scieee AI-readable full text Open interactive document viewer

The Heterogeneous Economic Consequences of Works Council Relations

Pfeifer, Christian

Abstract

EconStor is a publication server for scholarly economic literature, provided as a non-commercial public service by the ZBW.

Full text

Pfeifer, Christian Article The Heterogeneous Economic Consequences of Works Council Relations Schmollers Jahrbuch – Journal of Applied Social Science Studies. Zeitschrift für Wirtschaftsund Sozialwissenschaften Provided in Cooperation with: Duncker & Humblot, Berlin Suggested Citation: Pfeifer, Christian (2011) : The Heterogeneous Economic Consequences of Works Council Relations, Schmollers Jahrbuch – Journal of Applied Social Science Studies. Zeitschrift für Wirtschaftsund Sozialwissenschaften, ISSN 1865-5742, Duncker & Humblot, Berlin, Vol. 131, Iss. 1, pp. 59-71, https://doi.org/10.3790/schm.131.1.59 This Version is available at: https://hdl.handle.net/10419/292319 Standard-Nutzungsbedingungen: Die Dokumente auf EconStor dürfen zu eigenen wissenschaftlichen Zwecken und zum Privatgebrauch gespeichert und kopiert werden. Sie dürfen die Dokumente nicht für öffentliche oder kommerzielle Zwecke vervielfältigen, öffentlich ausstellen, öffentlich zugänglich machen, vertreiben oder anderweitig nutzen. Sofern die Verfasser die Dokumente unter Open-Content-Lizenzen (insbesondere CC-Lizenzen) zur Verfügung gestellt haben sollten, gelten abweichend von diesen Nutzungsbedingungen die in der dort genannten Lizenz gewährten Nutzungsrechte. Terms of use: Documents in EconStor may be saved and copied for your personal and scholarly purposes. You are not to copy documents for public or commercial purposes, to exhibit the documents publicly, to make them publicly available on the internet, or to distribute or otherwise use the documents in public. If the documents have been made available under an Open Content Licence (especially Creative Commons Licences), you may exercise further usage rights as specified in the indicated licence. https://creativecommons.org/licenses/by/4.0/ The Heterogeneous Economic Consequences of Works Council Relations By Christian Pfeifer* Abstract I use a question about works council relations from the 2006 wave of the IAB Establishment panel to analyze the heterogeneous effects of works councils on productivity, wages, and profits. The results indicate that the effects differ significantly between works council relationship types in a systematic pattern. The overall findings are in line with productivity-enhancing and rent-sharing functions of works councils. JEL Classification: J53, M54 Received: 5. November 2010 Accepted: 3. January 2011 1. Introduction In the last two decades, many empirical studies have been conducted that analyze the economic consequences of works councils in German firms (for detailed literature reviews see Frege, 2002; Addison / Schnabel / Wagner, 2001, 2004; Jirjahn, 2005, 2006; and articles by Jirjahn, 2011 and Mueller, 2011, in this special issue). Most of these studies use the IAB Establishment Panel or the Hannover Firm Panel and compare firms with and without works councils. The findings are mostly in line with theoretical considerations about productivity-enhancing and rent-sharing functions of works councils (Freeman / Lazear, 1995; Hübler / Jirjahn, 2003). For example, most studies report positive or at least non-negative effects of works councils on firms’labor productivity, positive effects on workers’wages, and negative or non-significant effects on profits. One shortcoming of previous econometric studies is that they only look at the existence of a works council, but cannot look into the black box of works Schmollers Jahrbuch 131 (2011), 59 –71 Duncker & Humblot, Berlin Schmollers Jahrbuch 131 (2011) 1 *I thank Knut Gerlach, Jens Mohrenweiser, and participants of the Workshop on Microdata and Codetermination at Leuphana University Lüneburg for their comments. The used data are protected and may only be accessed at the Research Data Center (FDZ) of the Federal Employment Service (BA) in the Institute for Employment Research (IAB, Nuremberg, Germany). The computer programs can be requested from the author. OPEN ACCESS | Licensed under | https://creativecommons.org/about/cclicenses/ DOI https://doi.org/10.3790/schm.131.1.59 | Generated on 2023-01-16 13:36:15 council-management-relations due to data limitations. The pioneering study of works council typology by Kotthoff (1981; 1994) follows a case study approach to analyze the social relationships between works councils and management in 63 German firms. Although Kotthoff does not explicitly analyze the economic consequences of different works council types, his study is important because it distinguishes between effective and ineffective works councils in terms of interest representation. Most firms in Kotthoff's sample have ineffective works councils that can be characterized as ignored by management or workers, as isolated by an authoritarian management, or as part of the management. On the other hand, effective works councils are characterized as respected regulators, as respected surveillance, or as cooperative counter-power. Nienhüser (2005) uses data from a telephone survey with HR officers in firms that have a works council. He identifies four types of works councils along the two dimensions bargaining power and willingness to cooperate and examines their impact on firm-level bargaining agreements. A recent study by Jirjahn / Mohrenweiser / Backes-Gellner (2009) uses a small works council survey to analyze the determinants of bad works council relations as perceived by management. Their main focus is on learning dynamics; economic consequences of works council relations are not considered. An exception in the evaluation of economic consequences of works council types are studies by Dilger (2002; 2006). Dilger connects data from the NIFAPanel with the Bochumer Survey of Works Councils. He uses perceptions of the relationship between management and works councils by both sides to identify different types such as antagonistic or cooperative works councils. The findings indicate that the effects of these works council types on labor turnover, flexible working time arrangements, product innovations, and profit situation differ significantly in size. For example, the cooperative type and a more positive perception of works council-management-relations reduces labor turnover by most. In this paper, I use a question about works council relations from the 2006 wave of the large-scale IAB Establishment Panel to analyze the heterogeneous effects of three works council types on economic outcomes such as labor productivity, wages, and profits. The main results are that the effects between the three works council types differ significantly in a systematic pattern, which is consistent with theoretical expectations of productivity-enhancing and rentsharing effects of works council types. For example, more cooperative works councils have stronger positive effects on productivity and more bargaining works councils have stronger positive effects on wages. Because I use a sample of firms with and without works councils as well as firms with different works council types, the results are important for the economic consequences of works councils as a whole, which is most important for policy, and for the efficient organization of industrial relations on the firm level. The remainder of the paper is structured as follows. The data set and variables are described in Section 2. Section 3 presents the econometric results. The 60 Christian Pfeifer OPEN ACCESS | Licensed under | https://creativecommons.org/about/cclicenses/ DOI https://doi.org/10.3790/schm.131.1.59 | Generated on 2023-01-16 13:36:15 paper concludes with a short summary and discussion of the results in Section 4. 2. Data and Descriptive Statistics The IAB Establishment Panel contains rich data on establishments from all sixteen German federal states (“Bundesländer”) and all industries (Kölling, 2000). Every year more than 15,000 firms with at least one employee covered by social security are interviewed in an unbalanced panel design survey. The sample is stratified according to ten establishment sizes and sixteen industries, with oversampling of larger firms. The observational unit is the establishment, i.e., the local unit in which major activities of an enterprise are carried out. Main concern of the survey is to gain insights into the firm’s most important parts of operation, decision-making, and more specifically employment. For the purpose of this study, I can only use the wave 2006 because it is the only year that contains a question about works council-management-relations. Since productivity and profit information are stated for the last business year, these economic outcome variables are taken from the wave 2007. Thus, firms in the estimation sample have to participate in surveys in both years 2006 and 2007. Further sample restrictions had to be applied. First, the sample includes only firms with at least five employees, because smaller firms do not need to establish works councils under the legislation of the Works Constitution Act (“Betriebsverfassungsgesetz”). Second, firms from agriculture, hotel, restaurant, education, health, social, non-profit, public, or financial sector are dropped, because they usually do not report profits or sales, which is important for the computation of the productivity variable. Third, only firms without missing values in the used variables are considered. Overall 4693 firms remain in my sample for the subsequent empirical analysis. The question about works council relations and the three possible answer categories, from which the explanatory variables of interests are generated, are worded as follows: How would you characterize the role of the works council in managerial decision making in your establishment? (1) Works council is in line with management in most decisions from the outset. )Type 1: “mostly in line with management” (2) Works council has often a different opinion, but in the end a consensus is reached. )Type 2: “different opinion but with consensus” (3) Decisions have often to be enforced against the works council. )Type 3: “different opinion without consensus” The Economic Consequences of Works Council Relations 61 OPEN ACCESS | Licensed under | https://creativecommons.org/about/cclicenses/ DOI https://doi.org/10.3790/schm.131.1.59 | Generated on 2023-01-16 13:36:15 If the works council complies with most management decisions from the outset, this relationship type can have two reasons. One the one hand, the management decisions might take already employees’interests into account so that the works council must not disagree with management. On the other hand, a works council might be simply too weak to bargain with management or might follow own interests instead of representing the workforce interests. The second type describes works councils that express different employee opinions and bargain successfully with management so that a consensus is reached in most cases. The third type is mostly against decisions of the management and agreements between works council and management are seldom reached so that decisions have to be enforced against the works council. The three works council types can be broadly characterized along the two dimensions cooperation and bargaining. Type 1 and type 2 can be seen as cooperative, whereas type 3 is likely to show few interests in cooperation with management or at least is unsuccessful in finding a consensus. Furthermore, type 1 is probably not strongly engaged in bargaining activities with management, whereas type 2 and type 3 are likely to bargain more and harder with management; the former mostly with and the latter mostly without a consensus. Because cooperation has productivity-enhancing effects and bargaining is often associated with rent-sharing activities, the effects of the three works council types on productivity, wages, and profits are expected to differ between each other, i.e., a homogeneous works council effect is rather unlikely. Table 1 informs about the frequency of works councils and their relationship types. The incidence of works councils is about 31 percent among all firms in my sample. Almost three out of four firms with a works council judge their works council of type 2 (“different opinion but with consensus”), about 23 percent of type 1 (“mostly in line with management”), and less than 4 percent of type 3 (“different opinion without consensus”). Thus, the large majority of works councils has a rather cooperative relationship with firms' management. An important determinant of works council relations seems to be the existence of a union bargained collective contract. Split samples for firms with and without collective contracts show that type 2 is more likely if the firm is bound to a collective contract. A rationale for this finding might be that unions strengthen works councils (e.g., advisory, financial, and personnel capacity) in their position to bargain with a firm's management. To estimate the economic consequences of the works council types, three outcome variables are used: productivity, wages, and profits. A firm's average labor productivity is proxied by the log value added per employee, which is sales minus inputs in the entire business year 2006 divided by the number of employees in June 2006. Firm's average wages are proxied by the log of total gross monthly salaries divided by the number of employees in June 2006. The data also includes the perceived profit situation in 2006, which could be answered on a five-point scale ranging from one for very good to five for very 62 Christian Pfeifer OPEN ACCESS | Licensed under | https://creativecommons.org/about/cclicenses/ DOI https://doi.org/10.3790/schm.131.1.59 | Generated on 2023-01-16 13:36:15 Table 1 Frequencies of works council relationship types Complete sample (n= 4693) Without union contract (n= 2374) With union contract (n= 2319) Works council existence 31.39 (100) 13.31 (100) 49.89 (100) Type 1: “mostly in line with management”7.35 (23.42) 3.96 (29.75) 10.82 (21.69) Type 2: “different opinion but with consensus” 22.97 (73.19) 9.01 (67.72) 37.26 (74.68) Type 3: “different opinion without consensus” 1.07 (3.39) 0.34 (2.53) 1.81 (3.63) Notes: Frequencies in percent. All four variables are dummies. The reference group for works council types is no works council so that the frequencies of the types sum up to the frequency of works council existence. The relative frequencies for firms with a works council are in parentheses. Source: IAB-Establishment Panel, 2006 / 2007 (controlled remote data access via FDZ); own computations. bad. In addition to this ordinal profit measure, a binary indicator for an at least good profit situation is generated. The basic econometric model looks as in equation (1), in which Ydenotes the outcome variable (log productivity, log wages, good profit dummy, or ordinal profit situation), WC indicates the existence of a works council as well as the three works council types (reference group are firms without works councils), Xincludes a set of control variables (union contract, mainly foreign capital owner, firm founded before 1990, state of technology, average working time, employment shares of qualification groups, part-time, fixed-term contracts and females, number of employees and squared term, 9 industry dummies, 16 regional dummies), Greek letters indicate parameters to be estimated, εthe usual residual term, and iis a firm index. Descriptive statistics for all variables are presented in Table 2. Yi¼þ1WCtype 1 iþ2WCtype 2 iþ3WCtype 3 iþXiþ"i ð1Þ For productivity as well as wages, linear regressions can be applied, whereas ordered probit estimates are the appropriate estimation technique for the ordered profit outcome and binary probit estimates for the binary profit situation. In all estimates, heteroskedasticity robust standard errors are computed. Also note that, though effects of works councils on outcomes are discussed, the applied regression analyses estimate in principal only correlations, which are not necessarily causal effects due to unobserved heterogeneity and reverse causality issues. The Economic Consequences of Works Council Relations 63 OPEN ACCESS | Licensed under | https://creativecommons.org/about/cclicenses/ DOI https://doi.org/10.3790/schm.131.1.59 | Generated on 2023-01-16 13:36:15 64 Christian Pfeifer Table 2 Descriptive statistics Mean Std. Dev. Min. Max. Economic outcome variables PROD: log productivity per employee (sales minus inputs divided by number of employees) 10.7360 0.8767 5.4146 14.3437 WAGE: log wage per employee (total monthly salaries divided by number of employees) 7.4890 0.5489 4.8721 8.9120 PROFIT_G: at least good profit situation (dummy) 0.4856 0.4998 0 1 PROFIT_O: ordinal profit situation (1: very good, 2: good, 3: satisfactory, 4: bad, 5: very bad) 2.6851 1.0389 1 5 Works council relations variables (reference group: no works council) Type 1: “mostly in line with management” dummy) 0.07351 0.26101 0 1 Type 2: “different opinion but with consensus” (dummy) 0.22970 0.42069 0 1 Type 3: “different opinion without consensus” (dummy) 0.01065 0.10268 0 1 Control variables Union bargained collective contract (dummy) 0.4941 0.5000 0 1 Mainly foreign capital owner (dummy) 0.0712 0.2571 0 1 Firm founded before the year 1990 (dummy) 0.5172 0.4998 0 1 State of technology is alright (dummy) 0.2742 0.4462 0 1 State of technology is new (dummy) 0.5152 0.4998 0 1 State of technology is very new (dummy) 0.1784 0.3828 0 1 Normal average working hours per week (hours) 39.1895 2.1619 15 60 Employees unskilled for easy tasks (share) 0.1734 0.2437 0 1 Employees with apprenticeships for qualified tasks (share) 0.6199 0.2458 0 1 Employees with college degrees for qualified tasks (share) 0.0888 0.1595 0 1 Apprentices (share) 0.0513 0.0709 0 0.5625 Employees with part-time work (share) 0.1669 0.2144 0 1 Employees with fixed-term contracts (share) 0.0445 0.1069 0 0.9745 Female employees (share) 0.3192 0.2588 0 1 Number of employees in June 2006 158.2924 791.6042 5 35019 Notes: Number of firms in the complete estimation sample is 4693. Source: IAB-Establishment Panel, 2006 / 2007 (controlled remote data access via FDZ); own computations. OPEN ACCESS | Licensed under | https://creativecommons.org/about/cclicenses/ DOI https://doi.org/10.3790/schm.131.1.59 | Generated on 2023-01-16 13:36:15 3. Econometric Results Table 3 presents the regression results for the complete estimation sample. The linear regression results for productivity (log value added per employee) are displayed in the second column. Firms with a works council of type 1 (“mostly in line with management”) have on average a significant larger productivity (about 20 percent) than firms without a works council; and firms with a works council of type 2 (“different opinion but with consensus”) have an even larger productivity (about 40 percent).1These findings are consistent with the consideration that cooperative works councils have a productivity-enhancing effect. But to enhance productivity, an exchange of information between works council and management in bargaining processes seems to have an additional positive effect. Moreover, it is a remarkable result that the existence of a works council does not harm productivity even if management decisions have to enforced against the works council (type 3). The estimated coefficients indicate that firms with a works council of type 3 (“different opinion without consensus”) are on average not less productive than firms without works councils. In fact, these firms have on average a productivity that is about 15 percent higher than in firms without works councils. The estimated effect for works councils of type 3 is however not statistically significant. The large standard error might be reasoned by effect heterogeneity, which can be seen in separate estimates for firms with and without collective contracts in the robustness section. The estimated coefficients for the wage function (log total salaries per employee) in the third column show that wages are significant larger in firms with works councils. Wages in firms with a works council of type 1 (“mostly in line with management”) are approximately 15 percent, in firms with a works council of type 2 (“different opinion but with consensus”) approximately 17 percent, and in firms with a works council of type 3 (“different opinion without consensus”) nearly 30 percent larger than in firms without works councils. The results are consistent with the rent-sharing function of works councils, i.e., works councils increase workers’share of the increased value added (Freeman / Lazear, 1995). That works councils of type 3 (“different opinion without consensus”) have the largest effect on wages might be reasoned by strong works councils, which mostly care about income of workers and not about firm performance in the longer run. It seems however questionable why firms can afford to pay such high wages, although the productivity estimates have shown no significant productivity-enhancing effect of this relationship type. An explanaThe Economic Consequences of Works Council Relations 65 1To compute the percentage change in productivity and wages measured in Euros from OLS coefficients of log-linear functions, the subsequent formula is applied: ðe1ÞThe estimated productivity effects of up to 40 percent are quite large but fall into the range of previous findings (e.g., Addison / Schnabel / Wagner, 2004; Jirjahn, 2011; Mueller, 2011). OPEN ACCESS | Licensed under | https://creativecommons.org/about/cclicenses/ DOI https://doi.org/10.3790/schm.131.1.59 | Generated on 2023-01-16 13:36:15 66 Christian Pfeifer Table 3 Impact on productivity, wages, and profit situation in complete sample OLS coefficients Binary probit marginal effects Ordered probit coefficients Marginal effects for ordered probit outcomes PROD WAGE PROFIT_G PROFIT_O Profit=1 Profit=2 Profit=3 Profit=4 Profit=5 Type 1: “mostly in line with management”0.1842*** (0.0468) 0.1444*** (0.0187) 0.0210 (0.0315) -0.0293 (0.0655) 0.0048 (0.0109) 0.0069 (0.0152) -0.0037 (0.0085) -0.0047 (0.0104) -0.0033 (0.0072) Type 2: “different opinion but with consensus”0.3564*** (0.0351) 0.1600*** (0.0153) -0.0607*** (0.0230) 0.1095** (0.0477) -0.0169** (0.0071) -0.0266** (0.0119) 0.0127** (0.0052) 0.0177** (0.0078) 0.0131** (0.0060) Type 3: “different opinion without consensus”0.1412 (0.1162) 0.2594*** (0.0434) -0.1401** (0.0691) 0.4011** (0.1981) -0.0490*** (0.0174) -0.1057* (0.0547) 0.0271*** (0.0025) 0.0659** (0.0318) 0.0618 (0.0392) Union bargained collective contract -0.0352 (0.0256) -0.0002 (0.0122) 0.0150 (0.0176) -0.0642* (0.0355) 0.0103* (0.0057) 0.0153* (0.0084) -0.0079* (0.0044) -0.0103* (0.0057) -0.0073* (0.0041) Mainly foreign capital owner 0.1707*** (0.0454) 0.0885*** (0.0167) 0.0361 (0.0303) -0.0407 (0.0704) 0.0067 (0.0119) 0.0095 (0.0162) -0.0052 (0.0093) -0.0065 (0.0112) -0.0045 (0.0076) Firm founded before the year 1990 0.0525** (0.0254) 0.0450*** (0.0128) -0.0347* (0.0179) 0.0672* (0.0369) -0.0108* (0.0060) -0.0160* (0.0088) 0.0083* (0.0046) 0.0108* (0.0059) 0.0077* (0.0042) State of technology is alright 0.0665 (0.0598) 0.0959*** (0.0308) 0.0062 (0.0448) -0.1282 (0.0954) 0.0215 (0.0166) 0.0297 (0.0214) -0.0168 (0.0131) -0.0203 (0.0149) -0.0140 (0.0100) State of technology is new 0.1814*** (0.0596) 0.0985*** (0.0302) 0.1145*** (0.0435) -0.3372*** (0.0941) 0.0541*** (0.0152) 0.0797*** (0.0220) -0.0408*** (0.0111) -0.0538*** (0.0149) -0.0392*** (0.0113) State of technology is very new 0.2588*** (0.0645) 0.1133*** (0.0323) 0.1998*** (0.0439) -0.5383*** (0.0996) 0.1082*** (0.0242) 0.1020*** (0.0135) -0.0846*** (0.0181) -0.0784*** (0.0129) -0.0473*** (0.0069) Normal average working hours -0.0098 (0.0077) -0.0033 (0.0036) -0.0052 (0.0038) 0.0121 (0.0082) -0.0020 (0.0013) -0.0029 (0.0019) 0.0015 (0.0010) 0.0019 (0.0013) 0.0014 (0.0009) Employees unskilled for easy tasks 0.3682** (0.1516) 1.0316*** (0.1006) 0.3003*** (0.1054) -0.5796*** (0.2030) 0.0933*** (0.0328) 0.1378*** (0.0498) -0.0717*** (0.0269) -0.0930*** (0.0327) -0.0663*** (0.0233) Employees with apprenticeships for qualified tasks 0.7915*** (0.1508) 1.4129*** (0.1015) 0.3265*** (0.1052) -0.6636*** (0.2037) 0.1068*** (0.0329) 0.1578*** (0.0504) -0.0821*** (0.0274) -0.1065*** (0.0329) -0.0759*** (0.0234) Employees with college degrees for qualified tasks 1.5263*** (0.1608) 2.1157*** (0.1024) 0.4245*** (0.1130) -0.7829*** (0.2168) 0.1260*** (0.0351) 0.1861*** (0.0537) -0.0969*** (0.0293) -0.1256*** (0.0351) -0.0896*** (0.0250) Apprentices -0.4605** (0.2094) 0.1176 (0.1283) 0.3165** (0.1506) -0.5205* (0.2962) 0.0838* (0.0477) 0.1238* (0.0714) -0.0644* (0.0378) -0.0835* (0.0477) -0.0596* (0.0339) Employees with part-time work -1.1755*** (0.0770) -0.9882*** (0.0435) -0.1973*** (0.0467) 0.3833*** (0.0899) -0.0617*** (0.0146) -0.0911*** (0.0216) 0.0474*** (0.0117) 0.0615*** (0.0145) 0.0439*** (0.0104) OPEN ACCESS | Licensed under | https://creativecommons.org/about/cclicenses/ DOI https://doi.org/10.3790/schm.131.1.59 | Generated on 2023-01-16 13:36:15