You are here: Eurofound > EIROnline > Comparative Information > Trade union strategies to recruit new groups of workers > Comparative study My Eurofound: Login or Sign Up   

Trade union strategies to recruit new groups of workers

Index | << Prev | Next >>

[ pdf version size 1470 kb ]

Recent trends in union membership and density

In recent years, the difficulties faced by trade unions to maintain their representation base and capacity to regulate employment relations has sparked off fierce debates. These debates also included topics such as the possibility to improve trade union membership and density levels, as well as to achieve significant improvements in pay and working conditions. The decline in trade union density, often going along with a decrease in membership, have been identified as clear signs of a weakening of the role of trade unions and industrial relations in both today’s society and economy. The emergence of forms of concession bargaining as well as the decline in collective bargaining coverage are also indicators for a weakening of the trade unions in European countries.

A number of structural transformations in the economy have been identified as possible reasons for this potential representation and regulatory crisis of trade unions:

  • sectoral shifts in European economies, with traditional union strongholds in manufacturing being severely affected by downsizing, while most of the employment creation takes place in much less organised industries, which are also more difficult to organise since workplaces are smaller and more dispersed;
  • public sector reform, characterised by extensive outsourcing of labour, liberalisation and privatisation procedures. This has resulted in significant workforce reductions in the public sector, which generally recorded high trade union density levels;
  • a renewed unilateralism in personnel issues pursued by company management, through the adoption of direct participation practices, incentive policies and career development measures which, in certain circumstances, may compete with union protections;
  • a tendency towards the individualisation of employment relations, resulting from the abovementioned management policies and also from growing demands by a more educated and qualified workforce, for whom collective identities and interests seem to become less relevant;
  • the globalisation of economic activities, which seems to weaken traditional industrial relations systems and exposes trade unions to offshoring threats and social regime competition.

These trends and challenges have spurred a wave of initiatives and analyses on trade union revitalisation. Although both the analyses on the weakening of the role of trade unions and possible responses in terms of unions’ organising initiatives have basically originated in the US, where such issues have been at the centre of debate for a long time, more recently attention has been growing around these developments in the EU.

In the EU, the institutional support as well as the fact that industrial relations function as an instrument for regulating the economy are an established and common feature of national systems, albeit with important differences between the countries and country specificities. Therefore, before addressing the issue of organising campaigns launched by European trade unions, which is the main focus of this report, it is important to investigate trends in trade union membership and density in European countries. Such an analysis will help to assess the varying relevance of the strategies implemented in each national case.

Long-term trends

An important data source regarding the basic characteristics of industrial relations systems in advanced economies is the Database on Institutional Characteristics of Trade Unions, Wage Setting, State Intervention and Social Pacts (ICTWSS Database), set up and maintained at the Amsterdam Institute for Advanced Labour Studies (AIAS) of the University of Amsterdam, under the responsibility of Professor Jelle Visser. The ICTWSS database collects information on 34 countries, including all EU Member States, covering a period of more than 40 years since 1960. Nonetheless, individual national time series may be shorter or discontinuous, depending on data availability or historical reasons, such as the presence of non-democratic political regimes like Francoist Spain or the former Soviet Bloc countries.

For those countries for which the entire or continuous time series are available since 1960, it is interesting to briefly analyse the long-term trends in both trade union membership and density levels. Such an analysis will help to understand the tendencies that have characterised such different decades in terms of economic conditions and industrial relations: the industrial golden age of the 1960s, oil shocks and stagflation in the 1970s, widespread restructuring in the 1980s and the internationalisation of markets in the 1990s, as well the mobilisation and periods of conflict which have accompanied these developments in some of the European countries.

The complete time series for membership trends and trade union density covering the period 1960–2007 are available for 12 European countries – Austria, Belgium, Germany, Denmark, Finland, France, Ireland, Italy, the Netherlands, Norway, Sweden and the UK. Shorter but still continuous time series are also available for Spain covering the years 1978–2007 corporate policy and for Malta from 1983 until 2007 (see Figure A1 in the Annex for trends in trade union membership and density levels in 13 EU Member States and Norway). When looking at relative developments between 1960 and 2007, it is possible to identify some important differences across countries in terms of trade union density patterns.

  • A minority of trade unions have managed to increase their density over the past decades: Finland (by 40 percentage points, from 32% in 1960 to 72% in 2006), Denmark (up 13 percentage points from 57% in 2006), Belgium (up 13 percentage points from 41% in 1960), Italy (up nine percentage points from 25% in 1960), Sweden (up three percentage points from 72% in 1960) and Malta (up 12 percentage points from 45% in 1983). Of these countries, only Belgium has confirmed this growing trend in recent years.
  • A small number of countries – including Finland, Italy, Ireland and Spain – have recently recorded an increase in trade union membership along with a decline in union density. Such a development indicates possible problems in organising workers in labour market segments where job creation takes place, while maintaining the capacity to extend membership. In Ireland, in particular, this seems to be a long-term trend which is possibly characteristic of the employment boost since the early 1990s.
  • Norway and the Netherlands show two mirroring trends. In Norway, trade unions have been able to considerably increase their membership while maintaining union density at a rather stable level; the number of trade union members has almost doubled between 1960 and 2007, while union density has declined by five percentage points to a still remarkable 55% in 2007. In the Netherlands, after a downturn that began in the 1970s, membership has recovered in the late 1990s and has remained relatively stable since then. However, this development appears not to have influenced unionisation in the country which remained at low density levels established since the mid-1980s: while the number of trade union members has increased by less than 20% over the period under examination, union density has almost halved from 40% to 21% in 2007.
  • France and Italy both recorded a peak in trade union membership in the 1970s, probably resulting from the similar mobilisation periods in the late 1960s. In France, on the one hand, the membership peak was smaller and did not bring about a significant increase in trade union density; on the contrary, it was quickly replaced by a substantial decrease in both indicators: union density more than halved from 20% in 1960 to 8% in 2007. At the time, it was already the lowest level reported among the 14 countries taken into consideration here. In Italy, on the other hand, trade union membership growth was significant, while union density steadily but moderately declined over the period under examination. The number of trade union members has almost doubled between 1960 and 2007, and union density increased by a third, from 25% in 1960 to 33% in 2007.
  • In the UK, trade union membership and density followed a similar pattern, even in the 1960s and 1970s, with no apparent influence of the demise of multi-employer collective bargaining in the 1980s. However, the levels of both indicators started to noticeably decline in 1980, following a peak in 1979; this decline could only be halted regarding membership in the late 1990s, while union density has continued to decrease ever since, although at a slower pace. The two turning points in union membership and density trends interestingly coincide with significant changes in the national political scene.
  • A steady declining trend can be found in the cases of Austria and Germany. In Austria, a constant decline in union density has been underway since the 1960s and more recently also in union membership. Over the reference period, trade union density more than halved from the second highest level of the 14 countries examined of 68% in 1960 to the average level of 32% in 2007. In Germany, the considerable injection of new members linked to the reunification in 1989 had a limited impact on union density, restoring levels of the early 1960s; after 1991, both trade union membership and density steadily declined, leading to a union density of 21% in 2007, down from 35% in 1960. Most of this decline is connected to the significant decrease in trade union membership and density in the federal states (Länder) of eastern Germany.

In fact, a significant decline in trade union density seems to be a common feature of eastern and central European countries over the period 1989–2006 (Table 1). While the level of union density in transition economies in the early 1990s was typically high and above the EU average, it rapidly decreased and now usually stands below this reference. Partial exceptions, with relatively higher affiliation rates, are Slovenia and Romania, while the most evident decline in union density occurred in the Baltic countries and Poland, considering that, when the transition started, union density was virtually 100% in the whole region – except for Slovenia.

Table 1: Trade union density in central and eastern European countries, 1989–2006 (%)
Country First observation (year) Last observation (year) Difference (% points) Difference

70 (1992)

21 (2006)




80 (1993)

21 (2006)




85 (1992)

13 (2006)




73 (1990)

18 (2005)




40 (1995)

14 (2006)




30 (1995)

16 (2006)




55 (1990)

14 (2006)




46 (1998)

34 (2006)




69 (1989)

41 (2003)




75 (1993)

24 (2006)



Source: ICTWSS Database

The remaining four EU Member States – Cyprus, Greece, Luxembourg and Portugal – show quite different situations regarding trade union membership and density. Cyprus and Luxembourg maintain a relatively strong union representation system, despite some decline in density levels. In Greece, and particularly in Portugal, the situation is more complex. In Portugal, there has been a steady decline in trade union density since the second half of the 1970s; Portugal recorded a drop in trade union density similar to that of Spain in the wake of the democratic transition, but without the limited signs of recovery that can be traced in the Spanish case. In Greece, the late 1970s and 1980s marked a small increase in union density, which then started to slowly decline in the early 1990s.

Table 2: Trade union density in Cyprus, Greece, Luxembourg and Portugal, 1970–2006 (%)
Country First observation (year) Last observation (year) Difference (% points) Difference

71 (1995)

62 (2006)




36 (1977)

23 (2005)




47 (1970

40 (2006)




61 (1978)

18 (2006)



Source: ICTWSS Database

Recent developments

When analysing the developments in trade union membership and density since 2000, a clearer picture emerges regarding the difficulties of representation. As illustrated in Figure 1 (below), in the majority of EU Member States and Norway the past decade has been marked by declining trade union density levels and also often decreasing membership numbers.

More specifically, it is possible to identify three main country clusters, including some subgroups of countries. The first cluster comprises a number of countries that show a positive trend in at least one of the two indicators – that is, trade union membership or density. For instance, in Belgium, trade unions succeeded to significantly increase their membership levels as well as raising their density. The same development occurred, although to a smaller extent, in Norway. In five more countries – Cyprus, Italy, Ireland, Luxembourg and Spain – trade union membership increased, while union density declined.

The second cluster consists of countries where losses in both dimensions have been fairly limited and the emerging picture is more or less one of stability: this includes the Nordic countries (Denmark, Finland and Sweden, with relatively high union density levels), France (with the lowest European union density level), and Malta, Slovenia and the UK. However, this cluster also comprises two subgroups of countries where the pace of erosion of trade union representation has been more significant: in Bulgaria, Greece and Portugal, the decline essentially occurred in union density, while in Austria and Germany membership has also been affected.

The third cluster regroups the eight central and eastern European countries where trade union membership and density have both decreased remarkably in recent years.

Figure 1: Trends in trade union membership and density, by country and country cluster (average annual % change)*

Figure 1: Trends in trade union membership and density, by country and country cluster (average annual % change)*

Notes: * Data for Cyprus, Latvia, Luxembourg, Poland and Romania cover the period 1998–2006, Greece and Hungary 1998–2005, Portugal 1997–2006 and Slovenia 1998–2003.

Source: ICTWSS Database

Table 3 includes information on employment growth and the most recent level of trade union density in all of the countries under examination, grouped into the three clusters. The rate of job creation may partly explain the decline in trade union density in countries where union membership continues to grow or remains rather stable. Such a decline in union density could underline possible problems in organising workers in the economic sectors where employment is increasing. However, the decline in union density is so widespread and significant that, regardless of the employment trend over the reference period, this correlation can be considered as weak, at most. Union density levels can complement the analysis of recent trends illustrated in Figure 1 above, thus providing a more comprehensive assessment of the situation of union representation in each country (Table 3, Index B).

Table 3: Trade union density, membership and dependent employment, 2000–2006*
Country Density (%) Union membership (average annual % change) Union density (average annual % change) Employment(average annual % change) Index A** Trend Index B** Situation
Cluster 1            

















































Cluster 2            



























































































Cluster 3            
























































Notes: * Data for Cyprus, Latvia, Luxembourg, Poland and Romania covers the period 1998–2006, Greece and Hungary 1998–2005, Portugal 1997–2006 and Slovenia 1998–2003.

** Index A = the sum of the average annual percentage change in trade union membership and density. Index B = the sum of the normalised Index A and normalised trade union density.

Source: ICTWSS Database

The two last columns in Table 3 present two combined indexes of the recent trend (the sum of the annual rates of membership and density change) and of the situation of union representation (the sum of the normalised union density and the normalised trend index). Calculating the recent trend and normalised situation of union representation for each country provides a rough picture of the overall situation of trade union representation. Understandably, a small erosion of union membership and density may be of limited significance if union density levels are above 70% in a country. On the other hand, if union density levels are below 20%, an increase in membership numbers combined with a decline in union density cannot be regarded as granting good prospects for the consolidation of trade union representation. Nonetheless, considerable losses in both indicators certainly represent a difficult context for trade union action.

Figure 2 (below) shows where the various national trade union representation systems are positioned in terms of trade union density in the most recent year available (2005 or 2006) and their combined trend index of union membership and density in the 2000s (Index A). Since both indicators have been ‘normalised’, positive values indicate above average trade union representation; this corresponds to a union density level above the simple average of 32.5% and recent developments above the overall negative average trend. The vertical dotted line, however, marks where the combined index of membership and density trend is close to zero – a position of relative stability in representation – so that countries to the right of the line are characterised by an increase in overall trade union representation capacity. The picture therefore helps to appreciate the strength (or weakness) of national union representation systems in terms of both union density (from top to bottom) and recent developments (from right to left). The closer to the upper right corner, the stronger appears the national union representation system.

Figure 2: Situation of national trade union representation systems (trade union density level, and combined union membership and density trend*)

Figure 2: Situation of national trade union representation systems (trade union density level, and combined union membership and density trend*)

Combined trade union membership and density trend corresponds to the sum of the average annual percentage change in membership and density (see Index A in Table 3 above). Both variables have been normalised (that is. transformed to have mean=0 and standard deviation=1) to allow a better representation. Since the original trend variable has a negative mean (-3.91). The vertical dotted line has been positioned at the positive value which corresponds to a relative stability in trade union membership and density (no changes or offsetting variations).

Source: ICTWSS Database (author’s calculations)

Notes: * Data for trade union density level from 2006, except for Hungary (2005), Greece (2005) and Slovenia (2003).

Specific groups

When looking at the membership composition of trade unions, a significant segmentation emerges in terms of sectors of activity and groups of workers. In general, the lower the national union density level, the higher the scope for the fragmentation of union representation and concentration of members in a few economic sectors and occupational groups. However, some situations cut across the various countries and seem to represent some common characteristics of the present structure of trade union representation in Europe.

In particular, the following four factors tend to influence the level of trade union density:

  • sector of activity – the public sector, including utilities such as public transport, electricity, gas and water supply, and traditional manufacturing activities still represent the core of unionised sectors. Union density is far lower in the private services sector;
  • age – young workers seem to be almost invariably the most problematic group of workers to unionise;
  • company size – small and medium-sized enterprises (SMEs) show systematically lower levels of unionisation;
  • employment contracts – workers on temporary contracts and those on other atypical work relationships are reportedly difficult to unionise.

It is not always easy to disentangle the effects of each individual factor, as conditions unfavourable to unionisation tend to cluster. For instance, young workers are often employed under non-standard employment contracts and often in private services. Moreover, these conditions are frequently combined with other variables, such as gender and nationality; for instance, women and migrant workers are sometimes associated with lower than average unionisation rates, mostly because they are concentrated in certain economic sectors and are often employed under atypical contracts.

Women, in particular, often show unionisation levels which are similar to those of men. In a number of countries, the unionisation rate is higher for women than for men. However, high unionisation rates among women are more the effect of the sector than of gender, with especially high female employment rates in the public sector, for example. The second annual ETUC report on the gender representation gap in trade unions and gender mainstreaming activities, which summarises the 8 March Survey results, confirms this tendency. The survey findings reveal that women are increasing their share in European trade union membership, often even in countries where total membership is declining. On the one hand, these findings reflect the increasing feminisation of the labour market and, on the other hand, they seem to suggest that organising initiatives are, to a certain extent, successful in the services sectors where women account for a large proportion of the workforce.

In many countries, the segmentation of trade union membership is perceived as a fundamental weakness. Sometimes, the trade union membership segmentation is considered being an even greater weakness than the erosion of union density levels, since it challenges the general representation of trade unions. With regard to the shrinking participation of young people in trade unions, the segmentation of union membership also seems to jeopardise the prospects of a generational renewal. In this regard, the divide between the private and the public sectors is becoming particularly important in many countries, including Portugal, Spain and the UK, as well as eastern and central European countries. In the latter countries, the membership gap is particularly wide between the private and public sectors, as trade unions are less present in private sector manufacturing. In Hungary, for instance, trade union members are concentrated in public administration and utilities, while in Poland the private sector is virtually union-free.

In some countries, however, the segmentation of union membership is not evident, as is the case in Belgium where no significant differences exist across sectors and generations. In Belgium, the unionisation rate of young workers is in line with the national average of unionisation; this situation may be linked to the Ghent system, as many young people become trade union members when they enter the labour market as unemployed. In fact, the labour market seems to be a crucial factor regarding union membership in Belgium, as two other uncommon characteristics emerge: unionisation is higher in the private than in the public sector (where job security is higher and therefore unemployment benefits do not play a crucial role) and company size does not affect the affiliation rates. Moreover, workers under non-standard contracts show greater awareness of the role of trade unions. In Finland, sectoral and generational differences in union membership are not high either, although more significant than in Belgium. Trade union membership remains lower in the private sector than in the public sector (67% compared with 86% in 2008), as well as for lower-educated workers and for workers with a foreign background. However, unionisation rates in Finland never reach levels below 63%, with the exception of young workers aged 15–24 years whose union density is 24%.

Index | << Prev | Next >>

Page last updated: 17 May, 2010
About this document
  • ID: TN0901028S
  • Author: Roberto Pedersini
  • Institution: University of Milan
  • Country: EU Countries
  • Language: EN
  • Publication date: 14-06-2010
  • EIRO Keywords: Trade unions