VoxEU Column EU policies Europe's nations and regions Welfare state and social Europe

Social Europe revisited

Almost 20 years after CEPR published “Social Europe: One for All?”, Social Europe has moved again to the top of the policy agenda. In this column, two of the authors revisit their report and argue that the challenges posed by the Global Crisis, the deepening of the inter­nal market, globalisation, technological progress, popu­lation ageing, and the refugee crisis now require a more effective strat­egy to strengthen the EU social acquis.

Almost 20 years ago, the eighth CEPR Monitoring European Integration report explored the interactions between social policy, broadly interpreted, and economic integration in Europe. The experience of the past enlargement of the European Community to include Greece, Portugal, and Spain from the early to the mid-1980s provided Social Europe: One for All? (Bean, Bentolila, Bertola, and Dolado, 1998) the hindsight to try to divine the future of social policy during the upcoming Eastern enlargement of 2004. Predictions were made about how the “social dimension” of the EU would evolve following the increasing economic integration in its various forms, from trade liberalisation to enhanced capital and labour mobility.

Download the publication for free here (create a VoxAccount if you don't already have one)

In a nutshell, the basic diagnosis of the report was that economic integration of economies at different levels of development, while generally a good thing, usually has adverse consequences for relatively inefficient producers, both in the former and newer member states. Hence, integration would lead to increasing demands for higher social protection. Yet, the effectiveness of national social policies would be affected by the degree of economic integration, so that the supply of these policies would decrease. As a result of these opposite forces, the ‘price’ of social protection would increase, whereas its ‘amount’ could move either way.

These trends in demand and supply were expected to impose further strain on the European model of social protection and labour market regulations. However, given the resilience of national systems of industrial relations and of social expenditure, and the extent to which they represent genuine differences in national preferences, a race to the bottom in social protection and labour market standards was neither expected nor advocated. Thus, harmonisation of minimum standards in those social policies that did not favour privileged population groups at the expense of disadvantaged members of society was deemed essential. For the richer member states, coordination could take the form of harmonisation of those standards which appear optimal within each of these economies. Yet, for countries with different tastes or lower levels of development, common standards would involve foregoing gains from trade and cause harm to the ‘furthest poor’, even if aimed at protecting the ‘nearest poor’ in the richer countries. Therefore, diversity was indeed desirable and the introduction of the full social acquis communautaire of the EU for the newer member states should only take place gradually, and after their living standards had approximated those in the West.

Indeed, taking stock from historical experience until the onset of the 21st century, the original debate in the 1950s as to whether social policy was a requirement or a consequence of economic integration was de facto resolved in favour of the latter. On the one hand, this happened because wage harmonisation was rejected as a prerequisite for a single market and, on the other hand, because working conditions (parental leave, lack of sex discrimination, equal treatment of part-time employees and consultation with worker representatives about management practices in large multinational firms) could be harmonised to prevent ‘social dumping’. Yet, even when EU-wide norms were implemented, enforcement has not been very rigorous. As a result, progress in developing an EU social policy has been much slower than in other areas such as trade or monetary policy.

The effects of trade and technology

Increasing globalisation in the form of trade in goods and services represents the first challenge to the EU equilibrium. As discussed in the report, traditional theories point to three reasons behind the gains from specialisation and trade: (i) comparative advantage due to technological differences, (ii) comparative advantage due to relative abundance of different factors of production, and (iii) economies of scale within industries leading to ‘intra-industry’ trade. In all these theories, the gains from international trade accrue through lower goods prices and changes in these prices affect the purchasing power of factor rewards. Since trade affects incomes as well as goods prices, economic integration would rarely make all individuals better off than under autarky. Thus, factor mobility and redistributive issues arise.

For instance, under technological differences, skilled workers in technologically backward (‘poor’) countries, who now have to specialise in the production of less sophisticated goods, bear retraining costs and the only way to improve their fortunes is to migrate to technologically advanced (‘rich’) countries. Thus, migration and trade become complements and the corresponding brain drain reduces inequality in poor economies, in parallel with a rise in poverty (unless remittances offset the loss of the more able workers). In the richer host countries, immigration may displace those workers who, prior to trade openness, were working in the least sophisticated sectors, leading to higher inequality. In a sense, these effects are similar to those of technological change. From the view point of displaced workers, imports are indeed ‘too cheap’, just as the new machines were ‘too productive’ to the old and new Luddites.

When it is based on differences in factor endowments rather than in technology, trade generally raises (lowers) the rewards of relatively abundant (scarce) factors. Since skilled labour becomes less scarce in the integrated economy than in the poor countries, and the opposite is true of unskilled labour, inequality decreases in these countries while it increases in rich countries. The same reasoning applies to capital and labour: the capitalists in the rich countries can take advantage of the scarcity of capital in the integrated economy. In addition, the degree of income inequality between rich and poor countries declines, so that factor mobility and trade become substitutes rather than complements. These effects, however, are bound to be much weaker if technological know-how and geographical and institutional features make richer member states in the EU more appealing locations for production than the poorer member states. Any absolute advantage may reverse the direction of factor flows regardless of the current endowment of (skilled-unskilled) labour and capital in those countries. Hence, the role of EU (cohesion/structural) funds to accommodate new members becomes essential. From the point of view of the incumbents, the potentially adverse effect on their less-skilled workers could be mitigated if these funds could help to raise living standards (and labour costs) in the new entrants, as well as reduce their incentives to migrate. From the point of view of newcomers, the potential loss of competitiveness arising from the adoption of the social acquis would be compensated by direct income transfers.

Twenty years after

Two decades after the report was published, Social Europe has moved again to the top of the policy agenda, but it has not been especially due to EU enlargement, as envisioned in the report. Rather, the debate on the pros and cons of the social acquis is now subject to four major challenges, which have simultaneously increased the demand for protection and limited the ability to supply it. First, inter-industry trade with developing countries, especially with China, has resurged. Second, technological change also plays a prominent role but it has changed shape, from its initial bias in favour of skilled workers to a wave of routinisation that progressively allows machines to carry out tasks that previously could only be performed by humans. This affects particularly workers in the middle of the skill distribution, i.e. job polarisation (Goos et al. 2014). Third, the global financial slump initiated in 2008 plus the subsequent debt crisis that hit in 2011 (i.e. the double-dip Great Recession) have shown the limits of the willingness of richer EU nations to mitigate the effects of the recession on the so-called ‘programme countries’. And fourth, the refugee crisis has severely tested the principle of free mobility of persons across the EU. Indeed, the current wave of nationalism in many European countries (and in the US) can be seen as resulting from these developments and their impact on welfare and the distribution of income.

To discuss how this set of forces has worked out in practice, it is useful to divide the EU28 (Brexit notwithstanding) into five blocks according to their social policy models (see Bertola et al. 2001): (i) Nordic (Denmark, Finland, and Sweden, plus the Netherlands, which can be assimilated to this area); (ii) Anglo-Saxon (Ireland and the United Kingdom); (iii) Continental (Austria, Belgium, France, Germany, and Luxembourg); (iv) Mediterranean (Greece, Italy, Portugal, and Spain); and (v) Central-Eastern Europe (Czech Republic, Estonia, Latvia, Lithuania, Poland, Slovak Republic, and Slovenia).1 For brevity, they will be labelled in what follows as the ND, AS, CT, MN, and CE blocks, respectively.

According to the OECD Income Distribution Database,2 the Gini coefficient – which measures inequality of household disposable income – in 2004 (the first available year in this harmonised database for a subset of EU-28 countries) was highest in the MN block (33.4%), followed by the AS block (32%), and the CE block (30.9%), with the CT and ND blocks ranked at the bottom (28.5% and 26.2%, respectively). A similar ranking holds regarding poverty (measured by the fraction of the total population with household income below the median): 13.1% (MN), 10.8% (AS), 10.6% (CE), 8.2% (CT), and 6.9% (ND). It may look surprising that inequality in the first half of the 21st century was higher in southern Europe that in the deregulated labour markets of the AS countries, but this is the outcome of their highly regulated dual (‘insider-outsider’) labour markets leading to high unemployment rates.

By 2014 (the latest available date), inequality had gone up in the MN block (34.0%), particularly in Greece and Spain, where the internal devaluation imposed by the Troika through the dismantling of their rigid wage-setting systems has meant an increase of almost 1.5 percentage points in their Gini indexes. By contrast, inequality has gone down in the AS and CT blocks (31.3% and 28%, respectively) and it has hardly changed in the remaining two blocks: 30.7% (CE) and 26.4% (ND). Poverty has gone up everywhere by about one percentage point, except in the AS block (9.0%). To the extent that ND, AS, and CT countries are richer that the CE and MN countries, at first sight falling inequality in the former and raising inequality in the latter seem supportive of a theory of comparative advantage based on relative factor abundance. Yet, given the increase in trade, the large migration flows from CE and MN blocks to the other blocks are seemingly inconsistent with this reasoning, except for the role played by the Great Recession in triggering larger migration flows, which were aggravated by the refugee crisis.

The slump has meant a big test for all those countries (especially those in the MN block) where labour market regulations were designed to favour specific groups of insiders (e.g. workers under permanent contracts in Spain or public employees and some groups of self-employed in Greece), as unemployment and poverty have surged. Since it is difficult to justify such policies on either distributional or efficiency grounds, the Great Recession has played a cleansing role in dismantling some (but not all) of these inefficient regulations. The direct cost, however, has been mainly borne by the weaker segments of society (youth, migrants, and older less-skilled workers). In turn, some of the offsetting policies that worked well in Germany and other CT countries – such as working time reductions, temporary wage cuts, and dual vocational training – are not fully applicable to the MN or CE countries precisely because their inefficient institutions led to specialisation in industries that were permanently hurt by the crisis (e.g. construction) or because of the large share of SMEs in these economies.

In this sense, it still seems paramount that a more stringent social acquis is not imposed too quickly on them. They should be allowed to clean their debris by fostering policy reforms that were discouraged by a common monetary policy that reduced the inflation bias (Sibert and Sutherland 2000). Reforms in the direction of more decentralised systems of wage determination are key, since centralised wage bargaining at the national level arguably becomes equivalent in the wider European context to decentralised wage bargains between employers and powerful unions. Likewise, highly dual employment protection should be ameliorated, allowing temporary jobs to be used exclusively for non-regular and replacement jobs. Recent reforms of this kind in the MN and CE blocks are steps in the right direction, although still far from being fully satisfactory. On the other hand, significant help from EU institutions to cope with the social impact of the recession has essentially come via country-specific bailout programs subject to strict conditionality rather than from any EU-wide income insurance mechanism.

Looking ahead

Despite the fact that the EU has gradually built up a social dimension that has somewhat improved living condi­tions and social convergence among its member states, the challenges posed by the crisis, the deepening of the inter­nal market, globalisation, technological progress, popu­lation ageing, and the refugee crisis now require a more effective strat­egy to strengthen the EU social acquis. Since convergence in employment rates and living standards has slowed down considerably in recent years, enhanced social cooperation ought to involve further harmonisation of basic standards – such as health and safety in the workplace, academic degrees recognition, retirement age, and, yes, corporate taxes – and gradualism in the implementation of these agreements in the less-advanced economies.

In the context of the design of a post-Brexit EU, the Commission has proposed a European Pillar of Social Rights (2017a). One of its goals is to increase the EU social budget, that currently represents only 0.3% of total public social expenditure in the EU. The proposal includes a Social Scoreboard to keep track of individual country progress that is already available online. A companion document (EU Commission 2017b) lists the existing acquis and the proposals included in the Social Pillar, though the latter remain quite underdeveloped and far from creating a situation in which the EU norms become more stringent than national regulations.

Concrete policy actions to push Europe towards a “social triple-A rating", in the words of its President, Jean-Claude Junker, could be as follows:

  • The implementation of an automatic stabiliser mecha­nism (e.g. an EU unemployment insurance fund with experience rating), to ensure that business cycle shocks do not lead to deep structural imbalances, as happened during the euro area crisis (Ábrahám et al. 2017).
  • An active role through structural/cohesion funds and recent initiatives such as the Youth Guarantee and the Social Investment Package (which currently are not working well enough) to reduce large disparities between member states as regards th­­eir human-capital investment policies (early childhood education and care, education policy, active labour market policies, etc.).
  • The enhancement of the exchange of information and administrative coopera­tion among countries to avoid social tourism and social dumping.
  • Limitations on the use of EU funds for those member states that break agreements on free labour mobility or the reception of refugee quotas, and the use of a ‘positive conditionality’ approach (Boeri and Jimeno 2016) ­­for countries that take steps to dismantle regulations that protect the privileged position of insiders rather than the disadvantaged members of society.
  • The use of qualified majority rather than unanimity in all issues subject to harmonisation of minimum standards.

We must acknowledge that, with heightened globalisation and technological progress, the challenge of increased demand for social protection and reduced scope for supplying is even stronger now than two decades ago. Still, as the European Commission (2017) stresses in a recent reflection paper on Europe's social dimension, there are two alternatives: “to embrace and direct change, or be driven by its current problems… however, every European country is striving for the same thing: to create a fairer society based on equal opportunity”. The time for new politics on the EU social dimension has arrived and a deep discussion on where and how our societies want to go cannot wait much longer.     


Ábrahám, A, J Brogueira de Sousa, R Marimon and L Mayr (2017), "On the Design of a European Unemployment Insurance Mechanism", European University Institute, mimeo.

Bean, C, S Bentolila, G Bertola, and J Dolado (1998), Social Europe: One for All?, Monitoring European Integration 8, London: CEPR. 

Bertola, G, T Boeri, and G Nicoletti (2001), Welfare and Employment in a United Europe, Cambridge, MA: MIT Press.

Boeri, T and J F Jimeno (2016), “Learning from the Great Divergence in Unemployment in Europe during the Crisis”, Labour Economics 41, 32-46.

European Commission (2017a), "Reflection Paper on the Social Dimension of Europe", COM(2017) 206 of 26 April.

European Commission (2017b), Commission Staff Working Paper accompanying the document "Establishing a European Pillar of Social Rights", SWD(2017) 201 final.

Goos, M, A Manning, and A Salomons (2014), "Explaining Job Polarization: Routine-Biased Technological Change and Offshoring", American Economic Review 104, 2509-2526.

Sibert, A and A Sutherland (2000), “Monetary Regimes and Labour Market Reforms”, Journal of International Economics 51, 421-435.


[1] The remaining member states of the EU28 (Bulgaria, Croatia, Cyprus, Hungary, Malta, and Romania) are not included since they only appear recently in IID (2016). 

[2] http://www.oecd.org/social/income-distribution-database.htm

1,680 Reads