1960s.) Between 1987 and 1989, and again in 1993, but not in other years, growth in gross wages did stay within the boundaries of overall productivity growth. In all years, however, wage increases exceeded the productivity gains in both private and public services (which were about four times lower than in manufacturing according to figures of the Belgian Planning Bureau). Moreover, responsiveness to unemployment and to economic shocks was limited. Regressions over the full period from 1970 to 1993 showed that in the case of Belgium a rise in unemployment by 2 percent was needed to obtain a 1 percent decline in real wages (less than half the elasticity of Austria, for example) (Heylen and Van Poeck 1995).

In the 1990s the old competitiveness problem resurfaced. One reason was that Belgium suffered more from the EMS troubles of 1992 and 1993 than Austria or the Netherlands. The new round of federalization in 1988 had not encouraged ‘the kind of fiscal behavior associated with restricted monetarism’ (Kurzer 1997: 120). International confidence in the franc remained low (in 1993 the BNB had to accept wider margins between the franc and Deutschmark). The other cause was wage growth. The CCÉ report for 1994 stated that Belgium had again developed a handicap of 6 percent vis‐à‐vis its trading partners. This figure was disputed by the unions, who called the statistics unreliable and blamed the recession on the (p.242) deflationary policies of the Bundesbank and the restrictive EMU membership criteria agreed upon in Maastricht (Serroyen and Delcroix 1996: 36). Formalized in its so‐called EMU convergence plan of 1992, the government had ushered in a fiscal consolidation package aiming at reducing the deficit from 7 percent to 3 percent of GDP by 1997 (OECD 1995: 7). Later a 3 percent EMU tax surcharge was introduced (abolished in 1999). During the EMS crisis a group of prominent Belgian economists called for unpegging the franc, abolishing indexation, and overhauling the social security system which they described as ‘wasteful’ and ‘inefficient’ (Kurzer 1997: 120). The proposal was hardly echoed in the press. Some unions might have favored a softer currency but not at the price of giving up indexation or lowering social security. Employers, choking under high costs, might have favored a softer franc, but they did not trust the unions to give up on indexation and deliver restraint. Instead, their primary objective was to lower government expenditure, taxes, and social security charges. The BNB was adamant in its defense of a franc fort; devaluation would mean higher imported inflation, diminished fiscal discipline, a rising public debt, and probably goodbye to EMU‐membership.

In comparison to the Netherlands, where a temporary wage stop was implemented early in 1993, followed by a new central agreement later that same year, Belgian wage negotiators were slow to adjust. The renewed Christian–Socialist coalition, led by the CVP's new leader Dehaene (1992–95), was unable to coax the unions into signing a similar pact for competitiveness and employment. Alarmed by sharply rising unemployment, the government had proposed a ‘Global Pact’ early in 1993. On the basis of a grim report on the Belgian economy, prepared under guidance of the National Bank, negotiations started in October. Infuriated by the proposal to trim the cost‐of‐living index, the socialist federation FGTB, under pressure from its powerful affiliates, walked out (Van Ruysseveldt and Visser 1996: 215). The government, including its socialist ministers, went ahead and excluded tobacco, alcohol, petrol, and diesel fuel from the calculation of the cost‐of‐living index. The new so‐called ‘health index’ lowered the adjustment of wages and benefits projected for 1994 and 1995 by 1.3 percentage points (with a projected 2% annual rate of inflation) (OECD 1995: 10). When the government's Global Plan became law in 1994, the eight‐year interlude of limited freedom in wage setting had ended (Blaise and Beaupain 1996). From now on, wage bargainers were placed under ‘house arrest’ (Vilrockx and Van Leemput 1997). Nominal wage increases, beyond the watered down price indexation, remained banned until the end of 1996.

The second Dehaene government (1995–99), again a Christian–Socialist coalition, started off proposing a ‘Pact for the Future of (p.243) Employment’ in pursuit of three objectives: to halve the current unemployment level of 12 percent by 2002, improve the sluggish rate of GDP growth (1.5%–2%), and secure the 3 percent budget deficit target required for EMU membership. The new government was under pressure from Europe and the regions. In Flanders in particular, employers and the regional government were showing impatience with the national government. Consultations over the Future Pact continued until April 1996, when the VEB reached a draft agreement with the unions. Its contents were (once again) wage moderation, a legal maximum working week of 39 hours in 1998, improved rights for leaves of absence without pay, part‐time retirement at age 58, and full‐time retirement at age 60, annualization of working time, and lowering of employers' social security contributions through the expansion of the Maribel subsidy schemes. The agreement failed, however, when the socialist union federation FGTB was unable to gain approval from its affiliates. The leadership of the Christian federation ACV slipped through with a slight majority (Van Ruysseveldt and Visser 1996: 217).

The government went ahead anyway, selectively borrowing from the failed agreement. The central wage standard was given a legal status with a new framework law. This 1996 Law on Safeguarding the Competitiveness of Enterprises prescribes that wage increase in Belgium must remain below the average wage increases of its three neighbors Germany, France, and the Netherlands. This was a clear tightening of the old law of 1989, which had entitled the government to intervene ex post. The new law introduced an ex ante maximum wage norm based on predicted wage increases in Germany, France, and the Netherlands. Under the law, industries and firms are punishable if they exceed the norm; they may settle for lower wage increases, for example, in exchange for extra job measures. The (revised) indexation mechanism remained in place. The Act also stipulates that multi‐industrial bargaining must take place every two years, on the basis of the CCÉ's technical report, and lead to an agreement on the precise margin for sectoral bargaining and additional job creation or working time matters. If no agreement is reached, as was the case in 1996, the government unilaterally sets the margin. For 1997 and 1998 it specified a maximum increase of 6.1 percent over two years, including a predicted 3.6 percent rise in consumer prices, 1 percent seniority related wage increases, and a real across‐the‐board improvement of 0.75 percent per year (Vilrokx and Van Leemput 1997: 339–40). Unlike the stalemate of 1996, a second round of bargaining for 1999–2000 did produce an agreement. This was eased by the government's decision to reduce social security contributions by 108 bn francs (2.5 bn euros) in five years and bring social charges down to the average level of Belgium's three neighbors. The FBE was under (p.244) pressure from its Flemish regional partner, which had announced that it would ‘go alone’ should no national agreement be reached. This time the federal leadership of the unions gained its member unions' consent. The deal itself was complex and linked future reductions in social insurance contributions to compliance with the wage norm or, failing that, extra spending on vocational training. Companies and sectors paying above the norm will not be penalized (as they should have been for exceeding the norm in 1997 and 1998) if they can show that there are no negative employment effects. Not only has this been extremely difficult to monitor, but it requires types of sectoral, regional, and company data on investment and employment that are not available.

5.4.5. Lackluster Growth in the 1990s

In the 1990s Belgium exemplified Europe's main illness: high structural unemployment and a lack of productivity and employment growth in private services. As elsewhere, the problem was magnified by the slowdown in international economic growth and restrictive policies in the run‐up to EMU. But this cannot explain the difference between Belgium and the two other countries discussed in this chapter. The causes of Belgium's low employment growth are probably a cluster of factors relating to problematic wage setting, a high tax and social security wedge, various unemployment traps, and inefficient active labor market policies, to mention just a few of the factors (Elmeskov, Martin, and Scarpetta 1998; Goubert and Heylen 1999). Past culprits were an aging industrial structure, full wage indexation, and excessive real wage growth. Once unemployment became high and persistent, social policies and labor market measures had to be expensive, even after various retrenchments. The weak social consensus and capacity for compromise delayed and weakened responses. The job leak due to deteriorating international competitiveness was only mended by placing wage negotiators under direct government control. But in international comparisons of wage costs, Belgium still comes up very high, second only to Germany. The OECD (1997: 1) maintains that ‘improving the wage formation process remains a largely unresolved issue’. This is also the view of the government's own advisers, like the newly established High Council for Employment, even though the Council's 1998 report on the matter was largely ignored.

Pattern bargaining has negative consequences, especially when productivity trends diverge. In Belgium differences in productivity gains across sectors are very large; from 1970 to 1996 average annual productivity gains were 4.7 percent in manufacturing industry, as against 1.0 percent in services (BFB 1998: 108), a gap that is almost twice as high as the European (p.245) average. If wages exceed the very small productivity margin, and wage bargaining is in real terms, profits and investments will decline. Consequently, job losses in manufacturing (between 1970 and 1994: −463,000) were hardly compensated by job increases in services (+682,000, of which only +114,000 were in non‐market services). With further declines in agriculture, energy, and construction, overall growth was just +35,000 jobs.

Another reason for concern is the compressed wage structure, especially at the lower end, with little variation across sectors, regions, and skill levels, despite very large differences in unemployment, labor demand, or productivity. Together with the strong bias towards passive labor market policies and an earnings‐compensating role for social security, this may explain why job turnover and labor mobility in Belgium is low, especially in Wallonia (De Grauwe 1998; Goubert and Heylen 1999; Marx 1999). There are a number of unemployment traps, related to the length rather than to the level of unemployment benefits. Net replacement rates for single parents and breadwinners, measured against the minimum wage for full‐time workers, are estimated at 107 percent and 91 percent respectively. For part‐time workers at 50 percent of the minimum wage, net replacement rates vary around 90 percent and 96 percent for breadwinners and unemployed spouses (Cantillon and Thirron 1997). Married women receive a small unemployment benefit, which often serves as a second income in the family. They have little monetary incentive to accept a job if it is not full‐time. This applies particularly to low‐income families. Moreover, the strong family bias in unemployment benefits discourages spouses of unemployed workers from seeking or holding on to part‐time jobs. Already in 1975 the female rate was twice the male unemployment rate, and in 1980 the ratio had increased to 3:1. This was partly determined by faster labor force growth for women, but unlimited unemployment insurance is another part of the story. In the 1990s the female/male disadvantage returned to its old ratio of 2:1, partly because of withdrawals from the labor market. Since 1987 administrators have been entitled to suspend unemployment benefits for second earners whenever unemployment spells were exceptionally long (i.e. 1.5 to 2 times the average for the industry or occupation).

Effective minimum wages, as defined by collective agreements, lie 20–30 percent above the statutory minimum. This would suggest that unemployment traps may be not as bad as they seem (Jadot 1998). However, the high level of wages at the lower end of the Belgian labor market, combined with a high level of wage costs and low profits, discourages job creation in services. This situation disadvantages workers with low levels of education, those with long spells of unemployment, or women who try to re‐enter the labor market. For exactly this reason, Dutch governments since 1993 have (p.246) successfully exerted pressure on unions and employers to lower negotiated minimum wage levels and provided wage cost subsidies for the low paid.

Wage cost subsidies to employers have remained the only answer, but until recently they were only targeted to industry. During the 1992–93 EMS crisis there was a rapid expansion, from 15 bn BEF in 1993 to 60 bn BEF in 1996 (there are roughly forty francs to the euro). As in the original Maribel scheme of 1981, subsidies were intended to compensate exporting firms for the hard currency policy and to slow down the decline of manual work jobs in manufacturing. On precisely these grounds the European Commission repeatedly criticized the subsidies as a distortion of competition, since they were not open to all firms. In response, the program was widened in 1997 and now lets all private firms, including services, reduce social insurance contributions proportionate to ‘labor intensity’. The revised Maribel scheme is valid until the year 2000 and applies to about 770,000 workers, as against 431,000 in the old program. Unions have criticized the subsidy as a handout to employers (Serroyen and Delcroix 1996: 37). But the proposal of Socialist ministers to make the subsidies contingent on explicit job targets was not implemented. Studies of the Belgian Planning Bureau suggest that a selective subsidy program, targeting workers with low earnings, would create more jobs but prove costlier and more difficult to monitor (CNT 1998). In 1995, however, the federal government phased in an additional ‘low earnings’ subsidy scheme, which applied to 782,000 workers at or around the minimum wage who get to have their social security charges lowered by between 2 percent and 12 percent of total wage costs. In 1996, 36 percent of total expenditures on tax‐ and charge‐reducing subsidies went to Maribel, 24 percent to the low earnings scheme, and 20 percent to special job creation plans for the long‐term unemployed (De Lathouwer 1999: 199). The remainder was set aside for contingency plans, negotiable with trade unions and employers.

As was mentioned in the introduction, Belgium spends more on ‘active’ labor market policies than either Austria or the Netherlands. Total expenditure nearly doubled from 64 bn BEF in 1985 to 112 bn BEF in 1994. Against the background of a rise in long‐term unemployment amounting to 6 percent of the active labor force, the spending increase is less impressive than it sounds and the effectiveness of policies more questionable. Placement offices and employment services are extremely overburdened, and in 1990 it took an average of 20 months before an unemployed person was contacted by a case officer (OECD 1994). Non‐participation in job placement programs is rarely penalized. Only 20 percent of total expenditure concerns training and education, and just 6 percent of the unemployed participate in training or work experience programs (De Lathouwer 1999: 200). Subsidies absorb most costs. Although there is perhaps no other government in Europe with so (p.247) many job plans, many quite creative, Belgium's top civil servant in the Ministry of Labor described his government's job policies in a recent policy review as ‘plainly unrealistic’ (Jadot 1998).

Many of the schemes are along the lines of the 5 + 3 + 3 schemes of the early 1980s. The government makes subsidies available, or lifts certain legal restrictions on work or working‐time regulation, if employers create extra jobs, and unions sign on to extra wage restraint and greater flexibility. Some of these schemes are specifically designed for young workers, the long‐term unemployed, small firms, or the non‐profit sector of charitable work. The take‐up for subsidy schemes, to be negotiated under company or plant‐level agreements, has been disappointing (Serroyen and Delcroix 1996: 43) and control over implementation is extremely difficult. A possible explanation is that Belgian unions and employers are not ready for the kind of decentralized bargaining required (Vilrokx and Van Leemput 1997: 341). Internally divided between their regional wings, Belgian employers' associations fear the strength of the union delegate system at the local level and the militant stance of the union rank‐and‐file in Wallonia. The Socialist union confederation dislikes decentralization for opposite reasons; left to their own devices, Flemish representatives would go along with more flexibility than the FGTB and Christian unions presently do, whereas Wallonian union representatives would probably accept even less change. Sectoral plans, negotiated under a central agreement of 1994, have been more numerous, but they mainly focused on early retirement. Once again, the agreement widened the opportunities for early retirement from age 58 to 55, under certain conditions. Many agreements allow workers who have been contributing for a minimum of 33 years to leave at age 55, and nearly all permit older workers to continue working half‐time and receive a supplementary unemployment benefit until they reach the legal retirement age of 65.

Part‐time employment has remained unpopular with the unions. Unlike in the Netherlands, it has not been (or become) the dominant choice of working mothers. The full‐time employment/population ratio of Belgian women is almost double that of the Netherlands, whereas the part‐time employment/population ratio is nearly three times lower (Visser 1999). Historically, a much larger number of Belgian married women continued to work full‐time, and childcare facilities, especially for very young children, have generally been available on a much wider scale (Daly, this volume). In Belgium married women used to stop working if their husbands earned enough, while in the Netherlands women continue working if they can buy or obtain private childcare. Hence the relationship among family income, women's work, and working hours is quite opposite in the two countries (Henkens, Siegers, and van den Bosch 1992).

(p.248) The policies of Belgian governments concerning part‐time work have been hesitant and inconsistent. Around 1980, part‐time work was encouraged. Combining part‐time work with part‐time unemployment benefits made it a very popular option at the time (Casey 1983). The program became too expensive and was stopped in 1982 because of its success. In recent years, the government has reinstated the possibility of combining part‐time work with benefits, especially for older workers. These options also exist in the case of part‐time career breaks. The regional government of Flanders has introduced additional incentive schemes for part‐time jobs and career breaks whose purpose is caring for young children or taking vocational training (Serroyen and Delcroix 1996: 47).

Part‐time work is treated with ambivalence in labor law and social security, as well as in statistics. Until 1987 anyone working two hours per week or less was not counted and remained outside the social security system. When the two‐hours rule was abolished, employment rose by 55,000 extra jobs. Many workers became self‐employed for tax and social security purposes (De Swert 1999: 44). In Belgium, mock self‐employment is the equivalent of small part‐time jobs amounting to less than 12 weekly hours in the Netherlands, except that since the 1990s the latter are covered by tax and social security. In Belgium, employment protection and minimum wage and vacation rights apply only to part‐time jobs that have at least one‐third of the hours of full‐time jobs, with a minimum of three hours per day. In the Netherlands the opposite move took place, to raise the status, rights, and popularity of small part‐time jobs, especially important in domestic services (Visser 1999).

5.4.6. The Efforts of a sur place

In professional cycling there is a game in which you have to work hard to stay in the same place and then leap to victory. This sur place appears to be a pertinent description of the Belgian case. There is no other country where governments have designed so many pacts, proposals, plans, and schemes to coax unions into accepting wage restraint and employers into creating jobs, and with so little success. There is also no other country where five Ministers of Labor, at the federal, regional, and communal levels, compete for attention and resources. The unanswered question is whether the new political situation will break the immobilism of Belgian politics —the Lib–Lab–Green coalition that gained power in 1999 has a clear mandate for change after the string of scandals in recent years and disastrous electoral results for the Christian and Social Democrats, especially in Flanders. In conjunction with the 1998 central agreement, this may be the time for a new start. One promising proposal by the new government (p.249) is to expand ‘social Maribel’ in 2000 and substantially reduce non‐wage labor costs by about 32,000 BEF per year per employee for ‘entry’‐level wages. In sum, the government wants to lower the ‘tax on labor’ by 50 bn BEF, to be financed with a tax reform announced for 2002 (but with a content yet unknown, except for the detail that the VAT on labor‐intensive services will be lowered from 21% to 6%).

Like the Netherlands, Belgium entered the 1970s with one of the lowest employment/population ratios in Europe. Thirty years later, unlike the Netherlands, the situation has deteriorated (Figure 5.3). Cutting the labor supply is still accepted as an alternative to overt unemployment, and governments have gone along with subsidizing early retirement for lack of alternatives. Unemployment remains high, hides large regional variations (with Flanders approaching full employment), and includes a large structural component, with the highest share of long‐term and youth unemployment in Europe outside Italy and Spain. Like its northern neighbor, Belgium went through a very difficult phase in the 1970s and sought to redress a rather desperate situation through a combination of wage restraint, fiscal consolidation, and job sharing. While some of this worked and Belgium's external competitiveness was restored, the outcome in terms of employment was less favorable. Despite a large amount of job sharing and job creation plans, employment growth remained sluggish. There was nothing compared to the growth in domestic services and part‐time jobs experienced later in the Netherlands. In contrast to the Dutch economy in the 1990s, the Belgian economy, in spite of its exposed sector's recovery, suffered from lackluster consumer spending, The savings rate of households reached an all‐time high and remained high in 1995–97. The OECD (1995: 7) speaks of a ‘healthy financial position of households’ and a show of ‘low confidence’.

The various subsidy schemes have mainly affected the distribution of jobs rather than their overall number. Unions criticize how many of the schemes targeted at low‐experience workers fail to requalify workers and instead trap them in dead‐end jobs. Economists have raised the criticism that subsidizing firms exposed to international competition has lowered employer resistance to union wage demands and slowed down the rate of innovation. There are remarkably few hard facts or impartial studies on these issues. The only observations that can be made with certainty are these: that there were many plans and schemes, that nearly all of them were contested, that many of these plans were designed to obtain cooperation from unions and employers but very hard to monitor, that many were short‐lived and depended upon complex implementation procedures, and finally that the actual job creation rate was disappointing. The narrow employment base of Belgium's economy and the absence of a jointly held (p.250) view on how that might be changed suggest continued vulnerability for this otherwise very wealthy country.

One may ask why there was not more policy change in response to stagnation and failure. Our main hypothesis is that organized actors in Belgium—inside and outside the government—failed to agree on the causes of the job crisis and its therapies, and that they continued to work at cross‐purposes. Hence, the kind of contrast in policies before and after 1982 found in the Netherlands is much less apparent in Belgium. Both countries suffered from a lack of consensus and coordination in the 1970s and went through a deep crisis in the early 1980s. In addition to a massive rise in unemployment, Belgium experienced a depletion of its national reserves, a crisis of international confidence, and currency devaluation. Yet Belgian trade unions did not accept that wage restraint and a recovery of profits were necessary conditions for economic recovery and job growth, as did the Dutch unions at the time. Was this a reflection of the continued strength of the Belgian economy? Did Dutch unions learn faster, and even shift paradigms, because they were so much weaker? The upshot was that Belgian governments had to impose conditions on trade unions and firms that were mutually negotiated in the Netherlands. In the Belgian case, this restricted decentralization and the flexibility of wage bargaining made linkages with other policies (especially to work‐sharing) more difficult, and rendered implementation‐failure more likely. Moreover, their ongoing attempts at begging for support of policy measures made successive Belgian governments far more vulnerable to various pressures (like having to repackage wage policies with subsidies to firms or promulgate protective measures for the low paid), and pleading for constituency support also compromised attempts at fiscal consolidation and welfare reform.

It is tempting to shift some blame onto the linguistic conflict and the cumbersome federalization process, discussed in the introduction to this chapter. The linguistic‐federal dimension overshadowed other pressing problems, like industrial decline, unemployment, public administration, justice, or environmental decline (Swyngedouw and Martiniello 1998). Michel Albert observed that the Belgians ‘were amusing themselves with their linguistic squabbles’ while the ‘ship was sinking’ (introduction in Hansenne 1985). The conflict was germane to government instability and produced a kind of ‘crisis consociationalism’ in which problems were combined until they became so big, numerous, or pressing as to require a large coalition in which each of the parties was allowed to deal with its particular clientele (Deschouwer 1998). The average tenure of postwar Belgian governments has been exceedingly short; in the ten crucial years between 1972 and 1982, when two major economic shocks needed a response, (p.251) Belgium had no less than thirteen governments (compared to five in the Netherlands and only three in Austria). The weakening of the state was compounded by the partisan use of the state, with recruitment practices not based on merit but on party membership and the right combination of language and region (Swyngedouw and Martiniello 1998; Dewachter 1989).

5.5. Conclusions

In the introduction we characterized Austria, Belgium, and the Netherlands as ‘birds of a feather’, albeit of different colors. Why did they not ‘flock together’? For the answer we have examined the external and internal conditions, perceptions of problems, policy choices, and any spillover from unsolved problems and unhelpful choices.

The Austrian economy was hit less severely by the oil shocks because it was (and is) less exposed to internationalization than the other two countries. Multinational firms were (and still are) lacking. Financial and stock markets were (and are) less developed, whereas Belgium hosts a number of very powerful financial holdings and the Netherlands traditionally had a highly developed stock market and a number of MNCs operating worldwide. For a much longer time and to a far greater degree, Austria has relied on all kinds of regulations, such as controls on prices, exports, and capital. Domestic‐oriented policies were maintained longer than they were abroad, and the public sector played a greater role in buffering external shocks. With the passing of time, however, Austrian exceptionalism subsided. But the country's geopolitical location and particular history help explain why political actors shared the same fears and the same perception of what the problems were: avoiding both mass unemployment and hyperinflation and restoring growth to an export‐oriented economy. The homogeneity of policy priorities is most prominently demonstrated by the amazing fact that income inequality was never a major topic in Austria, while wage moderation proved much easier to maintain than in Belgium and the Netherlands.

The reasons for the poor performance of the Belgian and Dutch economy in the 1970s were not dissimilar. High exposure to foreign trade was certainly a factor and meant that external shocks had a huge impact, but this was compounded by domestic factors, especially by a problematic method of wage setting fraught with diverging objectives and insufficient coordination. The result was that corporate incomes had to bear the brunt of adjustment in the terms of trade after the oil crises of 1973 (for both countries) and 1979 (for Belgium). In the Netherlands real wage growth (p.252) stopped in 1979, whereas between 1978 and 1982 Belgium was alone ‘among the seven small countries of Scandinavia, the Low Countries, Austria, and Switzerland, in having a real wage increase’ (Therborn 1986: 150). Since small countries are price‐takers in the international economy and the authorities maintained a fixed exchange rate policy, the export sector was unable to pass on the higher costs. The fixed exchange rate policy came at the price of very high real interest rates in the Belgium case. This worsened business conditions but did not attract sufficient foreign capital to offset the public sector's borrowing requirements. The export sector could only stay competitive through above‐average productivity growth and had to accept lower profit margins, which in turn led to a decline in investment. The combination of real wage growth, rising non‐wage labor costs and strong external constraints explains the large drop in manufacturing employment. The deterioration of public finances alarmed both countries' governments by 1976 at the latest, but attempts at reversing the trend were difficult in light of unstable and internally divided coalitions and growing compulsory expenditures to pay for rising unemployment and service the public debt. In the Belgian case we may add expensive solutions to linguistic conflict as well as the absence of any fundamental review of institutional mechanisms for manpower deployment and wage determination in the public sector.

Looking back on twenty‐five years of policy adjustment, one is struck by the ongoing importance of wage restraint for maintaining competitiveness and of a hard currency for fighting imported inflation. Apparently there were no alternative policy options in economies exposed to international competition, not even in Austria. Debate over alternative tools has been rare in each of the three countries. The contrast between Austria and the other two countries shows that the hard currency option, however unavoidable, is very harmful if not combined with wage restraint. The Austrian example also shows that wage restraint, if organized in cooperation with the social partners, facilitated opportunities for fiscal expansion and fighting unemployment that the other countries did not have.

In negotiated political systems like Austrian, Belgium, and the Netherlands, policy change is critically dependent upon the agreement of ruling coalition parties and support from the social partners. These systems may suffer from what Fritz Scharpf calls a ‘negotiator's dilemma’. Negotiators must simultaneously search for effective policy responses in terms of policy content while resolving distributive conflict, i.e. finding a ‘fair’ distribution of the social costs of adjustment between and among ruling coalition parties and the social partners (Scharpf 1997). Perhaps because Austria's problems were less pressing, but more likely because social partnership there was not as challenged from within, Vienna's governments (p.253) and social partners had little difficulty resolving the ‘negotiator's dilemma’ in the 1970s. All the relevant actors, in both the political and industrial arenas, agreed on a Keynesian definition of the crisis, a definition that was contested by key actors in Belgian and Dutch economic policymaking. Because of the conservative bias in Austrian society, there was no opposition to using foreign workers as a buffer by sending them home. This was not possible in the Netherlands and Belgium, and by the time of the second oil crisis it was also ruled out in Austria. Austrian unions were rather unique in giving priority to full employment over income redistribution. This amounted to an unfaltering policy choice for restraint. In the face of a much weaker Keynesian political consensus, Dutch and Belgian unions, partly as a result of union radicalization in the 1970s, were unprepared to forsake income redistribution for employment. The internal cohesion of the Austrian model was strikingly different from the situation in Belgium and the Netherlands, where the postwar consensus had in fact ended and was to be renegotiated within and between the different segments of society. Austria's legacy of a civil war between labor and capital in the 1930s and its position on the frontline between the communist East and the capitalist West bolstered the conviction that social conflict should be avoided at all costs. Moreover, Austrian social and economic policymaking—more encompassing, with fewer actors, and more informal in style—appeared easier to coordinate than in the other two countries (Unger 1998a). There are very few checks and balances in the Austrian political system. Institutional arrangements such as a weak bipartite system, modest federalism, and grand coalitions allow each governments to push through its agenda, provided the government has somehow engineered an agreement with the social partners. As demonstrated by numerous examples from the section on Austria—actions like privatizing the steel sector in the mid‐1980s, the 1988 tax reforms, joining the European Union, or the 1995 austerity package—agreements can be swiftly reached and promptly implemented. By comparison, the much more pluriform systems in the Netherlands and especially Belgium make for more cumbersome bargaining practices.

Lack of agreement in a negotiated policy system is very costly. The more the Belgian and Dutch social partners proved unable to organize restraint, the more important each country's hard currency stance became. The resultant negative side of declining competitiveness led to soaring unemployment in the 1970s and early 1980s, which in turn exacerbated the cost explosion in social security and the attendant potential for a damaging cycle of higher non‐wage costs, productivity hikes, unemployment, and a fiscal crisis. Austria experienced neither one of these two dire straits and was therefore in much better shape to stick to its incrementalist pattern of (p.254) corporatist adjustment. Before 1987 Austria maintained virtually full employment, thanks to a joint consensus between the political parties and social partners on the merits of Austro‐Keynesianism.

In the face of the acute unemployment crisis, a severe challenge to their power and pent‐up frustrations over recurrent government intervention, the Dutch unions finally conceded to the new realities of the world economy and returned to a strategy of wage moderation. Rather than continuing the prisoners' dilemma game, in which third party intervention would have become unavoidable, they redefined their position in terms of a ‘battle of the sexes', preferring a coordinated policy over distributive privilege. From 1982 on, Dutch unions have consistently placed jobs before income. This learning process on the part of the trade unions led to a revitalization of social partnership in the Netherlands during the 1980s. Over time, wage restraint allowed for a rather smooth interplay between wage setting and fiscal policy, stimulating economic growth while keeping inflation down.

This painful learning experience did not occur in Belgium, where the social partners remained stuck in a prisoners' dilemma and, moreover, suffered from internal fragmentation. Consequently, the government had to replace, and repay, the unions in the organization of wage restraint, accepting bargains with employers and unions over compensation through subsidies, make‐work programs, and measures for the low paid. Whereas the Austrian and Dutch social partners were successful in combining moderate wage increases with a high level of flexibility at the micro level, the Belgian strategy of imposed wage restraint sacrificed micro‐flexibility in the labor market for the purpose of macro‐adjustment. In addition, Belgian governments were faced with much larger implementation failures. Finally, they ran a much larger risk of being taken hostage and being forced to renegotiate their preferred policies of austerity and social security reform, policies that repeatedly overran their targets. Of course, the options of Belgian governments were also limited by the fact that more money had to be spent on interests and repayment of debts and that less was available for lowering taxes or propping up the purchasing power of workers and their families. Unlike the Netherlands, Belgium was unable to support wage moderation in the 1990s with tax rebates. Under conditions of permanent austerity the slightest of setbacks to economic growth had extremely disheveling consequences for the country's public finances.

Austria's welfare state demonstrates the highest degree of continuity but also some signs of obsolescence. So far, however, the political and institutional capacity for incremental reform and fast policy implementation has been high, and overall performance is good, though not as good as it was. The flexibilization of the labor market progresses slowly. Pressures for (p.255) changes towards a more liberal society and equal rights tend to come from abroad. The elections of 1999, turning the FPÖ into the second strongest party, demonstrate the persistence of Austrian conservatism and xenophobia, as well as some popular disenchantment with Austria's closed backroom politics. Austria's problems as a welfare state seem much less urgent than its difficulties transforming itself into a modern and open society. Belgium's welfare state reforms are much more severe because of larger and unsolved financial burdens, the federal conundrum, less homogeneity of preferences, and few learning experiences (so far) with successful, piecemeal reform. The Dutch welfare state certainly has experienced the greatest transformation and shown considerable capacity for reform and innovation. However, it now stands again at the crossroad of deciding how to spend its newly gained resources and reassess the balance between public and private investment and responsibilities.

References

Bibliography references:

Aarts, Leo, and de Jong, Philip (1992). Economic Aspects of Disability Behaviour. Amsterdam: North‐Holland.

—— —— (1996). Curing the Dutch Disease. Aldershot: Avebury.

Alber, Jens (1998). Recent Developments in Continental European Welfare States: Do Austria, Germany and the Netherlands Prove to be Birds of a Feather?. Paper for the 14th World Congress of Sociology, Montreal, 29 July 1998.

Bastian, Jens (1994). A Matter of Time: From Worksharing to Temporal Flexibility in Belgium, France and Britain. Aldershot: Avebury.

Béguin, J. M. (1985). Les Politiques d'emploi menées en Belgique de 1972 à 1984. Mimeo. Brussels: Vrije Universiteit.

BFP (Bureau Fédérale du Plan) (1998). L’Économie belge dans 1998. Brussels: BFP.

Blaise, Pierre, and Beaupain, Thérèse (1996). ‘La Concertation sociale 1993–95’. Courier hebdomaire du CRISP, No. 1497–98.

BMAGS (Bundesministerium für Arbeit, Gesundheit und Soziales) (1999). Der Arbeitsmarkt im Jahre 1998. Vienna: BMAGS.

Bottenburg, Maarten van (1995). Aan den Arbeid! In de wandelgangen van de Stichting van de Arbeid, 1945–1995. Amsterdam: Bert Bakker.

Braun, Anne Romanis (1976). ‘Indexation of Wages and Salaries in Developed Economies’. IMF Staff Papers, 23/ 1: 226–71.

Breuss, Fritz (1978). ‘Die Konjunktur 1978 im Zeichen des österreichischen Experiments der Zahlungsbilanzsanierung’. Wirtschaftspolitische Blätter, 4: 5–15.

Cantillon, Bea, and Thirron, A. (1997). Wegen naar een activerende verzorgingsstaat. Tussentijdse evaluatie van het BWA experiment. Paper voor Jaarboek Armoede en Sociale Uitsluiting, Brussels, Nov.

(p.256) Casey, Bernard (1983). ‘Staatliche Maßnahmen zur Förderung der Teilzeitarbeit: Erfahrungen in Belgien, Frankreich, Großbrittannien, den Niederlanden und der Bundesrepublik Deutschland’. Mitteilungen aus der Arbeitsmarkt und Berufsforschung, 16/4: 414–17.

CNT (Conseil National du Travail) (1998). Le Marché du travail dans 1998. Brussels: CNT.

Colpaert, Tom (1987). ‘Requiem voor de 32 uren week?’. Socialistische Standpunten, 17: 41–43.

CPB (Centraal Planbureau) (1997). Macro‐economische Verkenningen. The Hague: Centraal Planbureau.

Crepaz, Markus (1995). ‘An Institutional Dinosaur: Austrian Corporatism in the Post‐industrial Age’. West European Politics, 18/4: 64–88.

Crouch, Colin J., and Pizzorno, Alessandro (eds.) (1978). The Resurgence of Class Conflict in Western Europe since 1968. London: Macmillan.

Daalder, Hans (1966). ‘The Netherlands: Opposition in a Segmented Society’, in Robert A. Dahl (ed.), Political Opposition in Western Societies. New Haven: Yale University Press, 188–236.

Dancet, Guy (1988). ‘Wage Regulation and Complexity: The Belgian Experience’, in Robert Boyer (ed.), The Search for Labour Market Flexibility. Oxford: Clarendon Press, 212–37.

De Grauwe, Paul (1994). Onze schuld: Ontstaan en toekomst van werkloosheid en staatsschuld. Tielt: Lannoo.

—— (1998). Euro and Work. Paper presented at AIAS Workshop on Economic and Monetary Union, 1 July. Amsterdam: University of Amsterdam.

De Lathouwer, Lieve (1999). ‘Het Belgisch werkloosheidsstelsel: een evaluatie vanuit sociaal zekerheids‐ en arbeidsmarktperspectief’, in J. J. van Hoof and J. Mevissen (eds.), Loonvorming en werking van de arbeidsmarkt in België. Amsterdam: Elsevier, 185–210.

Dellen, H. van (1984). Een nieuw elan, de marktsector van de jaren tachtig. Deventer: Kluwer.

Denys, J., Hedebouw, G., and Lambertl, M. (1985). Nieuwe vormen van arbeidstijdregeling: De experimenten Hansenne. Mimeo. Louvain: KUL/Hoger Instituut voor de Arbeid.

Deschouwer, Kris (1998). Falling Apart Together: The Changing Nature of Belgian Consociationalism, 1961–1998. Paper delivered at conference ‘The Fate of Consociationalism in Western Europe, 1968–1998’, 29–31 May. Harvard University: Minda de Gunzberg Center for European Studies.

De Swert, Gilbert (1989). ‘Samen apart’, in BVVA Belgische Vereniging voor Arbeidsverhoudingen (ed.), Vijftig jaar Arbeidsverhoudingen. Bruges: De Keure and BVVA, 3–16.

—— (1999). ‘België in oude banen gebleven? Tien om te herzien’, in J. J. van Hoof and J. Mevissen (eds.), Loonvorming en werking van de arbeidsmarkt in België. Amsterdam: Elsevier, 43–54.

Dewachter, Wilfried (1996). ‘La Belgique d'aujourd'hui comme société politique’, in A. Dieckhoff (ed.), La Belgique: La Force de la désunion. Brussels: Complexe, 105–42.

(p.257) Deweerdt, M., and Smits, J. (1982). ‘Belgian Politics in 1981: Continuity and change in the crisis’. Res Publica, 24/2: 261–72.

Dolado, J., Kramarz, F., Machin, S., Manning, A., Margolis, D., and Teulings, C. N. (1996). ‘Minimum Wages: The European Experience’. Economic Policy, 23: 319–72.

Dyson, Ken (1994). Elusive Union. New York: Longman.

Ebbinghaus, Bernhard, and Visser, Jelle (1999). ‘When Institutions Matter: Union Growth and Decline in Western Europe, 1950–1995’. European Sociological Review, 15/2: 135–58.

Elmeskov, J., Martin, J. P., and Scarpetta, S. (1998). ‘Key Lessons for Labour Market Reform: Evidence from OECD Countries' Experiences’. Swedish Economic Policy Review, 2/2: 205–51.

Esping‐Andersen, Gøsta (1990). The Three Worlds of Welfare Capitalism. Princeton: Princeton University Press.

—— (1996). ‘Welfare States without Work: The Impasse of Labor Shedding en Familialism in Continental European Social Policy’, in Gøsta Esping‐Andersen (ed.), Welfare States in Transition. London: Sage, 66–87.

Farny, Otto (1996). ‘Analyse des Gewinnsteueraufkommens 1988–96’. Wirtschaft und Gesellschaft, 22/1: 39–55 (Orac publisher, Vienna).

Feuchtmüller, Wolfgang (1987). ‘Der österreichische Kapitalmarkt und die Privatisierung’. Wirtschaftspolitische Blätter, 5–6: 564–74.

Flora, Peter (ed.) (1986). Growth to Limits: The Western European Welfare States since World War II, 4 vols. Berlin: De Gruyter.

Goubert, Lucia, and Heylen, Freddy (1999). ‘Loonvorming en de werking van de arbeidsmarkt in België’, in J. J. van Hoof and J. Mevissen (eds.), Loonvorming en werking van de arbeidsmarkt in België. Amsterdam: Elsevier, 87–110.

Gros, David, and Thygesen, Niels (1994). European Monetary Integration. London: Macmillan.

Guger, Alois (1998). ‘Economic Policy and Social Democracy: The Austrian Experience’. Oxford Review of Economic Policy, 14/1: 40–58.

Hancké, Bob (1993). ‘Trade Union Membership in Europe 1960–90: Rediscovering Local Unions’. British Journal of Industrial Relations, 31/4: 593–613.

Hansenne, Michel (1985). Emploi: Les Scénarios du possible. Paris: Duculot.

Hartog, Joop (1999). Country Employment Policy Review: The Netherlands. Report Symposium on ‘Social Dialogue and Employment Success’, 2–3 March. Geneva: ILO.

—— and Theeuwes, Jules (1983). ‘De onstuitbare opkomst van de werkende gehuwde vrouw’. Economisch‐Statistische Berichten, 68: 1152–1157.

Hauth, Anton (1989). ‘Die Arbeitszeit der österreichischen Wirtschaft’. Wirtschaftspolitische Blätter, 1: 18–29.

Heertum‐Lemmen, Ankie, and Wilthagen, Ton (1996). De doorwerking van de aanbevelingen van de Stichting van de Arbeid. Den Haag: SDU.

Hemerijck, Anton (1995). ‘Corporatist Immobility in the Netherlands’, in Colin J. Crouch and Franz Traxler (eds.), Organized Industrial Relations in Europe: What Future? Aldershot: Avebury: 183–226.

(p.258) Henkens, Paul, Siegers, Jacques, and van den Bosch, Karel (1992). ‘Married Women on the Labor Market: A Comparative Study of Belgium and the Netherlands’. Bevolking en gezin, 24/1: 77–100.

Heylen, Freddy, and Van Poeck, André (1995). ‘National Labour Market Institutions and the European Economic and Monetary Integration Process’. Journal of Common Market Studies, 33/ 4: 573–95.

Hochreiter, Eduard, and Winckler, Georg (1996). ‘Advantages of Tying Austria's Hands: The Success of the Hard Currency Strategy’. European Journal of Political Economy, 11: 83–111.

Huyse, Luc (1970). Passiviteit, pacificatie en verzuiling in de Belgische politiek. Antwerp/Utrecht: Standaard.

Jadot, M. (1998). ‘La Politique fédérale de l'emploi, rapport d'évaluation 1998'. Brussels: Ministère Fédéral de l'Emploi, 13–28.

Jones, Erik (1995). ‘Changing the Political Formula: Economic Adjustment and Political Transformation in Belgium and the Netherlands’. Diss. Paul H. Nitze School of Advanced International Studies of the Johns Hopkins University, Baltimore.

Kam, Flip de (1996). ‘Tax policies in the 1980s and the 1990s: The Case of the Netherlands’, in Anthonie Knoester (ed.), Taxation in the United States and Europe. London: Macmillan, 259–301.

Katzenstein, Peter J. (1985). Small States in World Markets: Industrial Policy in Europe. Ithaca, NY: Cornell University Press.

Kersbergen, Kees van (1995). Social Capitalism: A Study of Christian Democracy and the Welfare State. Londen/New York: Routledge.

Kienzl, Heinz (1993). ‘Gesamtstabilität, der Weg und das Ziel—Einkommenspolitik und Währungspolitik seit 1951’, in F. Weber and Th. Venus (eds.), Austro‐Keynesianismus in Theorie und Praxis. Vienna: Stiftung Bruno Kreisky Archiv and Dachs‐Verlag, i. 63–72.

Koole, Rudy A. (1995). Politieke partijen in Nederland: Ontstaan en ontwikkeling van partijen en partijstelsel. Utrecht: Het Spectrum.

Kurzer, Paulette (1993). Business and Banking: Political Change and European Integration in Western Europe. Ithaca, NY: Cornell University Press.

—— (1997). ‘Placed in Europe: The Low Countries and Germany in the European Union’, in Peter J. Katzenstein (ed.), Tamed Power: Germany in Europe. Ithaca, NY: Cornell University Press, 108–41.

Layard, Richard, Nickel, Steve, and Jackman, Richard (1991). Unemployment: Macroeconomic Performance and the Labour Market. Oxford: Oxford University Press.

Lehmbruch, Gerhard (1967). Proporzdemokratie: Politisches System und politische Kultur in der Schweiz und Österreich. Tübingen: Mohr.

Lehner, Gerhard (1991). ‘Die steuerliche Wirtschaftsförderung’. Wirtschaftspolitische Blätter, 4: 472–81.

Lijphart, Arend (1968). The Politics of Accommodation: Pluralism and Democracy in the Netherlands. Berkeley: University of California Press.

Lipset, Seymour‐Martin, and Rokkan, Stein (1967). Party Systems and Voter Alignments. New York: Free Press.

(p.259) Lorwin, Val R. (1966). ‘Belgium: Religion, Class and Language in National Politics’, in Robert A. Dahl (ed.), Political Opposition in Western Democracies. New Haven: Yale University Press.

—— (1971). ‘Segmented Pluralism: Ideological Cleavages and Political Cohesion in the Smaller European Democracies’. Comparative Politics, 3/2: 141–75.

March, Jim, and Olsen, Johan P. (1989). Rediscovering Institutions: The Organizational Basis of Politics. New York: Free Press.

Marx, Ivo (1999). OESO‐recept voor meer werk: Is de weerstand echt gerechtsvaardigd?. Mimeo. Antwerp: Centrum voor Sociaal beleid.

Michel, E. (1994). ‘La Loi sur la compétitivité: Évaluation et proposition’. Reflets et Perspectives de la vie économique, XXXIII‐5: 395–417.

Moden, Jacques, and Sloover, Jean (1980) Le patronat belge. Discours et idéologie 1973–1980. Brussels: Centre de recherche et d'information socio‐politiques (CRISP).

Molitor, Michel (1978) ‘Social Conflict in Belgium’, in Colin J. Crouch and Alessandro Pizzorno (eds.), The Resurgence of Class Conflict, i. National Studies. London: Macmillan, 20–51.

Muehlberger, Ulrike (1998). ‘Atypische Beschärtigung in Österreich, sozial‐ und arbeitsmarktpolitische Implikationen atypischer Beschärtigungsverhältnisse’. Diplomawork, Vienna University of Economics and Business Administration.

NEI (Nederlands Economisch Instituut) (1999). Wordt succes bepaald door de vorm? Onderzoek naar de doorstroom van gesusidieerderde naar reguliere arbeid. Rotterdam: Nederlands Economisch Instituut

Nobelen, Paul W. M. (1983). ‘Nederland: Kwijnend corporatisme en stagnerende verzorgingsstaat’, in T. Akkermans and Paul W. M. Nobelen (eds.), Corporatism en Verzorgingsstaat. Leyden: Stenfert Kroese, 99–142.

Nowotny, Ewald (1989). ‘Die grosse Steuerreform 1988—Analyse und Bewertung’, in Andreas Khol (ed.), Oesterreichisches Jahrbuch für Politik 1988. Munich: Oldenburg, 571–89.

—— (1998). ‘Privatization, Deregulation, Reregulation: Experiences and Policy Issues in Austria’. IB Review, Journal for Institutional Innovation, Development and Transition, 2 (Slovenia): 35–48.

OECD (Organisation for Economic Co‐operation and Development) (1974). Economic Surveys 1973: The Netherlands. Paris: OECD.

—— (1980). Economic Surveys 1979: The Netherlands. Paris: OECD.

—— (1983). Economic Surveys 1981–82: Belgium. Paris: OECD.

—— (1986). Economic Surveys 1985: Belgium. Paris: OECD.

—— (1994). Employment Outlook 1994. Paris: OECD.

—— (1995). Economic Surveys 1994–1995: Belgium. Paris: OECD.

—— (1996). Employment Outlook 1996. Paris: OECD.

—— (1997). Economic Surveys 1996–1997: Belgium. Paris: OECD.

—— (1998). ‘Key Employment Policy Challenges Faced by OECD Countries’, in Labour Market and Social Policy Occasional Papers, 31. Paris: OECD.

—— (1999). Economics Outlook 1999. Paris: OECD.

(p.260) Peeters, Jan (1989). ‘De sociale zekerheid tien jaar later: Naar een crisismanagement’, in BVA (Belgische Vereniging voor Arbeidsverhoudingen) (ed.), Vijftig jaar Arbeidsverhoudingen. Bruges: De Keure and BVA, 197–228.

Pichelmann, Karl, and Hofer, Helmut (1999). Country Employment Policy Reviews: Austria. Report Symposium on ‘Social Dialogue and Employment Success’, 2–3 Mar. Geneva: ILO.

Przeworski, Adam, and Teune, Henry (1970). The Logic of Comparative Social Inquiry. New York: Wiley.

Renan, Paul (1990). La Sécurité sociale: Histoire, développement et perspectives. Brussels: Dossiers du Crisp, No. 38.

Roorda, Wouter, and Vogels, Els (1997). ‘Arbeidsmarkt: Bescherming en prestaties’. Economisch‐Statistische Berichten, 4099: 245–48.

Rothschild, Kurt (1993), ‘Austro‐Keynesianismus aus ökonomischer Sicht’, in F. Weber, and Th. Venus (eds.), Austro‐Keynesianismus in Theorie und Praxis. Vienna: Stiftung Bruno Kreisky Archiv and Dachs‐Verlag, i. 135–45.

Rothstein, Bo (1992). ‘Labor‐Market Institutions and Working‐Class Strength’, in Sven Steinmo, Kathleen Thelen, and Frank Longstreth (eds.), Structuring Politics. Historical Institutionalism in Comparative Analysis. Cambridge: Cambridge University Press, 33–56.

Salverda, Wiemer (1997). ‘Verdringing en arbeidsmarktbeleid voor de onderkant’, in F. Bergman, Joop Hartog, P. Hiemstra, Robert C. Kloosterman, Wiemer Salverda, and Jules Theeuwes, Creëren van werk aan de onderkant. Amsterdam: Welboom: 55–74.

Scharpf, Fritz W. (1988). ‘The Joint‐Decision Trap: Lessons from German Federalism and European Integration’. Public Administration, 66: 239–78.

—— (1991). Crisis and Choice in European Social Democracy. Ithaca, NY: Cornell University Press.

—— (1997). Games Real Actors Play. Boulder Colo.: Westview Press.

Schettkat, Roland (1999). ‘Small Economy Macroeconomics: The Economic Success of Ireland, Denmark, Austria and the Netherlands Compared’. Intereconomics, 34/4: 7–19 (Hamburg Institute for Economic Research: Review of International Trade and Development).

SCP (Sociaal Cultureel Planbureau) (1998). Sociaal Cultureel Plan 1998. Rijswijk: SCP.

Seidel, Hans (1993). ‘Austro‐Keynesianismus–Revisited’, in F. Weber and Th. Venus (eds.), Austro‐Keynesianismus in Theorie und Praxis. Vienna: Stiftung Bruno Kreisky Archiv and Dachs‐Verlag, i. 145–49.

Serroyen, Chris, and Delcroix, Jean‐Paul (1996). ‘Belgium’, in Giuseppe Fajertag (ed.), Collective Bargaining in Western Europe 1995–96. Brussels: European Trade Union Institute, 31–56.

Slomp, Hans (1983). ‘België: Naar “Nederlandse toestanden”?’, in T. Akkermans and Paul W. M. Nobelen (eds.), Corporatisme en verzorgingsstaat. Deventer: Stenfert Kroese, 143–69.

Smits, J. (1983). ‘Belgian Politics in 1982: Less Democracy for a Better Economy’. Res Publica, 25/2–3: 181–217.

Sneessens, Henri R., and Drèze, Jacques H. (1986). ‘A Discussion of Belgian (p.261) Unemployment Combining Traditional Concepts and Disequilibrium Economics’, in C. Bean, Richard Layard, and Stephen Nickell (eds.), The Rise in Unemployments. Oxford: Blackwell, 89–120.

Snels, Bert (1997). Political Mechanisms and Institutional Interaction: Politics in the Dutch Economy. Utrecht: Dissertation University of Utrecht.

Steiniger, Rudolf (1975). Polarisierung und Intergration: Eine vergleichende Untersuchung der strukturellen Versäulung der Gesellschaft in den Niederlanden und Österreich. Meidenheim a/G.: Anton Hain.

SWAV (Stweunpunbt Werkgelegenheid, Arbeid en Vorming) (1995). De arbeidsmarktonderzoekersdag 1995. Verslagboek, Louvain: SWAV, dossier 12.

Swyngedouw, Marc, and Martiniello, Marco (eds.) (1998). Belgische Toestanden: De lotgevallen van een kleine bi‐culturele demoncratie. Antwerp: Standaard.

SZW (Sociale Zaken en Werkgelegenheid) (1995). Sociale Nota 1996. The Hague: Ministerie van of SZW.

—— (1998). Evaluatie van het gebruik van de afdrachtskorting lage lonen, eindrapport. The Hague: Ministerie van of Sociale Zaken en Werkgelegenheid.

Tálos, Emmerich (1996). ‘Gewerkschaften und Sozialpartnerschaft’. Kurswechsel 4, Beirat für Gesellschafts‐, Wirtschafts‐ und Umweltpolitische Alternativen (Vienna), 59–72.

—— and Kittel, Bernhard (2000). ‘Austrocorporatism in the 1990s’, in Stefan Berger and Hugh Compston (eds.), Social Partnership in Europe. Oxford: Berghahn.

—— and Wörister, Karl (1998). ‘Soziale Sicherung in Österreich’, in Emmerich Tálos (ed.), Soziale Sicherung im Wande. Österreich und seine Nachbarstaaten. Ein Vergleich. Vienna: Böhlau Verlag, 209–88.

Teufelsbauer, Werner (1986). ‘Bremsen lockern statt Gasgeben!—Ein Plädoyer für eine seriöse Regulierungsdiskussion in Österreich’. Wirtschaftspolitische Blätter, 6: 708–20.

Therborn, Gøran (1986). Why Some Peoples Are More Unemployed Than Others. London: Verso.

Tichy, Gunther (1985). ‘Wie funktioniert die österreichische Wechselkurspolitik? Konzept, Voraussetzung und Handhabung’. Wirtschaftspolitische Blätter, 5: 493–506.

Tijdens, Kea (1998). ‘De balans van twee ronden arbeidsduurverkorting, 1982–85 en 1994–97’. Tijdschrift voor Arbeidsvraagstukken, 14/3: 212–25.

Toirkens, Jose (1988). Schijn en werkelijkheid van het bezuinigingsbeleid 1975–1986. Deventer: Kluwer.

Toren, Jan‐Peter van den (1996). Achter gesloten deuren? CAO‐overleg in de jaren negentig. Amsterdam: Welboom.

Traxler, Franz (1997). ‘Austria: Still the Country of Corporatism’, in Anthony Ferner and Richard Hyman (eds.), Changing Industrial Relations in Europe, Oxford: Basil Blackwell, 239–62.

Uher, Julian (1993). ‘Fortschritt und Machtspiel’, in F. Weber and Th. Venus (eds.), Austro‐Keynesianismus in Theorie und Praxis. Vienna: Stiftung Bruno Kreisky Archiv and Dachs‐Verlag, i. 78–100.

(p.262) Unger, Brigitte (1996). ‘Sozialpartnerschaft in der Midlife Crisis’. Kurswechsel 4, Beirat für Gesellschafts‐, Wirtschafts‐ und Umweltpolitische Alternativen. Vienna, 73–94.

—— (1998a). Beschäftigungspolitik in der Europäischen Union. Informationen zur Politischen Bildung, No. 15. Vienna: Forum für politische Bildung.

—— (1998b). Room for Manoeuvre. Choices Left for National Economic Policy. Vienna: Habilitationsschrift der Wirtschaftsuniversität.

Van Parijs, Philippe (1998). Should Europe be Belgian?. Paper presented at the Conference ‘The Historical Perspective of Republicanism and the Future of the European Union’, Siena, 23–27 Sept.

Van Ruysseveldt, Joris, and Visser, Jelle (1996). ‘Weak Corporatisms Going Different Ways? Industrial Relations in the Netherlands and Belgium’, in Joris Van Ruysseveldt and Jelle Visser (eds.), Industrial Relations in Europe: Traditions and transitions. London: Sage, 205–64.

Vilrokx, Jacques, and Van Leemput, Jim (1997). ‘Belgium: The Great Transformation’, in Anthony Ferner and Richard Hyman (eds.), Changing Industrial Relations in Europe. Oxford: Basil Blackwell, 315–47.

Visser, Jelle (1990). ‘Continuity and Change in Dutch Industrial Relations’, in Baglioni Guido and Colin J. Crouch (eds.), European Industrial Relations: The Challenge of Flexibility. London: Sage: 199–240.

—— (1995). ‘The Netherlands: From Paternalism to Representation’, in Joel Rogers and Wolfgang Streeck (eds.), Works Councils: Consultation, Representation, and Cooperation in Industrial Relations. Chicago: Chicago University Press, 79–114.

—— (1996). Trends and Variations in European Collective Bargaining. Amsterdam: CESAR Research Paper 96/2.

—— (1997). ‘The Netherlands: The Return of Responsive Corporatism’, in Anthony Ferner and Richard Hyman (eds.), Changing Industrial Relations in Europe. Oxford: Basil Blackwell, 283–314.

—— (1998). ‘Two Cheers for Corporatism, One for the Market: Industrial Relations, Unions, Wages, and Labour Markets in the Netherlands’. British Journal of Industrial Relations, 36/2: 269–82.

—— (1999). The First Part‐Time Economy in the World. Does it Work?. Paper presented at Euro‐Japan Symposium on the Development of Atypical Employment and Transformation of Labour Markets, Japan Productivity Center for Socio‐Economic Development, Tokyo, 24–25 Mar.

—— and Hemerijck, Anton (1997). A Dutch Miracle: Job Growth, Welfare Reform and Corporatism in the Netherlands. Amsterdam: Amsterdam University Press.

Walterskirchen, Ewald (1997). Austria's Road to Full Employment. WIFO Working Papers, 89. Vienna: Österreichisches Institut für Wirtschaftsforschung.

Werner, Heinz, and König, Ingeborg (1987). Maßnahmen zur Förderung der Beschäftigung und zur Bekämpfung der Arbeitslosigkeit in Belgien. Beiträge zur Arbeitsmarkt‐ und Berufsforschung, Nürnberg: IAB, 33–72.

(p.263) Wilthagen, Ton (1998). Flexicurity: A New Paradigm for Labour Market Policy Reform?. Discussion Paper FS 1 98–202. Berlin: Wissenschaftszentrum.

Winckler, Georg (1980). ‘Das Ende der aktiven Nominalzinspolitik in Österreich', Wirtschaftsanalysen, 4: 14–28.

Windmuller, John P. (1969). Labor Relations in the Netherlands. Ithaca, NY: Cornell University Press.

WRR (Wetenschappelijke Raad voor het Regeringsbeleid) (1980). Plaats en Toekomst van de Nederlandse Industrie. The Hague: WRR.

—— (1990). Een Werkend Perspectief: Arbeidsparticipatie in the jaren '90. The Hague: WRR.

Zanden, Jan‐Luiten van, and Griffiths, Richard T. (1989). Economische geschiedenis van Nederland in de 20e eeuw. Utrecht: Het Spectrum.

Notes:

Our joint paper has greatly benefited from comments and suggestions made by the directors of the ‘adjustment’ project, Fritz W. Scharpf and Vivien A. Schmidt. We thank Martin Schludi (MPIfG) and Friso Janssen (University of Amsterdam), who helped with preparing the tables and figures. We have divided our labor as follows: Brigitte Unger wrote Section 5.2 (Austria), Anton Hemerijck Section 5.3 (the Netherlands), and Jelle Visser Section 5.4 (Belgium). Section 5.1 (introduction) was written by Anton and Jelle; the conclusion by the three authors together. Jelle Visser did the final editing.