Conference Agenda

In their renowned study, “What do unions do?” Freeman and Medoff (1984) argued 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 vary according to the wage levels agreed upon in collective bargaining. Almost forty years later, however, little is known about which wages are set in collective bargaining agreements (CBAs). A few industrialized countries code the content of the CBAs, but the resulting databases are not comparable across countries. A large majority of countries does not collect the CBAs signed, let alone code their content. While information about wages and working conditions data is critical for monitoring their progress, many countries rely solely on data from labour force or enterprise surveys, on inventories of national bargaining systems or on reviews of legal regulations, but not on information about the wages agreed in CBAs. This absence of data on what exactly has been concluded in collective bargaining is most likely related to the fact that worldwide no person or institution is systematically collecting full texts of CBAs and coding their content. Our session highlights the first attempt to fill this gap, namely through the global CBA-Database maintained by WageIndicator. This database contains an increasing number of agreements from a growing number of countries, currently more than 2,000 agreements from 67 countries. Coding of CBAs is performed for twelve topics, namely Job titles, Wages, Working Hours/Schedules/Leaves, Employment Contracts, Work And Family Arrangements, Health And Safety And Medical Assistance, Sickness And Disability, Social Security And Pensions, Training, Gender Equality Issues, Coverage and General CBA data.

The outline of our session is as follows. The first contribution covers how CBAs are collected and coded according to a predefined coding scheme, as well as how a representative sample of CBAs relative to the labour forces is composed. The second contribution presents findings regarding the content of CBAs in agriculture and manufacturing in Africa. The third contribution outlines the findings regarding the CBAs in the textile, garment and leather industry in Indonesia. The fourth contribution focusses on the entire dataset and highlights which gender-related clauses have been agreed. The fifth and final paper details the content of collective bargaining in Turkey.

The CBA-Database was established in 2012 and, related to a range of projects, a growing number of CBAs has been added, of which approximately half come from European countries. Each agreement is collected, coded and analyzed according to a coding scheme with twelve macro topics with related sub-questions, resulting in around eight hundred variables. Almost all CBAs are published online in the WageIndicator national websites, all in national language and accessible for free by all users, who can browse CBAs online, select the clauses and topics of their interest and use the CBA comparison tool, which provides a unique opportunity to closely examine the variation across agreements and countries. More than 900 CBAs of this database are used to identify what share of the workers who are earning under 105% of the statutory minimum wage (or an imputed minimum wage for countries without minimum wages) are covered by sectoral or enterprise collective bargaining in 9 European countries.

Are wage settings up to the negotiation process between the employers and employees or at the level of the collective agreements or wages are set by the state or in individual contracts? Do pay scales differ across sectors in Africa CBAs? The CBA-Database has 424 CBAs from 20 African countries. One third are from manufacturing, followed by agriculture and transport. Analyses show that almost 90% have at least one clause related to wages. Exactly half of these state that wages are to be agreed in individual contracts, almost one in ten at sectoral level, one in ten at company level, three in ten unspecified. In manufacturing and education, wages are set the least in individual contracts. Only one in ten CBAs include pay scales, and these are most frequent noticed in human health and social work activities with 4 out of 5 CBAs. In manufacturing pay scales are hardly noticed with only 8 out of 108 CBAs.

Late 2020, Indonesia passed the Omnibus Law–Job Creation Law No.11, followed by its derivative regulations in 2021 aiming to stimulate foreign investment and a more flexible labour market. The CBA-Database holds 239 collective agreements from Indonesia, of which 163 from textile, garment, and leather. All CBAs minus 2 are signed by a single employer or company, of which 28 are multinational enterprises. In terms of wages, structural wage increases are more often agreed in the textile, garment and leather industry than in the remaining industries (88% versus 77%). Out of those 163 CBAs, 4 companies were found with one CBA negotiated before the Omnibus Law was implemented and one after. In all four CBAs the provisions deteriorated, specifically on severance pay, while 3 out of 4 CBAs stated increasing maximum overtime hours from 14 to 18 hours/week. An improvement in wage clauses is seen in 3 CBAs that regulate pay scales for those working one year or longer.

Women account for almost 50% of the global workforce. However, the Female Labour Force Participation varies across countries. Do countries with higher Female Labour Force Participation rates address gender equality and work/family balancing issues better through their CBAs? Or is it only the female dominated industries that have a higher percentage of female friendly clauses in their CBAs? Cross-Continental (Asia, Europe, Africa, Central and South America) analysis shows that almost eight in ten CBAs include clauses on work/family arrangements and four in ten do so on gender equality. With 30 weeks of maternity leave, CBAs in Croatia have the longest leave, whereas Mozambique has on average only 8.5 weeks. Other gender equality and work/family balancing clauses include health and safety, paternity leave, parental leave, and breastfeeding and childcare. More than 2000 agreements are analysed using 11 gender equality and 24 work/family balancing clauses to provide a comparative view of the female friendly CBAs across continents.

Collective bargaining in Turkey is subject to two different regimes, namely the public and private sector. Public sector CBAs are concluded at the national level to cover all sectors, especially regardıng wages and social rights. Additionally, CBAs are concluded at the national level for 11 different service branches. In the private sector, trade unions are organized at sector level and CBAs are concluded at the workplace and enterprise level. The CBA-Database has 68 CBAs for Turkey of which 25 were concluded by public servants' trade unions and 43 by the workers trade unions. It should also be noted that 13 of the relevant CBAs were concluded after the May 2020 period and are the renewed versions of the old CBAs in the CBA-Database. This situation has a very important place in terms of evaluating the reflection of the corona period on CBAs. Thus, the strengths and weaknesses of the collective bargaining system in Turkey will be made visible.