“The persistent differences between the Netherlands and Belgium illustrate precisely that the European Union does not create uniformity among its member states”, writes former Belgian minister and current professor at the University of Amsterdam Frank Vandenbroucke in an interesting article about these two small welfare states, that can “be successful in the European Union, but must not be ‘small’ in their view of the EU” .
This article, which you can read in full below, is a summary of the Pacification Lecture Vandenbroucke delivered on 11 November 2017 in Breda. The complete lecture can be found, in Dutch, on www.pacificatielezingen.org.
Translated by Lindsay Edwards / Photo © Jeroen Oerlemans
Are we the powerless playthings of forces that inevitably lead to more inequality: globalisation, digitalisation, European fiscal surveillance? Take Belgium and the Netherlands, two small, open economies. You would expect them to feel the impact of globalisation even more than other countries. But that is not the case. We see growing inequality, but less than elsewhere. The middle class is holding up reasonably well. Globalisation and Europeanisation have not turned us into clones. The Belgian and Dutch welfare states are very different – and will continue to be so. The observation is not new: when their small scale is combined with effective social dialogue, small states are highly adaptable. The Dutch Polder model is often praised as the perfect example of this quality.
Guy Tegenbos compared Belgium and the Netherlands in The Low Countries yearbook ten years ago. He pleaded for ‘constructive jealousy’, because the two countries can learn so much from their differences. This appeal has lost none of its relevance. Take poverty. The current domestic debate in Belgium has something ritualistic about it and induces despondency. A comparison with the Netherlands raises pertinent questions. Why is there less poverty among elderly people in the Netherlands than in Belgium? Likewise, why is there less poverty among young people in the Netherlands than in Belgium? And why is poverty increasing among young people in both countries?
As far as the elderly are concerned, I will be brief. The Dutch pension system is not without its problems, but it delivers better results than the Belgian system. But we cannot import it into Belgium, as it is the result of a long history. Belgium should improve pensions for employees in the long term by reforming the existing pay-as-you-go system. Unfortunately, though, constructive debate about the issue is just not getting off the ground in Belgium. This poses a huge threat to the Belgian model of social dialogue.
Why is there less poverty among the under-60s in the Netherlands? The explanation for this difference lies entirely in the specific category of households whose adult members are rarely (or never) in employment, notably households with children that are characterized by ‘very low work-intensity’. With or without children, households whose link with the labour market is weak run a bigger risk of poverty in Belgium than in the Netherlands. That is not surprising: the level of Dutch social assistance benefits compares favourably to the level of income protection benefits in Belgium. But not only is the risk of poverty higher in this group of households in Belgium, the group itself is also bigger. In Belgium there are more families with children whose parents are not in employment, or only to a very limited extent. That is the main reason why the poverty figures are better in the Netherlands. This Belgian labour market problem is much more salient in Wallonia and Brussels than in Flanders. But that is no reason for Flemings to be complacent. Benefits are low in Flanders too, and the income gap between families that have a strong position in the labour market and families whose position is weak is large by other countries’ standards. A broader approach to poverty, whereby not only income but also health, education, housing and non-material factors are taken into consideration, shows that increasing numbers of the youngest children in Flanders are disadvantaged.
The sad conclusion that there are a lot of Belgian households without working adults is linked to the position of low-skilled workers. In that, too, both countries differ. Low-skilled workers in the Netherlands are more likely to have a job, albeit very often a part-time job. So, should Belgium adopt the Dutch model?
The Dutch labour market: a moving target
The Polder model owes its fame to the spectacular drop in unemployment in the 80s and 90s. Negotiations, coordination and consensus building were crucial factors in this success. Besides, the Netherlands has a tradition of major reforms, whereas Belgium has always done things in small steps. This does not mean that there were no social conflicts in the Netherlands, nor that reforms always went according to plan. We should not idealise them. Persistent wage restraint, based on the 1982 Wassenaar Agreement, was key to the economic success. Restraint led to an unheard-of trade surplus, a mercantile success unrivalled in the rest of Europe. Minimum wages were reigned in even more than average salaries. Hence, labour’s share of national income fell, and wage inequality increased. However, the growing number of two-income households buttressed purchasing power at the level of households. Women went out to work massively in part-time jobs. After a rise in the 80s, income inequality at the family level remained limited.
Today, the prevalence of part-time work is not the only difference between the Netherlands and other countries. It also has one of the most flexible labour markets in Northwest Europe. In the 90s, flexibility was an important political theme, but the influence of the social partners was significant as well. Their 1996 Flexakkoord formed the basis of later legislation. There was to be more flexibility and less security for regular employees, and less flexibility and more security for flexible employees. Flexibility really took off, to the extent that in 2013 the social partners jointly concluded that it was ‘overshooting’. Despite the continuing importance attached to collective bargaining in the Netherlands, the poldermodel lost its grip on this dynamic.
The Netherlands has a sizable segment of low-paid jobs: 18% of Dutch wages are less than two-thirds of the average wage. In Belgium fewer than 4% fall into that bracket. There is a particularly big difference in the under-30s. In the Netherlands 47% of under-30s are in low-paid jobs; in Belgium only 13%. In the Netherlands minimum wages for young people are much lower than minimum wages for those aged 21 and over. In Belgium the difference between the minimum wage for 18-year-olds and the minimum wage for older workers was abolished in 2015. So, it is not surprising that the employment rate among young Dutch people – often students with jobs on the side – is extremely high, whereas in Belgium it is low.
Contrary to what is sometimes asserted, it is not the case that low-paid jobs invariably lead to ‘working poor’. In many cases these jobs are done by people who form a household with someone else who has a better paid job, so that the household income is adequate; whether or not someone is poor depends on the household income. This does not mean that this model has no side effects though. But more about that later.
Should Belgium copy this model? The reasons for the superior Dutch poverty record were already apparent in the early 90s, before flexibility got out of hand. There were fewer families where nobody was in employment, and the difference between the two countries increased steadily during the 90s, in favour of the Netherlands. Income protection for families without work was better in the Netherlands than in Belgium. However, while Dutch benefits are still better than Belgian benefits the difference has narrowed. In the 90s the Netherlands gave priority to employment promotion, while benefits became less generous. In Belgium, too, there was an erosion of benefits in the 90s, but it was less severe.
Dutch policy also contained the seeds of greater inequality between active workers. Ongoing wage restraint and the lowering of the lowest pay scales led to a bigger wage gap, because in practice there was less restraint on higher salaries. The combination of part-time jobs with full-time jobs ensured a redistribution across families, but at the same time the incomes of dual-income households grew more than those of single earners. The result was not that inequality increased sharply, at least not as it is traditionally measured, but if we concentrate on the highest and lowest incomes, we see growing inequality.
Limits to Dutch success
At a certain point any model hits a ceiling. It looks as if the percentage of Dutch families where nobody works reached its floor at the end of the 90s. Nowadays we see a similar pattern in the development of poverty in Belgium and in the Netherlands. It is systematically increasing, not in households with a strong or a relatively strong attachment to the labour market, but in households with a very low work-intensity. The Netherlands was in a better position to start with, but both countries are alas moving in the same direction now.
Increasing inequality between the highest and lowest incomes is not the only side effect of this growth model, based on wage restraint and hyper-flexibility. Cok Vrooman, a researcher at the Dutch Social and Cultural Planning Office, has charted how 35 years of change in social protection and the labour market has reduced security. ‘Participating in insecurity’ affects some social groups worse than others. It is also questionable whether allowing large numbers of students to work on the side is a good choice for society. I would not like to see Belgium move in this direction.
A flywheel effect emerges: as the sharing of social risks and incomes within households reduces the importance of traditional individual protection mechanisms, such as the minimum wage and the lowest pay scales in collective agreements, these classic protective mechanisms receive less attention. But that logic also has its limits: the erosion of traditional standards in the labour market does lead to greater inequality amongst workers. Despite the continuing importance of collective bargaining in the Netherlands, the Polder model has to a certain extent lost its grip on this dynamic. The conclusion in the 2013 social agreement, 17 years after the 1996 Flexakkoord, that there is now an ‘overshooting’ in flexibility, testifies in a certain sense of that loss of control.
The realisation that more and more people are confronted with an insecurity that they have not chosen, has become a policy issue – much more than inequality. After Rutte II, Rutte III has announced new measures to find a balance between security and flexibility. Whether these measures are strong enough, I cannot say. They do illustrate the widely-shared awareness that flexibility in the Dutch labour market has gone too far. But perhaps a deeper reflection on the underlying growth model is required. The conclusion of a number of Dutch researchers is that the capacities of the poldermodel are no longer in step with developments in the economy, the labour market and politics, and that it needs modernising.
‘The traditional storyline, that an internationally competitive business sector stimulated by wage restraint is the recipe for a well-functioning economy and labour market, is no longer sufficient. More attention needs to be paid to stimulating internal consumption, reversing the growth of inequality and dealing with private debt and vulnerability to shocks in the international financial system. More priority should also be given to raising job quality, productivity and innovation.’ (Quoteed from: Maarten Keune (ed.), Nog steeds een mirakel? De legitimiteit van het poldermodel in de eenentwintigste eeuw [Still a Miracle? The Legitimacy of the Polder Model in the 21st Century], Amsterdam University Press, 2016, download here)
Reformers in the North, caution in the South?
The Dutch example proves that it is possible to grant better income protection to the unemployed than Belgium does. Belgian benefits are too low. Not coincidentally, activation has been high on the Dutch agenda for many years. What is interesting is that this occurs in an emphatically decentralized framework and local governments have a major role to play in it. This experience may hold interesting lessons for Belgian policy-makers. In 2015, Belgium decentralized active labour market policy to its three Regions: Flanders, Wallonia and Brussels. At the same time, a growing portion of social policy is quietly being de facto decentralized to local municipalities. This quiet decentralization is the result of the increasing relative importance of social assistance in the income support for non-employed persons; social assistance (the so-called leefloon or revenue d’intégration) is managed by municipalities in Belgium.
Unfortunately, there is no debate in Flanders about how this new, decentralized landscape can be made successful in terms of the war on poverty and integration into the labour market, and how the policy capacity of municipalities can be reinforced and embedded in networks. The Belgian Regions have received more and more competences in recent decades. But Flanders now behaves like a particularly centralistic young nation state.
The Dutch are reformers. Sometimes they go too far, but one cannot deny that their reforms are always accompanied by much research and consultations. Reforms in Belgium are seldom systematic and thorough. Taboos are rarely discussed properly, with a true open mind on all sides of the debate. Could a more flexible labour market with lower wage costs for low-skilled workers in Belgium create opportunities for people who are unable to find work now, without such policy going too far, as has happened in the Netherlands? Could we, with a more ambitious tax shift, further reduce wage costs for low-skilled workers, without affecting the purchasing power of those doing low-skilled work?
It would be good to have an open debate about such questions. By way of contrast, giving people who already work four-fifths or more the opportunity to work flexibly on the side, as the Belgian government is proposing now, is an example of an ad hoc measure, which will not resolve the true problem at hand, that is, a lack of market-based jobs for people who are currently without work or without sufficient work. In addition, we need to invest more in jobs in the ‘social economy’ (initiatives in the realm of supported employment, ‘neighbourhood jobs’, etc.). The regular private market offers more opportunities than we valorise in Belgium, but the policy motto that everyone must transit to ‘the market’, which is catching on in Flanders, is naive.
A new European wind
Flexibility should not be disparaged, but neither should it be a goal in itself. Welfare states should ensure stability: macro-economic stability, but also a certain stability in people’s lives. The two are linked: income protection for people who lose their jobs addresses individual vulnerability but also functions as a macro-economic shock absorber.
How stability and flexibility are related to each other is key to the debate about the future of the European Monetary Union. When the European Monetary Union was launched, the idea that its members needed labour market flexibility was central to the European Employment Strategy, as was the belief that flexibility and individual security would go hand in hand if individuals’ skills were developed and activated. With regard to social dialogue, member states were told that they must, above all, decentralize their systems of collective bargaining.
Today, another wind is blowing in the European Commission, witness a series of recent communications. The currency union needs automatic stabilizers at the Eurozone level, for instance via a ‘reinsurance’ of national unemployment benefits systems. Stability is needed in labour’s share of national income; and German and Dutch wages need to rise faster. That is why coordinated wage negotiations are needed (rather than pure decentralization). Decent minimum wages and unemployment benefits are needed in all the member states, and every European who works should have access to social security. A very different set of messages!
So, what are the positions of Belgium and the Netherlands in this new debate? The Rutte III coalition agreement wants to exclude more or less all the concrete solutions that the European Commission suggests for stabilizing the monetary union based on any form of budgetary solidarity. Thus, the Netherlands is making it difficult for itself to play a creative role in European negotiations.
The current Belgian government’s reaction to the suggestion that stabilization of the currency union demands more solidarity is extremely cautious as well. The so-called eurorealism of the Flemish nationalist coalition partner, N-VA, limits the Michel government’s room for manoeuvre. But the problem in Belgium is broader than just one party. For the last ten years, Belgium’s position in Europe has mainly been to defend what the country already has, be it social rights or tax loopholes. This does not mean that the Netherlands and Belgium do not play a role in social debates in the European Union. Both countries were active in the reform of the Directive on Posted Workers, on which an agreement is now reached. Likewise, the Netherlands and Belgium certainly support the principles of the European Pillar of Social Rights, and the social model underlying it. And we expect Dutch and Belgian policymakers to endorse the conversion of the excellent principles of the Pillar into tangible policy – which is indeed a crucial challenge now. But when a demand for more European solidarity is attached to it, the Belgian and Dutch governments fail to respond.
A new balance
Yet, we will have to find a new balance in the European Union between the responsibilities of the individual member states and solidarity between them. In the 90s the Benelux countries acted together as bridge builders in Europe during such debates. Since 2000, that tradition has disappeared. Nonetheless the Dutch and Belgians remain well-placed to play this role and to play it together. Belgians often have some affinity with French positions in EU debates, while the Dutch understand the Germans. However, in order to valorise this bridge-building capacity rather than being defensive, they need to play a proactive role. One must also assess correctly how individual nations’ social challenges fit into a broader European debate. Sometimes, I am baffled by the way Dutch public intellectuals deal with this European debate, as if the identity of the Netherlands is endangered by every attempt to bring about more solidarity, and thereby more stability, in Europe.
The persistent differences between the Netherlands and Belgium illustrate precisely that the European Union does not create uniformity among its member states. Nor will it do so in the future. However, our debate about our social future is partly a European debate. The Union offers a framework within which we can learn from each other. We often fail to see that and to benefit from it. And the potential of European cooperation does not stop with mutual learning. Reforms are necessary for the monetary union to support its member states in a key social function: ensuring stability.
Today Belgium and the Netherlands are both, in their own way, acting mainly defensively in the European social debate. A defensive position with regard to the European Union is not a safe position, on the contrary, in the long term it is risky. Small welfare states can be successful in the European Union, but they must not be ‘small’ in their view of the European Union.