Representativeness of the European social partner organisations: Construction sector

This study provides information designed to aid the functioning of sectoral social dialogue in the construction sector. The study is divided into three parts: a summary of the sector’s economic background; an analysis of the social partner organisations in all EU Member States (apart from Croatia), including membership, role in collective bargaining, social dialogue and public policy, and national and European affiliations; and an overview of the relevant European organisations, in particular their membership composition and capacity to negotiate. The aim of Eurofound’s series of representativeness studies is to identify the relevant national and supranational social partner organisations in the field of industrial relations in selected sectors. The impetus for these studies comes from the European Commission’s aim to recognise the representative social partner organisations to be consulted under the provisions of the Treaty on the Functioning of the European Union (TFEU).

Download the full report [1.5 MB PDF]

See also the executive summary

Introduction

Objectives of the study

The aim of this representativeness study is to identify the relevant national and supranational social actors – the trade unions and employers’ organisations – in the field of industrial relations in the construction sector, and to show how these actors relate to the sector’s European interest associations of labour and business. The impetus for this study arises from the aim of the European Commission to identify the representative social partner associations to be consulted under the provisions of the Treaty on the Functioning of the European Union (TFEU). Hence, this study seeks to provide basic information needed to support sectoral social dialogue. The effectiveness of European social dialogue depends on whether its participants are sufficiently representative in terms of the sector’s relevant national actors across the EU Member States. Only associations which meet this precondition will be admitted to European social dialogue.

Concept and methodology

To accomplish these aims, the study identifies the sector-related national and European social partner organisations in the construction sector, via a top-down (listing the members of the European affiliations) and a bottom-up approach (through Eurofound’s Network of National Correspondents).

The study first identifies the relevant national social partner organisations in the construction sector and then analyses the structure of the sector’s relevant European organisations, in particular, their membership composition. This involves clarifying the unit of analysis at both the national and European level of interest representation.

The study includes only organisations whose membership domain is ‘sector-related’ (Table 1).

Table 1: Determining the ‘sector-relatedness’ of an organisation

Source: Standardised questionnaire sent to Eurofound’s Network of European Correspondents (2013–2014)

At both national and European levels, many associations are not considered to be social partner organisations as they do not essentially deal with industrial relations. Therefore, there is a need for criteria to define clearly the social partner organisations.

As regards national-level associations, classification as a sector-related social partner organisation implies fulfilling one of the following two criteria:

• be a party to ‘sector-related’ collective bargaining;
• be a member of a ‘sector-related’ European association of business or labour that is on the Commission’s list of European social partner organisations consulted under Article 154 of the TFEU and/or participates in the sector-related European social dialogue.

While affiliation to a European social partner organisation is sufficient to determine a national association as a social partner, this does not necessarily imply that the association is involved in industrial relations in its own country. Although this selection criterion may seem odd at first glance, a national association which is a member of a European social partner organisation will become involved in industrial relations matters at EU-level through its membership of the European organisation: through informal communication, consultation procedures and eventually the implementation of agreements concluded by the European social partners at national level.

It is also important to assess whether the national affiliates to the European social partner organisations are engaged in industrial relations in their respective country. Affiliation to a European social partner organisation and/or involvement in national collective bargaining are of the utmost importance to the European social dialogue, since they are the two constituent mechanisms that can systematically connect the national and European levels.

A European association is considered a relevant sector-related interest organisation if:

• it is on the European Commission’s list of interest organisations to be consulted on behalf of the sector under Article 154 TFEU;
• and/or it participates in the sector-related European social dialogue;
• and/or it has asked to be consulted under Article 154 TFEU.

In addition, this study considers any other European association with sector-related national social partner organisations – as defined above – under its umbrella.

Definitions

For the purpose of this study, the construction sector is defined in terms of the Statistical Classification of Economic Activities in the European Community (NACE) to ensure the cross-national comparability of the findings. The NACE code reflects the field of activities covered by the European Sectoral Social Dialogue Committee ‘Construction’ as demarcated by the social partners in agreement with the European Commission.

More specifically, the construction sector is defined as embracing NACE (Rev. 2) 41, 42 and 43. This includes the following activities:

• 41 Construction of buildings;
• 42 Civil engineering;
• 43 Specialised construction activities.

The domains of the trade unions and employers’ organisations and the scope of the relevant collective agreements are likely to vary from this precise NACE definition. The study therefore includes all trade unions, employers’ organisations and multi-employer collective agreements which are ‘sector-related’ in terms of any of the following four aspects or patterns:

• congruence – the domain of the organisation or purview of the collective agreement is identical to the NACE demarcation;
• sectionalism – the domain or purview covers only a certain part of the sector as demarcated by NACE classification, while no group outside the sector is covered;
• overlap – the domain or purview covers the entire sector together with (parts of) one or more other sectors. However, it is important to note that the study does not include general associations which do not deal with sector-specific matters;
• sectional overlap – the domain or purview covers part of the sector plus (parts of) one or more other sectors.

Organisations are considered to be ‘sector related’ if their membership domain relates to the sector in one of the ways displayed in Figure 1. Table 2 summaries the domain pattern and scope of the sector in terms of these four aspects or patterns.

Figure 1: Sector-relatedness of social partner organisations: possible domain patterns

figure1_ef1535.jpg

Table 2: Domain pattern and scope of the organisation’s domain

Note: The domain pattern is based on the answers to the questions on the scope of the domain illustrated in Table 1.

Sectoral Social Dialogue Committee – Construction

The Sectoral Social Dialogue Committee (SSDC) at European level for the construction sector was set up in 1999 in response to a joint request by the European Construction Industry Federation (FIEC) on the employers’ side and the European Federation of Building and Woodworkers (EFBWW) on the unions’ side.

Since 2007, the European Builders Confederation (EBC) has attended SSDC meetings in the FIEC delegation. At the time of drafting this report (spring 2015), EBC is not recognised by the European Commission as a European social partner but has requested recognition.

In line with the conceptualisation of this study as outlined above, affiliation to one of these three European organisations (FIEC, EBC and EFBWW) is a sufficient criterion for classifying a national association of one of the European Union Member States as a relevant social partner organisation for the purpose of this study.

However, the constituent criterion is one of sector-related membership. This is important, in particular, in the case of EFBWW due to its sector-overlapping membership domain. Thus, the study includes only those affiliates to EFBWW whose domain relates to the construction sector, as defined earlier.

Collection of data

The collection of quantitative data, such as those on membership, is essential for investigating the representativeness of the social partner organisations. Unless otherwise stated, this study draws on country studies provided by Eurofound’s European Industrial Relations Observatory (EIRO), a network of national industrial relations experts which became part of Eurofound’s Network of European Correspondents in April 2014. The national correspondents complete a standard questionnaire by contacting the sector-related social partner organisations in their countries. The contact is generally made via telephone interviews in the first place but might, in certain cases, be established via email. In case of the unavailability of any representative, the national correspondents are asked to fill out the relevant questionnaire based on secondary sources, such as information given on the social partner’s website, or derived from previous research studies.

The cut-off date for data collection was 20 March 2013. However, data provided at a later stage (for example, during the procedure of verification of the national reports by the social partners) have also been considered.

It is often difficult to find precise quantitative data. In such cases, the Eurofound correspondents are requested to provide rough estimates rather than leaving a question blank, given the practical and political relevance of the study. However, if there is any doubt over the reliability of an estimate, this is noted.

In principle, quantitative data may stem from the following three sources::

• official statistics and representative survey studies;
• administrative data, such as membership figures provided by the respective organisations, which are then used for calculating the density rate on the basis of available statistical figures on the potential membership of the organisation;
• personal estimates made by representatives of the respective organisations.

While the data sources of the economic sectoral figures cited in the report are generally statistics from Eurostat or national statistical offices, the figures in respect of the organisations are usually either administrative data or estimates. Furthermore, it should be noted that several country studies also present data on trade unions and business associations that do not meet the above definition of a sector-related social partner organisation, in order to give a complete picture of the sector’s associational ‘landscape’. For the above substantive reasons, as well as for methodological reasons of cross-national comparability, such trade unions and business associations are not considered in this overview report. These organisations can, however, still be found in the national contributions, which are available on demand from Eurofound.

Quality assurance

In order to assure the quality of the information gathered, several verification procedures and feedback loops have been employed.

• First, the study’s coordinators, in collaboration with Eurofound staff, check the consistency of the national contributions.
• Second, Eurofound sends the national contributions to the national members of its Governing Board, as well as to the European-level sector-related social partner organisations. The peak-level organisations then ask their affiliates to verify the information. Feedback received from the sector-related organisations is then taken into account, if it is in line with the methodology of the study.
• Third, the complete study is finally evaluated by the European-level sectoral social partners and Eurofound’s Advisory Committee on Industrial Relations, which consists of representatives from both sides of industry, governments and the European Commission.

Structure of report

The study consists of three main parts, beginning with a brief summary of the sector’s economic background, followed by an analysis of the relevant social partner organisations in all EU Member States (except for Croatia, which was not a Member State when the study was started). The third part of the study examines the representative associations at European level.

Each section contains a brief introduction explaining the concept of representativeness in greater detail, followed by the study findings. As representativeness is a complex issue, it requires separate consideration at national and European level for two reasons: First, the method applied by national regulations and practices to capture representativeness has to be taken into account; Second, the national and European organisations differ in their tasks and scope of activities. The concept of representativeness must therefore be suited to this difference.

Finally, it is important to note the difference between the research and political aspects of this study. While providing data on the representativeness of the organisations under consideration, the report does not reach any definite conclusion on whether the representativeness of the European social partner organisations and their national affiliates is sufficient for admission to the European social dialogue. The reason for this is that defining the criteria for adequate representativeness is a matter for political decision rather than an issue of research analysis.

Economic background

The construction sector, as defined for the purpose of this study, covers many different business activities such as residential housing building, construction of civil engineering projects (for example, roads, railways, tunnels and utility projects of various kinds), as well as plumbing, plastering, painting and other activities. These activities thus span both the public and the private sectors. 

According to the Eurostat Labour Force Survey (LFS), the European construction industry employed nearly 15 million people in the third quarter of 2014. Construction is the biggest industrial employer in the EU (4.34 MB PDF), representing about 7% of total employment in the EU27. But due to the relatively widespread practice of undeclared work in the sector in at least several Member States according to the EurWORK national reports, it seems to be likely that not all of the European construction employment is recorded in the Eurostat data.

Due to its economic importance and its role as ‘a major consumer of intermediate products (raw materials, chemicals, electrical and electronic equipment and so on) and related services’, the performance of the construction sector directly impacts on ‘the development of the overall economy’ according to a strategy for sustainable competitiveness of the construction sector (114 KB PDF) published by the European Commission in 2012.

Up to the mid-2000s, the European construction industry contributed significantly to job creation in particular, according to the European Commission’s strategy for its sustainable competitiveness, ‘in some very promising areas, such as the renovation of buildings and in infrastructure’. This was supported by favourable national policies in terms of public contracting and investments as well as the promotion of ‘low-energy’ building. However, the economic recession in 2007–2008 hit the construction sector particularly hard. This is partially attributable to the sector’s dependence on access to credit, which has increasingly been cut off during the crisis. Moreover, there have been severe drops in demand, particularly in the private residential but also in the infrastructure market, usually as a result of the constraints on public spending due to the crisis.

In general, one core reason for the pronounced decline in construction in many EU Member States may be found in the fact that (construction) investments can be postponed in a way that consumption cannot. Across the EU, the impact of the recession and the timing varied widely between Member States and not all construction subsectors have been equally affected. According to Eurostat structural business statistics, for example, construction activities in terms of both turnover and productivity declined more steeply in Ireland and Spain in the period between 2008 and 2013 than in the other Member States. While according to the European Commission’s strategy for the sustainable competitiveness of the construction sector, the residential housing construction subsector was hit harder than most other parts of construction.

According to the reports from Eurofound’s national correspondents, governments across the EU have taken various measures to mitigate the negative economic and employment effects of the recession on the construction sector. Such measures include:

• stimulating construction and low energy building programmes;
• advanced investment projects;
• tax concessions to stimulate demand in (private) building;
• subsidies for renovation.

Irrespective of the recent recession, the European construction sector faces a number of structural problems and challenges. As the most pressing among them, the Commission’s strategy for the sector’s sustainable competitiveness identified ‘a shortfall of skilled workers in many companies, low attractiveness to young people due to the working conditions, limited capacity for innovation and the phenomenon of undeclared work’. The strategy also noted the increasing pressure in the world market from competitors from non-European countries which may benefit from less tight regulations in terms of labour and environmental law as well as state aid.

The European construction sector’s business structure is highly fragmented, with a clear prevalence of small and medium-sized companies (SMEs), micro companies as well as an increasing number of self-employed people. However, there are also various forms of bogus self-employment. Since construction is a highly labour-intensive sector, the goal of minimising labour costs may induce employers to operate with bogus self-employed workers so as to save on direct pay and social security contributions. In a joint statement issued on 5 February 2010 (2MB PDF), the European sectoral social partners (EFBWW and FIEC) indirectly confirmed the widespread practice of undeclared work by recognising that:

unfair competition and social fraud are unacceptable in the construction industry and demand that these phenomena are eradicated, using a combination of prevention, information and enforcement. … [and that] bogus self-employment often occurs in a trilateral relationship between the contractor, an intermediary, and the worker.

Employment characteristics

Employment in the European construction industry is characterised by a clear prevalence of male workers and a relatively widespread practice of ‘atypical’ work. According to the LFS, men represented around 90% of total employment in the sector in the third quarter of 2014.

Many workers are on fixed-term or temporary employment contracts. In the third quarter of 2014, almost 12% of the total workforce was covered by temporary employment contracts according to the LFS. Apart from fixed-term employment, non-standard employment also involves part-time work and self-employment – 7.5% and more than 26%, respectively, in the third quarter of 2014 (LFS data). The share of non-standard employment in its various forms in construction is thus significantly higher than in most other economic sectors. With regard to self-employment, genuine self-employment needs to be distinguished from ‘bogus self-employment’. According to a European Commission proposal for a decision to establish a European Platform to enhance cooperation in the prevention and deterrence of undeclared work:

bogus self-employment … occurs when the worker is formally declared as self-employed on the basis of a service contract but the work he/she performs fulfils all the criteria that are used by national law and practice to characterize an employment relationship.

Workers with non-standard employment relationships tend to be more vulnerable in terms of job security than standard employees. They have therefore been more affected by workforce reductions caused by the recession, according to several national reports. The same holds true of low-skilled and migrant workers, who are commonly deployed at construction sites.

Long-term trends

Tables 3 and 4 give an overview of the development from approximately 2001 to approximately 2011 (that is, the situation just after the peak of the recession). They present figures on companies, employment and employees in the sector and in relation to the national economy, stemming from both national sources and Eurostat.

In all of the 17 Member States apart from three (Poland, Portugal and the UK) for which related data are available from the Eurofound correspondents, the number of companies more or less increased. However, it is uncertain whether this growth actually reflects a general expansion of the sector witnessed in these countries or just a process of fragmentation of the sector’s company structure and/or the emergence of a number of self-employed workers. In a few countries, such as Malta, Romania and Slovakia, the number of companies increased by 50% or more within the decade to the early 2010s (Table 3). The case of Sweden is unclear, since the data from the two reference years are not comparable.

Table 3: Total number of companies and employment in construction, 2001 and 2011 (approximately)

Notes: For a detailed description of the sources of these data, please refer to the national reports. ^a Without self-employed workers; ^b Figure questioned by FIEC and the French Building Federation (FFB); ^c Figure includes only employer companies; n.a. = not available.

Source: EIRO national correspondents (2013–2014), national statistics.

Twelve of the 20 countries with available data recorded an increase in overall employment within the sector in the same time period, while in eight countries employment fell. Losses in employment were most outstanding in Greece and Spain, recording declines of 19% and 26%, respectively (Table 3). In both countries, however, it is likely that the losses can be traced back almost exclusively to the period from 2008 onwards, when the global economic crisis arose.

In terms of the number of sectoral employees, nine countries recorded a decrease during the period of observance, while in 12 countries this indicator increased; no comparable data are available for six countries (Table 4). In at least seven Member States (Czech Republic, Germany, Greece, Hungary, Italy, Malta and Slovakia), the number of employees with a contractual relationship amounted to less than two-thirds of the total number of employment. One can infer from these findings that, at least in these countries (comparable data are not available for all Member States), the sector is characterised by a high incidence of non-standard employment arrangements.

Table 4: Total employees in construction, 2001 and 2011 (approximately)

Note: For a detailed description of the sources of these data, please refer to the national reports.

Source: EIRO national correspondents (2013–2014), national statistics.

Tables 3 and 4 also corroborate the earlier finding that men represent the vast majority of workers in the construction sector. In all countries with available data, male employees by far outnumber female employees, representing at least 80% or 90% of the sector’s total workforce.

The tables also indicate that the construction sector is very large. In terms of the share of employment, it proved quite dynamic during the decade to the early 2010s in most countries with available data, with 10 countries showing an upward trend and 10 countries showing a downward trend. In some countries, such as Estonia, Finland, Poland, Romania and Slovakia, the respective employment shares grew by at least 2 percentage points, while in other countries these shares fell considerably, in particular in Spain, were there was a 4.8 percentage point loss in the period between 2001 and 2011 (Table 4).

The construction sector’s share in aggregate employment ranges from 5.4% in Bulgaria to more than 11% in Cyprus (Table 3); no related data were reported for seven countries. In terms of absolute numbers of sectoral workers, six countries (France, Germany, Italy, Poland, Spain and the UK) recorded more than one million people who were gainfully employed in the sector in the early 2010s. Both Germany and the UK recorded far more than two million workers in the sector (Table 3).

Recent developments

The impact of the recession from 2008 onwards on the construction sector varied between countries. Overall, at least in terms of employment, the construction sector appears to have suffered more severely from the recession compared with most other industries.

Overall in the European Union, in terms of employment the construction sector was particularly severely hit by the recession: employment for the 15–64 age group declined steadily between 2008 and 2014 from more than 18.6 million in the third quarter of 2008 to less than 15 million in the third quarter of 2014 (Figure 2).

Figure 2: Total number of employees in construction during the recession (EU27)

figure2_ef1535.jpg

Note: Workforce aged 15–64 years.

Source: Eurostat, Labour Force Survey (LFS), 2015.

Figure 2 also shows that the steady decline in employment over the whole period of observance is coincidently overlaid by a cyclical development within each year. This indicates that employment variations in the sector within a year are caused by seasonal fluctuations, in that employment peaks are regularly observable in the second and third quarters of a year. This does not come as a surprise since construction activities tend to increase in the warmer seasons, dependent on weather conditions.

In contrast to Figure 2, which gives a view on the overall development of employment in the sector for the EU27, Figure 3 provides a picture of sectoral employment changes disaggregated by country. It shows the annual percentage changes of sectoral employment to the third quarter of the previous year for the period 2008 to 2014 for each individual Member State. Figure 3 indicates that, in all EU Member States, the sector declined, to at least a certain degree, in terms of employment in at least one of the six consecutive years from 2009 to 2014.

Figure 3: Development of employment in construction during the recession

figure3_ef1535.jpg

Notes: Workforce aged 15–64 years. Percentage change to quarter 3 of the previous year.

Source: Eurostat LFS, 2015, and authors’ own calculations on the basis of LFS data.

All countries except for the Czech Republic and Poland recorded a reduction in employment in the construction sector between 2008 and 2009 (Figure 3). In addition, a majority of countries recorded decreases in each of the subsequent years in relation to the respective previous years, although the number of countries recording a growth in employment in relation to the year before increased from only two in 2009 to 12 in 2013. However, this number fell again in 2014 and it remains to be seen whether this year marks a trend reversal or is just an isolated event.

The impact of the recession on the construction sector was particularly strong at the beginning of the crisis. However, the impact diminished steadily over the next few years – at least until 2013. No country recorded an increase in employment in the sector for all the six consecutive years from 2009 to 2014. Only one country, Germany, recorded increases for five years within the six-year period. Conversely, Greece, Italy and Spain all saw job losses within the sector in all six consecutive years of the observation period. Job losses occurred in five years within the six-year period in Cyprus, the Czech Republic, the Netherlands, Romania and Slovenia. Large-scale declines of more than 30% from one year to the next can be observed only in countries such as Cyprus, Ireland, Latvia and Lithuania.

Overall, Figure 3 confirms the finding above that both the impact of the recession on the construction sector and its timing varied considerably between Member States. In this context, it is not possible to link significant job losses to one single cause – the recent recession. Rather, it seems likely that changes in sectoral employment levels within a very short period of time are due to a number of factors including global economic trends and country- and sector-specific developments such as the property crash in Ireland or the real estate bubble burst in Spain.

National level of interest representation

In many Member States, the statutory regulations explicitly refer to the concept of representativeness when assigning certain rights of interest representation and public governance to trade unions and/or employers’ organisations. The most important rights addressed by such regulations include:

• formal recognition as a party to collective bargaining;
• extension of the scope of a multi-employer collective agreement to employers not affiliated to the signatory employers’ organisation;
• participation in public policy and tripartite consultation.

Under these circumstances, representativeness is normally measured by the membership strength of the organisations. In many countries, for instance, statutory provisions allow for the extension of collective agreements to unaffiliated employers only when the signatory trade union and employer association represent 50% or more of the employees within the agreement’s domain.

As outlined previously, the representativeness of the national social partner organisations is of interest to this study in terms of the capacity of their European umbrella organisations for participation in European social dialogue. Hence, the role of the national actors in collective bargaining and public policymaking constitutes another important component of representativeness. The relevance of the European sectoral social dialogue tends to increase with the growing ability of the national affiliates of the European organisations to regulate the employment terms and to influence national public policies affecting the sector.

A cross-national comparative analysis by Franz Traxler shows a generally positive correlation between the bargaining role of the social partners and their involvement in public policy. Social partner organisations that are engaged in multi-employer bargaining are incorporated in state policies to a significantly greater extent than their counterparts in countries where multi-employer bargaining is lacking. This can be attributed to the fact that only multi-employer agreements matter in macroeconomic terms; this in turn gives governments an incentive to persistently seek the cooperation of the social partner organisations. If single-employer bargaining prevails in a country, none of the collective agreements will have a noticeable effect on the economy due to their limited scope. As a result, the basis for generalised tripartite policy concertation will be limited.

In summary, representativeness is a multi-dimensional concept that embraces three basic elements:

• the membership domain and strength of the social partner organisations;
• their role in collective bargaining;
• their role in public policy making.

These elements are discussed in the section below.

Membership domains and strength

The membership domain of an organisation, as formally established by its constitution or name, distinguishes its potential members from other groups which the organisation does not claim to represent. As already explained, this study considers only organisations whose domain relates to the construction sector. However, there is insufficient room in this report to delineate the domain demarcations of all the organisations. Instead, the report notes how they relate to the sector by classifying them according to the four patterns of ‘sector-relatedness’, as specified in the section on concepts and methodology.

There is a difference between strength in terms of the absolute number of members and strength in relative terms. Research usually refers to relative membership strength as the density; in other words, the ratio of actual to potential members.

A difference also arises between trade unions and employers’ organisations in relation to measuring membership strength. Trade union membership simply means the number of unionised persons. Measuring the membership strength of employers’ organisations is more complex since they organise collective entities – companies that employ employees. In this case, there are two possible measures of membership strength – one referring to the companies themselves and the other to the employees working in the member companies of an employers’ organisation.

For a sector study such as this, measures of membership strength of trade unions and employers’ organisations generally also have to consider how the membership domains relate to the sector. If a domain is not identical with the sector demarcation, the organisation’s total density (that is, the density referring to its overall domain) may differ from sector-specific density (that is, the organisation’s density referring to the sector).

This report first presents data on the domains and membership strength of the trade unions and then considers those of the employers’ organisations. As far as sectoral membership numbers are concerned, sectoral densities can be calculated provided the number of employees within the sector is given.

Trade unions

Table 5 presents data on trade union domains and membership strength. It lists all trade unions which meet at least one of the two criteria for classification of a sector-related social partner organisation as defined earlier. The abbreviated and full names of trade unions in the construction sector in the EU27 are listed by country in Annex 1.

All the 27 Member States studied have at least one sector-related trade union. A total of 81 sector-related trade unions were identified. Of these 81 unions, only three have demarcated their domain in a way that is largely congruent relative to the sector definition. This is not a surprise, given that artificially defined demarcations of business activities for statistical purposes tend to differ from the lines along which employees identify common interests and gather in associations. Domains congruent relative to the sector can be found with SB-OGBL and LCGB-CA of Luxembourg and FNV Bouw of the Netherlands: all of them are specific construction trade unions, although it could be possible that that their membership domain would also cover smaller parts of sectors other than construction.

Table 5: Domain coverage, membership and density of trade unions in construction, 2011/2012/2013

Notes: * Domain overlap with other sector-related trade unions. ^a Domain coverage: C = Congruence; O = Overlap; SO = Sectional Overlap; S = Sectionalism (for details see Table 2); ^b Figure includes non-active members; ^c Union representative contacted refused to give (part of) the requested information; ^d Answer deliberately refused.n.a. = not available; n/a = not applicable.

Source: EIRO/EurWORK national correspondents (2013–2014), administrative data and estimates.

Domain demarcations resulting in overlap relative to the sector occur in 36% of the cases for which related information is available (Figure 4). In general, overlap arises from two different modes of demarcation:

• general (that is, cross-sectoral) domains (as is the case for GWU of Malta, CGSLB/ACLVB of Belgium, SIPTU of Ireland, IOZ of Slovakia, and GMB and UNITE of the UK);
• domains covering the broader construction (including manufacture of building materials) and woodworking sector, sometimes also including part of the metal, mining and public utilities industries (as is the case for FITUC and FCIW-Podkrepa of Bulgaria, OS STAVBA of the Czech Republic, IG BAU of Germany, MCA-UGT and FECOMA-CCOO of Spain, FO-Construction, FNCB-CFDT, BATI-MAT-TP CFTC and FNS Construction of France, EFEDOSZSZ of Hungary, LSPS of Lithuania, LCA of Latvia, Budowlani and SbiPD of Poland, SETTACOP and FEVICCOM of Portugal, FGS Familia of Romania and SDGD of Slovenia). Sectional overlaps prevail in the sector and occur in 53% of the cases for which information is available (Figure 4). This mode usually emanates from domain demarcations which focus on certain categories of employees which are then organised across several or all sectors. This mode can be found with trade unions representing employees in segments of the economy sectionalistically overlapping relative to the construction sector. Employee categories are specified by various parameters such as:
• distinct occupations, for example:

- electricians – see DEF of Denmark, FEWU of Finland, TEEU of Ireland and Elektrikerna of Sweden;

- painters – see Malerforbundet of Denmark and SPU of Sweden;

- managers – see YTN of Finland, CFE-CGC BTP of France and Ledarna of Sweden;

- graduate engineers – see SAGE of Sweden;

- draftspersons – see SQTD of Portugal.

• employment status, for example:

- white-collar workers, as is the case of GPA-djp of Austria, Denmark’s HK Privat, Finland’s Pro and Pardia, and Sweden’s Unionen;

- blue-collar workers, as is the case of GBH and PRO-GE of Austria, FGTB-CG/ABVV-AC/and CSC/ACV of Belgium, 3F of Denmark and HFOMD of Greece.

• geographic region, for example, ELA HAINBAT and FCM-CIG of Spain.

Other trade unions’ domains cover part of the construction sector in terms of business activities (rather than in terms of employee categories) in addition to (parts of) at least another sector. Such domains may, for instance, cover the following:

• part of the local government sector (for example, JHL of Finland);
• the private sector (for example, CWU and OOIMSEK of Cyprus);
• the metalworking sector (see EEAÜL of Estonia and IG Metall of Germany);
• the woodworking sector (see FILLEA CGIL, FILCA CISL and FENEAL UIL of Italy).

Last, but not least, sectionalism is also common in the sector, albeit with a relatively small share of 7.7% of trade unions for which related information is available recording this mode of domain demarcation relative to the sector (Figure 4). Sectionalism arises from the existence of sector-specific trade unions which represent, in terms of employee category, one or more particular building grades/professions (for example, plasterers as is the case of Ireland’s OPATSI, carpenters as is the case of DWUBC of Cyprus and bricklayers as is the case of Ireland’s BATU) or employment status (such as blue-collar workers – see Finland’s FCTU, Greece’s GFBRP and Sweden’s Byggnads), without any representational domain outside the sector.

Those trade unions whose membership domain does not cover the entire construction sector have delimited their domain primarily in terms of occupations and economic activities rather than (legal) form/size of enterprise and region: 42.5% of the trade unions with available information have a domain which does not cover all occupations, while 34.2% of the trade unions with available information have a domain which does not cover all economic activities within the sector.

But although there are several relatively highly specialised trade unions with a clear-cut and narrow membership domain focusing on a particular occupational subgroup within the sector’s workforce, there is no indication that this specialisation particularly fosters high unionisation rates in the sector. According to a number of national reports, unionisation rates tend to be lower in the construction sector than in other sectors.

Figure 4: Distribution of membership domain patterns of sector-related trade unions in construction

figure4_ef1535.jpg

Notes: Percentages are rounded. N = 78.

Source: EIRO/EurWORK national reports (2013–2014)

The construction sector is characterised by several features, most of which are generally deemed unfavourable to member recruitment. These are:

• relatively low average skill levels, in particular among onsite construction workers in the EU15 and in southern European countries (4.34 MB PDF);
• high labour turnover;
• high presence of non-standard (fixed-term, part-time) as well as undeclared work;
• high incidence of migrant work.

As the trade union domains often overlap with the demarcation in the sector, so do their sectoral domains with one another in the case of those countries with a pluralist trade union ‘landscape’ in the construction sector. In the pluralist trade union systems (recording more than one sector-related trade union) of Austria, Denmark, Estonia and Greece, no case of inter-union domain overlap within the sector can be observed. In all other countries with more than one sector-related trade union (Belgium, Bulgaria, Cyprus, Finland, France, Germany, Ireland, Italy, Latvia, Luxembourg, Malta, the Netherlands, Poland, Portugal, Spain, Sweden and the UK), the sectoral domain of at least one trade union overlaps with the sectoral domain of at least one other trade union (see Table 8 in the section on Collective bargaining). Depending on the scale of mutual overlap, this results in competition for members. Noticeable inter-union competition within the sector is recorded in five countries: Finland, France, Germany, Malta and Portugal.

Membership of the sector-related trade unions is voluntary in all cases.

The absolute numbers of trade union members differ widely, ranging from about 1.5 million (in the case of UNITE in the UK) to just over 600 (in the case of Ireland’s OPATSI) (Table 5). This considerable variation reflects differences in the size of the economy and the comprehensiveness of the membership domain rather than the ability to attract members. Hence, density is a more appropriate measure of membership strength for a comparative analysis. This holds true despite the fact that some of the density figures gathered and calculated for the purpose of this study may be unreliable.

Therefore this report considers densities referring to the sector (‘sectoral density’), given that a trade union’s membership within the sector and the number of employees in the sector are both provided. Moreover, some tentative information (without providing figures) on the trade unions’ ‘sectoral domain density’ (that is, the density referring only to that part of the sector as covered by the union’s domain) in relation to their overall ‘domain density’ (that is, the density referring to the union’s overall domain) is available for those unions with a domain (sectionalistically) overlapping with regard to the sector (see below).

The sectoral density figures provided in this section refer to net ratios, which means that they are calculated on the basis of active employees only rather than taking all union members (that is, those in work and those who are not) into account. This is mainly because research usually considers net union densities as more informative than gross densities, since the former measure tends to reflect unionisation trends among the active workforce more quickly and more appropriately than the latter, since only the active workforce is capable of taking industrial action.

More than 50% of the trade unions with available data have a sectoral density (calculated as the ratio of the number of members in the sector to the total number of employees in the sector) of less than 5%. Sectoral density is 30% or less in the case of more than 95% of the trade unions which document figures on density (Table 5).

There are two possible explanations for the overall very low sectoral densities of the sector-related trade unions:

• low densities with regard to the unions’ sectoral domain (sectoral domain densities);
• their generally small size (in terms of sectoral membership domain) in relation to the sector.

While only tentative information is available for the former issue (see below), the latter appears to apply to many of the sector-related trade unions. This is indicated by the fact that more than half of the unions have a membership domain which is sectionalist or sectionalistically overlapping with regard to the sector and thus covers only part of the sector. But because sectoral density data can be calculated for only slightly more than half of the 81 sector-related trade unions, the available figures on sectoral density should be treated with caution.

Comparing the trade unions’ overall domain densities with their sectoral domain densities provides an indication of whether the construction sector tends to be a stronghold of sector-related trade unions which also organise employees in sectors other than the construction industry, or not. The Eurofound national correspondents were asked to give a substantiated estimate of the relationship between these two densities, if possible, without providing exact figures. Almost the same number of trade unions (for which information is available) recorded a sectoral domain density that was either lower than or higher than their overall domain density. This result is astonishing given the outstandingly low sectoral densities of most trade unions which, in tandem with anecdotal evidence depicted in several national reports indicating low unionisation rates, would suggest a majority of unions record a sectoral domain density lower than their overall domain density. However, related information was provided for only a minority of the sector-related trade unions and, moreover, some of the answers to this question might eventually turn out to be unreliable. The results nevertheless show that, overall, the construction sector cannot be qualified as a stronghold of those trade unions with a membership domain (sectionalistically) overlapping with regard to the sector. Rather, in many cases, their core membership base supposedly lies in areas other than the construction sector.

In conclusion, the study reveals that there are many occupational trade unions in the construction sector which often record relatively narrow membership domains. This may favour a particularistic representation of collective interests on behalf of small professional groups. Nevertheless, despite the shortcomings in relation to data availability and the existing dataset, one can infer from the information collected that union density rates in the sector do not tend to be high. The relatively low densities within the sector can be explained by a large number of factors, such as relatively low average skill levels as well as non-standard (including bogus self-employment) and migrant work.

Employers’ organisations

Tables 6, 7 and 9 present information on employers’ and business organisations in the construction sector. The abbreviated and full names of the employers’ and business organisations in the construction sector in the EU27 are listed by country in Annex 1.

As is the case of the trade union side, all 27 Member States have at least one sector-related employers’ organisation. In eight countries (Bulgaria, Cyprus, Estonia, Latvia, Lithuania, Malta, Spain and Slovakia), only one sector-related employers’ organisation matching at least one of the two criteria for inclusion (see above) was identified. The remaining 19 countries have pluralist associational systems, meaning there are at least two sector-related employers’ organisations.

Generally, business interest organisations may also deal with interests other than those related to industrial relations. As explained in an earlier Eurofound report, Employers’ organisations in Europe, organisations specialised in matters other than industrial relations are commonly defined as ‘trade associations’. Such sector-related trade associations exist in the construction sector. In terms of the national scope of their activities, all the associations that are not involved in collective bargaining (see Table 9 in the section on Collective bargaining) either primarily or exclusively act as trade associations in their country. Put simply, the main reference of trade associations is the ‘product’ market (where business has interests in relation to customers and suppliers) rather than the labour market. It is only the conceptual decision to include all associational affiliates to FIEC and EBC, regardless of whether they have a role in national bargaining which gives them, as a work hypothesis, the status of a social partner organisation within the framework of this study.

Of the 104 employers’ or business organisations identified in this study, eight belong to this group of trade associations (no related information is available for three organisations, namely Luxembourg’s FEDIL, Poland’s KPB UNI-BUD and Romania’s UNPR). As outlined above, in eight countries of the EU27 only one single organisation (in the meaning of a social partner organisation as defined before) has been established. The incidence of pluralist associational systems on the employers’ side is therefore slightly lower (19 of the 27 Member States) than on the trade union side (21 of the 27 Member States). However, the number of sector-related employers’ or business organisations across the Member States (104) clearly outweighs the number of sector-related trade unions (81). Overall, the employers’ and business organisations are more unevenly distributed among the Member States than the trade unions, in that a number of countries that have only one sector-related employers’ or business organisation while a few countries (Austria, Italy and the UK) have 10 or more of such organisations.

Table 6: Domain coverage and membership of employers’ and business organisations in construction

Notes: Data for 2011, 2012 or 2013 as available.

* Domain overlap with other sector-related employer/business organisations; ** No information on domain overlaps provided; ^a Domain coverage: C = Congruence; O = Overlap; SO = Sectional Overlap; S = Sectionalism (for details see Table 2); ^b Compulsory until autumn 2013; ^c FIEC suggests 250,000 employees employed by member companies; ^d Figure doubtful; ^e Rough estimate provided by EBC; ^f Figure questioned by EBC; n.a. = not available.

Source: EIRO/EurWORK national correspondents (2013–2014), administrative data and estimates.

Table 7: Density and affiliations of employers’ and business organisations in construction

Notes: Data for 2011, 2012 or 2013 as available.

* Domain overlap with other sector-related employers’ or business organisations; ** No information on domain overlaps provided; *** Only cross-sectoral (that is, peak-level) associations are listed; **** Affiliation in parenthesis means indirect affiliation via higher order unit; ^a Until about 2012; ^b No FIEC member, but tied to FIEC by a cooperation agreement; ^c Figure questionable; ^d Figure questioned by EBC; n.a. = not available; n/a = not applicable

Source: EIRO/EurWORK national correspondents (2013–2014), administrative data and estimates

The membership domains of the employers’ and business organisations tend to be narrower than those of the sector-related trade unions. In contrast to organised labour, membership domains which are sectionalist relative to the sector – at least in relative terms – prevail among the employers’ organisations, with a share of 49% of the cases for which related information is available; 15.6 % and 26%, respectively, of the associations rest on overlapping and sectionalistically overlapping domains relative to the sector (Figure 5). No organisation in the sector has a domain that is cross-sectoral.

Most cases of domain overlaps (in the case of organisations with domains either overlapping or sectionalistically overlapping relative to the sector) are caused by:

• coverage of (part of) the broader defined building and construction sector, including also the manufacture of building materials (as is the case of BCC of Bulgaria, CNA UNIONE and ANIEM of Italy, LBA of Latvia, ZGIGM of Slovenia and CNC of Spain), architecture (as is the case of Portugal’s AICE and Slovakia’s ZSPS) and a number of distinct and very specialised activities related to construction in a broad sense, such as metal engineering (see Austria’s BIM), carpentry activities (see SDMSZS of the Czech Republic and TMF of Sweden), monument preservation (as is the case of Germany’s BV Farbe and BV Steinmetze) and recycling activities (see DA of Germany);
• coverage of (part of) the whole ‘industry’ sector (as is the case of Dansk Byggeri and DS H&I of Denmark and LSA of Lithuania).

There are also several employers’ or business organisations whose domain is focused on a particular segment of the economy which sectionally overlaps the construction sector. Such organisations may cover the cooperative sector (such as ANCPL, FEDERLAVORO and AGCI of Italy), all kinds of artisan activities (see IPOSZ of Hungary, CLAAI of Italy, FDA of Latvia and ZRP of Poland), training and education activities (as is the case of EEEL of Estonia, again LSA of Lithuania and again LBA of Latvia) and local/ regional government services (as is the case of LGA in the UK).

Sectionalism, the prevailing domain pattern relative to the construction sector among the sector-related employers’ and business organisations, is caused by domain demarcations that focus on a particular subsector or segment of the construction sector, without covering areas of business activity outside the sector. Such subsectors or segments may be defined by:

• size class of construction enterprises such as SMEs (as is the case of Austria’s BIB, Belgium’s Bouwunie, France’s CAPEB and SNSO, and FMB in the UK) or large companies (as is the case of FVBI of Austria and ASI of Italy);
• ownership structure of the construction enterprises such as private sector companies (as is the case of all employers’ organisations in Austria as well as Portugal’s AECOPS and FEPICOP);
• highly specialised activities within the construction sector, such as roofing (see Austria’s BIDGS, Germany’s ZVDH and Sweden’s PLR), sanitary and heating engineering activities (see Austria’s BISHL and Germany’s ZVSHK), electrical (engineering) activities (see Tekniq of Denmark, FFIE of France, ECA and AECI of Ireland, as well as ECA in the UK) and civil engineering activities (as is the case of FNTP of France and CECA in the UK).

Finally, 9.4% of the associations have a membership domain that is more or less congruent with the sector definition (Figure 5). This means that the domain of these organisations largely covers the construction sector as defined for the purpose of this study.

Figure 5: Distribution of membership domain patterns of sector-related employers’ organisations in construction

figure5_ef1535.jpg

Notes: Percentages are rounded; N = 96.

Source: EIRO/EurWORK national reports (2013–2014)

In several countries, the sectoral employers have managed to establish specific employers’ organisations as a particular voice of narrow and distinct business activities within the construction sector. Accordingly, almost 59% of the employers’ and business organisations for which information is available have delimited their domain in terms of business activities in that they do not cover all activities within the construction sector. Moreover, some 49% of the organisations for which information has been provided do not represent all (legal) forms of companies in the sector (in most cases focusing on particular size classes of enterprises), while domain demarcations in terms of territorial coverage play only a minor role.

In countries with a highly fragmented and differentiated associational ‘landscape’ on the employer side – such as Austria, Germany, France, Italy, Sweden and the UK – the associations’ domains tend to be tailor-made for a particular subgroup of employers and businesses within the sector. More strikingly than on the side of organised labour, this enables these associations to perform a very individualised interest representation on behalf of their members, although their membership strength may vary widely from one organisation to the other. Such a fragmented associational configuration tends to favour the (bargaining) power of organised business in small segments of the economy.  

Membership is compulsory for all 13 sector-related employers’ organisations in Austria and one in Germany (BI Gerüstbau) (Table 6). In the case of the Austrian associations, this is due to their public law status as chamber units. In case of BI Gerüstbau, membership is mandatory due to its public law status as a guild. As far as related information has been provided, membership of all the other sector-related employers’ or business organisations is voluntary (Table 6).

In those countries with a pluralist structure in relation to employers’ organisations, these associations have usually managed to arrive at non-competing and collaborative relationships; the exceptions are SPS v ČR and SDMSZS of the Czech Republic as well as several associations of Italy. Their activities are complementary to each other as a result of inter-associational differentiation by membership demarcation (as is the case, in particular, of Austria, France, Germany, Italy, Sweden and the UK) than by functions and tasks.

As the figures on membership totals (Table 6) and density (Table 7) indicate, membership strength in terms of both companies and employees varies widely in terms of the membership domain in general and the sector. Again, as outlined earlier in the context of the trade unions, density figures rather than absolute membership numbers are informative in terms of membership strength.

In the case of the sector-related employers’ and business organisations, the sectoral densities of both companies and employees (employed by these companies) can be calculated. However, the lack of absolute numbers of members, in particular in terms of employees, in the construction sector in the case of many associations means that sectoral densities can only be calculated only for some of them.

According to the figures available, about 70% and about 53%, respectively, of the employers’ and business organisations have a sectoral density in terms of companies and employees of 5% or less (Table 7). Whereas the median of the organisations’ sectoral densities in terms of companies lies at 1.4%, the corresponding median in terms of employees stands at 4%. This does not necessarily infer overall very low densities of the sector-related employers’ and business organisations in construction, since sectoral densities (in contrast to sectoral domain densities) tend to decline with increasing levels of associational fragmentation. Higher sectoral densities in terms of employees compared with those in terms of companies indicate a higher propensity by the larger companies to associate than their smaller counterparts.

As for the sector-related trade unions, some tentative information on the sectoral domain density of the employers’ and business organisations in relation to their overall domain density is available for those associations with a domain (sectionalistically) overlapping with regard to the sector. But because related information has only been provided for very few employers’ and business organisations, these data should be treated with caution.

In contrast to the situation on the trade union side, the sectoral domain densities, at least in terms of employees, of the employers’ and business organisations tend to be higher than or at least equal to their overall domain densities in the vast majority (that is, 83.3%) of cases for which related information has been provided. These, albeit very tentative, results indicate that, unlike the trade union side, the construction sector may constitute a stronghold of many of those employers’ and business organisations with a domain (sectionalistically) overlapping with regard to the sector. This corresponds with the fact that many sector-related employers’ and business organisations tailored their membership domain to the construction sector (or part of it) so as to align their policy of interest representation with the specific requirements of their members.

Collective bargaining and its actors

Table 8 lists all the trade unions engaged in sector-related collective bargaining. Despite numerous cases of inter-union domain overlap and some cases of unclear domain demarcation, inter-union rivalry and competition for bargaining capacities was identified in only a few countries such as Finland, France, Malta and Portugal.

Eight Member States (Czech Republic, Estonia, France, Italy, Latvia, Lithuania, Malta and Portugal) have an employers’ or business organisation that is not a party to collective bargaining (Table 9). Although this is the only employers’ or business organisation identified in the case of Estonia, Latvia and Malta, all three report being consulted regularly. No information on collective bargaining involvement was provided for three organisations (FEDIL of Luxembourg, KPB UNI-BUD of Poland and UNPR of Romania). Those associations not involved in sector-related collective bargaining are classified as social partner organisations in this report only because of their affiliation to at least one of the sector-related European-level employers’ organisations, FIEC and EBC. In the case of the sector-related employer organisations, cases of such rivalry have been reported from countries such as the Czech Republic, Finland, Ireland and Italy. Conversely, in at least 23 of the 27 Member States which record one or more sector-related employers’ or business organisations, at least one of them is engaged in sector-related collective bargaining.

Table 8: Collective bargaining, consultation and affiliations of trade unions in construction

Notes: Notes: Data for 2011, 2012 or 2013 as available. * Domain overlap with other sector-related trade unions; ** Only cross-sectoral (that is, peak level) associations are listed; *** Affiliation put in parenthesis means indirect affiliation via higher or lower order unit; ^a Collective bargaining involvement: S = single-employer bargaining; M = multi-employer bargaining; ^b Number of employees covered by collective agreements concluded by the union within the construction sector; ^c Union representative contacted refused to give (part of) the requested information; n.a. = not available

Source: EIRO/EurWORK national correspondents (2013–2014), administrative data and estimates

Table 9: Collective bargaining and consultation of employers’ and business organisations in construction