Clment Brbion (PSE) February 8th, 2019 1 What are works councils? - - PowerPoint PPT Presentation

The ‘strategic discrimination’ of works councilors in Germany: new evidence of the demise of a model?

Clément Brébion (PSE) February 8th, 2019

1

• Two main types of elected actors can represent the labour force at the shop

floor:

– Union delegates – Works councilors

• Union delegates:

– Take part in collective bargaining – Act on behalf of a union which dimensions and preferences exceed the scope of the firm

• Works councils:

– Scope of interests is restricted to the firm environment.

• Different entitlements

– Freeman and Lazear (1995, p. 29), “in contrast to plant-level unions, councils cannot call strikes nor negotiate wages […]. Their function, often specified in legislation, is to foster labor and management cooperation with the goal of increasing the size of the enterprise ‘pie’”

2

What are works councils?

• The theoretical economic literature on collective organisations

– Basic classical /neoclassical models under perfect competition, no collective organisation – Introduction of unions in the main neoclassical models as ‘rent-seaking’ (‘rent-distributing’)

• rganisations – mostly per the model of a monopoly able to set wages (Dunlop, 1944)

– Freeman (1976) et Pencavel (1977) emphasize that unions are also ‘rent-generating’

• rganisations – able to reduce some market failures (informational asymmetries, principal-

agent problems, free-riding problems)

3

Motivation 1 – In the economic literature, little is known on the workers’ representatives taking part in collective negotiations

• The theoretical economic literature on collective organisations fostered …
• … a large stream of empirical research estimating the impact of unions on

covered firms and their average worker

– In terms of wages, employment, productivity… (Lewis, 1986; Blanchflower and Bryson, 2004; Mayneris, Poncet and Zhang , 2015)

4

Motivation 1 – In the economic literature, little is known on the workers’ representatives taking part in collective negotiations

• The theoretical economic literature on collective organisations fostered …
• … a large stream of empirical research estimating the impact of unions on

covered firms and their average worker

• … research on works councils

– Legal entitlements of works councils arelikely to limit their ‘rent-seaking’ behaviours and to foster their ability to generate rents. – Freeman and Lazear (1995, p. 29), “in contrast to plant-level unions, councils cannot call strikes nor negotiate wages […]. Their function, often specified in legislation, is to foster labor and management cooperation with the goal of increasing the size of the enterprise ‘pie’” – Empirical research on councils’ impact on firm performance and wage (FitzRoy and Kraft 1985, 1987, 1990)

5

Motivation 1 – In the economic literature, little is known on the workers’ representatives taking part in collective negotiations

• The literature on collective organisations has traditionally been limited to

the estimation of their impacts on covered firms and workers.

• But very little is known on the actors of the negotiations.

– Stormer (2010) “The main body of research on work councils has been conducted on a collective institutional level, neglecting work council members at an individual level.” – Breda (2014) : “To my knowledge, there is no quantitative economic paper in French or English dealing with the role of union leaders, either at the level of the firm or at the national level”

6

Motivation 1 – In the economic literature, little is known on the workers’ representatives taking part in collective negotiations

• A context of generalized decentralization of the bargaining process

(Baccaro et Howell, 2011)

Motivation 2 – Focusing on negotiations at the firm level – the increasing strategic importance of WoCo for the firms

7

1 2 3 4 5 1975 1980 1985 1990 1995 2000 2005 2010

Note: The index goes from 1 (fully decentralized) to 5 (fully centralized). The shaded area displays the range between the first bottom and top deciles. Only countries with information over the full period are used. Source: ICTWSS database (Visser, 2015)

Figure 1 - Yearly average, bottom and top deciles of the 'predominant level of centralisation of wage bargaining' in the OECD

• A context of generalized decentralization of the bargaining process

(Baccaro et Howell, 2011)

Motivation 2 – Focusing on negotiations at the firm level – the increasing strategic importance of WoCo for the firms

8 1 2 3 4 5 1975 1980 1985 1990 1995 2000 2005 2010

Note: The index goes from 1 (fully decentralized) to 5.75 (fully centralized). It is calulated as: LEVEL - (fAEB+OCG)/4(=max value) + (Art+DR-1)/5(=max value), where LEVEL= the predominant level at which wage bargaining takes place, fAEB = the frequency or scope of additional enterprise bargaining, Art= the formality of additional enterprise bargaining and its control by union bodies, OCG = General Opening clauses in collective agreement,DR= Derogations. The shaded area displays the range between the first bottom and top deciles. Only countries with information over the full period are used. Source: ICTWSS database (Visser, 2015)

Figure 2 - Yearly average, bottom and top deciles of the 'actual level of centralisation of wage bargaining' in the OECD

• A context of generalized decentralization of the bargaining process

(Baccaro et Howell, 2011)

• The rising strategic importance of shop-floor representatives for the firm

– In particular, works councils have to engage in bargaining over an increasing range of issues Likely to deter the quality of employer-employee relations Employers are likely to pay increasing attention to the composition of works councils… … especially in the absence of sectoral agreement.

9

Motivation 2 – Focusing on negotiations at the firm level – the increasing strategic importance of WoCo for the firms

• A context of generalized decentralization of the bargaining process

(Baccaro et Howell, 2011)

• The rising strategic importance of shop-floor representatives for the firm
• An increased risk of ‘strategic discrimination’ (Breda, 2014; Bourdieu and

Breda, 2016), especially against works councilors

– Focus on the main workers’ representatives taking part in the collective negotiation at the firm level in France : the ‘délégués syndicaux’.

• The delegates play two bargaining games with their employer

– “Theoretically, the interaction between the employer and the ‘délégués syndicaux’ can lead to two Nash equilibria” (Breda, 2014)

• The non-cooperative equilibrium
• The cooperative equilibrium

 A risk of ‘strategic discrimination : Bully the most radical representatives, ‘buy’ the most cooperative ones

10

Motivation 2 – Focusing on negotiations at the firm level – the increasing strategic importance of WoCo for the firms

• A context of generalized decentralization of the bargaining process

(Baccaro et Howell, 2011)

• The rising strategic importance of shop-floor representatives for the firm
• An increased risk of ‘strategic discrimination’ (Breda, 2014; Bourdieu and

Breda, 2016), especially against works councilors

• Main result :

On average, union delegates suffer a wage penalty 10% larger than the other union members (whose penalty in null in Breda, 2014 and of 4-5% in Breda and Bourdieu, 2016). Several elements suggest that the most vehement representatives would suffer a stronger penalty

11

Motivation 2 – Focusing on negotiations at the firm level – the increasing strategic importance of WoCo for the firms

• A context of generalized decentralization of the bargaining process

(Baccaro et Howell, 2011)

• The rising strategic importance of shop-floor representatives for the firm
• An increased risk of ‘strategic discrimination’ (Breda, 2014; Bourdieu and

Breda, 2016), especially against works councilors

• … which is likely to depend on the institutional rules

12

Motivation 2 – Focusing on negotiations at the firm level – the increasing strategic importance of WoCo for the firms

• It can be expected that the welfare gains induced by works councils

are maximized in Germany

– Works councils benefit from the largest ‘co-determination rights’ in the Western world. – CBAs on most strategic issues are traditionally relegated to the branch level

13

Motivation 3 – A risk of ‘strategic discrimination’ theoretically limited in Germany

• Good quality of employer-employees relations in Germany

Source : World Economic Forum - The Global Competitiveness Index Historical Dataset Note : Self-declared estimation based on the question : “In your country, how do you characterize labor-employer relations?” [1 = generally confrontational; 7 = generally cooperative]

14

Motivation 3 – A risk of ‘strategic discrimination’ theoretically limited in Germany

0,00 1,00 2,00 3,00 4,00 5,00 6,00 7,00

France Korea, Rep. Italy Greece Turkey Spain Slovenia Poland Slovak Republic Portugal Hungary Belgium Latvia Mexico Australia Chile Czech Republic Israel United States Estonia Canada Germany United Kingdom Finland Ireland New Zealand Luxembourg Iceland Netherlands Sweden Japan Austria Norway Denmark Switzerland

Figure 3 - Cooperation in labor-employer relations in the OECD countries Averaged over 2007-2017

15

The German institutional context – The traditional dual system of industrial relations

• Sectoral collective bargaining

– Every four years – Between trade unions and business’ associations (single-employer bargaining also possible) – Typically includes questions of wages, working conditions, working time and job classifications – Agreements generally extended to the whole firm but rarely to the whole sector – Unions can call for strikes around these periods.

16

• Sectoral collective bargaining

– Every four years – Between trade unions and business’ associations (single-employer bargaining also possible) – Typically includes questions of wages, working conditions, working time and job classifications – Agreements generally extended to the whole firm but rarely to the whole sector – Unions can call for strikes around these periods.

• Firm-level bargaining between the employer and the WC

– In firms with more than 5 full-time employees, employers are compelled to facilitate the formation of a WC when referred to by the workers. – Professional elections on a 4-year basis – Works councilors can only discuss issues that are not bargained over at the sectoral level and their “intervention power in managerial decisions decreases with the proximity with essential business decisions” (Müller-Jentsch, 1995:60) – Forbidden to call for strikes. – About 2/3 of works councilors are union members.

The German institutional context – The traditional dual system of industrial relations

• A system under pressure

– Financialization of the economy (Goyer, 2007) – German reunification and the extension of the EU to Eastern European countries – Decreasing unionization rates – Privatization of former public services

17

The German institutional context – Recent evolutions

• A system under pressure

– Financialization of the economy – German reunification – Decreasing unionization rates – Privatization of former public services

• Works councils engage in ‘wildcat cooperation’

– Pre-existing feature of the traditional model which has been strengthening since the 1980s, in particular against the threat of outsourcing.

• Internal erosion

– Harship clauses, opening clauses, OT membership

18

The German institutional context – Recent evolutions

… bound to an opening clause … using an opening clause (when bound to an o.c.) 2005 39.7% 52.9% 2011 52.7% 77.0% Share of total employees working in firms….

Source : Addison et al (2017:46)

• External erosion

19

The German institutional context – Recent evolutions

Source : Oberfichtner and Schnabel (2017) . Database : IAB establishment panel

• The number of works councilor depends on the firm size
• Release from usual tasks

– The number of WoCo members benefitting from full-time release depends on the firm size. – The other works councilors « shall be released from their work duties […] to the extent necessary for the proper performance of their funcions » (WCA, section 37-2) – Extra-time on behalf of the mandate shall be converted into time-off rather than be paid.

• Earnings

– « The post of member of the works council shall be unpaid » (WCA, section 37-1). It is a honorary position rather than a job. – « During his term of office and for one year thereafter the remuneration of a member

• f the works council shall not be fixed at a lower rate than the remuneration paid to

workers in a comparable position who have followed the career that is usual in the

• establishment. The same shall apply to general benefits granted by the employer.”

(WCA, section 37-4)

20

The German institutional context – works councilors’ timetables and earnings

• German Socio Economic Panel

– Representative yearly panel of the German population

• I use the waves 2001, 2003, 2006, 2007, 2011 and 2015
• I approximate the firms’ WoCo status in 2003, 2007 and 2015 for the workers

– with firm-tenure larger than 5 years in the following wave. – whose firm has not changed status between the following and previous waves.

• I approximate the union status of 2006 by the one of 2007 for the workers

– with firm-tenure larger than 2 years in 2007.

21

Data

Information on… 2001 2003 2006 2007 2011 2015 2016 The presence of a WoCo in the firm     Works council membership       Union membership     

Variables of interest per chosen year in the GSOEP

• The main model is an OLS with individual and time fixed effects.

– I drop individuals observed only once – On average, a respondent is observed 3.45 times over 7.3 years

• The main sample is composed of workers on open-ended contracts, aged

between 25 and 64, working between 30 and 60 hours per week in firms with a works council

• Main dependent variable : logarithm of the hourly gross wage

– Calculated as the declared gross wage applying the month preceding the interview divided by 4.3*the declared number of hours worked the week preceding the interview – I trim the top and bottom 1% of the yearly wage distribution

22

Data

• Incidence of works councilors and union members in the main sample

23

Descriptive Statistics

Table 3: Incidence of WoCo and union memberships in the final sample Member of a Trade Union No Yes Total Member of the Works Council No 6379 2630 9009 65.3% 26.9% 92.2% Yes 262 498 760 2.7% 5.1% 7.8% Total 6641 3128 9769 68.0% 32.0% 100.0%

Source : German Socio-Economic Panel, own calculations

24

Descriptive Statistics

Table 2: Average values of different variables according to firm coverage and, within covered firm, according to the WoCo membership status.

Workers in uncovered firms Workers in covered firms WoCo members in covered firms WoCo non- members in covered firms Difference (3) – (4) (1) (2) (3) (4) (5) Hourly gross wage 13.76 18.19 17.17 18.18

• 1.01***

Monthly gross wage 2632.39 3343.19 3102.77 3336.97

• 234.20***

Actual working hours 44.37 42.43 42.08 42.40

• 0.31

Sex, m=1 fem=2 1.31 1.31 1.28 1.32

• 0.04**

Age of Individual 42.28 44.50 45.39 44.43 0.96*** No working hour agreement 0.06 0.03 0.01 0.03

• 0.02***

Region, W=1 E=2 1.31 1.23 1.24 1.23 0.01 Seniority 10.34 15.68 16.64 15.79 0.85** Education General Elementary 0.05 0.06 0.06 0.06

• 0.00

Middle vocational 0.58 0.49 0.56 0.48 0.07*** Vocational + Abitur 0.08 0.08 0.07 0.08

• 0.01

Higher Vocational 0.10 0.10 0.07 0.10

• 0.02**

Higher Education 0.18 0.26 0.23 0.27

• 0.04**

Inadequately or no Answer 0.01 0.01 0.01 0.01

• 0.00

Isco88 (1 digit) Legislators, senior officials and managers 0.08 0.06 0.06 0.06

• 0.00

Professionals 0.11 0.18 0.18 0.19

• 0.01

Technicians and associate professionals 0.22 0.27 0.27 0.27

• 0.00

Clerks 0.10 0.13 0.14 0.14 0.00 Service workers and shop and market sales workers 0.08 0.03 0.03 0.03

• 0.00

Craft and related workers 0.27 0.16 0.19 0.15 0.04*** Plant and machine operators and assemblers 0.09 0.11 0.08 0.11

• 0.03**

Elementary occupations 0.05 0.05 0.05 0.05 0.00 Firm size Ge 5 Lt 20 0.38 0.02 0.03 0.02 0.01 Ge 20 Lt 100 0.37 0.12 0.17 0.10 0.07*** Ge 100 Lt 200 0.10 0.11 0.12 0.11 0.01 Ge 200 Lt 2000 0.10 0.36 0.37 0.37 0.00 Ge 2000 0.05 0.38 0.32 0.40

• 0.09***

Source : German Socio-Economic Panel, own calculations

25

Estimations – The baseline regression

• The baseline model is an OLS with individual and time fixed effects.
• Controls:

– Gender, age, age², seniority, seniority², dummies for SES and firm sector at the 1-digit level, firm size (6 categories), the month of interview and dummies to control for the region (East, West) and for whether the individual has a contract specifying the number

• f working hours.

26

Estimations – The baseline regression

Table 5: Effect of works council and union memberships on the log hourly gross wage according to

the sector (baseline model) (1) (2) (3) (4) All sectors Manufacturing sector Private service sectors Public Sector (no civil servant) Member of the Works Council

• 0.004

0.048***

• 0.042**
• 0.008

(0.009) (0.018) (0.018) (0.016) Member of a Trade Union

• 0.014*

0.004

• 0.075***
• 0.001

(0.007) (0.015) (0.018) (0.014) Individual Fixed Effect Yes Yes Yes Yes Time Fixed Effect Yes Yes Yes Yes Observations 9,769 2,321 2,080 2,682 R-squared 0.896 0.893 0.933 0.879 Individuals 2833 734 687 820

Standard errors in parentheses. *** p<0.01, ** p<0.05, * p<0.1 Source : German Socio-Economic Panel, own calculations

27

Estimations – The baseline regression

Table 6: Differentiation of the baseline effects between entrance and exit from the works council. Dependent variable: log hourly gross wage (1) (2) (3) (4) Manufacturing sector Manufacturing sector Private service sectors Private service sectors Member of the Works Council 0.046** 0.069**

• 0.042*
• 0.032

(0.022) (0.033) (0.022) (0.035) Member of a Trade Union 0.005

• 0.002
• 0.082***
• 0.077***

(0.015) (0.016) (0.020) (0.022) Individual Fixed Effect Yes Yes Yes Yes Time Fixed Effect Yes Yes Yes Yes Group 1+   Group 2+   Group 3+ Group 4+ Group 5+     Observations 2,200 2,107 1,909 1,819 R-squared 0.900 0.900 0.939 0.932 Individuals 700 671 636 609

Standard errors in parentheses. *** p<0.01, ** p<0.05, * p<0.1 Source : German Socio-Economic Panel, own calculations

+ I separate agents who are observed at least once as a works councilor into four groups. Among the 'switchers', group 1 includes

respondents whose only change in status is to become a works councilor, group 2 includes respondents whose only change in status is to leave the works council, group 3 includes respondents who are observed both voted in and out of the organisation. Group 4 includes respondents always observed in office. Respondents never observed in office are part of the group 5

28

Estimations – The baseline regression

Table 7: Alternative dependent variable: log monthly gross wage and number of actual working hours Dependent variable : log of the monthy gross wage Dependent variable : number of actual working hours (1) (2) (3) (4) Manufacturing sector Private service sectors Manufacturing sector Private service sectors Member of the Works Council 0.044**

• 0.029*
• 0.263

0.564 (0.017) (0.017) (0.408) (0.461) Member of a Trade Union 0.003

• 0.066***
• 0.083

0.389 (0.014) (0.017) (0.328) (0.474) Individual Fixed Effect Yes Yes Yes Yes Time Fixed Effect Yes Yes Yes Yes Observations 2,321 2,080 2,321 2,080 R-squared 0.921 0.951 0.763 0.771 Individuals 734 687 734 687

Standard errors in parentheses. *** p<0.01, ** p<0.05, * p<0.1 Source : German Socio-Economic Panel, own calculations

29

Robustness checks (I) – Secondary sample 1 : large firms

• Possible measurement error of the WoCo coverage
• Secondary sample 1 : restrict on large firms workers rather than on

covered firms.

Table 8 : WoCo coverage according to the firm size WoCo at place of work Yes No Firm size 5-20 649 4 615 12.33 % 87.67% 20-100 3 405 4 392 43.67 % 56.33 % 100-200 3 335 1 104 75.13 % 24.87 % 200-2000 10 572 1 187 89.91 % 10.09 % >2000 11 536 546 95.48 % 4.52 % Total 29 995 12 147 71.18 % 28.82 %

Source : German Socio-Economic Panel, own calculations

30

Robustness checks (I) – Secondary sample 1 : large firms

Table 8: Alternative sample 1 - workers in firms with more than 200 employees. Dependent variable - log hourly gross wage (1) (2) (3) (4) All sectors Manufacturing sector Private service sectors Public Sector (no civil servant) Member of the Works Council

• 0.006

0.032*

• 0.060***
• 0.016

(0.010) (0.018) (0.022) (0.020) Member of a Trade Union

• 0.007
• 0.025*
• 0.009
• 0.002

(0.008) (0.014) (0.020) (0.017) Individual Fixed Effect Yes Yes Yes Yes Time Fixed Effect Yes Yes Yes Yes Observations 10,424 2,561 2,494 2,396 R-squared 0.900 0.894 0.931 0.879 Individuals 3262 847 875 792

Standard errors in parentheses. *** p<0.01, ** p<0.05, * p<0.1 Source : German Socio-Economic Panel, own calculations

31

Causality or adverse selection?

• Individual fixed effect => No inherent difference in the individual level of

productivity

• Agents experiencing differentiated wage evolution as compared to

‘normal’ trajectories more likely to run for elections because of this specific trend?

– Shirking employees who want to take advantage of the job protection right associated with works council membership – Employees who used to be invested but do not feel ‘thanked’ for their engagement (Artus, 2013)

• Compatible with the differentiated results according to the sector?

32

Causality or adverse selection?

• Do pre-trends in wage differ between respondents about to become

works councilor and workers who will never be in office?

• Strategy:

– Treated :

• Sequence in status
• I use observation and generate a variable equals to the trend in wage between

and – Control:

• Never observed in office
• I compute for all their observations the trend beween and

– Model :

• Probit
• Clustered standard errors

.

33

Causality or adverse selection?

Table 11: Test of differences in pre-trends between the treatment and control groups. Model : probit with clustered standard errors; the marginal effects are given. Dependent variable: dummy variable taking the value 1 if the respondent is works councilors in the following wave (see text) (1) (2) Manufacturing sector Private service sectors Pre-trend in the hourly gross wage

• .023

4.1e-03 (.026) (.028) Individual Fixed Effect No No Time Fixed Effect Yes Yes Observations 1,469 1,199 Individuals 607 532

Standard errors in parentheses, clustered by individual. *** p<0.01, ** p<0.05, * p<0.1 Source : German Socio-Economic Panel, own calculations

34

Causality or adverse selection?

• I cannot rule out the possibility of an evolution in post-election

productivity

– Job protection => increased likelihood to shirk – Rational employers may promote unelected workers who do benefit from release time.

• WCA (section 119) : “prejudicing or favouring a member or substitute

member of the works council […] by reason of his office [is] punishable by a term of imprisonment not exceeding one year or a fine, or both”

35

Estimations – The baseline regression

Table 5: Effect of works council and union memberships on the log hourly gross wage according to

the sector (baseline model) (1) (2) (3) (4) All sectors Manufacturing sector Private service sectors Public Sector (no civil servant) Member of the Works Council

• 0.004

0.048***

• 0.042**
• 0.008

(0.009) (0.018) (0.018) (0.016) Member of a Trade Union

• 0.014*

0.004

• 0.075***
• 0.001

(0.007) (0.015) (0.018) (0.014) Individual Fixed Effect Yes Yes Yes Yes Time Fixed Effect Yes Yes Yes Yes Observations 9,769 2,321 2,080 2,682 R-squared 0.896 0.893 0.933 0.879 Individuals 2833 734 687 820

Standard errors in parentheses. *** p<0.01, ** p<0.05, * p<0.1 Source : German Socio-Economic Panel, own calculations

36

Bringing back the context, a case for strategic discrimination?

• The manufacturing sector (Haipeter, 2011)

– Historically the sector at the core of the traditional model of industrial relations in Germany. Large density of employers’ and employee’s unions Large coverage of collective bargaining agreements – The ‘organised’ decentralisation of industrial relations first applied in this sector in the 1990s, to the benefit of firms in economic distress. – 2000s: employers constrain works councils to engage into negotiations over a new stream of derogations. Employees strongly opposed to further derogations considered as “violations of the norms of distributional justice” Works councils and unions proved to be cooperative partners.

37

Bringing back the context, a case for strategic discrimination?

• The manufacturing sector (Haipeter, 2011)

– 2 strategies to convince the workforce

• Gather employees to insist on the risk that some of them are to lose their job and

therefore on the need to stand unified

• Organize short-term strikes to “channel[l] the critical attitude towards derogations

into broad support for the works councilsin their negotiations with management”. – In the end,

• WoCo “were not helpless victims of structural constraints; on the contrary, they

actively tried to restructure the situation to their own advantage”

• “By demonstrating a new capacity to act, in most of the cases they have been able

to regain a great deal of the power lost through the structural constraints man-agement was able to impose on them” – Note the role of the relation WoCo-union

38

Bringing back the context, a case for strategic discrimination?

• The private service sectors

– More difficult to provide general elements on the private service sectors

• In the sub-sample: 25% work in trade, 16% in transport, 23% in banking or

insurance, 8% in service to industries and 28% in personal services

• Case studies on the relations between WoCo and the management are fewer

– In all these sectors , but the financial one, weak historical tradition of collective bargaining – Managers more often classify works agreements as increasing the rigidity in the firm than in the manufacturing sector (Nienhueser, 2009).

39

Bringing back the context, a case for strategic discrimination?

• The low-wage private service sectors (Artus, 2013)

– Sectors where “wage costs and flexible work-time are key dimensions of [the] competitive strategies” – Prevalence of a strong corporate identify. In these firms, “the absolute identification with the company […] is an explicit aim of personnel policy  Strong opposition of managers against the formation of a works council  When a works council is elected, “a climate that is hostile to co-determination still dominates

40

Bringing back the context, a case for strategic discrimination?

Table 13: Effect of the interaction between works council and union memberships on the log hourly gross wage, the log monthly gross wage and the actual number of working hours Dependent variable : log hourly gross wage Dependent variable : log monthly gross wage Dependent variable : number of actual working hours (1) (2) (3) (4) (5) (6) Manufacturing sector Private service sectors Manufacturing sector Private service sectors Manufacturing sector Private service sectors Member of the Works Council 0.012

• 0.036
• 0.001
• 0.017
• 0.654

0.837 (0.032) (0.022) (0.030) (0.021) (0.713) (0.585) Member of a Trade Union 0.001

• 0.072***
• 0.002
• 0.060***
• 0.122

0.533 (0.015) (0.020) (0.014) (0.019) (0.333) (0.511) WoCo Member * Union Member 0.052

• 0.015

0.065*

• 0.029

0.559

• 0.656

(0.037) (0.033) (0.035) (0.031) (0.838) (0.863) Individual Fixed Effect Yes Yes Yes Yes Yes Yes Time Fixed Effect Yes Yes Yes Yes Yes Yes Observations 2,321 2,080 2,321 2,080 2,321 2,080 R-squared 0.893 0.933 0.921 0.951 0.763 0.771 Individuals 734 687 734 687 734 687

Standard errors in parentheses. *** p<0.01, ** p<0.05, * p<0.1 Source : German Socio-Economic Panel, own calculations

41

Bringing back the context, a case for strategic discrimination?

Table 14: Effect of the interaction between works council and political steadfastness on the log hourly gross wage, the log monthly gross wage and the actual number of working hours Dependent variable : log hourly gross wage Dependent variable : log monthly gross wage Dependent variable : number of actual working hours (1) (2) (3) (4) (5) (6) Manufacturing sector Private service sectors Manufacturing sector Private service sectors Manufacturing sector Private service sectors Supports a Political Party 0.009

• 0.003

0.013 0.003 0.161 0.297 (0.011) (0.011) (0.010) (0.010) (0.240) (0.273) Member of the Works Council 0.015 0.001 0.023 0.008 0.349 0.283 (0.025) (0.023) (0.023) (0.022) (0.555) (0.593) Member of a Trade Union 0.004

• 0.073***

0.003

• 0.064***
• 0.075

0.356 (0.015) (0.018) (0.014) (0.017) (0.329) (0.473) WoCo Member * Supports a Pol. Party 0.063**

• 0.078***

0.039

• 0.069***
• 1.111

0.473 (0.031) (0.027) (0.030) (0.026) (0.701) (0.699) Individual Fixed Effect Yes Yes Yes Yes Yes Yes Time Fixed Effect Yes Yes Yes Yes Yes Yes Observations 2,321 2,080 2,321 2,080 2,321 2,080 R-squared 0.893 0.933 0.921 0.951 0.763 0.771 Individuals 734 687 734 687 734 687

Standard errors in parentheses. *** p<0.01, ** p<0.05, * p<0.1 Source : German Socio-Economic Panel, own calculations

• In most Western countries, increasing strategic importance of shop-floor labour

representatives for both employers and employees

• This is particularly true of works councilors in Germany where the traditional

model provide large codetermination rights on non-distributional issues.

• Yet the German traditional model of employment relations has largely evolved
• I show that, in the manufacturing sector, works councilors earn on average 5%

more than when they are out of office. In the private service sectors, a penalty of 5% is evidenced.

• I bring some elements suggesting that the relation is causal and, at least partly,

driven by some strategic discriminating behaviours of rational employers.

42

Conclusion