Labor market institutions and policies in old and new EU members Updated

After three recessions, a new emphasis on the importance of collective institutions and social dialogue is emerging

Università di Bologna and IZA, Germany

one-pager full article

Elevator pitch

Old and new EU member states still adopt quite different labor market institutions and policies: convergence has been partial and limited. Nevertheless, a new agreement is spreading on the importance of well-developed, coordinated institutions, supported by social dialogue, in view of the increasing challenges posed by the macro economy and by the increasing fragmentation of labor markets.

During each recession, the adjustment burden in the Euro area was more on hours worked, less on job losses

Key findings


In new EU member states, employment protection for temporary contracts has increased, converging towards that of older EU members.

Employment protection laws cushion the impact of negative macro shocks on workers’ incomes and reduce job losses.

If increased bargaining coverage is associated with greater bargaining coordination, this has positive effects on employment and price stability.

In fragmented labor markets, collective bargaining institutions may improve the functioning of labor markets.

Job retention schemes were widely promoted during the COVID-19 recession, providing a strong cushion against job losses.


Collective bargaining coverage has continued to decrease since 1985 in most European countries.

Expenditures on active labor market programs are contracting in old EU member states; they are constant but much lower in new members.

Until 2019, expenditures on unemployment benefits and related measures diminished throughout the EU.

Fragmentation of labor markets and increasing incidence of negative macro shocks pose hard challenges to labor market policies.

Some countries have not yet adopted provisions for adequate minimum wages nor for coordinating them with collective bargaining.

Author's main message

Differences in trade union density (the proportion of workers who are union members), collective bargaining coverage and expenditures on labor market policies between new and old member states have increased since 2000. On the other hand, with respect to employment protection laws there has been some convergence between the two groups. Especially since the COVID-19 recession, there is a new understanding for the importance of social dialogue and coordinating institutions for collective bargaining. Additionally, the macro-stabilization function of fiscal policies has evolved in this direction, by subsidizing temporary but widespread job retention schemes.


Labor market institutions as well as scholarly and policy debates about them have gradually evolved during the last three decades. One interesting lens through which to examine this topic is via differences in labor market institutions between (and occasionally within) EU “new” member states (that is, those that have transitioned from centrally planned to market economies) and older member states, and how they have evolved since the start of the 21st century.

It is likewise helpful to examine recent evidence and policy debates, up until the COVID-19 pandemic, on how labor market institutions may affect the functioning of labor markets.

Discussion of pros and cons

Labor market institutions in the new and old EU member states

Labor market institutions (LMI) influence the labor market through multiple channels. They affect wage determination, the pricing of different types of labor, firms’ hiring and firing decisions, the size and duration of unemployment, the search behavior of workers and their sectoral allocation.

In addition, by interacting with other structural characteristics of the economy, LMI may influence the evolution of many other economic variables: the determination of prices and inflation, the competitiveness of firms, the balance of net exports, the investment decisions of firms, the flow of foreign direct investments, and the duration and amplitude of business cycles.

Differences in LMI between new (former socialist) and old EU member states originate in the countries’ different histories and follow a clear pattern. New member states typically exhibit much lower wage coordination and union density, and smaller expenditures on labor programs. Only employment protection measures are, at least on paper, equivalent between old and new member states. In general, new member states show lower levels of institutional or policy interference with the functioning of labor markets, which also implies that displaced workers receive much less protection. The majority of labor contracts are bargained in a decentralized setting, meaning labor unions have limited roles and powers in new member states.

Changes in labor market institutions: a look at the facts

Union density and collective bargaining

Panel A of Figure 1 shows that, on average, trade union density has decreased from 2000 until 2018 in most EU countries. Main exceptions are France (roughly constant around 10.8%) and Iceland. The latter reached a “low” of 84.1% in 2005, but then increased to 90.7%, the highest density in the sample.

Four measures of labor market institutions

The decrease has been larger in new member states (Panel A, row 3), all of which have now reached densities between 5.9 and 13.4%. The only exception is Slovenia: despite having more than halved, it still has a membership rate of 23.8%.

In the same period, also collective bargaining coverage has continued to decrease, albeit with several exceptions (as in Austria, Belgium, Finland, France, Iceland, Italy, Switzerland and Sweden). The most dramatic fall in coverage took place in Greece (from 100% in 2011 to 14.2% in 2017). This can be attributed to the implementation of a series of labor market reforms, which opened the way for employers to defect from multi-employer bargaining arrangements [1].

Coverage has also decreased in new member states, with the exception of Latvia. In Slovenia, a marked decrease in the first part of the sample (-30%) has been partly compensated in the latter period (+8.6%); overall Slovenia remains among the countries with the highest coverage (78.6%), and is even more of an outlier among new member states.

Regulations and policies: trends until COVID-19

As shown in Panel B of Figure 1, Employment protection laws (EPL) for regular (or permanent) contracts have been moderately eased between 1998 and 2008, and then also again until 2019 (these indicators are a weighted sum of sub-indicators concerning regulations for both individual and collective dismissals for workers with a regular contract). In the latter period, more pronounced reductions have taken place in Portugal, Italy and Greece, while Belgium and Netherlands have moved in the opposite direction.

On average, new EU member states have eased protection laws comparatively more than other European countries: in this respect, they have converged to the same level of older EU members. Still, there are differences in this group: Estonia and Hungary are among the less protective countries, while Slovakia, Czechia and Latvia are among the more protective ones.

With respect to temporary contracts, EPL indicators measure the strictness of regulation on the use of fixed-term and temporary work contracts. Protection has moderately decreased in most old EU countries, at least until 2008. The greatest decrease has been in Greece, although it still remains above the group average.

On the other hand, new member states have converged from below average towards similar levels of protection of the older members. This applies to Czechia, Estonia, Hungary, Poland and Slovakia, while Lithuania has moved in the opposite direction. Estonia is now among the European countries with the most stringent protections for temporary workers.

Figure 2 shows public expenditures on labor market programs. The figure distinguishes between active and passive policies: the former include measures to foster labor market inclusion (such as training programs) while the latter include unemployment benefits. In the EU, expenditures on active policies have decreased on average (as a ratio to GDP) until 2008, and then oscillated between 0.58 and 0.75% thereafter. Denmark is by far the top country, with 1.86% in 2019. Among new EU member states, in 1998 four countries had zero or negligible expenditures (Czechia, Estonia, Latvia, Lithuania): they all reached levels between 0.2–0.3% of GDP in 2017, at par with Slovenia and Slovakia; above them are Poland (0.42%) and Hungary (0.85%).

Public expenditure on labor market programs

Among old EU members, expenditures on passive measures decreased on average (as a ratio to GDP) until 2007–2008. Then, after a sharp increase in 2009–2010 (not shown in the figure), they have slowly receded again towards the previous levels. New EU member states have followed a similar path, but with a constantly lower level of expenditures (approximately 40% of the old EU countries). The largest spenders in 2019 were France and Spain, both above 1.5% of GDP.

Summing up – until COVID-19

OECD data on labor market institutions and policy expenditures examined in the previous section have been, at the time of this writing, only updated to the pre-COVID-19 period. Summing up these data and their trends, one can observe:

  • Since 1980, a continued decline in workers' unionization rate and (in a more limited way) in the incidence of collective bargaining. The gap in these indicators between new and old EU member states has further increased. From the latest available data, union density is about 12% in new EU member states, and 32% in old ones; collective bargaining covers approximately 1/4 of the eligible labor force in new members, and 3/4 in old ones.

  • Fairly modest easing of employment protection laws for regular job contracts in both new and old EU member states; in this matter, the averages of the two groups have converged. Similar convergence is also observed for temporary contracts. Overall, employment protection laws are now rather similar, especially for regular contracts.

  • Decline in public expenditures on both active and passive labor market policies, both in old and (to a lower extent) new EU member states. Overall, new members spend considerably less (as a ratio to GDP) on both passive and active labor market policies.

Some of these facts are possibly related to other structural changes that have been taking place at the same time, such as the decline of manufacturing, new technological challenges and the expansion of non-standard forms of work. On the other hand, these and other structural changes (such as greater reliance on global value chains, population ageing and the energy transition) pose difficult economic and social challenges, which would call for an increase in social dialogue and require new policy actions.

Which labor market institutions work best?

Macroeconomic outcomes of wage coordination and union density

In the 1980s, two hypotheses competed to explain the macroeconomic consequences of alternative bargaining arrangements [2]. The “corporatist hypothesis” posited that centralization and wage bargaining coordination improved economic outcomes. This was based on the notion that induced wage restraint led to increased employment and lower inflation.

A counter-argument, known as the “hump-shaped hypothesis”, claims that in a highly decentralized labor market, equivalent moderating effects on wages could be realized via competitive pressures on the labor supply side. According to this view, the worst scenario (characterized by high inflation) would be reached with an intermediate level of centralization, that is with industry-based unions. In this case, unions do exert some market power but their intermediate position leads them not to internalize (that is, to ignore) the macro implications of their actions; at the same time, firms within such industries can more easily accommodate higher wages at low employment cost by raising prices [2].

However, two objections may weaken this conclusion. First, unions may not able to set higher wages in industries that participate to international trade. Second, wage coordination as it relates to wage bargaining extends beyond the national or sectoral levels. Informal coordination (though difficult to measure), may occur both across firms and between firm- and industry-level unions. Such informal coordination may lead to the same restraining effects as formal coordination mechanisms. Japan offers a prime example: if one does not account for informal coordination, Japan appears to be among the least coordinated countries in the OECD, whereas it is one of the most coordinated after considering informal mechanisms [2].

On the positive side, three conclusions may be drawn from studies conducted in the 1980s and 1990s: (i) after accounting for bargaining coverage and coordination, union density itself has little or no impact on the comparative performance of labor markets, except that it is associated with a reduction in earnings inequality due to wage distribution compression; (ii) while an increase in bargaining coverage per se might induce higher unemployment, to the extent that it is associated with an increase in the degree of bargaining coordination (whether formal or informal) these negative effects disappear and might be reversed; (iii) taking into account both formal and informal mechanisms (and as long as unions internalize the macro effects of their actions), then bargaining coordination can positively affect economic performance during periods when the economy is adapting to changing conditions, by helping eliminate the adverse effect of shocks. Support for these arguments became stronger after having observed the consequences of the recessions that hit Europe in the first two decades of the 21st century.

From “eurosclerosis” to “flexicurity”: debates and reforms

In the 1980s, the concept of “eurosclerosis” began to appear in policy discussions: the slow growth and high unemployment observed in continental Europe were imputed to overly rigid labor markets and restrictive regulations. These theses were echoed in the OECD Jobs Study, which examined the factors that had contributed to “ossifying the capacity of economies and the will of societies to adapt” [3]. The remedies suggested by the OECD, especially to increase working-time and wage flexibility, had a considerable influence in motivating the gradual changes that took place in the following years and have been described above: a reduction in the importance of collective bargaining, the easing of employment protection for permanent jobs, and reduced expenditures on passive labor market measures.

Around the same period, the European Community (since 2009, European Union) began to define its own Employment Strategy, which included the setting of (non-binding) guidelines for member states. Since 2005, this strategy became part of the broader Lisbon Strategy. At least on paper, the European strategy placed considerable importance on the dialogue between social partners and thus on a more coordinated approach to the formulation of labor policies. In addition, the European strategy explicitly linked labor market policies to social policies focused on reducing labor market segmentation, social exclusion and the risk of poverty. However, this strategy was not supported by community funds, and member states were not compelled to implement it in practice.

In practice, most European countries followed the OECD advise to reduce emphasis on passive measures (for instance, by lowering the duration of unemployment benefits and reducing entitlement criteria). Meanwhile, minimal progress, limited to only a few countries (Germany, Netherlands and Scandinavian countries) was achieved in respect of active policy measures, in spite of the accumulating evidence that they could contribute to reduce unemployment both in OECD [4] and in transition countries [5].

Between 1994 and 1996, the Danish experiment of “flexicurity” introduced a twist in this debate: the idea was to increase flexibility of labor market arrangements while at the same time providing insurance through generous welfare schemes and increasing spending on active policies aimed at improving workers’ employability. Soon, the EU adhered to this view, suggesting that flexicurity could provide a way to preserve the “European social model” (characterized by generous and generalized welfare benefits), while making the labor market more inclusive (through a greater reliance on active policies) and especially preserving firms’ competitiveness (through enhanced flexibility). In 2007, the European Commission published a Communication devoted to the “Common Principles of Flexicurity”, and since then the concept has become the mantra of the European Employment Strategy.

Soon, however, the debate on these issues was enriched by other considerations. Two in particular include: the continued decline of manufacturing, which brough about an expansion of non-standard forms of work; and new attention on monopsonistic labor markets, i.e. the idea that firms (alone or through employers’ organizations) may enjoy considerable market power in the setting of wages (as well as of other aspects of labor contracts). In 2021, Nobel Laureate David Card suggested that evidence on monopsony might “lead to some rethinking on policies such as minimum wages, the regulation of trade unions, and anti-trust” [6].

Similar considerations have contributed to shift attention from a (pre-2007) focus on the (monopolistic) market power of unions to the (monopsonistic) market power of employers (which is likelier to prevail in the new, more fragmented characterization of labor markets). Recently, more research has been devoted to study these issues, also in Europe. The largest study so far, conducted on six European countries, suggests that an increase in labor market concentration has relevant effects in reducing both wages and the probability of being hired on a permanent contract [7]. These findings provide indirect but rather strong support to the view that collective bargaining institutions may actually improve the functioning of labor markets, by balancing the market power of entrepreneurs.

This twist to the ongoing debate has been reinforced by the increased relevance of large (negative) macro shocks affecting all European economies and their labor markets. In such cases, collective centralized bargaining may well prove to be welfare improving, as it is likely to give a stronger voice to the unemployed and also to help solve an economy-wide coordination problem [8].

A tale of three recessions

The debates surveyed in the previous sections took place during a long period, in which both Europe and the US seemed to be immune from the worst booms and busts of the economic cycles – a period that was aptly named the Great Moderation. But that moderation did not continue for long, as the first two decades of the 21st century have been marked by three recessions. First, the Great Financial Crisis of 2007–2008 reached Europe (from the US) in 2008 and was the cause of the Great Recession. In the euro area, recession began in 2008. Q2 and lasted until 2009.Q2; in those fifteen months, seasonally adjusted real GDP fell by 5.7% and the employment rate by 1.9%. Unemployment continued to grow beyond the end of the recession, to reach 10.2% (+3%) in 2010.Q1. See Figure 3.

Three recessions, compared

In Europe, the lack of fiscal support and the early removal of monetary policy support after that first recession led to a second, milder recession between 2011.Q3 and 2013.Q1: in this case, the fall of GDP and of aggregate employment were less pronounced (-1.8 and -0.9%, respectively) whereas the increase in unemployment was proportionally larger (+2%, to reach 12%).

Across European regions, however, these two recessions look quite different: the first one had an even larger impact in new member states, whereas the second one went largely unnoticed in these countries: in fact, their employment rate continued to grow during the second recession.

At the end of the decade, Europe experienced a third recession, induced by COVID-19. Compared with the Great Recession, this was much sharper (on a quarterly basis, euro area GDP fell by 14.4%), but also shorter: it lasted only 2 quarters, thanks to the promptness of monetary and fiscal policy interventions. And this time the labor market was largely spared the strains of the recession: while in the first two recessions the fall in the employment rate had been (in the euro area) between one third and one half of the fall in GDP, this time employment fell by less than one fifth of the GDP fall. Similarly, the unemployment rate increased by only 1.1%, reaching 8.3% in 2020.Q3.

The experience of these recessions has contributed to swing the opinions of former supporters of outright economic liberalization toward a more nuanced approach. A new consensus has emerged, that some employment protection (including, in some cases, firing restrictions) may have contributed to cushion unemployment during the Great Recession. A large recent study focused on 36 OECD countries (most of them in Europe) concluded (contrary to many views that had been prevailing until 2007) that coordinated bargaining systems are associated with higher employment, better integration of vulnerable groups and lower wage inequality than fully decentralized systems [9]. And a detailed study on the “virtuous” case of Norway shows how, in the context of a well-coordinated bargaining system, increasing union density at the firm level may lead to a substantial increase in both productivity and wages [10].

As these and other structural changes (such as the greater reliance on global value chains, population ageing and the energy transition) are increasingly taken into account in policy debates, this naturally leads to stronger calls for an increase in social dialogue and new policy actions. In particular, there is by now an increasing agreement that, if appropriately designed, collective bargaining could be an important element of social dialogue – and thus should perhaps be reinforced rather than further dismissed [11].

After COVID-19: two policy innovations

The COVID-19 recession has left the strongest imprint on labor markets and, indirectly, on LMI. What made the consequences of this recession so different from the previous two? And what does this imply for LMI and policies?

The answer to the first question is that fiscal policy made all the difference. And the answer to the second question is similar: as innovations in fiscal policy have had remarkable effects, this experience will presumably continue to influence the format of LMI and policies. The COVID-19 pandemic has induced most countries (and the EU itself) to adopt sharp innovations in the size and type of policy interventions focused on labor markets. In the future, these innovations may remain as complements to, or even possibly substitutes of more traditional bilateral bargaining relations between employers and employees. Two such innovations deserve particular focus.

The main innovation is the adoption of job retention schemes. These schemes, borrowed from the long-standing German experience with Kurzarbeit (that is, “short time work”) have become “the main instrument used in most OECD countries for stemming the labour market impact of the COVID-19 crisis” [12]. The extent of their (temporary) unprecedented adoption was intended not only to counter the recession but more specifically to compensate for the government-imposed restrictions on many economic activities (so called “lockdowns” and “social distancing” measures). In short, “these schemes pay part of an employee’s normal wages if a business keeps them employed but reduces their hours. In this way, businesses can temporarily reduce labour costs in the face of weaker demand without laying off workers. At the same time, workers’ incomes are maintained even if the working hours are reduced. Moreover, by maintaining the worker-job bond, job-specific human capital is preserved and long-lasting economic scarring due to the loss of human capital is reduced.” [13]

Although job retention schemes seek to reduce employers’ costs or to maintain employees’ incomes, or both, employees’ contracts remain in force even when work hours are reduced to zero. Clearly, this radically affects, if only temporarily, the traditional relations between (un)employment and labor market bargaining [12].

Job retention schemes have become a major component of fiscal policies, and their macroeconomic impact has been quite large. “On average, EU countries are estimated to have spent nearly 2% of GDP on job retention schemes in 2020. The benefits of such schemes are readily apparent from the unemployment numbers. While hours worked in the second quarter of 2020 fell more than 15% below late-2019 levels in the EU, the employment rate fell less than two percentage points and the unemployment rate ticked up one percentage point” [13]. In other words, job retention schemes stimulate an adjustment to macro shock on the “intensive” (that is: hours worked) rather than on the “extensive” (that is: number of employed workers) margin, while at the same time providing quasi or full-insurance of workers’ incomes to the macro shock. This is clear from inspection of the illustration on page one: during the COVID-19 recession, the impact on the number of employed persons has been minimal when compared with the impact on output and on hours worked [14].

Although not related to the COVID-19 pandemic, a second important innovation in policy debates is the new emphasis on minimum wages. In October 2020, the European Commission filed a proposal for a directive, intended to “support Member States to set up a framework for minimum wages … either through collective arrangements of through statutory minimum wages” [15]. The Commission observes that “collective bargaining plays a key role for adequate minimum wage protection. The countries with high collective bargaining coverage tend to display a lower share of low-wage workers, higher minimum wages relative to the median wage, lower wage inequality and higher wages than the others” [15]. This statement suggests a new, hitherto unexplored link between the diffusion of collective bargaining practices and the adoption of minimum wages. The Directive proposed by the Commission has been approved by the European Parliament and, in October 2022, by the Council of the EU. Currently, 22 of the 27 EU member countries have adopted a minimum wage policy.

Limitations and gaps

While there have been considerable improvements in measuring LMI in both old and new EU member states as well as in understanding their effects, several constraints remain. First, enforcement of formal LMI may differ across countries and over time, particularly with respect to employment protection laws. Measures that account for actual enforcement would be useful in this regard. Second, wage bargaining coordination measures must reflect both formal and informal institutions and practices. While both types of institutions change over time, it is especially challenging to account for informal practices, which may also be influenced by the extent to which unions share macroeconomic objectives. Third, labor union power and labor union density do not necessarily go hand in hand. In addition to density, union power is influenced by the breadth of coverage of collective bargaining agreements. In some countries, such as France and Spain, coverage and density differ considerably, raising the question of how to best balance them to obtain more comprehensive measures. Fourth, understanding how the effectiveness of LMI is influenced by the prevailing institutional characteristics in each country would be insightful. Fifth, the fragmentation of labor markets induced by new technologies, new work practices and globalization of production pose strong, but to a large extent not yet understood, challenges. Better understanding of these challenges would contribute to improved design of institutions and policies, the evaluation of their effects and the quality of social dialogue.

Summary and policy advice

European countries are facing an increasing array of economic and social challenges: negative macro shocks, the decline of manufacturing, the diffusion of new technologies and nonstandard work practices, the initial expansion and subsequent regionalization of global value chains, population ageing, migration flows, the energy transition and more. This immensely complex situation calls for an increase in social dialogue and requires policy action, either at the national or the supra-national level. Labor markets are at the center of many of these challenges.

To respond, the decline in public expenditures on both active and passive labor market policies, which has taken place at least until 2019, must be reversed. New EU member states must enhance their efforts in this direction: to date, new EU members spend considerably less (as a ratio to GDP) than other EU countries on passive and especially on active labor market policies. Employment protection measures may help cushion the negative impact of shocks, thereby reducing the duration and amplitude of economic recessions, while at the same time strengthening social cohesion. During the COVID-19 recession, governments contributed to the adoption and financing of vast job retention schemes; it is important to understand to what extent similar measures should become institutionalized going forward, while increasing their flexibility and strengthening their links with active labor market policies.

A new consensus has formed that stronger economy-wide bargaining coordination over wages and work conditions (which often goes together with higher union density) may lead to lower unemployment, better integration of vulnerable groups and lower wage inequality, and possibly also higher productivity and external competitiveness. Wage coordination is especially important within a monetary union such as the eurozone. Policymakers should encourage the establishment of this pattern of wage settlements, by favoring opportunities for labor market participants to share information and to coordinate their objectives and negotiation goals. Finally, it is important to consider to what extent the adoption or strengthening of minimum wage provisions may contribute to pursue these goals.


The author thanks Hartmut Lehmann, an anonymous referee and the IZA World of Labor editors for many helpful suggestions on an earlier draft. Version 2 revises the discussion on the evolution of labor market institutions across different EU countries and adds new Key references [1], [3], [4], [5], [6], [7], [8], [9], [10], [11], [12], [13], [14] and [15].

Competing interests

The IZA World of Labor project is committed to the IZA Guiding Principles of Research Integrity. The author declares to have observed these principles.

© Riccardo Rovelli

evidence map

Labor market institutions and policies in old and new EU members

Full citation

Full citation

Data source(s)

Data type(s)