How national employment systems relate to employee involvement: a decomposition analysis of Germany, the UK and Sweden

How national employment systems relate to employee involvement: a decomposition analysis of... Abstract We investigate general claims that national institutional conditions relate to employee involvement across countries. Using a decomposition analysis, we examine how much of the key domains of national employment systems contribute to differences in employee involvement in Germany, the UK and Sweden. Drawing on the 2015 European Working Conditions Survey (EWCS), our decomposition analysis explains between 40% and 65% of the cross-national differences. These differences stem from key national employment systems domains, namely the management system, information and communication technology use, as well as training and education. We show that these domains contribute simultaneously and with different weights to cross-national differences, and conclude that domains’ contributions reflect the specific institutional characteristics of the investigated national employment systems. 1. Introduction Employee involvement is enjoying renewed interest as a key element of job quality (e.g. Boxall and Winterton, 2015; Marchington, 2015). Since it allows for direct employee integration and influence on decisions about the wider work organization, employee involvement influences working conditions and well-being (e.g. Kalleberg et al., 2009; Gallie, 2013). From a cross-national comparative perspective, several empirical studies have shown that employees’ opportunities for involvement in organizations differ significantly across countries (e.g. Dobbin and Boychuk, 1999; Olsen et al., 2010; Esser and Olsen, 2012; Holman, 2013; Brewster et al., 2014). An explanation of these differences can be found in national institutional conditions, which are assumed to shape organizational practices (e.g. Maurice et al., 1980; Kern and Schumann, 1984; Piore and Sabel, 1985; Sorge and Streeck, 1988; Sorge, 1991; Streeck, 1991; Appelbaum and Batt, 1994). In particular, building on Fligstein and Byrkjeflot, (1996), Dobbin and Boychuk (1999) argue that specific national employment systems govern how work and employment is organized in a given country. However, to date, we do not fully understand how national employment systems relate to organizational practices generally, nor how they relate to employee involvement in particular (Delbridge et al., 2011; Almond and Gonzalez Menendez, 2014; Wood et al., 2014). In comparative studies on organizational practices and employee involvement, Germany, the UK and Sweden represent core countries (e.g. Maurice et al., 1980; Hall and Soskice, 2001; Amable, 2003; Lorenz and Valeyre, 2005; Gallie, 2007; Croucher et al., 2014).1 According to recent empirical studies, opportunities for involvement in the workplace are lower in Germany as compared to the UK and Sweden (Eurofound, 2013, p. 63; similar findings for autonomy by Esser and Olsen (2012), and job quality types by Holman (2013)). This is surprising because, based on international comparative research, we would expect an opposite ranking for Germany and the UK. Several researchers assume that national institutional conditions in Germany foster higher job quality, including higher employee involvement levels (Gallie, 2007; Frege and Godard, 2014). In contrast, the UK is usually considered an example of low employee involvement, stressing the low-road approach taken by UK firms (Danford et al., 2008; Kalleberg, 2009; Kelly, 2013) and the focus on managerial workplace authority (Dobbin and Boychuk, 1999). Accordingly, the considerably lower employee involvement levels in Germany compared to the UK present a puzzle that has not yet been solved in the literature. A possible explanation for this puzzle could be that, in a direct comparison to national employment systems, several institutional conditions vary simultaneously. For instance, countries differ in the ways management shares authority with regular employees, the extent to which employees use new technology, which education levels dominate, as well as the ways in which employee representatives can influence working conditions. Several empirical studies refer to institutional differences in order to explain cross-national differences in employee involvement (e.g. Dobbin and Boychuk, 1999; Olsen et al., 2010; Holman, 2013). However, these studies usually focus on the overall differences between national employment systems and do not disentangle the different domains’ distinct effects and relative weights. Thus, we do not yet know the extents to which single domains contribute to the overall cross-national differences or whether a given domain contributes at all. However, it is necessary to disentangle the effects of national employment systems domains on employee involvement if we are to better understand how national institutional conditions relate to employee involvement opportunities. In turn, this could also help us to understand the puzzling differences between Germany and the UK. We address this research gap by asking how much distinct institutional domains of national employment systems contribute to the overall cross-national differences in employee involvement. To answer this question, we first discuss five key domains of national employment systems—namely management system, information and communication technology (ICT) use, training and education, employee representation and employment conditions—and their relationships to employee involvement. Building on this, we characterize Germany, the UK and Sweden along these domains. Our empirical analysis utilizes employee data from the 2015 European Working Conditions Survey (EWCS) from these three countries. We conduct a decomposition analysis (Jann, 2008) to identify the institutional domains that contribute to cross-national differences in employee involvement. With this analytic strategy, we statistically show—for the first time—how much particular institutional domains contribute to cross-national differences in employee involvement. 2. Theoretical background 2.1 National employment systems The comparative capitalisms (CC) literature (Jackson and Deeg, 2008) provides general arguments for the relationship between national employment systems and organizational practices. Proponents of CC posit that organizational patterns and workplaces differ, since they are embedded in distinct national institutional frameworks (Marsden, 1999; Whitley, 1999, 2003; Hall and Soskice, 2001; Amable, 2003; Delbridge et al., 2011; Frege and Kelly, 2013; Hauptmeier and Vidal, 2014; Morgan and Hauptmeier, 2014). Thus, several authors argue that national institutional frameworks generate a more or less coherent general logic of economic action in a given country (Jackson and Deeg, 2008; Almond and Gonzalez Menendez, 2014). Such logics manifest as ‘national employment systems carry different logics of work control that influence how work is governed in a wide range of settings’ (Dobbin and Boychuk, 1999, p. 262). Thus, national employment systems are characterized by dominant institutional conditions that shape dominant logics of appropriateness (March, 1994). These logics influence actors’ behaviors by establishing identities and matching rules to recognized situations (Fligstein and Byrkjeflot, 1996; Frege and Godard, 2014). Following this literature, we argue that national employment systems enable certain organizational practices while constraining others. Specifically, we assume that the cross-national differences in specific domains account for the empirical differences in organizational practices across countries, which in our case is employee involvement across Germany, the UK and Sweden. In particular, we focus on five key domains of national employment systems, namely management system, ICT use, training and education, employee representation and employment conditions (Fligstein and Byrkjeflot, 1996; Dobbin and Boychuk, 1999; Fligstein, 2001): (a) Management system describes the ways in which control is organized in the workplace (Dobbin and Boychuk, 1999). Management systems differ in the prevalence of managerial positions and the ways by which managers share authority in the workplace with regular employees (Fligstein and Byrkjeflot, 1996; Whitley, 2003). The sharing of responsibilities in the workplace is usually implemented through different forms of job design, which can be described along two basic dimensions: task variability and uncertainty (Appelbaum et al., 2000). Low task variety and uncertainty represent the power of rules and routines. In contrast, high task variety and uncertainty are indicators of employee-oriented management practices. A particular measure to shift managerial control to regular employees via job design is discretionary teamwork (Appelbaum and Batt, 1994; Appelbaum et al., 2000). (b) Another aspect closely related to the management system is ICT use. As Dobbin and Boychuk (1999) noted, employment systems face the challenges of new technologies (similar Fligstein, 2001), and national employment systems differ substantially in their abilities to adopt new technologies in the workplace (Castells, 2000; Hall and Soskice, 2001; Amable, 2003). (c) Training and education refers to the ways employees usually acquire skills as well as the formal education levels employees attain (Gallie, 2007). Key differences are the necessity of continuing training and the foci on either vocational or university education (Fligstein and Byrkjeflot, 1996; Gallie, 2007; Goergen et al., 2012; Goergen et al., 2014). (d) Employee representation describes mechanisms through which employees can collectively influence work and employment conditions. These mechanisms of collective influence are seen as a key difference between national employment systems, because countries differ in the prevalence of employee representation and the power exerted by employee representatives (Fligstein and Byrkjeflot, 1996; Dobbin and Boychuk, 1999; Korpi, 2006; Gallie, 2007). (e) Employment conditions comprise key aspects that regulate employment conditions and unemployment benefits (Dobbin and Boychuk, 1999; Amable, 2003; Gallie, 2007). Lightly regulated employment systems favor market mechanisms, which lead to high labor turnover rates and precarious employment conditions. In contrast, highly regulated employment systems enable longer job tenure and mitigate the precariousness of employment conditions. 2.2 Employee involvement and national employment systems domains Employee involvement—also termed consultative involvement (Felstead et al., 2010), consultative participation (Gallie, 2013), high-involvement management (Wood et al., 2012) or organizational empowerment (Wall et al., 2004)—describes employees’ opportunities to personally influence decisions about the work organization or other aspects of the work environment. Thus, in contrast to job autonomy or task discretion, employee involvement goes beyond the confines of an immediate task. It can range from information, consultation in workplace meetings or more localized briefing groups, suggestion schemes, problem-solving groups, to employee participation in decisions about organizational issues (Wood et al., 2012; Eurofound, 2013; Gallie, 2013). Employee involvement can be influenced by several aspects. We argue that employee involvement relates to national employment systems domains: Differences within these domains should generally relate to differences in employee involvement levels across countries. To substantiate this assumption, we will first show how national employment systems domains relate to employee involvement, since it is only where such a general relationship exists that differences between domains can eventually account for cross-national differences in employee involvement levels. We will then characterize Germany, the UK and Sweden along national employment systems domains. Management system should be relevant, since the prevalence of managerial positions and responsibility-sharing through job design should affect employee involvement. Employees in managerial positions experience more employee involvement. Also, if jobs are usually characterized by low task variety and uncertainty, i.e. if they are governed by rules and routines, there are fewer employee involvement opportunities (Dobbin and Boychuk, 1999; Appelbaum et al., 2000). In contrast, high task variety and high uncertainty should not only foster job autonomy but also employee involvement because, in such a situation, ‘it is seldom practical for managers to have unilateral control over decisions: efficiency requires a more consensus-based approach to decision making’ (Soskice, 1999, p. 115). Teamwork should also increase employee involvement for regular employees (Appelbaum et al., 2000). However, this increase requires one to grant teams de facto discretion (Lawler, 1986, p. 108), and teams are found to differ substantially in their discretion (Pruijt, 2003), for instance, to change tasks, to alter working schedules or to appoint a team leader. ICT use is often perceived as being closely interrelated with management systems and specific job designs that allow for more employee involvement (Appelbaum et al., 2000). Following Castells (2000), ICT use enables new forms of decentralized work organization that in turn allow and require a substantial involvement of employees in work processes and organizational matters. Ample empirical evidence supports this general claim about a positive relationship between ICT use and employee involvement (e.g. Hempell and Zwick, 2008; Green, 2012; Bayo-Moriones et al., 2017).2 Concerning training and education, we assume that employee involvement increases with continuing training opportunities and high education levels, because better-qualified employees should be granted more opportunities for involvement in the organization (Jackson and Schuler, 1995). Again, the underlying idea is that employee knowledge influences the effectiveness of participation and that highly qualified employees should increase performance more than those with little knowledge (Glew et al., 1995). The forms of employee representation should also relate to employee involvement levels. However, the literature provides ambiguous predictions. First, some assume that employee representatives seek to improve employee work and employment conditions, including more employee involvement opportunities (Jackson and Schuler, 1995; Doellgast et al., 2009; Esser and Olsen, 2012). Second, stronger individual employee involvement could challenge collective employee representations, because the former might decrease the power of the latter. Thus, employee representatives might be inclined to limit direct employee involvement (Baron and Kreps, 1999; Hauff et al., 2014). Third, Dobbin and Boychuk (1999) argue that there could be no effect, because practices quickly spread from workplaces with representation to workplaces without representation. A further effect can be expected from general employment conditions. Long-term employment increases employees’ firm-specific experience and mutual trust. This should in turn increase employee involvement (Dobbin and Boychuk, 1999; similarly, Streeck, 1991). Conversely and similarly, precarious employment conditions should negatively affect involvement. To advance our understanding about the relationship between national employment systems domains and employee involvement, we will analyze how these key national employment systems domains empirically relate to employee involvement in the cases of Germany, the UK and Sweden. First, we will briefly characterize our three cases’ national employment systems along these key domains. 2.3 The national employment systems of Germany, the UK and Sweden 2.3.1 The case of Germany Germany is often seen as a role model of highly regulated market economies (Hall and Soskice, 2001). Traditionally, national institutional conditions should foster longer job tenure, which enables employees to build firm-specific skills (Streeck, 1991; Amable, 2003). However, the traditional regulated logic of Germany’s employment system is increasingly shaped by a subnotion of dualism in the labor market (Thelen, 2012). In Germany’s management system, the distribution of workplace authority traditionally focuses on a strong position of highly skilled blue-collar workers (Fligstein and Byrkjeflot, 1996; Whitley, 2003). Skilled workers are granted high discretion, and management partially shares authority with skilled employees in a decentralized organizational model (Streeck, 1991). This partial authority-sharing in a decentralized model is accompanied by a specific job design. An international comparative study (Lorenz and Valeyre, 2005) found that, in Germany, employees work in workplaces with comparably high variability and high uncertainty. ICT usage levels should turn out to be moderate for German employees. The CC literature stresses that German firms favor incremental changes (Hall and Soskice, 2001; Amable, 2003). This decreases the adoption speed of new technologies, such as ICT. However, recent developments—such as the ‘Industrie 4.0’ discourse (Pfeiffer, 2017) and the broader ‘Arbeiten 4.0’ process (Bundesministerium für Arbeit und Soziales, 2015)—promote increased ICT usage as part of an overarching digital agenda in Germany’s employment system. Training and education revolves around the traditionally strong German vocational training system (Amable, 2003; Gallie, 2007). Employers and employees should have incentives to invest in firm-specific skills (Streeck, 1991). However, comparative studies reported comparably low participation rates in continuing vocational training for German employees (Gallie, 2007). This indicates potential limitations to continuing vocational training. In Germany, a strong employee representation system has traditionally been a key pillar of the national employment system (Gallie, 2007). However, more recent developments highlight a substantial declining coverage that limits the scope of representation (Jackson and Deeg, 2012). While some economic sectors still have a strong employee representation system, other sectors exist outside the traditional system (Thelen, 2012). Traditionally, employment conditions in Germany were characterized by high job security (Streeck, 1991). However, recent liberalization of employment regulations have led to a growing dualization (Thelen, 2012; Hassel, 2014). Dualization widens the gap between a secure core worforce and a peripheral workforce with insecure and more precarious employment conditions. 2.3.2 The case of the UK The UK is often considered a key example of a lightly regulated market economy (Hall and Soskice, 2001). In contrast to Germany, market mechanisms shape the relationships between firms and employees. A lack of regulation has led researchers to assume that UK firms follow a low-road approach to workforce management that leads to overall poorer working conditions (Danford et al., 2008; Kalleberg, 2009, p. 12; Kelly, 2013). The UK’s employment system is therefore more market-based and more manager-focused. The UK’s management system is characterized by managers who seize workplace authority and extend authority to skilled workers less often (Dobbin and Boychuk, 1999; Whitley, 2003). This focus on managers in the UK also shapes job design, which exhibits not only high variety levels, but also lower uncertainty levels (Lorenz and Valeyre, 2005). This particular job design, which is often called the lean model, allows for more variety compared to traditional job designs, yet there are fewer learning opportunities. Research results report that the introduction of teamwork in the UK closely followed the lean model (Danford et al., 2008; Kelly, 2013). UK employees’ ICT use should turn out to be high, because the institutional framework supports more radical innovation and should also foster a more extensive implementation of new technologies in the workplace (Hall and Soskice, 2001; Amable, 2003). Concerning training and education, the UK’s vocational training system is comparatively weak (Amable, 2003). A very competitive university system provides highly skilled graduates. The education system emphasizes general skills. Employee representation is also traditionally considered weak in the UK (Amable, 2003; Gallie, 2007). This deprives UK employees of substantial power resources to influence their working conditions. UK employment conditions are shaped by market forces that govern labor market regulations and social security. Thus, higher job insecurity and precarious employment are common employment practices. Recently, further deregulation has increased this tendency (Thelen, 2012). 2.3.3 The case of Sweden For some authors (Hall and Soskice, 2001), Sweden and Germany fall into the same country category of highly regulated market economies. These authors view firms’ high skills and long-term orientation as commonalities between the two countries. However, more recent approaches in the CC literature emphasize dissimilarities (Gallie, 2003, 2007; Thelen, 2012). Compared to Germany, the Swedish employment system provides more support for marginalized employee groups. Thus, the Swedish employment system follows a regulated and more inclusive approach. In the Swedish management system, workplace authority is traditionally shared with regular employees in a decentralized organizational model (Appelbaum and Batt, 1994). Sweden’s workers traditionally enjoyed high involvement levels in semi-autonomous teams. This team approach was a key element of the so-called socio-technical model (Appelbaum and Batt, 1994). The country’s workplace’s development followed this tradition, as high discretionary teamwork prevailed. This allows for a job design with high variability and high uncertainty. Unsurprisingly, international comparisons show that Swedish employees enjoy the highest learning organization rates (Lorenz and Valeyre, 2005). Swedish employees should also exhibit a high ICT usage level, indicating a particular path of technological transformation. Firms receive support for extensive implementation of ICT combined with targeted education policies (Schnyder, 2012; Ornston, 2013). Concerning training and education, Sweden is an example of a combination of a strong vocational system and an extensive higher education system. Thus, the general emphasis is on high skills and education levels (Amable, 2003). In contrast to Germany and the UK, Sweden’s education system is more inclusive and enables lifelong retraining. This also applies to marginal employee groups, who receive support to raise their skill levels (Gallie, 2007; Schnyder, 2012). Employees in Sweden enjoy powerful employee representation in the workplace, backed by the unions’ substantial influence on national policies (Amable, 2003; Korpi, 2006; Gallie, 2007). This provides Swedish employees with substantial power resources to influence their working conditions. Concerning employment conditions, Sweden follows an inclusive approach that supports marginal employee groups (Gallie, 2007; Thelen, 2012). This moderates insecure labor market positions and precarious employment conditions. At the same time, job security is relaxed, so as to foster labor market flexibility. However, through training programs, employees receive support to quickly find new jobs. In sum, we lay out substantial reasons to assume that Germany, the UK and Sweden differ significantly across national employment systems domains (i.e. management system, ICT use, training and education, employee representation and employment conditions). Table 1 summarizes the national employment systems domains, their theorized relationships to employee involvement and the differences between Germany, the UK and Sweden across these domains. Following our argumentation above, there are sound theoretical reasons to expect that all these domains relate to employee involvement. Accordingly, we assume that all five domains contribute to the differences in employee involvement across these countries. We use these insights drawn from the literature as a general framework to organize our empirical analysis. Table 1. Employment systems domains, relationships to employee involvement and cross-national differences Domains of national employment systems Theorized relationship with employee involvement DE UK Sweden (SE) (a) Management system Increases with managerial position, discretionary teamwork as well as with job design variety and uncertainty Vocational employee focus, traditional and learning job design: moderate variety and uncertainty Managerial focus, lean job design: lower variety and lower uncertainty Discretionary teamwork-based, learning job design: high variety and high uncertainty (b) ICT use Increases with ICT use Moderate High High (c) Training and education Increases with higher education levels and training activity Vocational training focus Focus on competitive higher education Focus on higher education, general and inclusive training (d) Employee representation Ambiguous (possibly increases, decreases or unrelated) Medium (potential sectoral focus, dualism) Low High (e) Employment conditions Increases with job security and non-precarious employment High and low (dualism) Medium (traditionally liberal market forces) High (traditionally inclusive) Domains of national employment systems Theorized relationship with employee involvement DE UK Sweden (SE) (a) Management system Increases with managerial position, discretionary teamwork as well as with job design variety and uncertainty Vocational employee focus, traditional and learning job design: moderate variety and uncertainty Managerial focus, lean job design: lower variety and lower uncertainty Discretionary teamwork-based, learning job design: high variety and high uncertainty (b) ICT use Increases with ICT use Moderate High High (c) Training and education Increases with higher education levels and training activity Vocational training focus Focus on competitive higher education Focus on higher education, general and inclusive training (d) Employee representation Ambiguous (possibly increases, decreases or unrelated) Medium (potential sectoral focus, dualism) Low High (e) Employment conditions Increases with job security and non-precarious employment High and low (dualism) Medium (traditionally liberal market forces) High (traditionally inclusive) Source: Own depiction of employment systems domains (adopted by especially drawing on Fligstein and Byrkjeflot, 1996; Dobbin and Boychuk, 1999; Gallie, 2007). Table 1. Employment systems domains, relationships to employee involvement and cross-national differences Domains of national employment systems Theorized relationship with employee involvement DE UK Sweden (SE) (a) Management system Increases with managerial position, discretionary teamwork as well as with job design variety and uncertainty Vocational employee focus, traditional and learning job design: moderate variety and uncertainty Managerial focus, lean job design: lower variety and lower uncertainty Discretionary teamwork-based, learning job design: high variety and high uncertainty (b) ICT use Increases with ICT use Moderate High High (c) Training and education Increases with higher education levels and training activity Vocational training focus Focus on competitive higher education Focus on higher education, general and inclusive training (d) Employee representation Ambiguous (possibly increases, decreases or unrelated) Medium (potential sectoral focus, dualism) Low High (e) Employment conditions Increases with job security and non-precarious employment High and low (dualism) Medium (traditionally liberal market forces) High (traditionally inclusive) Domains of national employment systems Theorized relationship with employee involvement DE UK Sweden (SE) (a) Management system Increases with managerial position, discretionary teamwork as well as with job design variety and uncertainty Vocational employee focus, traditional and learning job design: moderate variety and uncertainty Managerial focus, lean job design: lower variety and lower uncertainty Discretionary teamwork-based, learning job design: high variety and high uncertainty (b) ICT use Increases with ICT use Moderate High High (c) Training and education Increases with higher education levels and training activity Vocational training focus Focus on competitive higher education Focus on higher education, general and inclusive training (d) Employee representation Ambiguous (possibly increases, decreases or unrelated) Medium (potential sectoral focus, dualism) Low High (e) Employment conditions Increases with job security and non-precarious employment High and low (dualism) Medium (traditionally liberal market forces) High (traditionally inclusive) Source: Own depiction of employment systems domains (adopted by especially drawing on Fligstein and Byrkjeflot, 1996; Dobbin and Boychuk, 1999; Gallie, 2007). 3. Data, measures and method 3.1 Data: The EWCS The following analyses are based on data from the EWCS conducted in 2015 (Eurofound, 2016). The EWCS provides comparable data on working conditions across Europe. The target population is people aged 15 and older who were employed at the time of the survey. Data was gathered through face-to-face interviews using a multistage, stratified random sample. Questionnaire development included expert reviews and real-life tests in order to design a reliable and valid instrument that is easy to understand. The interviews were done by interviewers with substantial experience and training (Ipsos, 2015). Thus, the EWCS provides a unique, high-quality data source for cross-national comparative research. For our analysis, we only used data from employees in Germany (DE), the United Kingdom (UK) and Sweden (SE). Further, we included only employees from the manufacturing and service sectors.3 Our final sample comprised 3522 cases (the n for each country is DE = 1564; UK = 1165; SE = 793). 3.2 Dependent variable: the employee involvement factor Employee involvement is our dependent variable. In line with our definition of employee involvement, and following other seminal contributions (Wood et al., 2012; Eurofound, 2013; Gallie, 2013), we used several items to capture different aspects of employment involvement. We computed a factor analysis using the pooled dataset which revealed a single factor.4Table 2 reports the five variables and their factor loadings. The factor’s eigenvalue was 2.42 and the Cronbach’s alpha was 0.81, indicating a viable factor solution. We used the predicted factor scores of the computed factor to generate a new variable for employee involvement. Finally, we normalized the variable values so that it ranged between 0 and 1. Table 2. Variables and their loadings with employee involvement factor Variables Loadings for employee involvement factor Consulted before work targets are set 0.63 Involved in improving the work organization or work processes of the department or organization 0.71 A say in the choice of work partners 0.50 The ability to apply own ideas in work 0.71 Influence on decisions that are important for work 0.87 Variables Loadings for employee involvement factor Consulted before work targets are set 0.63 Involved in improving the work organization or work processes of the department or organization 0.71 A say in the choice of work partners 0.50 The ability to apply own ideas in work 0.71 Influence on decisions that are important for work 0.87 Notes: Answer categories: 1 = never to 5 = always; n = 3458. Source: EWCS, 2015, own calculations. Table 2. Variables and their loadings with employee involvement factor Variables Loadings for employee involvement factor Consulted before work targets are set 0.63 Involved in improving the work organization or work processes of the department or organization 0.71 A say in the choice of work partners 0.50 The ability to apply own ideas in work 0.71 Influence on decisions that are important for work 0.87 Variables Loadings for employee involvement factor Consulted before work targets are set 0.63 Involved in improving the work organization or work processes of the department or organization 0.71 A say in the choice of work partners 0.50 The ability to apply own ideas in work 0.71 Influence on decisions that are important for work 0.87 Notes: Answer categories: 1 = never to 5 = always; n = 3458. Source: EWCS, 2015, own calculations. 3.3 Independent variables Our independent variables correspond to the national employment systems domains introduced above. We subsumed independent variables that account for single employee characteristics into the various national employment systems domains. Management system: Our variables on management system include variables concerning managerial positions and job design. For our analysis, we subdivided the two aspects of management systems into two subsets: For management system positions, a binary variable measures a position of workplace authority, indicating the respondent’s supervisory authority (1 = no, 0 = yes). We also distinguished employees in managerial occupations using the International Standard Classification of Occupations (ISCO).5 For job design in the management system, we refer to the two major characterizations (Appelbaum et al., 2000): task variety and uncertainty. We measured task variety by two items (1 = no, 0 = yes) relating to the question whether a job involves monotony (no task variety) and job complexity. We measured uncertainty by two binary items (1 = no, 0 = yes) related to the question whether a job involves solving unforeseen problems and learning new things. For teamwork, as a specific form of job design, we grouped respondents into three categories: no teamwork, teamwork without influence and teamwork with influence. The latter indicates discretionary teamwork with influence on tasks, on the team leader selection and/or on time. ICT use: One variable accounts for the usage of ICT technology at the workplace, asking respondents how often they work with computers, laptops, and smartphones, on a seven-item scale (1 = never to 7 = all of the time). Training and education: We included variables on training and educational attainment. For training activities, we used two variables to measure whether employees received training paid for by the employer (1 = yes, 0 = no) or on-the-job training (1 = yes, 0 = no). We included education levels, using an international classification (ISCED). We collapsed the original seven-step classification into three categories: low (ISCED 0–2: up to lower secondary education), medium (ISCED 3–4: up to post-secondary non-tertiary education) and high (ISCED 5–6: university degree and beyond). Employee representation: Here, we used a binary variable stating if there is a trade union, works council or a similar committee that represents employees at the respondent’s company or organization (1 = yes, 0 = no). Employment conditions: We included several variables that account for job security and precarious employment conditions. As a first measure, we used the contract type: indefinite contract, fixed-term contract, temporary employment agency contract or other contract. We also included a subjective evaluation of job insecurity and the labor market position as well as a variable on job tenure. Job insecurity was measured as agreement with the statement I may lose my job in the next 6 months. We determined labor market positioning by the statement If I were to lose or quit my current job, it would be easy for me to find a job of similar salary. (The five-item scale for both items ranged from 1 = strongly disagree to 5 = strongly agree.) Job tenure was measured in terms of years with the current employer. Table 3 reports the distribution of all variables across the three countries. Table 3. Distribution of dependent and independent variables by country DE UK SE Min. Max. Employee involvement 0.55 0.67 0.66 0 1 Factor employee involvement (normalized) Management system: positions 10 26 13 0 1 Supervisory function Occupation: managers 1 15 7 0 1 Management system: job design 28 58 22 0 1 Job design: monotonous tasks Job design: complex tasks 62 66 69 0 1 Job design: learning new things 61 83 91 0 1 Job design: solving unforeseen problems on your own 79 84 96 0 1 Teamwork: none 47 22 27 0 1 Teamwork: with influence 38 54 59 0 1 Teamwork: without influence 15 24 14 0 1 ICT usage 3.09 4.34 4.40 1 7 Working with computers, laptops, smartphones Training and education 37 56 46 0 1 Training: employer paid Training: on-the-job 38 58 52 0 1 Education level: low (ISCED 0–2) 8 34 7 0 1 Education level: medium (ISCED 3–4) 78 22 49 0 1 Education level: high (ISCED 5+) 14 44 44 0 1 Employee representation 48 46 84 0 1 Employee representative exists Employment conditions 84 87 85 0 1 Contract: indefinite contract Contract: fixed-term contract 10 4 10 0 1 Contract: temporary employment agency contract 1 3 2 0 1 Contract: other 5 6 3 0 1 Job insecurity 4.15 4.04 4.20 1 5 Easy to find a new job 2.88 3.17 3.31 1 5 Job tenure (years) 9.71 7.79 9.89 0 64 DE UK SE Min. Max. Employee involvement 0.55 0.67 0.66 0 1 Factor employee involvement (normalized) Management system: positions 10 26 13 0 1 Supervisory function Occupation: managers 1 15 7 0 1 Management system: job design 28 58 22 0 1 Job design: monotonous tasks Job design: complex tasks 62 66 69 0 1 Job design: learning new things 61 83 91 0 1 Job design: solving unforeseen problems on your own 79 84 96 0 1 Teamwork: none 47 22 27 0 1 Teamwork: with influence 38 54 59 0 1 Teamwork: without influence 15 24 14 0 1 ICT usage 3.09 4.34 4.40 1 7 Working with computers, laptops, smartphones Training and education 37 56 46 0 1 Training: employer paid Training: on-the-job 38 58 52 0 1 Education level: low (ISCED 0–2) 8 34 7 0 1 Education level: medium (ISCED 3–4) 78 22 49 0 1 Education level: high (ISCED 5+) 14 44 44 0 1 Employee representation 48 46 84 0 1 Employee representative exists Employment conditions 84 87 85 0 1 Contract: indefinite contract Contract: fixed-term contract 10 4 10 0 1 Contract: temporary employment agency contract 1 3 2 0 1 Contract: other 5 6 3 0 1 Job insecurity 4.15 4.04 4.20 1 5 Easy to find a new job 2.88 3.17 3.31 1 5 Job tenure (years) 9.71 7.79 9.89 0 64 Notes: Figures of dummy variables in % of yes answers; for all other variables, means are depicted. Values are represented as percentage, unless otherwise mentioned. Source: EWCS, 2015, own calculations. Table 3. Distribution of dependent and independent variables by country DE UK SE Min. Max. Employee involvement 0.55 0.67 0.66 0 1 Factor employee involvement (normalized) Management system: positions 10 26 13 0 1 Supervisory function Occupation: managers 1 15 7 0 1 Management system: job design 28 58 22 0 1 Job design: monotonous tasks Job design: complex tasks 62 66 69 0 1 Job design: learning new things 61 83 91 0 1 Job design: solving unforeseen problems on your own 79 84 96 0 1 Teamwork: none 47 22 27 0 1 Teamwork: with influence 38 54 59 0 1 Teamwork: without influence 15 24 14 0 1 ICT usage 3.09 4.34 4.40 1 7 Working with computers, laptops, smartphones Training and education 37 56 46 0 1 Training: employer paid Training: on-the-job 38 58 52 0 1 Education level: low (ISCED 0–2) 8 34 7 0 1 Education level: medium (ISCED 3–4) 78 22 49 0 1 Education level: high (ISCED 5+) 14 44 44 0 1 Employee representation 48 46 84 0 1 Employee representative exists Employment conditions 84 87 85 0 1 Contract: indefinite contract Contract: fixed-term contract 10 4 10 0 1 Contract: temporary employment agency contract 1 3 2 0 1 Contract: other 5 6 3 0 1 Job insecurity 4.15 4.04 4.20 1 5 Easy to find a new job 2.88 3.17 3.31 1 5 Job tenure (years) 9.71 7.79 9.89 0 64 DE UK SE Min. Max. Employee involvement 0.55 0.67 0.66 0 1 Factor employee involvement (normalized) Management system: positions 10 26 13 0 1 Supervisory function Occupation: managers 1 15 7 0 1 Management system: job design 28 58 22 0 1 Job design: monotonous tasks Job design: complex tasks 62 66 69 0 1 Job design: learning new things 61 83 91 0 1 Job design: solving unforeseen problems on your own 79 84 96 0 1 Teamwork: none 47 22 27 0 1 Teamwork: with influence 38 54 59 0 1 Teamwork: without influence 15 24 14 0 1 ICT usage 3.09 4.34 4.40 1 7 Working with computers, laptops, smartphones Training and education 37 56 46 0 1 Training: employer paid Training: on-the-job 38 58 52 0 1 Education level: low (ISCED 0–2) 8 34 7 0 1 Education level: medium (ISCED 3–4) 78 22 49 0 1 Education level: high (ISCED 5+) 14 44 44 0 1 Employee representation 48 46 84 0 1 Employee representative exists Employment conditions 84 87 85 0 1 Contract: indefinite contract Contract: fixed-term contract 10 4 10 0 1 Contract: temporary employment agency contract 1 3 2 0 1 Contract: other 5 6 3 0 1 Job insecurity 4.15 4.04 4.20 1 5 Easy to find a new job 2.88 3.17 3.31 1 5 Job tenure (years) 9.71 7.79 9.89 0 64 Notes: Figures of dummy variables in % of yes answers; for all other variables, means are depicted. Values are represented as percentage, unless otherwise mentioned. Source: EWCS, 2015, own calculations. In addition to these variables, we added general controls (see Appendix): We included respondent’s sex, age and weekly working hours, since these basic characteristics might substantially vary between the three countries. We also included industry, coded according to the current classification (NACE, see Eurostat, 2008). A single categorical variable measures establishment size.6 Further, we included the remaining seven non-managerial occupations as control variables: professionals; technicians and associate professionals; clerical support workers; service and sales workers; craft and related trades workers; plant and machine operators and assemblers; and elementary occupations. 3.4 Method: decomposition analysis To investigate cross-national differences, we computed a decomposition analysis (Jann, 2008) of employee involvement. This analysis revealed how much the differences in independent variables contribute to mean differences of the dependent variable (employee involvement) between distinct groups (employees in different countries), that is, our decomposition analysis revealed the characteristics that contribute to the cross-national differences in employee involvement. Our decomposition analysis also showed how much these characteristics contribute to overall cross-national differences. We reported this contribution of combined or individual variables by the estimated coefficients in the tables shown below. This is not possible with common regression analysis. Specifically, a decomposition analysis divides mean differences between two groups into two parts—an explained part and an unexplained part: The explained part accounts for the group differences in the independent variables and thus describes the mean differences accounted for by the model. The unexplained part accounts for the remaining group differences owing to unobserved influences. Thus, the unexplained part comprises all the cross-national differences that the model cannot directly account for using the included independent variables. Thus, we cannot statistically determine what generates the unexplained part. Because a decomposition analysis compares only two groups at a time, we computed two separate decomposition models: Model A (Germany vs. the UK) and Model B (Germany vs. Sweden). In both decomposition models, Germany serves as common reference country. Finally, we took statistical precautions that all results for categorical variables with more than two categories would not depend on chosen reference categories (for details, see Jann, 2008). 4. Results We progressed our decomposition analysis in three steps from general to more specific results: 4.1 Step 1: Overall results of the decomposition models In a first step, we computed the overall results. Table 4 reports the overall results of decomposition models A and B. The decomposition models report the values from the perspective of Swedish or UK employees: positive values in the table indicate that Swedish or UK employees report more employee involvement compared with their German counterparts. The total computed difference in employee involvement between Germany and the UK amounts to 0.12; between Germany and Sweden, it is 0.11. German employees have significantly less employee involvement. Table 4. Overall results of decomposition models of employee involvement Model A Model B DE vs. UK DE vs. SE Estimated coefficients Estimated coefficients Estimated group values Value for employees in Germany (DE) 0.56 0.56 Value for employees in the UK (UK) 0.68 . Value for employees in Sweden (SE) . 0.67 Overall decomposition results Total difference to Germany 0.12*** 0.11*** Explained difference 0.05*** 0.07*** Unexplained difference 0.07*** 0.04*** Explained difference share of total difference (%) ∼42 ∼64 Subsets—only explained part Management system: positions 0.03*** 0.01** Management system: job design 0.02*** 0.05*** ICT use 0.01* 0.02*** Training and education −0.00 0.01* Representation −0.00 −0.00 Employment conditions −0.00 −0.00 Control variables −0.01** −0.01* n 2122 1865 Model A Model B DE vs. UK DE vs. SE Estimated coefficients Estimated coefficients Estimated group values Value for employees in Germany (DE) 0.56 0.56 Value for employees in the UK (UK) 0.68 . Value for employees in Sweden (SE) . 0.67 Overall decomposition results Total difference to Germany 0.12*** 0.11*** Explained difference 0.05*** 0.07*** Unexplained difference 0.07*** 0.04*** Explained difference share of total difference (%) ∼42 ∼64 Subsets—only explained part Management system: positions 0.03*** 0.01** Management system: job design 0.02*** 0.05*** ICT use 0.01* 0.02*** Training and education −0.00 0.01* Representation −0.00 −0.00 Employment conditions −0.00 −0.00 Control variables −0.01** −0.01* n 2122 1865 Notes: Results (estimated coefficients) reported as seen from Swedish or British employees; * P < 0.05, ** P < 0.01, *** P < 0.001. Source: EWCS, 2015, own calculations. Table 4. Overall results of decomposition models of employee involvement Model A Model B DE vs. UK DE vs. SE Estimated coefficients Estimated coefficients Estimated group values Value for employees in Germany (DE) 0.56 0.56 Value for employees in the UK (UK) 0.68 . Value for employees in Sweden (SE) . 0.67 Overall decomposition results Total difference to Germany 0.12*** 0.11*** Explained difference 0.05*** 0.07*** Unexplained difference 0.07*** 0.04*** Explained difference share of total difference (%) ∼42 ∼64 Subsets—only explained part Management system: positions 0.03*** 0.01** Management system: job design 0.02*** 0.05*** ICT use 0.01* 0.02*** Training and education −0.00 0.01* Representation −0.00 −0.00 Employment conditions −0.00 −0.00 Control variables −0.01** −0.01* n 2122 1865 Model A Model B DE vs. UK DE vs. SE Estimated coefficients Estimated coefficients Estimated group values Value for employees in Germany (DE) 0.56 0.56 Value for employees in the UK (UK) 0.68 . Value for employees in Sweden (SE) . 0.67 Overall decomposition results Total difference to Germany 0.12*** 0.11*** Explained difference 0.05*** 0.07*** Unexplained difference 0.07*** 0.04*** Explained difference share of total difference (%) ∼42 ∼64 Subsets—only explained part Management system: positions 0.03*** 0.01** Management system: job design 0.02*** 0.05*** ICT use 0.01* 0.02*** Training and education −0.00 0.01* Representation −0.00 −0.00 Employment conditions −0.00 −0.00 Control variables −0.01** −0.01* n 2122 1865 Notes: Results (estimated coefficients) reported as seen from Swedish or British employees; * P < 0.05, ** P < 0.01, *** P < 0.001. Source: EWCS, 2015, own calculations. A simple juxtaposition of employee involvement means in Table 3 produces the same conclusion. However, going well beyond simple mean differences, the upper rows of Table 4 reveal that the explained difference in model A (for Germany and the UK) accounts for about 42% of the total differences, while model B (for Germany and Sweden) explains around 64% of the differences. These findings have two key implications: First, because the explained parts score around 40% and 60%, the models explain the cross-national differences well. This indicates that the variables in the model capture a substantial share of the differences and underscores our assumptions that the various national employment systems domains are associated with the employee involvement levels across countries. Second, while the models account for some differences with statistical certainty, the model cannot determine a remaining share. We will return to this issue in our discussion. 4.2 Step 2: Aggregated decomposition models and the analysis of subsets In a second step, we investigated the specific contributions of the employment system domains. We computed aggregated decomposition models that combine the various variables of each domain into several subsets. Using the domain subsets, the models revealed the combined differences from all independent variables in a given national employment systems domain. The computed coefficients in the models denote the contribution of the subset variables to the overall difference in employee involvement: a positive coefficient signifies that the cross-national difference in the subset variables increase the difference in employee involvement between Germany and the UK or Sweden. Our results reported in the lower rows of Table 4 revealed statistically significant relationships for several subsets, including both aspects of management system positions and job design as well as ICT use and training and education. The models estimated no statistically significant relationships for the subsets representation and employment conditions. Thus, only some domains statistically fed into the share of the explained difference. Interestingly, the control variables showed significant relationships in both models. This means that specific structural conditions (mostly establishment size) are associated with slightly higher employee involvement in Germany. However, the much larger contributions by the other domain subsets fully consumed this small counteracting effect. Figure 1 depicts the results from the aggregated decomposition models and illustrates the emerging—nuanced—picture.7 This figure shows only domain subsets that statistically significantly contributed to the cross-national differences alongside the unexplained part. For an interpretation, we need to combine two questions: First, does the subset relate to the cross-national differences in employment involvement? Second, how do levels of subset variables differ across Germany to the UK or Sweden? With the descriptive statistics from Table 3, we can now interpret the displayed differences: Figure 1. View largeDownload slide Aggregated decomposition model results: subsets with significant relationships and unexplained part. Domain variables combined in subsets; figure displays only results for four subsets with significant contributions as well as unexplained part displayed; not displayed subsets: representation and employment conditions. Source: EWCS, 2015, own calculations and depiction of decomposition results. Figure 1. View largeDownload slide Aggregated decomposition model results: subsets with significant relationships and unexplained part. Domain variables combined in subsets; figure displays only results for four subsets with significant contributions as well as unexplained part displayed; not displayed subsets: representation and employment conditions. Source: EWCS, 2015, own calculations and depiction of decomposition results. For Germany and the UK (model A), our results showed a substantial association with managerial positions in the management system (0.03). In addition, aspects of job design in the management systems (0.02) and ICT use (0.01) are associated with the higher employee involvement level in the UK. As reported above, this comes with a remaining large unexplained difference (0.07) owing to unknown factors. Thus, overall higher employee involvement levels in the UK relate to a substantially larger share of managerial positions, moderately more discretionary job design and higher ICT use in the workplace. For Germany and Sweden (model B), the results also revealed the importance of the management system. In contrast, here we see a comparably small association with positions (0.01) alongside a more substantial association with job design (0.05). The model reports slightly higher differences owing to ICT use (0.02) and an additional yet relatively small association with training and education (0.01). Thus, overall higher employee involvement levels in Sweden relate to slightly more managerial positions, substantially more discretionary job design, higher ICT use in the workplace and higher training and education levels. 4.3 Step 3: Focused analysis of management system variables In a third step, we highlighted selected variables that are particularly responsible for the cross-national differences in employee involvement. Table 5 reports only the detailed properties of the management system variables, because they contribute substantially to the cross-national differences in employee involvement in both models. Here, the computed coefficients in the models denote the contribution of the specific variables to the overall difference in employee involvement: a positive coefficient signifies that the cross-national difference in the variable increases the difference in employee involvement between Germany and the UK or Sweden. Again, we used the descriptive statistics from Table 3 to interpret the computed differences. Table 5. Detailed decomposition models of employee involvement—explained difference focusing on the management system variables Model A Model B DE vs. UK DE vs. SE Estimated coefficients for explained difference Estimated coefficients for explained difference Overall 0.05*** 0.07*** Management system: positions 0.01*** 0.00** Supervisory function Occupation: managers 0.02*** 0.00* Management system: job design −0.01*** 0.00** Job design: Monotonous tasks Job design: Complex tasks 0.00* 0.00* Job design: Learning new things 0.01*** 0.02*** Job design: Solving unforeseen problems on your own 0.00* 0.01*** Teamwork: none 0.01*** 0.01*** Teamwork: with influence 0.01*** 0.01*** Teamwork: without influence 0.00 −0.00 n 2122 1865 Model A Model B DE vs. UK DE vs. SE Estimated coefficients for explained difference Estimated coefficients for explained difference Overall 0.05*** 0.07*** Management system: positions 0.01*** 0.00** Supervisory function Occupation: managers 0.02*** 0.00* Management system: job design −0.01*** 0.00** Job design: Monotonous tasks Job design: Complex tasks 0.00* 0.00* Job design: Learning new things 0.01*** 0.02*** Job design: Solving unforeseen problems on your own 0.00* 0.01*** Teamwork: none 0.01*** 0.01*** Teamwork: with influence 0.01*** 0.01*** Teamwork: without influence 0.00 −0.00 n 2122 1865 Notes: Results (estimated coefficients) reported as seen from Swedish or British employees: * P < 0.05, ** P < 0.01, *** P < 0.001; control variables are not displayed; minor deviations to overall subset values owing to rounding. Source: EWCS, 2010, own calculations. Table 5. Detailed decomposition models of employee involvement—explained difference focusing on the management system variables Model A Model B DE vs. UK DE vs. SE Estimated coefficients for explained difference Estimated coefficients for explained difference Overall 0.05*** 0.07*** Management system: positions 0.01*** 0.00** Supervisory function Occupation: managers 0.02*** 0.00* Management system: job design −0.01*** 0.00** Job design: Monotonous tasks Job design: Complex tasks 0.00* 0.00* Job design: Learning new things 0.01*** 0.02*** Job design: Solving unforeseen problems on your own 0.00* 0.01*** Teamwork: none 0.01*** 0.01*** Teamwork: with influence 0.01*** 0.01*** Teamwork: without influence 0.00 −0.00 n 2122 1865 Model A Model B DE vs. UK DE vs. SE Estimated coefficients for explained difference Estimated coefficients for explained difference Overall 0.05*** 0.07*** Management system: positions 0.01*** 0.00** Supervisory function Occupation: managers 0.02*** 0.00* Management system: job design −0.01*** 0.00** Job design: Monotonous tasks Job design: Complex tasks 0.00* 0.00* Job design: Learning new things 0.01*** 0.02*** Job design: Solving unforeseen problems on your own 0.00* 0.01*** Teamwork: none 0.01*** 0.01*** Teamwork: with influence 0.01*** 0.01*** Teamwork: without influence 0.00 −0.00 n 2122 1865 Notes: Results (estimated coefficients) reported as seen from Swedish or British employees: * P < 0.05, ** P < 0.01, *** P < 0.001; control variables are not displayed; minor deviations to overall subset values owing to rounding. Source: EWCS, 2010, own calculations. Overall, the detailed decomposition model results in Table 5 revealed similar relationships as well as country-specific patterns: major effects in the Germany–UK comparison stem from managerial positions and supervisory functions more often held by UK employees. Also, more favorable job design (learning and teamwork) add to the cross-national differences. However, the UK model also revealed specific effects of more monotonous work that reflect particular characteristics of the UK employment system. Here, the higher monotonous work levels in the UK partially mitigate the difference to German employees. Overall, this small counteracting effect of monotony is easily absorbed by the remaining positive influences of positions and job design. Comparing Germany and Sweden revealed similar relationships alongside specific patterns. Here, better job design (especially less monotonous tasks, more learning opportunities and more problem-solving opportunities) as well as higher education levels increase employee involvement for Swedish employees. However, similar to the UK but to a much lesser extent, the slightly higher share of managerial positions in Sweden also increases the overall opportunities for employee involvement. Returning to Table 1, which summarized our considerations, we can now conclude that our models support some but not all of our theorized relationships. Our analyses revealed substantial statistical contributions by three of the five investigated domains. This includes management system, ICT use and training and education. These domains contribute to the cross-national differences in employee involvement, since the characteristics along these domains vary substantially among employees from Germany, the UK and Sweden. 5. Discussion We started with the puzzle that opportunities for involvement in the workplace are lower in Germany compared to the UK and Sweden (Eurofound, 2013). Building on previous approaches (Fligstein and Byrkjeflot, 1996; Dobbin and Boychuk, 1999), we assumed that institutional domains of national employment systems (management system, ICT use, training and education, employee representation and employment conditions) relate to employee involvement and therefore could help explain the different employee involvement levels across countries. To analyze how much each of the different domains contributes to cross-national differences in employee involvement, we performed a decomposition analysis using employee data from Germany, the UK and Sweden. In line with previous findings (Eurofound, 2013), our more recent results from 2015 also showed that employee involvement levels in Sweden and the UK exceeded those in Germany. Following the research, we would expect higher employee involvement levels in Germany owing to more favorable national institutional conditions. In contrast to these perspectives, our empirical findings point to a German employee involvement gap. Going beyond existing studies, our decomposition analysis disentangles the effects behind these puzzling differences and helps us to understand why employee involvement in Germany is significantly lower than in the UK. In particular, our empirical investigation shows the extent to which differences in the key national employment systems domains contribute to the cross-national differences. We can now characterize the differences related to the five key national employment systems domains: Management systems contribute substantially to the cross-national differences in employee involvement. We found general relationships alongside distinct national patterns. Compared to Germany, differences to the UK emerged owing to the increased prevalence of managerial positions, reflecting the strong management focus in the UK employment system (Fligstein and Byrkjeflot, 1996; Dobbin and Boychuk, 1999; Whitley, 2003). A higher share of managerial positions in the UK increases the involvement level. If we were to disregard the difference owing to managerial positions, the overall difference in employee involvement to German employees would shrink considerably. We also found that job design contributes to the differences between Germany and the UK. More learning opportunities and more teamwork in the UK increase employee involvement there. While this finding conflicts with the general assumption that the UK embarked on a low-road model, this result is in line with recent findings by Frege and Godard (2014), who argue that firms in liberal countries improve job design elements to compensate for trust problems in weak institutional environments. However, more monotonous work in the UK partially counteracts this general pattern, which—in turn—partially supports conflicting assumptions about the low road and managerial focus in the UK. Overall, the UK results show a combination of a strong management focus combined with an effectively compensating job design. In contrast, in the comparison between Germany and Sweden, we found that good job design accounts for cross-national differences, reflecting the long tradition of discretionary job designs in Swedish workplaces (Appelbaum and Batt, 1994). However, also in Sweden, a slightly higher share of managerial positions increases employee involvement levels. Overall, our results show the substantial importance of management system for the difference between Germany and the UK. This underlines the key role of managerial positions in understanding employee involvement. To some extent, this also applies to the differences between Germany and Sweden. Based on our findings, we highlight that studies of employees should include not only regular employees. Empirically, managers count as employees too, since they also form part of the sample population in general employee surveys. Thus, scholars of employee involvement should account for employees in managerial positions in order to correctly address differences between countries. Generally, in our view, theoretical and empirical approaches could advance if they would more explicitly address the importance of managerial positions for job quality. Turning to ICT use, our results point to an overall positive relationship between ICT use and employee involvement. The lower ICT usage levels of German employees contribute to the German involvement gap. This finding again underscores the incremental dynamics of Germany’s economy in the face of technological shifts. In light of these findings, recent developments that promote ICT use in Germany’s economy—such as the Industrie 4.0 discourse (Pfeiffer, 2017) and the broader Arbeiten 4.0 process (Bundesministerium für Arbeit und Soziales, 2015)—might enable a catch-up process that could increase opportunities for more employee involvement in Germany. Taken together, our findings regarding ICT use underline the importance of the use of new technologies as a national employment systems domain. This importance will grow further as the digital economy expands. Concerning training and education, our results show that differences in education levels also contribute to the German employee involvement gap, compared to Sweden. This underlines the particular characteristics of Sweden’s employment system, reflecting its general emphasis on high education levels (Amable, 2003). Concerning employee representation and employment conditions, we found no conclusive statistical evidence that they relate to increased or decreased employee involvement. Thus, the finding on employee representation tends to support positions in the literature that assume no direct effect at the firm level (Dobbin and Boychuk, 1999). However, our analysis only relies on one item for firm-level employee representation. Because this item is limited, we cannot rule out that there might be a relationship between employee representation and employee involvement levels across countries. This would require more elaborate measures (see limitations, below). While both aspects might have shaped employee involvement in past processes, our findings provide no evidence for a current statistical relationship. Overall, our decomposition models explain the differences in national employee involvement levels well. The differences in the key national employment systems domains account for 40–65% of the cross-national differences in employee involvement. Our results point out that different national employment systems domains (i.e. management system, ICT use and training and education) contribute simultaneously to the cross-national differences. Accordingly, a comprehensive account of the cross-national differences in employee involvement requires several domains at the same time. Building on these differentiated results, we interpret the different domains’ contributions as a result of the diverse characteristics that underlie the national employment systems of Germany, the UK and Sweden. Thus, our results indicate that the national institutional conditions and the related organizational practices in Germany prove less favorable for employee involvement than researchers often believed them to be. Further, our results contribute to the literature, since they inform theoretical concepts (and future empirical studies) that specific domains contribute with different weights to cross-national differences in organizational practices. Thus, depending on the issue in question, some institutional domains may prove more influential than others. This is highly relevant for the study and explanation of cross-national differences in organizational practices. For instance, in a case of institutional transformation of a domain, organizational practices may prove stable if the domain only marginally contributes to the differences. Also, transformations in one domain could compensate for changes in another domain: for instance, an increase in ICT use could be counteracted by a decrease in job design. Thus, the de facto influence of institutional transformations on organizational practices depends on the relative weight exerted by the involved domains. Finally, our decomposition models leave some differences unexplained. Here, unexplained denotes that differences remain that cannot be determined by the variables in the decomposition model. This especially pertains to large parts of the differences between Germany and the UK. Thus, researchers should treat absolute mean differences between countries with caution. While there might be good theoretical foundations that help to interpret cross-national differences, researchers should empirically decompose relationships and should statistically determine underlying patterns, if empirically possible. This would allow for more substantiated claims. Methodological sources inflating the unexplained part could simply derive from measurement errors (nationally specific understandings, response patterns, questionnaire designs). Theoretical sources feeding the unexplained part might stem from overlooked additional theoretical approaches (e.g. cultural differences). Future research should propose additional and complementary theoretical explanations to resolve the remaining cross-national differences. Our findings and interpretations should be seen in light of several limitations: First, since we analyzed cross-sectional data, our statistical analysis does not present evidence for a causal argument. However, our findings shed light on the current associations that underlie cross-national differences in employee involvement. In our view, the empirical investigation of the current associations takes an important step in the empirical foundations of cross-national comparative research. Nonetheless, we maintain that future research should undertake the possible empirical steps in order to advance our understanding of cross-national differences based on longitudinal data. Second, there might be answering patterns across countries. This concerns, for instance, information on occupation, which might be biased by national standards and interpretations (e.g. in the UK, the title manager is used differently to Germany and Sweden). However, the EWCS implements the common international standard classifications (ISCO), with considerable efforts to ensure comparability across countries (Ipsos, 2015). Other major studies (e.g. by Eurostat) rely on the same international standards. Third, the EWCS only includes a general question about an on-site workplace representative. This limits our model’s capability to reveal possible effects of employee representation. Here, more elaborate measures might provide additional empirical insights. Besides limitations, there are several avenues for future research: We limited our analysis to Germany, the UK and Sweden, which represent core countries in comparative studies. Future studies should also include other countries. This could deepen our understandings of how the key national employment systems domains shape organizational practices. By going beyond employee involvement, future research should apply decomposition analysis to other dependent variables. Such extensions could more clearly show whether or not key national employment systems domains also relate to cross-national differences in other organizational practices. In conclusion, our empirical analysis reveals how national institutional conditions relate to differences in employee involvement in Germany, UK and Sweden. We statistically show, for the first time, that different institutional domains of national employment systems (i.e. management system, ICT use and training and education) contribute simultaneously and with different weights to cross-national differences in employee involvement. Our study also contributes to the cross-national comparative research, since it integrates long-held assumptions with current empirical data, employing innovative statistical techniques. Thus, our analysis underscores the high relevance of cross-national comparative research to understand differences in employee involvement, in particular as well as differences in organizational practices, in general. Footnotes 1 The UK is often considered the same country category as the USA. So, it is usually assumed that characteristics of the USA mostly also apply to the UK. While we acknowledge that there are differences between these two countries, we follow the general custom to relate insights for the UK or the USA to the general country category. 2 Some empirical studies (e.g. Bayo-Moriones et al., 2017) highlight specific conditions under which ICT use may affect employee involvement differently. However, in this article, we limit our argument to the question whether or not there is an average positive relationship between ICT use and employee involvement. 3 Owing to low frequencies, we also excluded the remaining 18 respondents from skilled agricultural, forestry and fish workers occupations. 4 We computed a series of alternative factor models (e.g. models for each country independently). Across the different specifications, the models all show a high consistency with the pooled model. 5 To obtain information regarding occupations, EWCS respondents answered two open-ended questions about their job title and their main activities at the workplace. This ensures that respondents do not falsely assign themselves to particular occupations. Also, the EWCS employed a centralized coding interface and multiple coders per country to make the coding process as uniform as possible across countries. In addition, training sessions were held for all coders and materials, with detailed coding instructions provided (Ipsos, 2015). 6 We included industry and company size to control for possible effects that derive from countries’ industry specializations. The literature suggests that Germany, the UK and Sweden differ in their industries’ compositions, and dominant company sizes (Dobbin and Boychuk, 1999; Amable, 2003; Kelly, 2013). 7 Note that in both models, the control variable subset has a significant mitigating effect of around −0.01, accounting for the minor discrepancy between depicted differences in Figure 1 and the differences gap in Table 4; for further details, also see the Appendix. References Almond P. , Gonzalez Menendez M. C. ( 2014 ) ‘Cross-National Comparative Human Resource Management and the Ideational Sphere: A Critical Review’ , The International Journal of Human Resource Management , 25 , 2591 – 2607 . Google Scholar CrossRef Search ADS Amable B. ( 2003 ) The Diversity of Modern Capitalism , Oxford , Oxford University Press . Google Scholar CrossRef Search ADS Appelbaum E. , Bailey T. , Berg P. , Kalleberg A. L. ( 2000 ) Manufacturing Advantage: Why High-Performance Work Systems Pay Off , Ithaca, NY , Cornell University Press . Appelbaum E. , Batt R. ( 1994 ) The New American Workplace. Transforming Work Systems in the United States , Ithaca, NY , ILR Press . Baron J. N. , Kreps D. M. ( 1999 ) Strategic Human Resources: Frameworks for General Managers , New York , Wiley . Bayo-Moriones A. , Billon M. , Lera-López F. ( 2017 ) ‘ Are New Work Practices Applied Together with Ict and Amt?’, The International Journal of Human Resource Management , 28 , 553 – 580 . Google Scholar CrossRef Search ADS Boxall P. , Winterton J. ( 2015 ) ‘Which Conditions Foster High-Involvement Work Processes? A Synthesis of the Literature and Agenda for Research’, Economic and Industrial Democracy, 1 – 21 . doi: 10.1177/0143831X15599584. Brewster C. , Brookes M. , Johnson P. , Wood G. ( 2014 ) ‘ Direct Involvement, Partnership and Setting: A Study in Bounded Diversity’, The International Journal of Human Resource Management , 25 , 795 – 809 . Google Scholar CrossRef Search ADS Bundesministerium für Arbeit und Soziales ( 2015 ) Grünbuch Arbeiten 4.0, Berlin, Bundesministerium für Arbeit und Soziales Abteilung Grundsatzfragen des Sozialstaats, der Arbeitswelt und der sozialen Marktwirtschaft. Castells M. ( 2000 ) The Rise of the Network Society: The Information Age: Economy, Society, and Culture , Malden, MA , Blackwell . Croucher R. , Brookes M. , Wood G. , Brewster C. ( 2014 ) ‘Is There Convergence Towards Individual Voice in Europe?’. In Wood G. , Brewster C. , Brookes M. (eds) Human Resource Management and the Institutional Perspective , New York, London , Routledge , pp. 59 – 77 . Danford A. , Richardson M. , Stewart P. , Tailby S. , Upchurch M. ( 2008 ) ‘ Partnership, High Performance Work Systems and Quality of Working Life’, New Technology, Work and Employment , 23 , 151 – 166 . Google Scholar CrossRef Search ADS Delbridge R. , Hauptmeier M. , Sengupta S. ( 2011 ) ‘ Beyond the Enterprise: Broadening the Horizons of International HRM’, Human Relations , 64 , 483 – 505 . Google Scholar CrossRef Search ADS Dobbin F. , Boychuk T. ( 1999 ) ‘ National Employment Systems and Job Autonomy: Why Job Autonomy Is High in the Nordic Countries and Low in the United States, Canada, and Australia’, Organization Studies , 20 , 257 – 291 . Google Scholar CrossRef Search ADS Doellgast V. , Batt R. , Sørensen O. H. ( 2009 ) ‘ Introduction: Institutional Change and Labour Market Segmentation in European Call Centres’, European Journal of Industrial Relations , 15 , 349 – 371 . Google Scholar CrossRef Search ADS Esser I. , Olsen K. M. ( 2012 ) ‘ Perceived Job Quality: Autonomy and Job Security within a Multi-Level Framework’, European Sociological Review , 28 , 443 – 454 . Google Scholar CrossRef Search ADS Eurofound ( 2013 ) Work Organisation and Employee Involvement in Europe (Authors: Duncan Gallie and Ying Zhou) , Luxembourg , Publications Office of the European Union . Eurofound ( 2016 ) Sixth European Working Conditions Survey – Overview Report , Luxembourg , Publications Office of the European Union . Eurostat ( 2008 ) Nace Rev. 2. Statistical Classification of Economic Activities in the European Community , Luxembourg , Office for Official Publications of the European Communities . Felstead A. , Gallie D. , Green F. , Zhou Y. ( 2010 ) ‘ Employee Involvement, the Quality of Training and the Learning Environment: An Individual Level Analysis’, The International Journal of Human Resource Management , 21 , 1667 – 1688 . Google Scholar CrossRef Search ADS Fligstein N. ( 2001 ) The Architecture of Markets. An Economic Sociology of Twenty-First-Century Capitalist Societies, Princeton, Princeton University Press. Fligstein N. , Byrkjeflot H. ( 1996 ) ‘The Logic of Employment Systems’. In Baron J. N. , Grusky D. B. , Treiman D. J. (eds) Social Differentiation and Social Inequality , Boulder, CO , Westview Press , pp. 11 – 35 . Frege C. , Godard J. ( 2014 ) ‘ Varieties of Capitalism and Job Quality: The Attainment of Civic Principles at Work in the United States and Germany’ , American Sociological Review , 79 , 942 – 965 . Google Scholar CrossRef Search ADS Frege C. , Kelly J. ( 2013 ) ‘Theoretical Perspectives on Comparative Employment Relations’. In Frege C. , Kelly J. (eds) Comparative Employment Relations in the Global Economy , London , Routledge , pp. 8 – 26 . Gallie D. ( 2003 ) ‘ The Quality of Working Life: Is Scandinavia Different?’, European Sociological Review , 19 , 61 – 79 . Google Scholar CrossRef Search ADS Gallie D. ( 2007 ) ‘Production Regimes, Employment Regimes, and the Quality of Work’. In Gallie D. (ed.) Employment Regimes and the Quality of Work , Oxford , Oxford University Press , pp. 1 – 33 . Google Scholar CrossRef Search ADS Gallie D. ( 2013 ) ‘ Direct Participation and the Quality of Work’, Human Relations , 66 , 453 – 473 . Google Scholar CrossRef Search ADS Glew D. J. , O’Leary-Kelly A. M. , Griffin R. W. , Van Fleet D. D. ( 1995 ) ‘ Participation in Organizations: A Preview of the Issues and Proposed Framework for Future Analysis’, Journal of Management , 21 , 395 – 421 . Google Scholar CrossRef Search ADS Goergen M. , Brewster C. , Wood G. , Wilkinson A. ( 2012 ) ‘ Varieties of Capitalism and Investments in Human Capital’, Industrial Relations: A Journal of Economy and Society , 51 , 501 – 527 . Google Scholar CrossRef Search ADS Goergen M. , Wood G. , Brewster C. , Brookes M. ( 2014 ) ‘Corporate Governance Systems and Investments in Human Capital’. In Wood G. , Brewster C. , Brookes M. (eds) Human Resource Management and the Institutional Perspective , New York, London , Routledge , pp. 78 – 107 . Green F. ( 2012 ) ‘ Employee Involvement, Technology and Evolution in Job Skills: A Task-Based Analysis’, ILR Review , 65 , 36 – 67 . Google Scholar CrossRef Search ADS Hall P. A. , Soskice D. ( 2001 ) ‘An Introduction to Varieties of Capitalism’. In Hall P. , Soskice D. (eds) Varieties of Capitalism: The Institutional Foundations of Comparative Advantage , Oxford , Oxford University Press , pp. 1 – 68 . Google Scholar CrossRef Search ADS Hassel A. ( 2014 ) ‘ The Paradox of Liberalization — Understanding Dualism and the Recovery of the German Political Economy’, British Journal of Industrial Relations , 52 , 57 – 71 . Google Scholar CrossRef Search ADS Hauff S. , Alewell D. , Hansen N. K. ( 2014 ) ‘ HRM Systems between Control and Commitment: Occurrence, Characteristics and Effects on HRM Outcomes and Firm Performance’, Human Resource Management Journal , 24 , 424 – 441 . Google Scholar CrossRef Search ADS Hauptmeier M. , Vidal M. (eds) ( 2014 ) Comparative Political Economy of Work , Houndmills , Palgrave Macmillan . Google Scholar CrossRef Search ADS Hempell T. , Zwick T. ( 2008 ) ‘New Technology, Work Organization and Innovation’ , Economics of Innovation and New Technology , 17 , 331 – 354 . Google Scholar CrossRef Search ADS Holman D. ( 2013 ) ‘ Job Types and Job Quality in Europe’, Human Relations , 66 , 475 – 502 . Google Scholar CrossRef Search ADS Ipsos ( 2015 ) 6th European Working Conditions Survey: Technical Report , Dublin , European Foundation for the Improvement of Living and Working Conditions . Jackson G. , Deeg R. ( 2008 ) ‘ From Comparing Capitalisms to the Politics of Institutional Change’, Review of International Political Economy , 15 , 680 – 709 . Google Scholar CrossRef Search ADS Jackson G. , Deeg R. ( 2012 ) ‘ The Long-Term Trajectories of Institutional Change in European Capitalism’, Journal of European Public Policy , 19 , 1109 – 1125 . Google Scholar CrossRef Search ADS Jackson S. E. , Schuler R. S. ( 1995 ) ‘ Understanding Human Resource Management in the Context of Organizations and Their Environments’, Annual Review of Psychology , 46 , 237 – 264 . Google Scholar CrossRef Search ADS PubMed Jann B. ( 2008 ) ‘The Blinder-Oaxaca Decomposition for Linear Regression Models’ , Stata Journal , 8 , 453 – 479 . Kalleberg A. L. ( 2009 ) ‘ Precarious Work, Insecure Workers: Employment Relations in Transition’, American Sociological Review , 74 , 1 – 22 . Google Scholar CrossRef Search ADS Kalleberg A. L. , Nesheim T. , Olsen K. M. ( 2009 ) ‘ Is Participation Good or Bad for Workers?: Effects of Autonomy, Consultation and Teamwork on Stress among Workers in Norway’, Acta Sociologica , 52 , 99 – 116 . Google Scholar CrossRef Search ADS Kelly J. ( 2013 ) ‘The United Kingdom’. In Frege C. , Kelly J. (eds) Comparative Employment Relations in the Global Economy , London , Routledge , pp. 170 – 186 . Kern H. , Schumann M. ( 1984 ) Das Ende Der Arbeitsteilung?: Rationalisierung in Der Industriellen Produktion: Bestandsaufnahme, Trendbestimmung , München , Beck . Korpi W. ( 2006 ) ‘ Power Resources and Employer-Centered Approaches in Explanations of Welfare States and Varieties of Capitalism: Protagonists, Consenters, and Antagonists’, World Politics , 58 , 167 – 206 . Google Scholar CrossRef Search ADS Lawler E. E. ( 1986 ) High-Involvement Management: Participative Strategies for Improving Organizational Performance , San Francisco, CA , Jossey-Bass . Lorenz E. , Valeyre A. ( 2005 ) ‘Organisational Innovation, Human Resource Management and Labour Market Structure: A Comparison of the Eu-15’ , Journal of Industrial Relations , 47 , 424 – 442 . Google Scholar CrossRef Search ADS March J. G. ( 1994 ) Primer on Decision Making: How Decisions Happen , New York , Free Press . Marchington M. ( 2015 ) ‘ Analysing the Forces Shaping Employee Involvement and Participation (Eip) at Organisation Level in Liberal Market Economies (LMEs)’, Human Resource Management Journal , 25 , 1 – 18 . Google Scholar CrossRef Search ADS Marsden D. ( 1999 ) A Theory of Employment Systems: Micro-Foundations of Societal Diversity: Micro-Foundations of Societal Diversity , Oxford, Oxford University Press . Google Scholar CrossRef Search ADS Maurice M. , Sorge A. , Warner M. ( 1980 ) ‘ Societal Differences in Organizing Manufacturing Units: A Comparison of France, West Germany, and Great Britain’, Organization Studies , 1 , 59 – 86 . Google Scholar CrossRef Search ADS Morgan G. , Hauptmeier M. ( 2014 ) ‘Varieties of Institutional Theory in Comparative Employment Relations’. In Wilkinson A. , Wood G. , Deeg R. (eds) Oxford Handbook of Employment Relations - Comparative Employment Systems , Oxford, Oxford University Press , pp. 190 – 214 . Olsen K. M. , Kalleberg A. L. , Nesheim T. ( 2010 ) ‘Perceived Job Quality in the United States, Great Britain, Norway and West Germany, 1989-2005’ , European Journal of Industrial Relations , 16 , 221 – 240 . Google Scholar CrossRef Search ADS Ornston D. ( 2013 ) ‘ Creative Corporatism: The Politics of High-Technology Competition in Nordic Europe’, Comparative Political Studies , 46 , 702 – 729 . Google Scholar CrossRef Search ADS Pfeiffer S. ( 2017 ) ‘The Vision of “Industrie 4.0” in the Making—A Case of Future Told, Tamed, and Traded’ , NanoEthics , 11 , 107 – 121 . Google Scholar CrossRef Search ADS PubMed Piore M. J. , Sabel C. F. ( 1985 ) Das Ende Der Massenproduktion: Studie Über Die Requalifizierung Der Arbeit Und Die Rückkehr Der Ökonomie in Die Gesellschaft , Berlin , Wagenbach . Pruijt H. ( 2003 ) ‘ Teams between Neo-Taylorism and Anti-Taylorism’, Economic and Industrial Democracy , 24 , 77 – 101 . Google Scholar CrossRef Search ADS Schnyder G. ( 2012 ) ‘ Like a Phoenix from the Ashes? Reassessing the Transformation of the Swedish Political Economy Since the 1970s’, Journal of European Public Policy , 19 , 1126 – 1145 . Google Scholar CrossRef Search ADS Sorge A. ( 1991 ) ‘ Strategic Fit and the Societal Effect: Interpreting Cross-National Comparisons of Technology, Organization and Human Resources’, Organization Studies , 12 , 161 – 190 . Google Scholar CrossRef Search ADS Sorge A. , Streeck W. ( 1988 ) ‘New Technology and Industrial Relations: The Case for an Extended Perspective’. In Hyman R. , Streeck W. (eds) New Technology and Industrial Relations , Oxford , Basil Blackwell , pp. 19 – 47 . Soskice D. ( 1999 ) ‘Divergent Production Regimes: Coordinated and Uncoordinated Market Economies in the 1980s and 1990s’. In Kitschelt H. , Lange P. , Marks G. , Stephens J. D. (eds) Continuity and Change in Contemporary Capitalism , New York , Cambridge University Press , pp. 101 – 134 . Google Scholar CrossRef Search ADS Streeck W. ( 1991 ) ‘On the Institutional Conditions of Diversified Quality Production’. In Matzner E. , Streeck W. (eds) Beyond Keynesianism. The Socio-Economics of Production and Full Employment , Aldershot , Edward Elgar , pp. 21 – 61 . Thelen K. ( 2012 ) ‘ Varieties of Capitalism: Trajectories of Liberalization and the New Politics of Social Solidarity’, Annual Review of Political Science , 15 , 137 – 159 . Google Scholar CrossRef Search ADS Wall T. D. , Wood S. J. , Leach D. J. ( 2004 ) ‘Empowerment and Performance’. In Cooper C. L. , Robertson I. (eds) International Review of Industrial and Organizational Psychology , Chichester , John Wiley , pp. 1 – 46 . Whitley R. ( 1999 ) Divergent Capitalisms: The Social Structuring and Change of Business Systems , Oxford , Oxford University Press . Whitley R. ( 2003 ) ‘ The Institutional Structuring of Organizational Capabilities: The Role of Authority Sharing and Organizational Careers’, Organization Studies , 24 , 667 – 695 . Google Scholar CrossRef Search ADS Wood G. , Brewster C. , Brookes M. ( 2014 ) ‘Institutions and Firm Level Hrm Practice’. In Wood G. , Brewster C. , Brookes M. (eds) Human Resource Management and the Institutional Perspective , New York, London , Routledge , pp. 1 – 14 . Wood S. , van Veldhoven M. , Croon M. , de Menezes L. M. ( 2012 ) ‘ Enriched Job Design, High Involvement Management and Organizational Performance: The Mediating Roles of Job Satisfaction and Well-Being’ , Human Relations , 65 , 419 – 445 . Google Scholar CrossRef Search ADS Appendix Table A1. Distribution of control variables by country (percentages and means) DE UK SE Min. Max. Control variables Sex: male 54 52 54 0 1 Age: years 44.29 42.16 44.34 15 84 Weekly working hours 32.18 35.03 37.57 1 84 Industry: manufacturing 22 13 15 0 1 Industry: wholesale and retail trade; repair of … 20 15 10 0 1 Industry: transportation and storage 8 7 6 0 1 Industry: accommodation and food service … 5 5 3 0 1 Industry: information and communication 3 4 5 0 1 Industry: financial and insurance activities 3 4 3 0 1 Industry: real estate activities 1 1 1 0 1 Industry: professional, scientific and technical … 5 4 4 0 1 Industry: administrative and support service … 8 6 6 0 1 Industry: education 4 16 16 0 1 Industry: human health and social work … 17 21 24 0 1 Industry: arts, entertainment and recreation 2 2 3 0 1 Industry: other service activities 3 2 3 0 1 Size of establishment 3.09 3.48 3.45 1 4 Occupation: managers (for comparison) 1 15 7 0 1 Occupation: professionals 12 22 32 0 1 Occupation: technicians and associate professionals 14 10 18 0 1 Occupation: clerical support workers 15 9 6 0 1 Occupation: service and sales workers 27 25 23 0 1 Occupation: craft and related trades workers 9 4 4 0 1 Occupation: plant and machine operators … 10 7 6 0 1 Occupation: elementary occupations 11 8 5 0 1 DE UK SE Min. Max. Control variables Sex: male 54 52 54 0 1 Age: years 44.29 42.16 44.34 15 84 Weekly working hours 32.18 35.03 37.57 1 84 Industry: manufacturing 22 13 15 0 1 Industry: wholesale and retail trade; repair of … 20 15 10 0 1 Industry: transportation and storage 8 7 6 0 1 Industry: accommodation and food service … 5 5 3 0 1 Industry: information and communication 3 4 5 0 1 Industry: financial and insurance activities 3 4 3 0 1 Industry: real estate activities 1 1 1 0 1 Industry: professional, scientific and technical … 5 4 4 0 1 Industry: administrative and support service … 8 6 6 0 1 Industry: education 4 16 16 0 1 Industry: human health and social work … 17 21 24 0 1 Industry: arts, entertainment and recreation 2 2 3 0 1 Industry: other service activities 3 2 3 0 1 Size of establishment 3.09 3.48 3.45 1 4 Occupation: managers (for comparison) 1 15 7 0 1 Occupation: professionals 12 22 32 0 1 Occupation: technicians and associate professionals 14 10 18 0 1 Occupation: clerical support workers 15 9 6 0 1 Occupation: service and sales workers 27 25 23 0 1 Occupation: craft and related trades workers 9 4 4 0 1 Occupation: plant and machine operators … 10 7 6 0 1 Occupation: elementary occupations 11 8 5 0 1 Notes: Figures of dummy variables in % of yes answers; for all other variables means are depicted. Values are represented as percentage, unless otherwise mentioned. Source: EWCS 2015, own calculations. View Large Table A1. Distribution of control variables by country (percentages and means) DE UK SE Min. Max. Control variables Sex: male 54 52 54 0 1 Age: years 44.29 42.16 44.34 15 84 Weekly working hours 32.18 35.03 37.57 1 84 Industry: manufacturing 22 13 15 0 1 Industry: wholesale and retail trade; repair of … 20 15 10 0 1 Industry: transportation and storage 8 7 6 0 1 Industry: accommodation and food service … 5 5 3 0 1 Industry: information and communication 3 4 5 0 1 Industry: financial and insurance activities 3 4 3 0 1 Industry: real estate activities 1 1 1 0 1 Industry: professional, scientific and technical … 5 4 4 0 1 Industry: administrative and support service … 8 6 6 0 1 Industry: education 4 16 16 0 1 Industry: human health and social work … 17 21 24 0 1 Industry: arts, entertainment and recreation 2 2 3 0 1 Industry: other service activities 3 2 3 0 1 Size of establishment 3.09 3.48 3.45 1 4 Occupation: managers (for comparison) 1 15 7 0 1 Occupation: professionals 12 22 32 0 1 Occupation: technicians and associate professionals 14 10 18 0 1 Occupation: clerical support workers 15 9 6 0 1 Occupation: service and sales workers 27 25 23 0 1 Occupation: craft and related trades workers 9 4 4 0 1 Occupation: plant and machine operators … 10 7 6 0 1 Occupation: elementary occupations 11 8 5 0 1 DE UK SE Min. Max. Control variables Sex: male 54 52 54 0 1 Age: years 44.29 42.16 44.34 15 84 Weekly working hours 32.18 35.03 37.57 1 84 Industry: manufacturing 22 13 15 0 1 Industry: wholesale and retail trade; repair of … 20 15 10 0 1 Industry: transportation and storage 8 7 6 0 1 Industry: accommodation and food service … 5 5 3 0 1 Industry: information and communication 3 4 5 0 1 Industry: financial and insurance activities 3 4 3 0 1 Industry: real estate activities 1 1 1 0 1 Industry: professional, scientific and technical … 5 4 4 0 1 Industry: administrative and support service … 8 6 6 0 1 Industry: education 4 16 16 0 1 Industry: human health and social work … 17 21 24 0 1 Industry: arts, entertainment and recreation 2 2 3 0 1 Industry: other service activities 3 2 3 0 1 Size of establishment 3.09 3.48 3.45 1 4 Occupation: managers (for comparison) 1 15 7 0 1 Occupation: professionals 12 22 32 0 1 Occupation: technicians and associate professionals 14 10 18 0 1 Occupation: clerical support workers 15 9 6 0 1 Occupation: service and sales workers 27 25 23 0 1 Occupation: craft and related trades workers 9 4 4 0 1 Occupation: plant and machine operators … 10 7 6 0 1 Occupation: elementary occupations 11 8 5 0 1 Notes: Figures of dummy variables in % of yes answers; for all other variables means are depicted. Values are represented as percentage, unless otherwise mentioned. Source: EWCS 2015, own calculations. View Large © The Author 2017. Published by Oxford University Press and the Society for the Advancement of Socio-Economics. All rights reserved. For Permissions, please email: journals.permissions@oup.com http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Socio-Economic Review Oxford University Press

How national employment systems relate to employee involvement: a decomposition analysis of Germany, the UK and Sweden

Loading next page...
 
/lp/ou_press/how-national-employment-systems-relate-to-employee-involvement-a-2WAEyfJaDg
Publisher
Oxford University Press
Copyright
© The Author 2017. Published by Oxford University Press and the Society for the Advancement of Socio-Economics. All rights reserved. For Permissions, please email: journals.permissions@oup.com
ISSN
1475-1461
eISSN
1475-147X
D.O.I.
10.1093/ser/mwx053
Publisher site
See Article on Publisher Site

Abstract

Abstract We investigate general claims that national institutional conditions relate to employee involvement across countries. Using a decomposition analysis, we examine how much of the key domains of national employment systems contribute to differences in employee involvement in Germany, the UK and Sweden. Drawing on the 2015 European Working Conditions Survey (EWCS), our decomposition analysis explains between 40% and 65% of the cross-national differences. These differences stem from key national employment systems domains, namely the management system, information and communication technology use, as well as training and education. We show that these domains contribute simultaneously and with different weights to cross-national differences, and conclude that domains’ contributions reflect the specific institutional characteristics of the investigated national employment systems. 1. Introduction Employee involvement is enjoying renewed interest as a key element of job quality (e.g. Boxall and Winterton, 2015; Marchington, 2015). Since it allows for direct employee integration and influence on decisions about the wider work organization, employee involvement influences working conditions and well-being (e.g. Kalleberg et al., 2009; Gallie, 2013). From a cross-national comparative perspective, several empirical studies have shown that employees’ opportunities for involvement in organizations differ significantly across countries (e.g. Dobbin and Boychuk, 1999; Olsen et al., 2010; Esser and Olsen, 2012; Holman, 2013; Brewster et al., 2014). An explanation of these differences can be found in national institutional conditions, which are assumed to shape organizational practices (e.g. Maurice et al., 1980; Kern and Schumann, 1984; Piore and Sabel, 1985; Sorge and Streeck, 1988; Sorge, 1991; Streeck, 1991; Appelbaum and Batt, 1994). In particular, building on Fligstein and Byrkjeflot, (1996), Dobbin and Boychuk (1999) argue that specific national employment systems govern how work and employment is organized in a given country. However, to date, we do not fully understand how national employment systems relate to organizational practices generally, nor how they relate to employee involvement in particular (Delbridge et al., 2011; Almond and Gonzalez Menendez, 2014; Wood et al., 2014). In comparative studies on organizational practices and employee involvement, Germany, the UK and Sweden represent core countries (e.g. Maurice et al., 1980; Hall and Soskice, 2001; Amable, 2003; Lorenz and Valeyre, 2005; Gallie, 2007; Croucher et al., 2014).1 According to recent empirical studies, opportunities for involvement in the workplace are lower in Germany as compared to the UK and Sweden (Eurofound, 2013, p. 63; similar findings for autonomy by Esser and Olsen (2012), and job quality types by Holman (2013)). This is surprising because, based on international comparative research, we would expect an opposite ranking for Germany and the UK. Several researchers assume that national institutional conditions in Germany foster higher job quality, including higher employee involvement levels (Gallie, 2007; Frege and Godard, 2014). In contrast, the UK is usually considered an example of low employee involvement, stressing the low-road approach taken by UK firms (Danford et al., 2008; Kalleberg, 2009; Kelly, 2013) and the focus on managerial workplace authority (Dobbin and Boychuk, 1999). Accordingly, the considerably lower employee involvement levels in Germany compared to the UK present a puzzle that has not yet been solved in the literature. A possible explanation for this puzzle could be that, in a direct comparison to national employment systems, several institutional conditions vary simultaneously. For instance, countries differ in the ways management shares authority with regular employees, the extent to which employees use new technology, which education levels dominate, as well as the ways in which employee representatives can influence working conditions. Several empirical studies refer to institutional differences in order to explain cross-national differences in employee involvement (e.g. Dobbin and Boychuk, 1999; Olsen et al., 2010; Holman, 2013). However, these studies usually focus on the overall differences between national employment systems and do not disentangle the different domains’ distinct effects and relative weights. Thus, we do not yet know the extents to which single domains contribute to the overall cross-national differences or whether a given domain contributes at all. However, it is necessary to disentangle the effects of national employment systems domains on employee involvement if we are to better understand how national institutional conditions relate to employee involvement opportunities. In turn, this could also help us to understand the puzzling differences between Germany and the UK. We address this research gap by asking how much distinct institutional domains of national employment systems contribute to the overall cross-national differences in employee involvement. To answer this question, we first discuss five key domains of national employment systems—namely management system, information and communication technology (ICT) use, training and education, employee representation and employment conditions—and their relationships to employee involvement. Building on this, we characterize Germany, the UK and Sweden along these domains. Our empirical analysis utilizes employee data from the 2015 European Working Conditions Survey (EWCS) from these three countries. We conduct a decomposition analysis (Jann, 2008) to identify the institutional domains that contribute to cross-national differences in employee involvement. With this analytic strategy, we statistically show—for the first time—how much particular institutional domains contribute to cross-national differences in employee involvement. 2. Theoretical background 2.1 National employment systems The comparative capitalisms (CC) literature (Jackson and Deeg, 2008) provides general arguments for the relationship between national employment systems and organizational practices. Proponents of CC posit that organizational patterns and workplaces differ, since they are embedded in distinct national institutional frameworks (Marsden, 1999; Whitley, 1999, 2003; Hall and Soskice, 2001; Amable, 2003; Delbridge et al., 2011; Frege and Kelly, 2013; Hauptmeier and Vidal, 2014; Morgan and Hauptmeier, 2014). Thus, several authors argue that national institutional frameworks generate a more or less coherent general logic of economic action in a given country (Jackson and Deeg, 2008; Almond and Gonzalez Menendez, 2014). Such logics manifest as ‘national employment systems carry different logics of work control that influence how work is governed in a wide range of settings’ (Dobbin and Boychuk, 1999, p. 262). Thus, national employment systems are characterized by dominant institutional conditions that shape dominant logics of appropriateness (March, 1994). These logics influence actors’ behaviors by establishing identities and matching rules to recognized situations (Fligstein and Byrkjeflot, 1996; Frege and Godard, 2014). Following this literature, we argue that national employment systems enable certain organizational practices while constraining others. Specifically, we assume that the cross-national differences in specific domains account for the empirical differences in organizational practices across countries, which in our case is employee involvement across Germany, the UK and Sweden. In particular, we focus on five key domains of national employment systems, namely management system, ICT use, training and education, employee representation and employment conditions (Fligstein and Byrkjeflot, 1996; Dobbin and Boychuk, 1999; Fligstein, 2001): (a) Management system describes the ways in which control is organized in the workplace (Dobbin and Boychuk, 1999). Management systems differ in the prevalence of managerial positions and the ways by which managers share authority in the workplace with regular employees (Fligstein and Byrkjeflot, 1996; Whitley, 2003). The sharing of responsibilities in the workplace is usually implemented through different forms of job design, which can be described along two basic dimensions: task variability and uncertainty (Appelbaum et al., 2000). Low task variety and uncertainty represent the power of rules and routines. In contrast, high task variety and uncertainty are indicators of employee-oriented management practices. A particular measure to shift managerial control to regular employees via job design is discretionary teamwork (Appelbaum and Batt, 1994; Appelbaum et al., 2000). (b) Another aspect closely related to the management system is ICT use. As Dobbin and Boychuk (1999) noted, employment systems face the challenges of new technologies (similar Fligstein, 2001), and national employment systems differ substantially in their abilities to adopt new technologies in the workplace (Castells, 2000; Hall and Soskice, 2001; Amable, 2003). (c) Training and education refers to the ways employees usually acquire skills as well as the formal education levels employees attain (Gallie, 2007). Key differences are the necessity of continuing training and the foci on either vocational or university education (Fligstein and Byrkjeflot, 1996; Gallie, 2007; Goergen et al., 2012; Goergen et al., 2014). (d) Employee representation describes mechanisms through which employees can collectively influence work and employment conditions. These mechanisms of collective influence are seen as a key difference between national employment systems, because countries differ in the prevalence of employee representation and the power exerted by employee representatives (Fligstein and Byrkjeflot, 1996; Dobbin and Boychuk, 1999; Korpi, 2006; Gallie, 2007). (e) Employment conditions comprise key aspects that regulate employment conditions and unemployment benefits (Dobbin and Boychuk, 1999; Amable, 2003; Gallie, 2007). Lightly regulated employment systems favor market mechanisms, which lead to high labor turnover rates and precarious employment conditions. In contrast, highly regulated employment systems enable longer job tenure and mitigate the precariousness of employment conditions. 2.2 Employee involvement and national employment systems domains Employee involvement—also termed consultative involvement (Felstead et al., 2010), consultative participation (Gallie, 2013), high-involvement management (Wood et al., 2012) or organizational empowerment (Wall et al., 2004)—describes employees’ opportunities to personally influence decisions about the work organization or other aspects of the work environment. Thus, in contrast to job autonomy or task discretion, employee involvement goes beyond the confines of an immediate task. It can range from information, consultation in workplace meetings or more localized briefing groups, suggestion schemes, problem-solving groups, to employee participation in decisions about organizational issues (Wood et al., 2012; Eurofound, 2013; Gallie, 2013). Employee involvement can be influenced by several aspects. We argue that employee involvement relates to national employment systems domains: Differences within these domains should generally relate to differences in employee involvement levels across countries. To substantiate this assumption, we will first show how national employment systems domains relate to employee involvement, since it is only where such a general relationship exists that differences between domains can eventually account for cross-national differences in employee involvement levels. We will then characterize Germany, the UK and Sweden along national employment systems domains. Management system should be relevant, since the prevalence of managerial positions and responsibility-sharing through job design should affect employee involvement. Employees in managerial positions experience more employee involvement. Also, if jobs are usually characterized by low task variety and uncertainty, i.e. if they are governed by rules and routines, there are fewer employee involvement opportunities (Dobbin and Boychuk, 1999; Appelbaum et al., 2000). In contrast, high task variety and high uncertainty should not only foster job autonomy but also employee involvement because, in such a situation, ‘it is seldom practical for managers to have unilateral control over decisions: efficiency requires a more consensus-based approach to decision making’ (Soskice, 1999, p. 115). Teamwork should also increase employee involvement for regular employees (Appelbaum et al., 2000). However, this increase requires one to grant teams de facto discretion (Lawler, 1986, p. 108), and teams are found to differ substantially in their discretion (Pruijt, 2003), for instance, to change tasks, to alter working schedules or to appoint a team leader. ICT use is often perceived as being closely interrelated with management systems and specific job designs that allow for more employee involvement (Appelbaum et al., 2000). Following Castells (2000), ICT use enables new forms of decentralized work organization that in turn allow and require a substantial involvement of employees in work processes and organizational matters. Ample empirical evidence supports this general claim about a positive relationship between ICT use and employee involvement (e.g. Hempell and Zwick, 2008; Green, 2012; Bayo-Moriones et al., 2017).2 Concerning training and education, we assume that employee involvement increases with continuing training opportunities and high education levels, because better-qualified employees should be granted more opportunities for involvement in the organization (Jackson and Schuler, 1995). Again, the underlying idea is that employee knowledge influences the effectiveness of participation and that highly qualified employees should increase performance more than those with little knowledge (Glew et al., 1995). The forms of employee representation should also relate to employee involvement levels. However, the literature provides ambiguous predictions. First, some assume that employee representatives seek to improve employee work and employment conditions, including more employee involvement opportunities (Jackson and Schuler, 1995; Doellgast et al., 2009; Esser and Olsen, 2012). Second, stronger individual employee involvement could challenge collective employee representations, because the former might decrease the power of the latter. Thus, employee representatives might be inclined to limit direct employee involvement (Baron and Kreps, 1999; Hauff et al., 2014). Third, Dobbin and Boychuk (1999) argue that there could be no effect, because practices quickly spread from workplaces with representation to workplaces without representation. A further effect can be expected from general employment conditions. Long-term employment increases employees’ firm-specific experience and mutual trust. This should in turn increase employee involvement (Dobbin and Boychuk, 1999; similarly, Streeck, 1991). Conversely and similarly, precarious employment conditions should negatively affect involvement. To advance our understanding about the relationship between national employment systems domains and employee involvement, we will analyze how these key national employment systems domains empirically relate to employee involvement in the cases of Germany, the UK and Sweden. First, we will briefly characterize our three cases’ national employment systems along these key domains. 2.3 The national employment systems of Germany, the UK and Sweden 2.3.1 The case of Germany Germany is often seen as a role model of highly regulated market economies (Hall and Soskice, 2001). Traditionally, national institutional conditions should foster longer job tenure, which enables employees to build firm-specific skills (Streeck, 1991; Amable, 2003). However, the traditional regulated logic of Germany’s employment system is increasingly shaped by a subnotion of dualism in the labor market (Thelen, 2012). In Germany’s management system, the distribution of workplace authority traditionally focuses on a strong position of highly skilled blue-collar workers (Fligstein and Byrkjeflot, 1996; Whitley, 2003). Skilled workers are granted high discretion, and management partially shares authority with skilled employees in a decentralized organizational model (Streeck, 1991). This partial authority-sharing in a decentralized model is accompanied by a specific job design. An international comparative study (Lorenz and Valeyre, 2005) found that, in Germany, employees work in workplaces with comparably high variability and high uncertainty. ICT usage levels should turn out to be moderate for German employees. The CC literature stresses that German firms favor incremental changes (Hall and Soskice, 2001; Amable, 2003). This decreases the adoption speed of new technologies, such as ICT. However, recent developments—such as the ‘Industrie 4.0’ discourse (Pfeiffer, 2017) and the broader ‘Arbeiten 4.0’ process (Bundesministerium für Arbeit und Soziales, 2015)—promote increased ICT usage as part of an overarching digital agenda in Germany’s employment system. Training and education revolves around the traditionally strong German vocational training system (Amable, 2003; Gallie, 2007). Employers and employees should have incentives to invest in firm-specific skills (Streeck, 1991). However, comparative studies reported comparably low participation rates in continuing vocational training for German employees (Gallie, 2007). This indicates potential limitations to continuing vocational training. In Germany, a strong employee representation system has traditionally been a key pillar of the national employment system (Gallie, 2007). However, more recent developments highlight a substantial declining coverage that limits the scope of representation (Jackson and Deeg, 2012). While some economic sectors still have a strong employee representation system, other sectors exist outside the traditional system (Thelen, 2012). Traditionally, employment conditions in Germany were characterized by high job security (Streeck, 1991). However, recent liberalization of employment regulations have led to a growing dualization (Thelen, 2012; Hassel, 2014). Dualization widens the gap between a secure core worforce and a peripheral workforce with insecure and more precarious employment conditions. 2.3.2 The case of the UK The UK is often considered a key example of a lightly regulated market economy (Hall and Soskice, 2001). In contrast to Germany, market mechanisms shape the relationships between firms and employees. A lack of regulation has led researchers to assume that UK firms follow a low-road approach to workforce management that leads to overall poorer working conditions (Danford et al., 2008; Kalleberg, 2009, p. 12; Kelly, 2013). The UK’s employment system is therefore more market-based and more manager-focused. The UK’s management system is characterized by managers who seize workplace authority and extend authority to skilled workers less often (Dobbin and Boychuk, 1999; Whitley, 2003). This focus on managers in the UK also shapes job design, which exhibits not only high variety levels, but also lower uncertainty levels (Lorenz and Valeyre, 2005). This particular job design, which is often called the lean model, allows for more variety compared to traditional job designs, yet there are fewer learning opportunities. Research results report that the introduction of teamwork in the UK closely followed the lean model (Danford et al., 2008; Kelly, 2013). UK employees’ ICT use should turn out to be high, because the institutional framework supports more radical innovation and should also foster a more extensive implementation of new technologies in the workplace (Hall and Soskice, 2001; Amable, 2003). Concerning training and education, the UK’s vocational training system is comparatively weak (Amable, 2003). A very competitive university system provides highly skilled graduates. The education system emphasizes general skills. Employee representation is also traditionally considered weak in the UK (Amable, 2003; Gallie, 2007). This deprives UK employees of substantial power resources to influence their working conditions. UK employment conditions are shaped by market forces that govern labor market regulations and social security. Thus, higher job insecurity and precarious employment are common employment practices. Recently, further deregulation has increased this tendency (Thelen, 2012). 2.3.3 The case of Sweden For some authors (Hall and Soskice, 2001), Sweden and Germany fall into the same country category of highly regulated market economies. These authors view firms’ high skills and long-term orientation as commonalities between the two countries. However, more recent approaches in the CC literature emphasize dissimilarities (Gallie, 2003, 2007; Thelen, 2012). Compared to Germany, the Swedish employment system provides more support for marginalized employee groups. Thus, the Swedish employment system follows a regulated and more inclusive approach. In the Swedish management system, workplace authority is traditionally shared with regular employees in a decentralized organizational model (Appelbaum and Batt, 1994). Sweden’s workers traditionally enjoyed high involvement levels in semi-autonomous teams. This team approach was a key element of the so-called socio-technical model (Appelbaum and Batt, 1994). The country’s workplace’s development followed this tradition, as high discretionary teamwork prevailed. This allows for a job design with high variability and high uncertainty. Unsurprisingly, international comparisons show that Swedish employees enjoy the highest learning organization rates (Lorenz and Valeyre, 2005). Swedish employees should also exhibit a high ICT usage level, indicating a particular path of technological transformation. Firms receive support for extensive implementation of ICT combined with targeted education policies (Schnyder, 2012; Ornston, 2013). Concerning training and education, Sweden is an example of a combination of a strong vocational system and an extensive higher education system. Thus, the general emphasis is on high skills and education levels (Amable, 2003). In contrast to Germany and the UK, Sweden’s education system is more inclusive and enables lifelong retraining. This also applies to marginal employee groups, who receive support to raise their skill levels (Gallie, 2007; Schnyder, 2012). Employees in Sweden enjoy powerful employee representation in the workplace, backed by the unions’ substantial influence on national policies (Amable, 2003; Korpi, 2006; Gallie, 2007). This provides Swedish employees with substantial power resources to influence their working conditions. Concerning employment conditions, Sweden follows an inclusive approach that supports marginal employee groups (Gallie, 2007; Thelen, 2012). This moderates insecure labor market positions and precarious employment conditions. At the same time, job security is relaxed, so as to foster labor market flexibility. However, through training programs, employees receive support to quickly find new jobs. In sum, we lay out substantial reasons to assume that Germany, the UK and Sweden differ significantly across national employment systems domains (i.e. management system, ICT use, training and education, employee representation and employment conditions). Table 1 summarizes the national employment systems domains, their theorized relationships to employee involvement and the differences between Germany, the UK and Sweden across these domains. Following our argumentation above, there are sound theoretical reasons to expect that all these domains relate to employee involvement. Accordingly, we assume that all five domains contribute to the differences in employee involvement across these countries. We use these insights drawn from the literature as a general framework to organize our empirical analysis. Table 1. Employment systems domains, relationships to employee involvement and cross-national differences Domains of national employment systems Theorized relationship with employee involvement DE UK Sweden (SE) (a) Management system Increases with managerial position, discretionary teamwork as well as with job design variety and uncertainty Vocational employee focus, traditional and learning job design: moderate variety and uncertainty Managerial focus, lean job design: lower variety and lower uncertainty Discretionary teamwork-based, learning job design: high variety and high uncertainty (b) ICT use Increases with ICT use Moderate High High (c) Training and education Increases with higher education levels and training activity Vocational training focus Focus on competitive higher education Focus on higher education, general and inclusive training (d) Employee representation Ambiguous (possibly increases, decreases or unrelated) Medium (potential sectoral focus, dualism) Low High (e) Employment conditions Increases with job security and non-precarious employment High and low (dualism) Medium (traditionally liberal market forces) High (traditionally inclusive) Domains of national employment systems Theorized relationship with employee involvement DE UK Sweden (SE) (a) Management system Increases with managerial position, discretionary teamwork as well as with job design variety and uncertainty Vocational employee focus, traditional and learning job design: moderate variety and uncertainty Managerial focus, lean job design: lower variety and lower uncertainty Discretionary teamwork-based, learning job design: high variety and high uncertainty (b) ICT use Increases with ICT use Moderate High High (c) Training and education Increases with higher education levels and training activity Vocational training focus Focus on competitive higher education Focus on higher education, general and inclusive training (d) Employee representation Ambiguous (possibly increases, decreases or unrelated) Medium (potential sectoral focus, dualism) Low High (e) Employment conditions Increases with job security and non-precarious employment High and low (dualism) Medium (traditionally liberal market forces) High (traditionally inclusive) Source: Own depiction of employment systems domains (adopted by especially drawing on Fligstein and Byrkjeflot, 1996; Dobbin and Boychuk, 1999; Gallie, 2007). Table 1. Employment systems domains, relationships to employee involvement and cross-national differences Domains of national employment systems Theorized relationship with employee involvement DE UK Sweden (SE) (a) Management system Increases with managerial position, discretionary teamwork as well as with job design variety and uncertainty Vocational employee focus, traditional and learning job design: moderate variety and uncertainty Managerial focus, lean job design: lower variety and lower uncertainty Discretionary teamwork-based, learning job design: high variety and high uncertainty (b) ICT use Increases with ICT use Moderate High High (c) Training and education Increases with higher education levels and training activity Vocational training focus Focus on competitive higher education Focus on higher education, general and inclusive training (d) Employee representation Ambiguous (possibly increases, decreases or unrelated) Medium (potential sectoral focus, dualism) Low High (e) Employment conditions Increases with job security and non-precarious employment High and low (dualism) Medium (traditionally liberal market forces) High (traditionally inclusive) Domains of national employment systems Theorized relationship with employee involvement DE UK Sweden (SE) (a) Management system Increases with managerial position, discretionary teamwork as well as with job design variety and uncertainty Vocational employee focus, traditional and learning job design: moderate variety and uncertainty Managerial focus, lean job design: lower variety and lower uncertainty Discretionary teamwork-based, learning job design: high variety and high uncertainty (b) ICT use Increases with ICT use Moderate High High (c) Training and education Increases with higher education levels and training activity Vocational training focus Focus on competitive higher education Focus on higher education, general and inclusive training (d) Employee representation Ambiguous (possibly increases, decreases or unrelated) Medium (potential sectoral focus, dualism) Low High (e) Employment conditions Increases with job security and non-precarious employment High and low (dualism) Medium (traditionally liberal market forces) High (traditionally inclusive) Source: Own depiction of employment systems domains (adopted by especially drawing on Fligstein and Byrkjeflot, 1996; Dobbin and Boychuk, 1999; Gallie, 2007). 3. Data, measures and method 3.1 Data: The EWCS The following analyses are based on data from the EWCS conducted in 2015 (Eurofound, 2016). The EWCS provides comparable data on working conditions across Europe. The target population is people aged 15 and older who were employed at the time of the survey. Data was gathered through face-to-face interviews using a multistage, stratified random sample. Questionnaire development included expert reviews and real-life tests in order to design a reliable and valid instrument that is easy to understand. The interviews were done by interviewers with substantial experience and training (Ipsos, 2015). Thus, the EWCS provides a unique, high-quality data source for cross-national comparative research. For our analysis, we only used data from employees in Germany (DE), the United Kingdom (UK) and Sweden (SE). Further, we included only employees from the manufacturing and service sectors.3 Our final sample comprised 3522 cases (the n for each country is DE = 1564; UK = 1165; SE = 793). 3.2 Dependent variable: the employee involvement factor Employee involvement is our dependent variable. In line with our definition of employee involvement, and following other seminal contributions (Wood et al., 2012; Eurofound, 2013; Gallie, 2013), we used several items to capture different aspects of employment involvement. We computed a factor analysis using the pooled dataset which revealed a single factor.4Table 2 reports the five variables and their factor loadings. The factor’s eigenvalue was 2.42 and the Cronbach’s alpha was 0.81, indicating a viable factor solution. We used the predicted factor scores of the computed factor to generate a new variable for employee involvement. Finally, we normalized the variable values so that it ranged between 0 and 1. Table 2. Variables and their loadings with employee involvement factor Variables Loadings for employee involvement factor Consulted before work targets are set 0.63 Involved in improving the work organization or work processes of the department or organization 0.71 A say in the choice of work partners 0.50 The ability to apply own ideas in work 0.71 Influence on decisions that are important for work 0.87 Variables Loadings for employee involvement factor Consulted before work targets are set 0.63 Involved in improving the work organization or work processes of the department or organization 0.71 A say in the choice of work partners 0.50 The ability to apply own ideas in work 0.71 Influence on decisions that are important for work 0.87 Notes: Answer categories: 1 = never to 5 = always; n = 3458. Source: EWCS, 2015, own calculations. Table 2. Variables and their loadings with employee involvement factor Variables Loadings for employee involvement factor Consulted before work targets are set 0.63 Involved in improving the work organization or work processes of the department or organization 0.71 A say in the choice of work partners 0.50 The ability to apply own ideas in work 0.71 Influence on decisions that are important for work 0.87 Variables Loadings for employee involvement factor Consulted before work targets are set 0.63 Involved in improving the work organization or work processes of the department or organization 0.71 A say in the choice of work partners 0.50 The ability to apply own ideas in work 0.71 Influence on decisions that are important for work 0.87 Notes: Answer categories: 1 = never to 5 = always; n = 3458. Source: EWCS, 2015, own calculations. 3.3 Independent variables Our independent variables correspond to the national employment systems domains introduced above. We subsumed independent variables that account for single employee characteristics into the various national employment systems domains. Management system: Our variables on management system include variables concerning managerial positions and job design. For our analysis, we subdivided the two aspects of management systems into two subsets: For management system positions, a binary variable measures a position of workplace authority, indicating the respondent’s supervisory authority (1 = no, 0 = yes). We also distinguished employees in managerial occupations using the International Standard Classification of Occupations (ISCO).5 For job design in the management system, we refer to the two major characterizations (Appelbaum et al., 2000): task variety and uncertainty. We measured task variety by two items (1 = no, 0 = yes) relating to the question whether a job involves monotony (no task variety) and job complexity. We measured uncertainty by two binary items (1 = no, 0 = yes) related to the question whether a job involves solving unforeseen problems and learning new things. For teamwork, as a specific form of job design, we grouped respondents into three categories: no teamwork, teamwork without influence and teamwork with influence. The latter indicates discretionary teamwork with influence on tasks, on the team leader selection and/or on time. ICT use: One variable accounts for the usage of ICT technology at the workplace, asking respondents how often they work with computers, laptops, and smartphones, on a seven-item scale (1 = never to 7 = all of the time). Training and education: We included variables on training and educational attainment. For training activities, we used two variables to measure whether employees received training paid for by the employer (1 = yes, 0 = no) or on-the-job training (1 = yes, 0 = no). We included education levels, using an international classification (ISCED). We collapsed the original seven-step classification into three categories: low (ISCED 0–2: up to lower secondary education), medium (ISCED 3–4: up to post-secondary non-tertiary education) and high (ISCED 5–6: university degree and beyond). Employee representation: Here, we used a binary variable stating if there is a trade union, works council or a similar committee that represents employees at the respondent’s company or organization (1 = yes, 0 = no). Employment conditions: We included several variables that account for job security and precarious employment conditions. As a first measure, we used the contract type: indefinite contract, fixed-term contract, temporary employment agency contract or other contract. We also included a subjective evaluation of job insecurity and the labor market position as well as a variable on job tenure. Job insecurity was measured as agreement with the statement I may lose my job in the next 6 months. We determined labor market positioning by the statement If I were to lose or quit my current job, it would be easy for me to find a job of similar salary. (The five-item scale for both items ranged from 1 = strongly disagree to 5 = strongly agree.) Job tenure was measured in terms of years with the current employer. Table 3 reports the distribution of all variables across the three countries. Table 3. Distribution of dependent and independent variables by country DE UK SE Min. Max. Employee involvement 0.55 0.67 0.66 0 1 Factor employee involvement (normalized) Management system: positions 10 26 13 0 1 Supervisory function Occupation: managers 1 15 7 0 1 Management system: job design 28 58 22 0 1 Job design: monotonous tasks Job design: complex tasks 62 66 69 0 1 Job design: learning new things 61 83 91 0 1 Job design: solving unforeseen problems on your own 79 84 96 0 1 Teamwork: none 47 22 27 0 1 Teamwork: with influence 38 54 59 0 1 Teamwork: without influence 15 24 14 0 1 ICT usage 3.09 4.34 4.40 1 7 Working with computers, laptops, smartphones Training and education 37 56 46 0 1 Training: employer paid Training: on-the-job 38 58 52 0 1 Education level: low (ISCED 0–2) 8 34 7 0 1 Education level: medium (ISCED 3–4) 78 22 49 0 1 Education level: high (ISCED 5+) 14 44 44 0 1 Employee representation 48 46 84 0 1 Employee representative exists Employment conditions 84 87 85 0 1 Contract: indefinite contract Contract: fixed-term contract 10 4 10 0 1 Contract: temporary employment agency contract 1 3 2 0 1 Contract: other 5 6 3 0 1 Job insecurity 4.15 4.04 4.20 1 5 Easy to find a new job 2.88 3.17 3.31 1 5 Job tenure (years) 9.71 7.79 9.89 0 64 DE UK SE Min. Max. Employee involvement 0.55 0.67 0.66 0 1 Factor employee involvement (normalized) Management system: positions 10 26 13 0 1 Supervisory function Occupation: managers 1 15 7 0 1 Management system: job design 28 58 22 0 1 Job design: monotonous tasks Job design: complex tasks 62 66 69 0 1 Job design: learning new things 61 83 91 0 1 Job design: solving unforeseen problems on your own 79 84 96 0 1 Teamwork: none 47 22 27 0 1 Teamwork: with influence 38 54 59 0 1 Teamwork: without influence 15 24 14 0 1 ICT usage 3.09 4.34 4.40 1 7 Working with computers, laptops, smartphones Training and education 37 56 46 0 1 Training: employer paid Training: on-the-job 38 58 52 0 1 Education level: low (ISCED 0–2) 8 34 7 0 1 Education level: medium (ISCED 3–4) 78 22 49 0 1 Education level: high (ISCED 5+) 14 44 44 0 1 Employee representation 48 46 84 0 1 Employee representative exists Employment conditions 84 87 85 0 1 Contract: indefinite contract Contract: fixed-term contract 10 4 10 0 1 Contract: temporary employment agency contract 1 3 2 0 1 Contract: other 5 6 3 0 1 Job insecurity 4.15 4.04 4.20 1 5 Easy to find a new job 2.88 3.17 3.31 1 5 Job tenure (years) 9.71 7.79 9.89 0 64 Notes: Figures of dummy variables in % of yes answers; for all other variables, means are depicted. Values are represented as percentage, unless otherwise mentioned. Source: EWCS, 2015, own calculations. Table 3. Distribution of dependent and independent variables by country DE UK SE Min. Max. Employee involvement 0.55 0.67 0.66 0 1 Factor employee involvement (normalized) Management system: positions 10 26 13 0 1 Supervisory function Occupation: managers 1 15 7 0 1 Management system: job design 28 58 22 0 1 Job design: monotonous tasks Job design: complex tasks 62 66 69 0 1 Job design: learning new things 61 83 91 0 1 Job design: solving unforeseen problems on your own 79 84 96 0 1 Teamwork: none 47 22 27 0 1 Teamwork: with influence 38 54 59 0 1 Teamwork: without influence 15 24 14 0 1 ICT usage 3.09 4.34 4.40 1 7 Working with computers, laptops, smartphones Training and education 37 56 46 0 1 Training: employer paid Training: on-the-job 38 58 52 0 1 Education level: low (ISCED 0–2) 8 34 7 0 1 Education level: medium (ISCED 3–4) 78 22 49 0 1 Education level: high (ISCED 5+) 14 44 44 0 1 Employee representation 48 46 84 0 1 Employee representative exists Employment conditions 84 87 85 0 1 Contract: indefinite contract Contract: fixed-term contract 10 4 10 0 1 Contract: temporary employment agency contract 1 3 2 0 1 Contract: other 5 6 3 0 1 Job insecurity 4.15 4.04 4.20 1 5 Easy to find a new job 2.88 3.17 3.31 1 5 Job tenure (years) 9.71 7.79 9.89 0 64 DE UK SE Min. Max. Employee involvement 0.55 0.67 0.66 0 1 Factor employee involvement (normalized) Management system: positions 10 26 13 0 1 Supervisory function Occupation: managers 1 15 7 0 1 Management system: job design 28 58 22 0 1 Job design: monotonous tasks Job design: complex tasks 62 66 69 0 1 Job design: learning new things 61 83 91 0 1 Job design: solving unforeseen problems on your own 79 84 96 0 1 Teamwork: none 47 22 27 0 1 Teamwork: with influence 38 54 59 0 1 Teamwork: without influence 15 24 14 0 1 ICT usage 3.09 4.34 4.40 1 7 Working with computers, laptops, smartphones Training and education 37 56 46 0 1 Training: employer paid Training: on-the-job 38 58 52 0 1 Education level: low (ISCED 0–2) 8 34 7 0 1 Education level: medium (ISCED 3–4) 78 22 49 0 1 Education level: high (ISCED 5+) 14 44 44 0 1 Employee representation 48 46 84 0 1 Employee representative exists Employment conditions 84 87 85 0 1 Contract: indefinite contract Contract: fixed-term contract 10 4 10 0 1 Contract: temporary employment agency contract 1 3 2 0 1 Contract: other 5 6 3 0 1 Job insecurity 4.15 4.04 4.20 1 5 Easy to find a new job 2.88 3.17 3.31 1 5 Job tenure (years) 9.71 7.79 9.89 0 64 Notes: Figures of dummy variables in % of yes answers; for all other variables, means are depicted. Values are represented as percentage, unless otherwise mentioned. Source: EWCS, 2015, own calculations. In addition to these variables, we added general controls (see Appendix): We included respondent’s sex, age and weekly working hours, since these basic characteristics might substantially vary between the three countries. We also included industry, coded according to the current classification (NACE, see Eurostat, 2008). A single categorical variable measures establishment size.6 Further, we included the remaining seven non-managerial occupations as control variables: professionals; technicians and associate professionals; clerical support workers; service and sales workers; craft and related trades workers; plant and machine operators and assemblers; and elementary occupations. 3.4 Method: decomposition analysis To investigate cross-national differences, we computed a decomposition analysis (Jann, 2008) of employee involvement. This analysis revealed how much the differences in independent variables contribute to mean differences of the dependent variable (employee involvement) between distinct groups (employees in different countries), that is, our decomposition analysis revealed the characteristics that contribute to the cross-national differences in employee involvement. Our decomposition analysis also showed how much these characteristics contribute to overall cross-national differences. We reported this contribution of combined or individual variables by the estimated coefficients in the tables shown below. This is not possible with common regression analysis. Specifically, a decomposition analysis divides mean differences between two groups into two parts—an explained part and an unexplained part: The explained part accounts for the group differences in the independent variables and thus describes the mean differences accounted for by the model. The unexplained part accounts for the remaining group differences owing to unobserved influences. Thus, the unexplained part comprises all the cross-national differences that the model cannot directly account for using the included independent variables. Thus, we cannot statistically determine what generates the unexplained part. Because a decomposition analysis compares only two groups at a time, we computed two separate decomposition models: Model A (Germany vs. the UK) and Model B (Germany vs. Sweden). In both decomposition models, Germany serves as common reference country. Finally, we took statistical precautions that all results for categorical variables with more than two categories would not depend on chosen reference categories (for details, see Jann, 2008). 4. Results We progressed our decomposition analysis in three steps from general to more specific results: 4.1 Step 1: Overall results of the decomposition models In a first step, we computed the overall results. Table 4 reports the overall results of decomposition models A and B. The decomposition models report the values from the perspective of Swedish or UK employees: positive values in the table indicate that Swedish or UK employees report more employee involvement compared with their German counterparts. The total computed difference in employee involvement between Germany and the UK amounts to 0.12; between Germany and Sweden, it is 0.11. German employees have significantly less employee involvement. Table 4. Overall results of decomposition models of employee involvement Model A Model B DE vs. UK DE vs. SE Estimated coefficients Estimated coefficients Estimated group values Value for employees in Germany (DE) 0.56 0.56 Value for employees in the UK (UK) 0.68 . Value for employees in Sweden (SE) . 0.67 Overall decomposition results Total difference to Germany 0.12*** 0.11*** Explained difference 0.05*** 0.07*** Unexplained difference 0.07*** 0.04*** Explained difference share of total difference (%) ∼42 ∼64 Subsets—only explained part Management system: positions 0.03*** 0.01** Management system: job design 0.02*** 0.05*** ICT use 0.01* 0.02*** Training and education −0.00 0.01* Representation −0.00 −0.00 Employment conditions −0.00 −0.00 Control variables −0.01** −0.01* n 2122 1865 Model A Model B DE vs. UK DE vs. SE Estimated coefficients Estimated coefficients Estimated group values Value for employees in Germany (DE) 0.56 0.56 Value for employees in the UK (UK) 0.68 . Value for employees in Sweden (SE) . 0.67 Overall decomposition results Total difference to Germany 0.12*** 0.11*** Explained difference 0.05*** 0.07*** Unexplained difference 0.07*** 0.04*** Explained difference share of total difference (%) ∼42 ∼64 Subsets—only explained part Management system: positions 0.03*** 0.01** Management system: job design 0.02*** 0.05*** ICT use 0.01* 0.02*** Training and education −0.00 0.01* Representation −0.00 −0.00 Employment conditions −0.00 −0.00 Control variables −0.01** −0.01* n 2122 1865 Notes: Results (estimated coefficients) reported as seen from Swedish or British employees; * P < 0.05, ** P < 0.01, *** P < 0.001. Source: EWCS, 2015, own calculations. Table 4. Overall results of decomposition models of employee involvement Model A Model B DE vs. UK DE vs. SE Estimated coefficients Estimated coefficients Estimated group values Value for employees in Germany (DE) 0.56 0.56 Value for employees in the UK (UK) 0.68 . Value for employees in Sweden (SE) . 0.67 Overall decomposition results Total difference to Germany 0.12*** 0.11*** Explained difference 0.05*** 0.07*** Unexplained difference 0.07*** 0.04*** Explained difference share of total difference (%) ∼42 ∼64 Subsets—only explained part Management system: positions 0.03*** 0.01** Management system: job design 0.02*** 0.05*** ICT use 0.01* 0.02*** Training and education −0.00 0.01* Representation −0.00 −0.00 Employment conditions −0.00 −0.00 Control variables −0.01** −0.01* n 2122 1865 Model A Model B DE vs. UK DE vs. SE Estimated coefficients Estimated coefficients Estimated group values Value for employees in Germany (DE) 0.56 0.56 Value for employees in the UK (UK) 0.68 . Value for employees in Sweden (SE) . 0.67 Overall decomposition results Total difference to Germany 0.12*** 0.11*** Explained difference 0.05*** 0.07*** Unexplained difference 0.07*** 0.04*** Explained difference share of total difference (%) ∼42 ∼64 Subsets—only explained part Management system: positions 0.03*** 0.01** Management system: job design 0.02*** 0.05*** ICT use 0.01* 0.02*** Training and education −0.00 0.01* Representation −0.00 −0.00 Employment conditions −0.00 −0.00 Control variables −0.01** −0.01* n 2122 1865 Notes: Results (estimated coefficients) reported as seen from Swedish or British employees; * P < 0.05, ** P < 0.01, *** P < 0.001. Source: EWCS, 2015, own calculations. A simple juxtaposition of employee involvement means in Table 3 produces the same conclusion. However, going well beyond simple mean differences, the upper rows of Table 4 reveal that the explained difference in model A (for Germany and the UK) accounts for about 42% of the total differences, while model B (for Germany and Sweden) explains around 64% of the differences. These findings have two key implications: First, because the explained parts score around 40% and 60%, the models explain the cross-national differences well. This indicates that the variables in the model capture a substantial share of the differences and underscores our assumptions that the various national employment systems domains are associated with the employee involvement levels across countries. Second, while the models account for some differences with statistical certainty, the model cannot determine a remaining share. We will return to this issue in our discussion. 4.2 Step 2: Aggregated decomposition models and the analysis of subsets In a second step, we investigated the specific contributions of the employment system domains. We computed aggregated decomposition models that combine the various variables of each domain into several subsets. Using the domain subsets, the models revealed the combined differences from all independent variables in a given national employment systems domain. The computed coefficients in the models denote the contribution of the subset variables to the overall difference in employee involvement: a positive coefficient signifies that the cross-national difference in the subset variables increase the difference in employee involvement between Germany and the UK or Sweden. Our results reported in the lower rows of Table 4 revealed statistically significant relationships for several subsets, including both aspects of management system positions and job design as well as ICT use and training and education. The models estimated no statistically significant relationships for the subsets representation and employment conditions. Thus, only some domains statistically fed into the share of the explained difference. Interestingly, the control variables showed significant relationships in both models. This means that specific structural conditions (mostly establishment size) are associated with slightly higher employee involvement in Germany. However, the much larger contributions by the other domain subsets fully consumed this small counteracting effect. Figure 1 depicts the results from the aggregated decomposition models and illustrates the emerging—nuanced—picture.7 This figure shows only domain subsets that statistically significantly contributed to the cross-national differences alongside the unexplained part. For an interpretation, we need to combine two questions: First, does the subset relate to the cross-national differences in employment involvement? Second, how do levels of subset variables differ across Germany to the UK or Sweden? With the descriptive statistics from Table 3, we can now interpret the displayed differences: Figure 1. View largeDownload slide Aggregated decomposition model results: subsets with significant relationships and unexplained part. Domain variables combined in subsets; figure displays only results for four subsets with significant contributions as well as unexplained part displayed; not displayed subsets: representation and employment conditions. Source: EWCS, 2015, own calculations and depiction of decomposition results. Figure 1. View largeDownload slide Aggregated decomposition model results: subsets with significant relationships and unexplained part. Domain variables combined in subsets; figure displays only results for four subsets with significant contributions as well as unexplained part displayed; not displayed subsets: representation and employment conditions. Source: EWCS, 2015, own calculations and depiction of decomposition results. For Germany and the UK (model A), our results showed a substantial association with managerial positions in the management system (0.03). In addition, aspects of job design in the management systems (0.02) and ICT use (0.01) are associated with the higher employee involvement level in the UK. As reported above, this comes with a remaining large unexplained difference (0.07) owing to unknown factors. Thus, overall higher employee involvement levels in the UK relate to a substantially larger share of managerial positions, moderately more discretionary job design and higher ICT use in the workplace. For Germany and Sweden (model B), the results also revealed the importance of the management system. In contrast, here we see a comparably small association with positions (0.01) alongside a more substantial association with job design (0.05). The model reports slightly higher differences owing to ICT use (0.02) and an additional yet relatively small association with training and education (0.01). Thus, overall higher employee involvement levels in Sweden relate to slightly more managerial positions, substantially more discretionary job design, higher ICT use in the workplace and higher training and education levels. 4.3 Step 3: Focused analysis of management system variables In a third step, we highlighted selected variables that are particularly responsible for the cross-national differences in employee involvement. Table 5 reports only the detailed properties of the management system variables, because they contribute substantially to the cross-national differences in employee involvement in both models. Here, the computed coefficients in the models denote the contribution of the specific variables to the overall difference in employee involvement: a positive coefficient signifies that the cross-national difference in the variable increases the difference in employee involvement between Germany and the UK or Sweden. Again, we used the descriptive statistics from Table 3 to interpret the computed differences. Table 5. Detailed decomposition models of employee involvement—explained difference focusing on the management system variables Model A Model B DE vs. UK DE vs. SE Estimated coefficients for explained difference Estimated coefficients for explained difference Overall 0.05*** 0.07*** Management system: positions 0.01*** 0.00** Supervisory function Occupation: managers 0.02*** 0.00* Management system: job design −0.01*** 0.00** Job design: Monotonous tasks Job design: Complex tasks 0.00* 0.00* Job design: Learning new things 0.01*** 0.02*** Job design: Solving unforeseen problems on your own 0.00* 0.01*** Teamwork: none 0.01*** 0.01*** Teamwork: with influence 0.01*** 0.01*** Teamwork: without influence 0.00 −0.00 n 2122 1865 Model A Model B DE vs. UK DE vs. SE Estimated coefficients for explained difference Estimated coefficients for explained difference Overall 0.05*** 0.07*** Management system: positions 0.01*** 0.00** Supervisory function Occupation: managers 0.02*** 0.00* Management system: job design −0.01*** 0.00** Job design: Monotonous tasks Job design: Complex tasks 0.00* 0.00* Job design: Learning new things 0.01*** 0.02*** Job design: Solving unforeseen problems on your own 0.00* 0.01*** Teamwork: none 0.01*** 0.01*** Teamwork: with influence 0.01*** 0.01*** Teamwork: without influence 0.00 −0.00 n 2122 1865 Notes: Results (estimated coefficients) reported as seen from Swedish or British employees: * P < 0.05, ** P < 0.01, *** P < 0.001; control variables are not displayed; minor deviations to overall subset values owing to rounding. Source: EWCS, 2010, own calculations. Table 5. Detailed decomposition models of employee involvement—explained difference focusing on the management system variables Model A Model B DE vs. UK DE vs. SE Estimated coefficients for explained difference Estimated coefficients for explained difference Overall 0.05*** 0.07*** Management system: positions 0.01*** 0.00** Supervisory function Occupation: managers 0.02*** 0.00* Management system: job design −0.01*** 0.00** Job design: Monotonous tasks Job design: Complex tasks 0.00* 0.00* Job design: Learning new things 0.01*** 0.02*** Job design: Solving unforeseen problems on your own 0.00* 0.01*** Teamwork: none 0.01*** 0.01*** Teamwork: with influence 0.01*** 0.01*** Teamwork: without influence 0.00 −0.00 n 2122 1865 Model A Model B DE vs. UK DE vs. SE Estimated coefficients for explained difference Estimated coefficients for explained difference Overall 0.05*** 0.07*** Management system: positions 0.01*** 0.00** Supervisory function Occupation: managers 0.02*** 0.00* Management system: job design −0.01*** 0.00** Job design: Monotonous tasks Job design: Complex tasks 0.00* 0.00* Job design: Learning new things 0.01*** 0.02*** Job design: Solving unforeseen problems on your own 0.00* 0.01*** Teamwork: none 0.01*** 0.01*** Teamwork: with influence 0.01*** 0.01*** Teamwork: without influence 0.00 −0.00 n 2122 1865 Notes: Results (estimated coefficients) reported as seen from Swedish or British employees: * P < 0.05, ** P < 0.01, *** P < 0.001; control variables are not displayed; minor deviations to overall subset values owing to rounding. Source: EWCS, 2010, own calculations. Overall, the detailed decomposition model results in Table 5 revealed similar relationships as well as country-specific patterns: major effects in the Germany–UK comparison stem from managerial positions and supervisory functions more often held by UK employees. Also, more favorable job design (learning and teamwork) add to the cross-national differences. However, the UK model also revealed specific effects of more monotonous work that reflect particular characteristics of the UK employment system. Here, the higher monotonous work levels in the UK partially mitigate the difference to German employees. Overall, this small counteracting effect of monotony is easily absorbed by the remaining positive influences of positions and job design. Comparing Germany and Sweden revealed similar relationships alongside specific patterns. Here, better job design (especially less monotonous tasks, more learning opportunities and more problem-solving opportunities) as well as higher education levels increase employee involvement for Swedish employees. However, similar to the UK but to a much lesser extent, the slightly higher share of managerial positions in Sweden also increases the overall opportunities for employee involvement. Returning to Table 1, which summarized our considerations, we can now conclude that our models support some but not all of our theorized relationships. Our analyses revealed substantial statistical contributions by three of the five investigated domains. This includes management system, ICT use and training and education. These domains contribute to the cross-national differences in employee involvement, since the characteristics along these domains vary substantially among employees from Germany, the UK and Sweden. 5. Discussion We started with the puzzle that opportunities for involvement in the workplace are lower in Germany compared to the UK and Sweden (Eurofound, 2013). Building on previous approaches (Fligstein and Byrkjeflot, 1996; Dobbin and Boychuk, 1999), we assumed that institutional domains of national employment systems (management system, ICT use, training and education, employee representation and employment conditions) relate to employee involvement and therefore could help explain the different employee involvement levels across countries. To analyze how much each of the different domains contributes to cross-national differences in employee involvement, we performed a decomposition analysis using employee data from Germany, the UK and Sweden. In line with previous findings (Eurofound, 2013), our more recent results from 2015 also showed that employee involvement levels in Sweden and the UK exceeded those in Germany. Following the research, we would expect higher employee involvement levels in Germany owing to more favorable national institutional conditions. In contrast to these perspectives, our empirical findings point to a German employee involvement gap. Going beyond existing studies, our decomposition analysis disentangles the effects behind these puzzling differences and helps us to understand why employee involvement in Germany is significantly lower than in the UK. In particular, our empirical investigation shows the extent to which differences in the key national employment systems domains contribute to the cross-national differences. We can now characterize the differences related to the five key national employment systems domains: Management systems contribute substantially to the cross-national differences in employee involvement. We found general relationships alongside distinct national patterns. Compared to Germany, differences to the UK emerged owing to the increased prevalence of managerial positions, reflecting the strong management focus in the UK employment system (Fligstein and Byrkjeflot, 1996; Dobbin and Boychuk, 1999; Whitley, 2003). A higher share of managerial positions in the UK increases the involvement level. If we were to disregard the difference owing to managerial positions, the overall difference in employee involvement to German employees would shrink considerably. We also found that job design contributes to the differences between Germany and the UK. More learning opportunities and more teamwork in the UK increase employee involvement there. While this finding conflicts with the general assumption that the UK embarked on a low-road model, this result is in line with recent findings by Frege and Godard (2014), who argue that firms in liberal countries improve job design elements to compensate for trust problems in weak institutional environments. However, more monotonous work in the UK partially counteracts this general pattern, which—in turn—partially supports conflicting assumptions about the low road and managerial focus in the UK. Overall, the UK results show a combination of a strong management focus combined with an effectively compensating job design. In contrast, in the comparison between Germany and Sweden, we found that good job design accounts for cross-national differences, reflecting the long tradition of discretionary job designs in Swedish workplaces (Appelbaum and Batt, 1994). However, also in Sweden, a slightly higher share of managerial positions increases employee involvement levels. Overall, our results show the substantial importance of management system for the difference between Germany and the UK. This underlines the key role of managerial positions in understanding employee involvement. To some extent, this also applies to the differences between Germany and Sweden. Based on our findings, we highlight that studies of employees should include not only regular employees. Empirically, managers count as employees too, since they also form part of the sample population in general employee surveys. Thus, scholars of employee involvement should account for employees in managerial positions in order to correctly address differences between countries. Generally, in our view, theoretical and empirical approaches could advance if they would more explicitly address the importance of managerial positions for job quality. Turning to ICT use, our results point to an overall positive relationship between ICT use and employee involvement. The lower ICT usage levels of German employees contribute to the German involvement gap. This finding again underscores the incremental dynamics of Germany’s economy in the face of technological shifts. In light of these findings, recent developments that promote ICT use in Germany’s economy—such as the Industrie 4.0 discourse (Pfeiffer, 2017) and the broader Arbeiten 4.0 process (Bundesministerium für Arbeit und Soziales, 2015)—might enable a catch-up process that could increase opportunities for more employee involvement in Germany. Taken together, our findings regarding ICT use underline the importance of the use of new technologies as a national employment systems domain. This importance will grow further as the digital economy expands. Concerning training and education, our results show that differences in education levels also contribute to the German employee involvement gap, compared to Sweden. This underlines the particular characteristics of Sweden’s employment system, reflecting its general emphasis on high education levels (Amable, 2003). Concerning employee representation and employment conditions, we found no conclusive statistical evidence that they relate to increased or decreased employee involvement. Thus, the finding on employee representation tends to support positions in the literature that assume no direct effect at the firm level (Dobbin and Boychuk, 1999). However, our analysis only relies on one item for firm-level employee representation. Because this item is limited, we cannot rule out that there might be a relationship between employee representation and employee involvement levels across countries. This would require more elaborate measures (see limitations, below). While both aspects might have shaped employee involvement in past processes, our findings provide no evidence for a current statistical relationship. Overall, our decomposition models explain the differences in national employee involvement levels well. The differences in the key national employment systems domains account for 40–65% of the cross-national differences in employee involvement. Our results point out that different national employment systems domains (i.e. management system, ICT use and training and education) contribute simultaneously to the cross-national differences. Accordingly, a comprehensive account of the cross-national differences in employee involvement requires several domains at the same time. Building on these differentiated results, we interpret the different domains’ contributions as a result of the diverse characteristics that underlie the national employment systems of Germany, the UK and Sweden. Thus, our results indicate that the national institutional conditions and the related organizational practices in Germany prove less favorable for employee involvement than researchers often believed them to be. Further, our results contribute to the literature, since they inform theoretical concepts (and future empirical studies) that specific domains contribute with different weights to cross-national differences in organizational practices. Thus, depending on the issue in question, some institutional domains may prove more influential than others. This is highly relevant for the study and explanation of cross-national differences in organizational practices. For instance, in a case of institutional transformation of a domain, organizational practices may prove stable if the domain only marginally contributes to the differences. Also, transformations in one domain could compensate for changes in another domain: for instance, an increase in ICT use could be counteracted by a decrease in job design. Thus, the de facto influence of institutional transformations on organizational practices depends on the relative weight exerted by the involved domains. Finally, our decomposition models leave some differences unexplained. Here, unexplained denotes that differences remain that cannot be determined by the variables in the decomposition model. This especially pertains to large parts of the differences between Germany and the UK. Thus, researchers should treat absolute mean differences between countries with caution. While there might be good theoretical foundations that help to interpret cross-national differences, researchers should empirically decompose relationships and should statistically determine underlying patterns, if empirically possible. This would allow for more substantiated claims. Methodological sources inflating the unexplained part could simply derive from measurement errors (nationally specific understandings, response patterns, questionnaire designs). Theoretical sources feeding the unexplained part might stem from overlooked additional theoretical approaches (e.g. cultural differences). Future research should propose additional and complementary theoretical explanations to resolve the remaining cross-national differences. Our findings and interpretations should be seen in light of several limitations: First, since we analyzed cross-sectional data, our statistical analysis does not present evidence for a causal argument. However, our findings shed light on the current associations that underlie cross-national differences in employee involvement. In our view, the empirical investigation of the current associations takes an important step in the empirical foundations of cross-national comparative research. Nonetheless, we maintain that future research should undertake the possible empirical steps in order to advance our understanding of cross-national differences based on longitudinal data. Second, there might be answering patterns across countries. This concerns, for instance, information on occupation, which might be biased by national standards and interpretations (e.g. in the UK, the title manager is used differently to Germany and Sweden). However, the EWCS implements the common international standard classifications (ISCO), with considerable efforts to ensure comparability across countries (Ipsos, 2015). Other major studies (e.g. by Eurostat) rely on the same international standards. Third, the EWCS only includes a general question about an on-site workplace representative. This limits our model’s capability to reveal possible effects of employee representation. Here, more elaborate measures might provide additional empirical insights. Besides limitations, there are several avenues for future research: We limited our analysis to Germany, the UK and Sweden, which represent core countries in comparative studies. Future studies should also include other countries. This could deepen our understandings of how the key national employment systems domains shape organizational practices. By going beyond employee involvement, future research should apply decomposition analysis to other dependent variables. Such extensions could more clearly show whether or not key national employment systems domains also relate to cross-national differences in other organizational practices. In conclusion, our empirical analysis reveals how national institutional conditions relate to differences in employee involvement in Germany, UK and Sweden. We statistically show, for the first time, that different institutional domains of national employment systems (i.e. management system, ICT use and training and education) contribute simultaneously and with different weights to cross-national differences in employee involvement. Our study also contributes to the cross-national comparative research, since it integrates long-held assumptions with current empirical data, employing innovative statistical techniques. Thus, our analysis underscores the high relevance of cross-national comparative research to understand differences in employee involvement, in particular as well as differences in organizational practices, in general. Footnotes 1 The UK is often considered the same country category as the USA. So, it is usually assumed that characteristics of the USA mostly also apply to the UK. While we acknowledge that there are differences between these two countries, we follow the general custom to relate insights for the UK or the USA to the general country category. 2 Some empirical studies (e.g. Bayo-Moriones et al., 2017) highlight specific conditions under which ICT use may affect employee involvement differently. However, in this article, we limit our argument to the question whether or not there is an average positive relationship between ICT use and employee involvement. 3 Owing to low frequencies, we also excluded the remaining 18 respondents from skilled agricultural, forestry and fish workers occupations. 4 We computed a series of alternative factor models (e.g. models for each country independently). Across the different specifications, the models all show a high consistency with the pooled model. 5 To obtain information regarding occupations, EWCS respondents answered two open-ended questions about their job title and their main activities at the workplace. This ensures that respondents do not falsely assign themselves to particular occupations. Also, the EWCS employed a centralized coding interface and multiple coders per country to make the coding process as uniform as possible across countries. In addition, training sessions were held for all coders and materials, with detailed coding instructions provided (Ipsos, 2015). 6 We included industry and company size to control for possible effects that derive from countries’ industry specializations. The literature suggests that Germany, the UK and Sweden differ in their industries’ compositions, and dominant company sizes (Dobbin and Boychuk, 1999; Amable, 2003; Kelly, 2013). 7 Note that in both models, the control variable subset has a significant mitigating effect of around −0.01, accounting for the minor discrepancy between depicted differences in Figure 1 and the differences gap in Table 4; for further details, also see the Appendix. References Almond P. , Gonzalez Menendez M. C. ( 2014 ) ‘Cross-National Comparative Human Resource Management and the Ideational Sphere: A Critical Review’ , The International Journal of Human Resource Management , 25 , 2591 – 2607 . Google Scholar CrossRef Search ADS Amable B. ( 2003 ) The Diversity of Modern Capitalism , Oxford , Oxford University Press . Google Scholar CrossRef Search ADS Appelbaum E. , Bailey T. , Berg P. , Kalleberg A. L. ( 2000 ) Manufacturing Advantage: Why High-Performance Work Systems Pay Off , Ithaca, NY , Cornell University Press . Appelbaum E. , Batt R. ( 1994 ) The New American Workplace. Transforming Work Systems in the United States , Ithaca, NY , ILR Press . Baron J. N. , Kreps D. M. ( 1999 ) Strategic Human Resources: Frameworks for General Managers , New York , Wiley . Bayo-Moriones A. , Billon M. , Lera-López F. ( 2017 ) ‘ Are New Work Practices Applied Together with Ict and Amt?’, The International Journal of Human Resource Management , 28 , 553 – 580 . Google Scholar CrossRef Search ADS Boxall P. , Winterton J. ( 2015 ) ‘Which Conditions Foster High-Involvement Work Processes? A Synthesis of the Literature and Agenda for Research’, Economic and Industrial Democracy, 1 – 21 . doi: 10.1177/0143831X15599584. Brewster C. , Brookes M. , Johnson P. , Wood G. ( 2014 ) ‘ Direct Involvement, Partnership and Setting: A Study in Bounded Diversity’, The International Journal of Human Resource Management , 25 , 795 – 809 . Google Scholar CrossRef Search ADS Bundesministerium für Arbeit und Soziales ( 2015 ) Grünbuch Arbeiten 4.0, Berlin, Bundesministerium für Arbeit und Soziales Abteilung Grundsatzfragen des Sozialstaats, der Arbeitswelt und der sozialen Marktwirtschaft. Castells M. ( 2000 ) The Rise of the Network Society: The Information Age: Economy, Society, and Culture , Malden, MA , Blackwell . Croucher R. , Brookes M. , Wood G. , Brewster C. ( 2014 ) ‘Is There Convergence Towards Individual Voice in Europe?’. In Wood G. , Brewster C. , Brookes M. (eds) Human Resource Management and the Institutional Perspective , New York, London , Routledge , pp. 59 – 77 . Danford A. , Richardson M. , Stewart P. , Tailby S. , Upchurch M. ( 2008 ) ‘ Partnership, High Performance Work Systems and Quality of Working Life’, New Technology, Work and Employment , 23 , 151 – 166 . Google Scholar CrossRef Search ADS Delbridge R. , Hauptmeier M. , Sengupta S. ( 2011 ) ‘ Beyond the Enterprise: Broadening the Horizons of International HRM’, Human Relations , 64 , 483 – 505 . Google Scholar CrossRef Search ADS Dobbin F. , Boychuk T. ( 1999 ) ‘ National Employment Systems and Job Autonomy: Why Job Autonomy Is High in the Nordic Countries and Low in the United States, Canada, and Australia’, Organization Studies , 20 , 257 – 291 . Google Scholar CrossRef Search ADS Doellgast V. , Batt R. , Sørensen O. H. ( 2009 ) ‘ Introduction: Institutional Change and Labour Market Segmentation in European Call Centres’, European Journal of Industrial Relations , 15 , 349 – 371 . Google Scholar CrossRef Search ADS Esser I. , Olsen K. M. ( 2012 ) ‘ Perceived Job Quality: Autonomy and Job Security within a Multi-Level Framework’, European Sociological Review , 28 , 443 – 454 . Google Scholar CrossRef Search ADS Eurofound ( 2013 ) Work Organisation and Employee Involvement in Europe (Authors: Duncan Gallie and Ying Zhou) , Luxembourg , Publications Office of the European Union . Eurofound ( 2016 ) Sixth European Working Conditions Survey – Overview Report , Luxembourg , Publications Office of the European Union . Eurostat ( 2008 ) Nace Rev. 2. Statistical Classification of Economic Activities in the European Community , Luxembourg , Office for Official Publications of the European Communities . Felstead A. , Gallie D. , Green F. , Zhou Y. ( 2010 ) ‘ Employee Involvement, the Quality of Training and the Learning Environment: An Individual Level Analysis’, The International Journal of Human Resource Management , 21 , 1667 – 1688 . Google Scholar CrossRef Search ADS Fligstein N. ( 2001 ) The Architecture of Markets. An Economic Sociology of Twenty-First-Century Capitalist Societies, Princeton, Princeton University Press. Fligstein N. , Byrkjeflot H. ( 1996 ) ‘The Logic of Employment Systems’. In Baron J. N. , Grusky D. B. , Treiman D. J. (eds) Social Differentiation and Social Inequality , Boulder, CO , Westview Press , pp. 11 – 35 . Frege C. , Godard J. ( 2014 ) ‘ Varieties of Capitalism and Job Quality: The Attainment of Civic Principles at Work in the United States and Germany’ , American Sociological Review , 79 , 942 – 965 . Google Scholar CrossRef Search ADS Frege C. , Kelly J. ( 2013 ) ‘Theoretical Perspectives on Comparative Employment Relations’. In Frege C. , Kelly J. (eds) Comparative Employment Relations in the Global Economy , London , Routledge , pp. 8 – 26 . Gallie D. ( 2003 ) ‘ The Quality of Working Life: Is Scandinavia Different?’, European Sociological Review , 19 , 61 – 79 . Google Scholar CrossRef Search ADS Gallie D. ( 2007 ) ‘Production Regimes, Employment Regimes, and the Quality of Work’. In Gallie D. (ed.) Employment Regimes and the Quality of Work , Oxford , Oxford University Press , pp. 1 – 33 . Google Scholar CrossRef Search ADS Gallie D. ( 2013 ) ‘ Direct Participation and the Quality of Work’, Human Relations , 66 , 453 – 473 . Google Scholar CrossRef Search ADS Glew D. J. , O’Leary-Kelly A. M. , Griffin R. W. , Van Fleet D. D. ( 1995 ) ‘ Participation in Organizations: A Preview of the Issues and Proposed Framework for Future Analysis’, Journal of Management , 21 , 395 – 421 . Google Scholar CrossRef Search ADS Goergen M. , Brewster C. , Wood G. , Wilkinson A. ( 2012 ) ‘ Varieties of Capitalism and Investments in Human Capital’, Industrial Relations: A Journal of Economy and Society , 51 , 501 – 527 . Google Scholar CrossRef Search ADS Goergen M. , Wood G. , Brewster C. , Brookes M. ( 2014 ) ‘Corporate Governance Systems and Investments in Human Capital’. In Wood G. , Brewster C. , Brookes M. (eds) Human Resource Management and the Institutional Perspective , New York, London , Routledge , pp. 78 – 107 . Green F. ( 2012 ) ‘ Employee Involvement, Technology and Evolution in Job Skills: A Task-Based Analysis’, ILR Review , 65 , 36 – 67 . Google Scholar CrossRef Search ADS Hall P. A. , Soskice D. ( 2001 ) ‘An Introduction to Varieties of Capitalism’. In Hall P. , Soskice D. (eds) Varieties of Capitalism: The Institutional Foundations of Comparative Advantage , Oxford , Oxford University Press , pp. 1 – 68 . Google Scholar CrossRef Search ADS Hassel A. ( 2014 ) ‘ The Paradox of Liberalization — Understanding Dualism and the Recovery of the German Political Economy’, British Journal of Industrial Relations , 52 , 57 – 71 . Google Scholar CrossRef Search ADS Hauff S. , Alewell D. , Hansen N. K. ( 2014 ) ‘ HRM Systems between Control and Commitment: Occurrence, Characteristics and Effects on HRM Outcomes and Firm Performance’, Human Resource Management Journal , 24 , 424 – 441 . Google Scholar CrossRef Search ADS Hauptmeier M. , Vidal M. (eds) ( 2014 ) Comparative Political Economy of Work , Houndmills , Palgrave Macmillan . Google Scholar CrossRef Search ADS Hempell T. , Zwick T. ( 2008 ) ‘New Technology, Work Organization and Innovation’ , Economics of Innovation and New Technology , 17 , 331 – 354 . Google Scholar CrossRef Search ADS Holman D. ( 2013 ) ‘ Job Types and Job Quality in Europe’, Human Relations , 66 , 475 – 502 . Google Scholar CrossRef Search ADS Ipsos ( 2015 ) 6th European Working Conditions Survey: Technical Report , Dublin , European Foundation for the Improvement of Living and Working Conditions . Jackson G. , Deeg R. ( 2008 ) ‘ From Comparing Capitalisms to the Politics of Institutional Change’, Review of International Political Economy , 15 , 680 – 709 . Google Scholar CrossRef Search ADS Jackson G. , Deeg R. ( 2012 ) ‘ The Long-Term Trajectories of Institutional Change in European Capitalism’, Journal of European Public Policy , 19 , 1109 – 1125 . Google Scholar CrossRef Search ADS Jackson S. E. , Schuler R. S. ( 1995 ) ‘ Understanding Human Resource Management in the Context of Organizations and Their Environments’, Annual Review of Psychology , 46 , 237 – 264 . Google Scholar CrossRef Search ADS PubMed Jann B. ( 2008 ) ‘The Blinder-Oaxaca Decomposition for Linear Regression Models’ , Stata Journal , 8 , 453 – 479 . Kalleberg A. L. ( 2009 ) ‘ Precarious Work, Insecure Workers: Employment Relations in Transition’, American Sociological Review , 74 , 1 – 22 . Google Scholar CrossRef Search ADS Kalleberg A. L. , Nesheim T. , Olsen K. M. ( 2009 ) ‘ Is Participation Good or Bad for Workers?: Effects of Autonomy, Consultation and Teamwork on Stress among Workers in Norway’, Acta Sociologica , 52 , 99 – 116 . Google Scholar CrossRef Search ADS Kelly J. ( 2013 ) ‘The United Kingdom’. In Frege C. , Kelly J. (eds) Comparative Employment Relations in the Global Economy , London , Routledge , pp. 170 – 186 . Kern H. , Schumann M. ( 1984 ) Das Ende Der Arbeitsteilung?: Rationalisierung in Der Industriellen Produktion: Bestandsaufnahme, Trendbestimmung , München , Beck . Korpi W. ( 2006 ) ‘ Power Resources and Employer-Centered Approaches in Explanations of Welfare States and Varieties of Capitalism: Protagonists, Consenters, and Antagonists’, World Politics , 58 , 167 – 206 . Google Scholar CrossRef Search ADS Lawler E. E. ( 1986 ) High-Involvement Management: Participative Strategies for Improving Organizational Performance , San Francisco, CA , Jossey-Bass . Lorenz E. , Valeyre A. ( 2005 ) ‘Organisational Innovation, Human Resource Management and Labour Market Structure: A Comparison of the Eu-15’ , Journal of Industrial Relations , 47 , 424 – 442 . Google Scholar CrossRef Search ADS March J. G. ( 1994 ) Primer on Decision Making: How Decisions Happen , New York , Free Press . Marchington M. ( 2015 ) ‘ Analysing the Forces Shaping Employee Involvement and Participation (Eip) at Organisation Level in Liberal Market Economies (LMEs)’, Human Resource Management Journal , 25 , 1 – 18 . Google Scholar CrossRef Search ADS Marsden D. ( 1999 ) A Theory of Employment Systems: Micro-Foundations of Societal Diversity: Micro-Foundations of Societal Diversity , Oxford, Oxford University Press . Google Scholar CrossRef Search ADS Maurice M. , Sorge A. , Warner M. ( 1980 ) ‘ Societal Differences in Organizing Manufacturing Units: A Comparison of France, West Germany, and Great Britain’, Organization Studies , 1 , 59 – 86 . Google Scholar CrossRef Search ADS Morgan G. , Hauptmeier M. ( 2014 ) ‘Varieties of Institutional Theory in Comparative Employment Relations’. In Wilkinson A. , Wood G. , Deeg R. (eds) Oxford Handbook of Employment Relations - Comparative Employment Systems , Oxford, Oxford University Press , pp. 190 – 214 . Olsen K. M. , Kalleberg A. L. , Nesheim T. ( 2010 ) ‘Perceived Job Quality in the United States, Great Britain, Norway and West Germany, 1989-2005’ , European Journal of Industrial Relations , 16 , 221 – 240 . Google Scholar CrossRef Search ADS Ornston D. ( 2013 ) ‘ Creative Corporatism: The Politics of High-Technology Competition in Nordic Europe’, Comparative Political Studies , 46 , 702 – 729 . Google Scholar CrossRef Search ADS Pfeiffer S. ( 2017 ) ‘The Vision of “Industrie 4.0” in the Making—A Case of Future Told, Tamed, and Traded’ , NanoEthics , 11 , 107 – 121 . Google Scholar CrossRef Search ADS PubMed Piore M. J. , Sabel C. F. ( 1985 ) Das Ende Der Massenproduktion: Studie Über Die Requalifizierung Der Arbeit Und Die Rückkehr Der Ökonomie in Die Gesellschaft , Berlin , Wagenbach . Pruijt H. ( 2003 ) ‘ Teams between Neo-Taylorism and Anti-Taylorism’, Economic and Industrial Democracy , 24 , 77 – 101 . Google Scholar CrossRef Search ADS Schnyder G. ( 2012 ) ‘ Like a Phoenix from the Ashes? Reassessing the Transformation of the Swedish Political Economy Since the 1970s’, Journal of European Public Policy , 19 , 1126 – 1145 . Google Scholar CrossRef Search ADS Sorge A. ( 1991 ) ‘ Strategic Fit and the Societal Effect: Interpreting Cross-National Comparisons of Technology, Organization and Human Resources’, Organization Studies , 12 , 161 – 190 . Google Scholar CrossRef Search ADS Sorge A. , Streeck W. ( 1988 ) ‘New Technology and Industrial Relations: The Case for an Extended Perspective’. In Hyman R. , Streeck W. (eds) New Technology and Industrial Relations , Oxford , Basil Blackwell , pp. 19 – 47 . Soskice D. ( 1999 ) ‘Divergent Production Regimes: Coordinated and Uncoordinated Market Economies in the 1980s and 1990s’. In Kitschelt H. , Lange P. , Marks G. , Stephens J. D. (eds) Continuity and Change in Contemporary Capitalism , New York , Cambridge University Press , pp. 101 – 134 . Google Scholar CrossRef Search ADS Streeck W. ( 1991 ) ‘On the Institutional Conditions of Diversified Quality Production’. In Matzner E. , Streeck W. (eds) Beyond Keynesianism. The Socio-Economics of Production and Full Employment , Aldershot , Edward Elgar , pp. 21 – 61 . Thelen K. ( 2012 ) ‘ Varieties of Capitalism: Trajectories of Liberalization and the New Politics of Social Solidarity’, Annual Review of Political Science , 15 , 137 – 159 . Google Scholar CrossRef Search ADS Wall T. D. , Wood S. J. , Leach D. J. ( 2004 ) ‘Empowerment and Performance’. In Cooper C. L. , Robertson I. (eds) International Review of Industrial and Organizational Psychology , Chichester , John Wiley , pp. 1 – 46 . Whitley R. ( 1999 ) Divergent Capitalisms: The Social Structuring and Change of Business Systems , Oxford , Oxford University Press . Whitley R. ( 2003 ) ‘ The Institutional Structuring of Organizational Capabilities: The Role of Authority Sharing and Organizational Careers’, Organization Studies , 24 , 667 – 695 . Google Scholar CrossRef Search ADS Wood G. , Brewster C. , Brookes M. ( 2014 ) ‘Institutions and Firm Level Hrm Practice’. In Wood G. , Brewster C. , Brookes M. (eds) Human Resource Management and the Institutional Perspective , New York, London , Routledge , pp. 1 – 14 . Wood S. , van Veldhoven M. , Croon M. , de Menezes L. M. ( 2012 ) ‘ Enriched Job Design, High Involvement Management and Organizational Performance: The Mediating Roles of Job Satisfaction and Well-Being’ , Human Relations , 65 , 419 – 445 . Google Scholar CrossRef Search ADS Appendix Table A1. Distribution of control variables by country (percentages and means) DE UK SE Min. Max. Control variables Sex: male 54 52 54 0 1 Age: years 44.29 42.16 44.34 15 84 Weekly working hours 32.18 35.03 37.57 1 84 Industry: manufacturing 22 13 15 0 1 Industry: wholesale and retail trade; repair of … 20 15 10 0 1 Industry: transportation and storage 8 7 6 0 1 Industry: accommodation and food service … 5 5 3 0 1 Industry: information and communication 3 4 5 0 1 Industry: financial and insurance activities 3 4 3 0 1 Industry: real estate activities 1 1 1 0 1 Industry: professional, scientific and technical … 5 4 4 0 1 Industry: administrative and support service … 8 6 6 0 1 Industry: education 4 16 16 0 1 Industry: human health and social work … 17 21 24 0 1 Industry: arts, entertainment and recreation 2 2 3 0 1 Industry: other service activities 3 2 3 0 1 Size of establishment 3.09 3.48 3.45 1 4 Occupation: managers (for comparison) 1 15 7 0 1 Occupation: professionals 12 22 32 0 1 Occupation: technicians and associate professionals 14 10 18 0 1 Occupation: clerical support workers 15 9 6 0 1 Occupation: service and sales workers 27 25 23 0 1 Occupation: craft and related trades workers 9 4 4 0 1 Occupation: plant and machine operators … 10 7 6 0 1 Occupation: elementary occupations 11 8 5 0 1 DE UK SE Min. Max. Control variables Sex: male 54 52 54 0 1 Age: years 44.29 42.16 44.34 15 84 Weekly working hours 32.18 35.03 37.57 1 84 Industry: manufacturing 22 13 15 0 1 Industry: wholesale and retail trade; repair of … 20 15 10 0 1 Industry: transportation and storage 8 7 6 0 1 Industry: accommodation and food service … 5 5 3 0 1 Industry: information and communication 3 4 5 0 1 Industry: financial and insurance activities 3 4 3 0 1 Industry: real estate activities 1 1 1 0 1 Industry: professional, scientific and technical … 5 4 4 0 1 Industry: administrative and support service … 8 6 6 0 1 Industry: education 4 16 16 0 1 Industry: human health and social work … 17 21 24 0 1 Industry: arts, entertainment and recreation 2 2 3 0 1 Industry: other service activities 3 2 3 0 1 Size of establishment 3.09 3.48 3.45 1 4 Occupation: managers (for comparison) 1 15 7 0 1 Occupation: professionals 12 22 32 0 1 Occupation: technicians and associate professionals 14 10 18 0 1 Occupation: clerical support workers 15 9 6 0 1 Occupation: service and sales workers 27 25 23 0 1 Occupation: craft and related trades workers 9 4 4 0 1 Occupation: plant and machine operators … 10 7 6 0 1 Occupation: elementary occupations 11 8 5 0 1 Notes: Figures of dummy variables in % of yes answers; for all other variables means are depicted. Values are represented as percentage, unless otherwise mentioned. Source: EWCS 2015, own calculations. View Large Table A1. Distribution of control variables by country (percentages and means) DE UK SE Min. Max. Control variables Sex: male 54 52 54 0 1 Age: years 44.29 42.16 44.34 15 84 Weekly working hours 32.18 35.03 37.57 1 84 Industry: manufacturing 22 13 15 0 1 Industry: wholesale and retail trade; repair of … 20 15 10 0 1 Industry: transportation and storage 8 7 6 0 1 Industry: accommodation and food service … 5 5 3 0 1 Industry: information and communication 3 4 5 0 1 Industry: financial and insurance activities 3 4 3 0 1 Industry: real estate activities 1 1 1 0 1 Industry: professional, scientific and technical … 5 4 4 0 1 Industry: administrative and support service … 8 6 6 0 1 Industry: education 4 16 16 0 1 Industry: human health and social work … 17 21 24 0 1 Industry: arts, entertainment and recreation 2 2 3 0 1 Industry: other service activities 3 2 3 0 1 Size of establishment 3.09 3.48 3.45 1 4 Occupation: managers (for comparison) 1 15 7 0 1 Occupation: professionals 12 22 32 0 1 Occupation: technicians and associate professionals 14 10 18 0 1 Occupation: clerical support workers 15 9 6 0 1 Occupation: service and sales workers 27 25 23 0 1 Occupation: craft and related trades workers 9 4 4 0 1 Occupation: plant and machine operators … 10 7 6 0 1 Occupation: elementary occupations 11 8 5 0 1 DE UK SE Min. Max. Control variables Sex: male 54 52 54 0 1 Age: years 44.29 42.16 44.34 15 84 Weekly working hours 32.18 35.03 37.57 1 84 Industry: manufacturing 22 13 15 0 1 Industry: wholesale and retail trade; repair of … 20 15 10 0 1 Industry: transportation and storage 8 7 6 0 1 Industry: accommodation and food service … 5 5 3 0 1 Industry: information and communication 3 4 5 0 1 Industry: financial and insurance activities 3 4 3 0 1 Industry: real estate activities 1 1 1 0 1 Industry: professional, scientific and technical … 5 4 4 0 1 Industry: administrative and support service … 8 6 6 0 1 Industry: education 4 16 16 0 1 Industry: human health and social work … 17 21 24 0 1 Industry: arts, entertainment and recreation 2 2 3 0 1 Industry: other service activities 3 2 3 0 1 Size of establishment 3.09 3.48 3.45 1 4 Occupation: managers (for comparison) 1 15 7 0 1 Occupation: professionals 12 22 32 0 1 Occupation: technicians and associate professionals 14 10 18 0 1 Occupation: clerical support workers 15 9 6 0 1 Occupation: service and sales workers 27 25 23 0 1 Occupation: craft and related trades workers 9 4 4 0 1 Occupation: plant and machine operators … 10 7 6 0 1 Occupation: elementary occupations 11 8 5 0 1 Notes: Figures of dummy variables in % of yes answers; for all other variables means are depicted. Values are represented as percentage, unless otherwise mentioned. Source: EWCS 2015, own calculations. View Large © The Author 2017. Published by Oxford University Press and the Society for the Advancement of Socio-Economics. All rights reserved. For Permissions, please email: journals.permissions@oup.com

Journal

Socio-Economic ReviewOxford University Press

Published: Dec 2, 2017

There are no references for this article.

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create lists to
organize your research

Export lists, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off