Europe-Wide Collective Agreements Data Collection Feasibility Study

Abstract

In their renowned study “What do unions do?”, Freeman and Medoff (1984) argue that trade unions bargain for higher wages, equal pay and fair working conditions, implying that collective bargaining is central to wage setting processes and that wage outcomes will vary according to the wage levels agreed in collective bargaining. Almost forty years later, little is known about the wage outcomes agreed in collective bargaining agreements (CBAs) in the European Union (EU). The CAWIE report (Van Gyes and Schulten, 2015) showed that some knowledge of agreed wages is available from National Statistical Offices or Central Banks for ten EU countries only. Even less is known about the working conditions agreed in CBAs in the EU whereas, in view of the European Commission’s (EC) re-launched dialogue with social partners at European level (European Commission, 2016), such data is critical for monitoring progress in wage-setting and in setting standards for working conditions. The knowledge gap remains evident in the ESDE 2020 report (European Commission, 2020), in the OECD ‘Negotiating Our Way Up’ report (2019), and in EUROFOUND’s flagship report ‘Industrial relations: Developments 2015–2019’ (Eurofound 2020). Knowledge about the impact of collective bargaining is based on survey data of individuals or companies, on inventories of national bargaining systems or on legal regulations, but not on details about what exactly has been agreed in the large number of CBAs concluded throughout Europe. A main reason for this knowledge gap is that a European-level registry that collects and codes CBAs is lacking. In the EU knowledge about what exactly is concluded in collective bargaining remains a blind spot; therefore, neither cross-sectional nor longitudinal knowledge can be accumulated.

This article aims to explore the feasibility of building a database of CBAs in Europe to tackle this blind spot in the body of knowledge in industrial relations research, taking as a starting point that the use of the Internet and advances in natural language processing may accelerate data collection and coding (Askitas and Zimmermann, 2015). In section 2 the body of knowledge regarding the impact of CBA clauses on wages and working conditions is reviewed. Section 3 explores which EU countries maintain a CBA registry, how large the stock of CBAs is in Europe, whether an EU-wide registry s feasible, and what the possibilities are for representative sampling. Section 4 details the building blocks of an EU-wide CBA registry-in-the-making, including the requirements for gathering, coding and annotating CBAs with the WageIndicator CBA Database as an example and the options for machine-reading of CBA texts. Section 5 draws conclusions on how to establish a continuous, Europe-wide data collection of coded CBAs and thus fullfilling the main condition for contributing to the body of knowledge about the outcomes of collective bargaining throughout the EU.

This article is based on desk research and on three experiences with CBA Databases. A first one regards the coded database of collective agreements in the Netherlands maintained by the trade union confederation FNV and the employers’ association AWVN (Tijdens and Van Klaveren, 2003; Schreuder and Tijdens, 2004; Yerkes and Tijdens, 2010). A second one concerns the coding of collective agreements by means of survey questions in Europe in an EC-funded Social Dialogue project (WIBAR-3 VS/2014/0533). A third one concerns experiences with the WageIndicator CBA Database in two consecutive EC-funded Social Dialogue projects (BARCOM VS/2016/0106, COLBAR-EUROPE VS/2019/0077), BARCOM VS/2016/0106, COLBAR-EUROPE VS/2019/0077, Ceccon et al., 2016).