Democracy Economic and Industrial

Economic and Industrial Democracy http://eid.sagepub.com/

Employee attitudes to the distribution of organizational influence: Who should have the most influence on which issues? Hans Jeppe Jeppesen, Thomas Jønsson and Mark Shevlin Economic and Industrial Democracy 2011 32: 69 originally published online 2 August 2010 DOI: 10.1177/0143831X10372432 The online version of this article can be found at: http://eid.sagepub.com/content/32/1/69

Published by: http://www.sagepublications.com

On behalf of: Department of Economic History, Uppsala University, Sweden

Additional services and information for Economic and Industrial Democracy can be found at: Email Alerts: http://eid.sagepub.com/cgi/alerts Subscriptions: http://eid.sagepub.com/subscriptions Reprints: http://www.sagepub.com/journalsReprints.nav Permissions: http://www.sagepub.com/journalsPermissions.nav Citations: http://eid.sagepub.com/content/32/1/69.refs.html

>> Version of Record - Jan 19, 2011 OnlineFirst Version of Record - Aug 10, 2010 OnlineFirst Version of Record - Aug 2, 2010 What is This? Downloaded from eid.sagepub.com at Moshi University on October 11, 2013

Article

Employee attitudes to the distribution of organizational influence: Who should have the most influence on which issues?

Economic and Industrial Democracy 32(1) 69–86 © The Author(s) 2011 Reprints and permission: sagepub. co.uk/journalsPermissions.nav DOI: 10.1177/0143831X10372432 eid.sagepub.com

Hans Jeppe Jeppesen and Thomas Jønsson Aarhus University, Denmark

Mark Shevlin

University of Ulster, UK

Abstract This article investigates the relationship between how individuals actually experience influence in organizational issues and which actors in the organizational structure are preferred to have the most influence over different issues. In this way the article also highlights links between different forms of direct and indirect influence. Four different Danish companies applying teamwork, varying in size and type, participated in the study. The total sample size included 526 employees involved in teamwork who returned the questionnaire (response rate of 72 percent). The results indicated that employees would like the actors who already have responsibilities for a given area to control that area. The results of the regression analyses show significant associations between the degree to which influence was experienced and desires for which organizational actors should have most influence. Furthermore, employees who experience the highest degree of influence within a domain want similarly placed colleagues to have high levels of influence.The results emphasize the significance of considering the level of employee influence in understanding employee attitudes to the distribution of organizational influence and leadership.

Keywords employee participation, employee representation, internal democracy, psychology, work teams

Different themes, ideas and questions concerning employee influence and involvement in organizational decision-making and leadership tasks at different structural levels in organizations have been a focus for a range of initiatives and policies. In their wake, new

Corresponding author: Hans Jeppe Jeppesen, Aaarhus University, Jens Chr. Skous Vej 4, 8000 Aarhus C, Denmark. Email: [email protected]

70

Economic and Industrial Democracy 32(1)

forms of leadership, different forms of work organization, organizational structures and labour market regulations have emerged. Furthermore, participation programmes have been developed with the aim of stimulating organizational productivity and efficiency via better application of employees’ resources and improving the quality of employees’ working life. Examples of this are the Danish Agreement on Cooperation between the Organization of Employers and the Federation of Unions and the German Legislation on Works Councils. Such initiatives may also arise from policies aimed at improving organizational democracy and employees’ social rights (e.g. European Commission, 1990), and improved conditions for human development. One common feature of participative activities is that, to some degree, they change the conditions for interaction and relations between employees and management. A central theme in the debate about the introduction of organizational participation programmes has been the employees’ attitudes to influence, and the controversial question of how much influence employees actually desire. An early study by Hespe and Wall (1976) found that those employees who felt more involved were also more satisfied. A cross-national study in 12 European countries (IDE, 1981) found that direct influence had a strong correlation with desired influence and that the strongest desire for influence was related to organizational domains closest to the employees’ daily work operations. From the 1980s onwards management tended to initiate forms of organizational influence based on human resource management (HRM) ideas, along with schemes for financial participation. Simultaneously, participatory legal or agreement-driven regulatory bodies such as Works Councils, Cooperation Committees and Health and Safety Committees were developed and encompassed new organizational areas. The Industrial Democracy in Europe study (IDE, 1981) was replicated in the early 1990s, but the results were largely unchanged. A later study by Kahnweiler and Thompson (2000) noted the importance of more research on the desired versus actual levels of influence in a variety of areas, and they questioned the assumption that employees per se want to be more involved. Over the last decades teamwork has increased, followed by aims to establish flatter organizational structures. Teamwork holds the potential for increased employee influence concerning issues connected to work performance and organization of work tasks and contributes to experiencing increased influence in interaction with fellow workers depending on the kind of team and teamwork (Weber, 2000). Participating in teamwork may thus affect employee attitudes to organizational influence and its distribution. As Gollan et al. (2006) state, jobs are designed to be broader and combine planning and implementation, and teams rather than individuals are the organizational units accountable for performance. Thus the widespread use of teams in organizing work performance makes teamwork an additional agent in studying employee influence.

Background Studies have indicated (Dawkins and Frass, 2005; Gollan et al., 2006; IDE, 1981) that the function of the different participatory programmes, and the attitudes of the employees involved in these programmes, cannot be understood in isolation of an awareness and knowledge of the organizational context and the labour market traditions and culture where they take place. The Danish labour market follows the Scandinavian model and

Jeppesen et al.

71

has (1) well-established and extended organizations for employers and employees; (2) widespread cooperation between the employers and trade unions which is actively supported by the state; and (3) a comprehensive system of conflict resolution and extended local participatory systems. A formal system of organizational participatory joint committees is in place both for agreement issues like personnel policies, organizational work conditions and technology development (Cooperation Committees) and for health and safety legislation (Health and Safety Committees). The history of the Cooperation Committees dates back to 1947 and the Health and Safety Committees date back to 1976. Furthermore, it should be noted that Denmark has a union density of between 70 and 80 percent. The IDE (1981) group investigated how much influence employees desired to have on a number of issues at the proximal (work performance and organization), medium (employment and health and safety policies) and distal (economic and development strategies) organizational levels. With regard to proximal issues, the employees wanted ‘to give their opinion and have their opinion taken into account’; on medium issues, ‘being informed and being able to give their opinion’ was desired; and on distal issues, the employees had a desire to ‘be informed’. In summary, employees wanted to be involved but not in control with regard to medium issues and just kept informed about distal issues. They wanted the representative bodies to have greater influence than the employees themselves. Hespe and Wall (1976) compared the level of desired influence of employees in different organizations and concluded that ‘only an extremely small minority of individuals feel that “workers” should have greater influence over medium and distant decisions than management’ (Hespe and Wall, 1976: 426). Similarly, Liverpool (1990) reported that employees’ desired level of influence for medium and distant issues was between no and a little say. Delbridge and Whitfield (2001) reported that the employees operating in organizations with participation schemes experienced greater influence over their jobs, particularly in how they performed their duties. Kahnweiler and Thompson (2000) found in their study of employees in 60 US organizations that age had a positive effect on desire for influence, whereas gender had no effect. They also reported that a higher level of education was an indicator of wanting to be more involved in a variety of job and organizational issues. Based on a manager survey in 10 European Union countries, Poutsma et al. (2003) argued that companies that implemented direct participation tended to be more innovative and have more highly qualified workers, whereas representative participation is driven by cultural climate and institutionalized employee influence. Dawkins and Frass (2005) argued that it was important that management supported employee involvement programmes in order for them to be successful. In particular, it was noted that employees should be provided with sufficient documentation explaining to them that they will receive the necessary training and operational support. Concerning the reasons behind the development of organizing work in teams, Zwick (2004) found, looking at a three-year period from 1997 to 2000, that the introduction of teamwork and autonomous work groups and a reduction of hierarchies significantly increased average productivity. Delarue et al. (2008) reviewed the literature on team design, team development and team performance and found that teamwork is likely to have a positive impact on operational performance. Similar to Zwick (2004), they state

72


that teamwork had a positive effect on financial outcome. In relation to employee outcome, they stated that most studies have reported positive implications for workers, such as increased job satisfaction, empowerment and task involvement. In their review on psychological factors in teamwork, Rasmussen and Jeppesen (2006) stated that job satisfaction and intra-team support was associated with team-related attitudes, and also that types of psychological variables were related to team factors such as autonomy and interdependence, and contextual support variables. This indicated that team autonomy may be important for the psychological variables associated with teamwork. Mueller et al. (2000) emphasized that key variables in teamwork and its effects were autonomy and interdependence. In this way it is shown that teamwork potentially contains conditions that may offer fertile soil for new approaches to arise about the interaction of the individual and the group. The dimension of team autonomy can in this way be seen as an employee participation arrangement. Theories and research on agency, autonomy and empowerment may increase understanding of the importance of having these needs fulfilled and the wider potential interactions between different forms of participation. Greasley et al. (2005) stated that the more autonomy in initiation and continuation of work behaviour and processes, and the more the employee can influence certain outcomes at work, the more the sense of empowerment. As such it is not only the single employee that can be empowered but also representative bodies (depending on their function). The study emphasizes that competences are necessary for empowerment to occur. Furthermore, it argues that empowerment is not a fixed reality but rather something that varies from individual to individual. Bandura’s (2001) social cognitive theory focuses on the individual’s beliefs about his or her amount of influence and ability to interact efficiently with the social environment and achieve desired goals, i.e. self-efficacy. Self-efficacy is the central tenet in personal agency. The theory also identifies beliefs about other, socially mediated modes of agency, namely collective and proxy agency. Bandura (2001) argues that in many spheres of functioning, people do not have control over the conditions or institutional practices that affect their everyday lives, and under these circumstances they seek well-being, security and valued outcomes through other agents. In these socially mediated modes of agency people look to those who have access to resources or expertise or who wield influence and power to act in their best interest to secure the outcome they desire. In this way, a person’s self-efficacy is extended through beliefs about the efficacy of a collective or a proxy agent. On realizing his or her own limitations of efficacy, the individual may attempt to join with other persons with the purpose of executing a joint action, i.e. a collective agency strategy. Alternatively, the person may turn to another person, perceived to be willing and able to safeguard his or her interests and to achieve the desired goals, i.e. a proxy agency strategy (Bandura, 2001). Social cognitive theory may be applied as a framework for understanding employees’ agential behaviour in terms of personal, proxy and collective agency in different organizational areas. However, little is known about who, among the many possibilities in an organization, is desired to be an agent and in which areas, and what variables are associated with whom one may desire to be one’s agent. Moreover, Bandura’s (2000, 2001) theory does not account for the fact that the socially mediated agents may be closer to or further away from the individual, for example

Jeppesen et al.

73

that an employee is more often and more directly interacting with the team than with the employee representative on the company board. This study is timely as the tasks and areas for employee participation at local levels have expanded since the aforementioned earlier studies in the area. Furthermore, representative functions have become more important for and more visible to employees. Together with the increase in the use of teams as a way of organizing performance and work tasks, this presents new conditions for the interplay between the representative and direct form of organizational participation and between the employee and the group. Another point of interest is to gather information about the desired distribution of influence in organizations, i.e. who employees would prefer to have most influence in particular areas. It is proposed that such knowledge will provide a more detailed understanding of intrinsic work motivation, which in turn yields understanding of the processes that create individual involvement and desires for influence as a state of autonomy (Ryan and Deci, 2002). Specifically, this study aims to assess employees’ desire for influence and to investigate the following research questions: RQ1: Do employees want the organizational agents who have responsibilities in a certain domain to have the most influence over that domain? RQ2: Will the degree of influence that an employee desires for an agent be consistent with the amount of influence they themselves experience at a given organizational level?

Methods Participants The data for this study came from four Danish companies that use teamwork in their organizational structure. They were selected to provide variability in terms of size and organizational type in order to get a broad description of the role of autonomous teams as a possible way of enhancing employee participation and distribution of power and leadership. The four companies in the sample were as follows. ‘Company A’ was a centre for labour market training in a variety of trades combining theory and practice. The teacher trainers were organized in both permanent and project teams. The company had regulatory based participation programmes. At this centre 66 out of 89 employees filled out the questionnaire (response rate of 74 percent). ‘Company B’ was part of a large multinational organization within the telecommunications industry. The organization’s primary task was to develop software guidance-systems for 3rd generation cell-phones and employees were generally highly educated with graduate and postgraduate qualifications. The teams in the organization were all project teams, of differing size, tasks, composition and duration (from three months to two years). There were between eight and 12 functioning teams at any given time. The rest of the company was structured as a matrixorganization with few stable structures and many ad hoc arrangements. The company had a health and safety committee that was established according to legislation but was informal and infrequent in practice. There was no cooperation committee. Seventy percent of

74


the employees (53 out of 76) completed the questionnaire. ‘Company C’ was a department in the Danish postal service. The work consisted of receiving, sorting and delivering all postal items. The work was organized in permanent teams but there was often cover by temporary workers. The organizational structure had participatory structures stipulated by regulations and typical of a traditional public organization. The employees have varying levels of education, and 73 of 87 employees filled out the questionnaire (84 percent). ‘Company D’ was a production company that makes fitted kitchens on assembly lines. The production was continuous and organized into day, evening and night shifts. Work was organized in permanent teams and this system had existed in the company since 1997. The company had a traditional hierarchical structure. There was a high union membership among the employees in production, who comprised both unskilled and skilled workers. There was a tradition of strong employee–management cooperation through the 10 shop-stewards and the company had the regulatory based participation programmes. Our questionnaire was completed by 334 of the 480 employees in production (70 percent). In total, for the whole sample, 526 out of 732 filled out the questionnaire – a response rate of 72 percent.

Measurement Respondents were asked if they had either management (0 = no, 1 = yes) or employee representative responsibilities (0 = no, 1 = yes). Educational level was measured as ‘unskilled’, ‘skilled’, ‘middle range education’, ‘academic’ and a residual category, ‘other’ (see Table 1). Educational level was then dummy coded (1 = having the particular educational level, 0 = not having the education). The respondents also stated their gender, age, seniority in company and seniority in the job function in years. Experienced influence was measured by asking: ‘How much influence do you experience that you have on . . . ?’ The issues close to the employee’s environment (e.g. how to perform the daily work, organizing the daily work and the working time) constituted the variables that express experienced influence on proximal-level issues (Cronbach’s alpha = .82). Influence on decisions about employment policies and health and safety constitute medium-level issues (Cronbach’s alpha = .73), and influence on financial decision-making and company strategies and plans constitute distal-level issues (Cronbach’s alpha = .80). The responses were scored on a scale from 1 (none) to 5 (very much). Desires for influence were measured by asking: ‘Whom do you want to have most influence on . . . ?’ The statement was then followed by the same organizational Table 1. Distribution of education in the companies (Ns) Unskilled Skilled Middle-range education Academic education Other

Company A 1 22 28 7 8

Company B 1 0 12 41 0

Company C

Company D

29 21 9 6 6

145 140 16 0 21

Jeppesen et al.

75

domains and issues used for the assessment of ‘experienced influence’ and the reply categories were: (1) ‘yourself’, (2) ‘your team or work group’, (3) ‘employees in common’, (4) ‘employee representatives, (5) ‘works council or health and safety committees’, or (6) ‘management’. These categories correspond with actual actors in Danish organizations. The response scale was ordered according to how close an agent formally was to the employee. The answers expressed the respondents’ desires about how close the agent in control of the issue should be to the respondent. The closest form of control is the desire to control the issue oneself. The team/work group and employees in common refer to agents where the respondent him or herself is a member of the agent in control. At the mid points of the scale, the employee wishes colleagues and/or representatives to control the issue. The employee is not necessarily a member him or herself but might influence the composition of the agent. Both employee representatives and management participate in the works council and health and safety committee, though management has the final say in case of parity of votes. At the other end of the scale, the desire for management to control the particular issue expresses a wish for no employee control over the issue. Thus, the scale ranges from 1 to 6, and the lower the score, the closer the agent desired to be in control.

Analytic strategy Based on prior research, specific decision issues can be grouped into proximal, medium and distal issues (e.g. IDE, 1993). To test if these three subdimensions applied to the measures of desired influence, two confirmatory factor models were specified and estimated using LISREL 8.72 (Jöreskog and Sörbom, 2005). The first model was a onefactor model that specified all the indicators of the desired influence scale to load onto one factor. The second model comprised three factors. The first factor was labelled ‘desired proximal influence’ and was measured by three items. The second and third factors were labelled ‘desired medium influence’ and ‘desired distal influence’, and each was measured by two items. The seven indicators of the desired influence scale are ordinal in nature. The most appropriate method of analysis is based on a matrix of polychoric correlations estimated using Diagonally Weighted Least Squares (DWLS) along with a weight matrix of asymptotic variances (Jöreskog, 1990, 1994; Jöreskog and Sörbom, 2006). All factors were allowed to correlate and no correlated errors were included in the model. Following the guidelines suggested by Hoyle and Panter (1995) the goodness of fit for each model was assessed using a range of fit indices (acceptance criteria in parentheses); Satorra–Bentler scaled chi-square (S-Bc2, p > .05), Incremental Fit Index (IFI, > .95), Comparative Fit Index (CFI, > .95), Root Mean Square Error of Approximation (RMSEA, < .08) and Standardized Root Mean Square Residual (SRMR, < .08). The comparative fit of the models was assessed using the Expected Cross Validation Index (ECVI), with the smallest value being indicative of the best fitting model. Factor scores from the above factor analysis were applied as continuous variables that express the degree of desired closeness of control on proximal, medium range and distal issues. Variables such as gender, educational level, attending to management tasks and attending to employee representative tasks were coded as binary variables and included

76


in the regression analyses. The four companies differed in terms of their size and the organizational type. The aim of the study was to estimate the association between variables while controlling for differences in these companies. Hierarchical multiple regression was an appropriate method of analysis as it allowed the influence of variables to be estimated while controlling, or holding constant, any effect attributable to company type. To control for these differences dummy variables representing the companies were entered in the first block of a hierarchical multiple regression. The regression coefficients for variables in subsequent blocks of the model can be interpreted in terms of their unique association with the dependent variable, holding constant the variables in the first block.

Results Confirmatory factor analysis The fit indices from the confirmatory factor analyses are reported in Table 2 and suggested that the three-factor model was an acceptable model, and better than the onefactor model. On the basis of meeting the criteria associated with the RMSEA, the IFI, the CFI and the SRMR Model 2 was judged to exhibit reasonable model fit, and Model 1 failed to meet any of the criteria. Although the chi-square for Model 2 was large relative to the degrees of freedom, and statistically significant, this should not lead to the rejection of the model as the large sample size increases the power of the test (Tanaka, 1987). The increased power of the chi-square test can result in models with no serious misspecification being rejected as minor discrepancies between the sample and implied covariance matrix are detected. Model 2 also had a lower ECVI value compared to Model 1. On the basis of this, Model 2 was judged to be an acceptable description of the data, and better than Model 1. The standardized factor loadings and factor correlations are presented in Table 3. Table 2. Fit indices for the alternative models of the desired influence scale Item

Model 1 One-factor model

Model 2 Three-factor model

S-Bχ2 515.50 22.23 d.f. (p) 14 (.00) 11 (.02) RMSEA .26 .044 90% CI (.24-.28) (.016-.07) ECVI 1.02 .11 90% CI (.89-1.17) (.08-.14) IFI .65 .99 CFI .64 .99 SRMR .22 .045 Notes: S-Bχ2 = Satorra–Bentler scaled chi-square; RMSEA = Root Mean Square Error of Approximation; CI = confidence intervals; ECVI = Expected Cross Validation Index; IFI = Incremental Fit Index; CFI = Comparative Fit Index; SRMR = Standardized Root Mean Residual.

77

Jeppesen et al. Table 3. Factor loadings for three-factor model of the desired influence scale ‘Whom do you want to have the most influence on . . .’

Desired proximal influence

Desired medium influence

How the daily work is performed? .86* How the daily work tasks .92* are organized? How working time is organized .59* and scheduled? The employment policies of .66* the organization? How health and safety is managed? .67* The financial decision-making by the organization? Plans and strategies for the development of the organization? Correlation 1.00 Desired medium influence .14 1.00 Desired distant influence -.02 .68*

Desired distal influence

.98* .78*

1.00

*p < .05.

Descriptive results Table 4 shows the distribution of answers, and it is evident that employees desire to control performance and organization of work by themselves personally or through their team. The majority (86 percent) desired either the team or the single employee to control work performance; while 84 percent desired to organize work and 62 percent desired to control working time, either personally or through the team. With regard to the mediumlevel issues, the majority desired that these issues should be controlled by employees in common, via representatives, or together with managers in the committees; 72 percent of the respondents desired that these agents should be in control of employment policies and 80 percent desired that these agents should be in control of health and safety matters. It was also desired by a majority of respondents that management should control financial decisions (69 percent) and plans and strategies for the development of the organization (63 percent). Table 5 show the means, standard deviations and zero-order correlations between the variables. The means of the factor scores of the desired influence variables are centred around zero and the desired influence variables are raw scores with answers on a scale from 1 to 5.

Results of analyses In order to test for associations, regression analyses were conducted that controlled for company membership (by using dummy variables in the first block of a hierarchical regression model). The second block of the model included the variables representing age, seniority in company, present job function and education.

a. How the daily work 28% 58% is performed? b. How the daily work tasks 17% 67% are organized? c. How working time is 14% 48% organized and scheduled? d. The employment policies 1% 11% of the organization? e. How health and safety 2% 11% is managed? f. The financial decision-making 8% by the organization? g. Plans and strategies for the 7% development of the organization?

Cooperation comm./ health and safety committees

Management

63%

11%

13% 7%

41% 7%

16%

69%

12%

27%

13%

10% 7% 6%

22%

37%

22% 6% 2% 8%

13% 1% 2%

12% 1% 1%

Whom do you want to have most Yourself Your team Employees Employee influence on . . . (put only one cross or your in common representatives in each line) work group

Table 4. Description of whom employees desire should have most control over different issues

78 Economic and Industrial Democracy 32(1)

79

Jeppesen et al. Table 5. Means, standard deviations and Pearson’s correlations between the variables

Mean

SD

1.

2.

3.

4.

5.

1. Desired, proximal 2. Desired, medium 3. Desired, distal 4. Experienced, proximal 5. Experienced, medium 6. Experienced, distal

.02 -.05 -.11 3.42 2.38 1.57

.86 .83 1.01 1.01 .94 .82

.28*** .03 -.45*** .04 -.02

.84*** -.17*** -.14*** -.08

-.06 -.11** -.13***

.38*** .29***

.59***

Notes: Pearson’s correlation coefficients. *p < .05; **p < .01; ***p < .001. Since lower scores on the desired influence scale express desires to have most influence closer to oneself, negative coefficients mean that the higher the negative score on the dependent variable, the closer to oneself the desire of most influence.

Seniority in company was significantly related to desired influence on medium and distal areas but not to desired proximal control. Neither age nor seniority in job function significantly predicted desired influence. Educational levels were not significantly connected to desired proximal influence. Education did not add significantly to the variance of desired proximal influence (R2 change = .01, NS), though a tendency for more highly educated employees to desire closer control with proximal issues was identified (b = -.13, p = .06). Educational level was not significantly associated with desired medium influence either (R2 change = .01, NS). Education did explain additional variance on desired distal influence (R2 change = .02, p < .05). In particular, the results show that ‘academics’ desired to have most influence on distal issues further away from themselves, i.e. more management control over financial and strategic decisions, than employees in the ‘unskilled’, ‘skilled’ and ‘middle range’ education groups (see Table 6). Managerial responsibility was not significantly linked to desired closeness of influence for any area; desired proximal control: b = .01, NS; desired medium control: b = .00, NS; desired distal control: b = .02, NS. Employees responsible for employee representative tasks desired to control medium and distal issues more closely than employees without such duties (desired medium influence: b = -.12, p < .01; desired distal influence: b = -.11, p < .05). This was not found for desired proximal influence (b = -.03, NS). All regression models were controlled for company membership.

Table 6. Regression analyses: relationships between education and desired influence

Desired proximal influence

Desired medium influence

Desired distal influence

Company A Company B Company D Skilled workers Medium education Academics Other education

-.28*** -.09 -.22** -.03 -.05 -.13 .03

-.16** .11 .11 .06 .04 .08 .08

-.10 .17* .19** .07 .01 .15* .10*

*p < .05; **p < .01; ***p < .001.

80


In order to test if higher experienced influence is positively associated with desires to have the control closer to oneself, regression analyses were applied with the factor scores of the desired influence dimensions and the experienced influence dimensions. Table 7 shows the analyses of experienced influence and desired closeness of control. These analyses showed that higher experienced influence on proximal issues was associated with desires to control these issues more closely. Pertaining to the medium-level issues, the experienced influence was associated with desired agent in control in a similar fashion as for proximal issues. Finally, for the distal issues, higher experienced influence was associated with desires for closer control with these issues. In sum, the results support the thesis that higher experienced influence was associated with desires of control by an agent closer to oneself within the same range of issues. It is worth noticing that on the mediumlevel issues and distal issues, the agents closer to oneself tend to be employees in common or representatives, and not oneself. Thus, the results do not show that higher influence experienced by oneself is associated with desires to control any issue on one’s own.

Discussion The distribution of scores on the desired influence measure showed that employees desired that the agents who already have responsibilities for a given area should continue to control that area. Employees desire that they should have most influence directly on proximal work areas, either individually or via teams and collectives. The employees desire that the collective or representatives, together with management in joint committees, should have most influence on the medium-range issues (employment policies and health and safety). Finally, management should have the most influence on distal issues (finance and strategy). The results of the regression analyses showed that, when we controlled for company membership, gender, age, seniority in job function, management tasks were not related to desires of influence on proximal issues. More highly educated employees desired distal issues to be controlled by actors who were more distant to themselves (i.e. more management control over finance and strategy) than employees with other educational backgrounds. The higher the seniority in the company, the closer to themselves the Table 7. Regression analyses of experienced influence (independent variables) and desired influence (dependent variables) controlled for company membership Company A Company B Company C Experienced influence proximal Experienced influence medium Experienced influence distal

Desired control, proximal issues -.15*** .02 -.18*** -.38***

Desired control, medium issues

Desired control, distal issues

-.14* .01 .08

-.21*** .15** -.12**

-.13**

-.15**

Notes: All numbers are beta values. *p < .05; **p < .01; ***p < .001. On negative coefficients see footnote to Table 5.

Jeppesen et al.

81

employees desire the control of medium and distal areas to be. Employee representatives desired medium and distal issues to be controlled more closely to themselves than nonrepresentative employees did. The most coherent result was that the more influence experienced, the stronger the desire for agents who are closer to have control within the particular decision area. In his discussion on human agency, Bandura (2001) proposed that social cognitive theory distinguished three modes of human agency: personal, proxy and collective. In the interpretation of our findings in relation to these modes of human agency, the results have supported two extensions to these notions: first, the finding that the organizational agents have different degrees of closeness to the individual employee and, second, that the desired closeness of agents is positively associated with the degree of influence experienced by the employee. If we look at the results from the approach of social cognitive theory, they show in general that the employees prefer that those agents who perform the particular tasks as their daily job should also have most influence in or be in control of the issues at that level. These results may indicate that both influence and competence with regard to an issue are considered in employees’ desires of agency. More precisely, the results found that the employees desired that they themselves or their team should control the work issues (proximal issues). At the level of medium issues, a different group of desired agents can be identified, i.e. employees in common, their representatives and the works council/health and safety committee, which both also include employee representatives together with managerial representatives. At the distal level, management tasks comprise the financial and strategic issues of the organization and the majority of employees desire that management should stay in control of these issues. In ranging the issues or domains as proximal, medium and distal, it also implies that the agents acting at those same levels are supposed to have the same degree of closeness to the individual employee. In self-efficacy terms, the majority of employees show a preference for the team to be in control of work-related issues, which may have different implications. One line of interpretation could be that the team offers the individual employee more autonomy than he or she would otherwise have and by that will also increase the employee’s opportunity and ability to exercise control and in turn increase ‘collective self-efficacy’. This is in line with the results from a survey of health care teams by Van Mierlo et al. (2006), where it was found that individual autonomy was related to the level of team autonomy but also to self-efficacy and social support. The teams thus provide their members with the possibility of more autonomy. It may also be the case that the employee considers that the team can act more effectively in his or her best interest – be that with or without the employee’s own involvement – and charges the team with the particular responsibilities, demands and efforts associated with work-related issues. Different interpretations and combinations are possible and constitute an agenda for future research. At the medium-level issues, different agents involving employees – i.e. employees in common, or employee representatives alone or together with management – are desired more as agents. While this may express the experience of the actual functioning in the organization, it also indicates a desire for closeness – these agents, though not as close as with one’s fellow team workers, include colleagues and employee representatives who are already dealing with employment and health and safety issues as part of their organizational role

82


via the established regulatory structures. In a cross-national study on shift work and prevention, Jeppesen et al. (2006) reported that the agent preferred by the employees as being best able to take care of health and safety issues was ‘employees in common’ irrespective of country (Italy, Norway, the Netherlands, Sweden and Germany) or, accordingly, national labour market regulations. Management as the preferred agent at the distal level may indicate that the majority of employees desire to have an agent here whom they can feel confident will be best to secure the continuance of the company, as they think of this as being in the common interest. This is probably strongly related and most sensitive to the actual functioning of management. In this way we find that our first thesis is supported. Furthermore, the results showed that those employees that experience highest influence within a domain desire agents closer to themselves than employees who experience lower influence (Table 4). These results are in line with previous findings in the IDE (1981, 1993) projects, but it could be argued that this study offers a supplementary picture of this as the employees have been asked about which agent they want to have the most influence instead of being asking about the desired degree of influence. Here, the employees do indeed state that at proximal and medium levels they want employees to have most influence either alone or together with management concerning health and safety issues. Experienced influence is linked to desires for control or most influence within the same level (see Table 4). The findings showed associations between high levels of experienced influence and desires to have an agent closer to oneself to control the issues, while this is not the case for employees with less experienced influence. To experience and exercise influence, and take part in participative programmes, demands the relevant competences and resources together with trust. Self-determination theory describes such conditions as those that offer autonomy, relatedness and competences along with an autonomous causality orientation (Deci and Ryan, 2000; Ryan and Deci, 2002). Heller (2003) states that before effective participation can take place employees must have the relevant experience and competences. He argues that the distribution of influence in an organization is the way power engages with human capital and the ideal type requires a support structure for competence utilization and motivation. He underlines that trust and competence are intrinsic to exercising influence effectively (Heller, 1998). Here the article presumes that experiencing influence in a given domain already implies the necessary competences to be able to exercise influence, so that experiencing influence is interconnected with competence. Following this line it can be argued from a social cognitive theory and self-efficacy approach that in order to be in direct control and to experience self-efficacy or ‘collective self-efficacy’ in teams (as this article proposes) the employee needs to have or to develop competences and from this experiences influence. This has implications for possible reciprocal dynamics at both organizational and individual levels. As Allport (1945) argued, writing about the personal level, when participating the individual becomes interested in shaping many of the events that control his or her life and also that participation sinks a shaft in the inner subjective regions of the personality. The results of this study demonstrate clearly that it is the degree of experienced influence that is connected to the desire for the most influential agent at the same level. Experienced influence is

Jeppesen et al.

83

also a precondition for employees taking an active part in utilizing the influence. Furthermore, it is also related to desires for agents at the next level. As such the article finds that the self-efficacy concept and the modes of human agency are useful in interpreting the results of experienced and desired influence at different organizational levels, but we still need more multilevel research on self-efficacy (Chen and Bliese, 2002) to improve our knowledge. The article also contributes to an increased understanding of the conditions needed for an interaction between direct and indirect participation, and as Strauss (2006) wrote, we need to pay more attention to the dynamics across the different forms and levels of participation. The results from the questions about perceived influence and desires for influence concerning medium issues show a slight increase in employee influence compared to the IDE (1981, 1993) studies and to Liverpool’s (1990) study, too. Some of the results cannot be directly compared as, contrary to the IDE studies, we also asked respondents about their desires for influence of the different organizational actors, but the findings may be an indication of a tendency in the development of employees’ organizational influence in such issues as health and safety and employment policies. It could be assumed that owing to the introduction of regulations concerning employee involvement in those issues they have become more institutionalized since the earlier studies and therefore it is more natural and obvious for employees to participate in these areas. Additionally, the results could be interpreted as a manifestation of the fact that when employees gradually become more and more familiar with having influence concerning issues that affect their daily working lives, they also want to apply these opportunities. The results also indicate that this may readily be in cooperation with management. It cannot be ignored, however, that the longer and stronger traditions in the ‘Scandinavian model’ for employee organizational participation also to a certain extent affect our results. In considering the validity of the results it can be argued that differences in the employees’ education and work tasks contribute to create the results when comparisons are made about the level of experienced influence and between the organizational actors at the different levels, but this is not really a substantial argument. The regression analyses controlled for effects of company, and as education level and company were merged the results will not be an effect of education. In the same way, job tasks and type of teams were merged with company, and thus controlled for. The results have their origin in the Scandinavian labour market, with its long tradition for cooperation at different organizational levels. It is a central belief of the present study that the participative practices affect both desired and experienced influence and the results are connected to this context. The emphasis on cooperation between management and labour and the practices associated with this system must be considered when generalizing to other types of labour market systems.

Conclusion The results have elucidated the research questions we advanced. First, employees want those organizational actors who already have responsibilities for a certain organizational domain to control that domain. Second, regression analyses show significant links between experienced influence and desires for which actor should have most

84


influence. Furthermore, the employees who experience high influence within a domain desire agents closer to themselves to be in control, than employees who experience low levels of influence. Interpreting the results from the approach of social cognitive theory about modes of human agency and the concepts of self- and collective efficacy has proven useful concerning our understanding of the interaction between different forms of organizational participation. Together with self-determination theory, these approaches provide a clearer picture of the necessary preconditions for being able to take an active part in participatory programmes. It may also be claimed that the results show a tendency to a general increase over the years in employees’ desire for control in the medium-level organizational domains, in issues such as employment policies and health and safety. References Allport GW (1945) The psychology of participation. Psychological Reviews 53: 117–132. Bandura A (2000) Exercise of human agency through collective efficacy. Current Directions in Psychological Science 3: 75–78. Bandura A (2001) Social cognitive theory: An agentic perspective. Annual Review Psychology 52: 1–26. Chen G, Bliese PD (2002) The role of different levels of leadership in predicting self- and collective efficacy: Evidence for discontinuity. Journal of Applied Psychology 87: 549–556. Dawkins CE, Frass JW (2005) Decision of union workers to participate in employee involvement: An application of the theory of planned behavior. Employee Relations 27: 511–531. Deci EL, Ryan RM (2000) The ‘what’ and ‘why’ of goal pursuits: Human needs and the selfdetermination of behavior. Psychological Inquiry 11: 227–268. Delarue A, Van Hootegem G, Procter S, and Burridge M (2008) Team working and organizational performance: A review of survey-based research. International Journal of Management Reviewers 10: 127–148. Delbridge R, Whitfield K (2001) Employee perception of job influence and organizational participation. Industrial Relations 40: 472–489. European Commission. (1990) Community Charter of the Fundamental Social Rights of Workers. Luxembourg: Office for Official Publications of the European Communities. Gollan PJ, Poutsma E, and Veersma U (2006) Editor’s introduction: New roads in organizational participation. Industrial Relations 45: 499–512. Greasley K, Bryman A, Dainty A, Price A, Soetanto R, and King N (2005) Employee perception of empowerment. Employee Relations 27: 354–368. Heller F (1998) Influence at work: A 25-year program of research. Human Relations 51: 1425–1456. Heller F (2003) Participation and power: A critical assessment. Applied Psychology: An International Review 52: 144–163. Hespe G, Wall T (1976) The demand for participation among employees. Human Relations 29: 411–428. Hoyle RH, Panter AT (1995) Writing about structural equation models. In: Hoyle RH (ed.) Structural Equation Modeling: Concepts, Issues and Applications. London: Sage, 158–198. IDE (1981) Industrial Democracy in Europe. Oxford: Clarendon Press. IDE (1993) Industrial Democracy in Europe Revisited. Oxford: Oxford University Press.

Jeppesen et al.

85

Jeppesen HJ, Kleiven M, Bøggild H, and Gill C (2006) Participation and Prevention when Organizing Shift Work at Company Level in Various European Countries. Aarhus: Aarhus University Press. Jöreskog K (1990) New developments in LISREL: Analysis of ordinal variables using polychoric correlations and weighted least squares. Quality and Quantity 24: 387–404. Jöreskog KG (1994) On the estimation of polychoric correlations and their asymptotic covariance matrix. Psychometrica 59: 381–389. Jöreskog KG, Sörbom D (2005) LISREL 8.72. Chicago, IL: Scientific Software Inc. Jöreskog KG, Sörbom D (2006) LISREL 8.80 for Windows. Lincolnwood, IL: Scientific Software International, Inc. Kahnweiler WM, Thompson MA (2000) Levels of desired, actual, and perceived control of employee involvement in decision making: An empirical investigation. Journal of Business and Psychology 14: 407–427. Liverpool PR (1990) Employee participation in decision-making: An analysis of the perceptions of members and nonmembers of quality circles. Journal of Business and Psychology 4: 411–422. Mueller F, Procter S, and Buchanan D (2000) Team working in its context(s): Antecedents, nature and dimensions. Human Relations 53: 1387–1423. Poutsma E, Hendrickx J, and Huigen F (2003) Employee participation in Europe: In search of the participative workplace. Economic and Industrial Democracy 24: 45–76. Rasmussen T, Jeppesen HJ (2006) Teamwork and associated psychological factors: A review. Work and Stress 20: 105–128. Ryan RM, Deci EL (2002) Overview of self-determination theory: An organismic dialectical perspective. In: Deci EL, Ryan RM (eds) Handbook of Self-Determination Research. New York: The University of Rochester Press, 3–33. Strauss G (2006) Worker participation: Some under-considered issues. Industrial Relations 45: 778–803. Tanaka JS (1987) How big is big enough? Sample size and goodness of fit in structural equation models with latent variables. Child Development 48: 134–146. Van Mierlo H, Rutte CG, Vermunt JK, Kompier MAJ, and Dorewaard MC (2006) Individual autonomy in work teams: The role of team autonomy, self-efficacy, and social support. European Journal of Work and Organizational Psychology 15: 281–299. Weber WG (2000) Organizational conditions fostering prosocial work orientations in teams. In: Vartiainen M, Avallone F, and Anderson N (eds) Innovative Theories, Tools, and Practices in Work and Organizational Psychology. Seattle, WA: Hogrefe and Huber, 75–96. Zwick T (2004) Employee participation and productivity. Labour Ergonomics 11: 715–740.

Hans Jeppe Jeppesen is a professor in work and organizational psychology at Aarhus University and head of the research unit Leadership and Influence in Organizations (LINOR). His research interests include organizational participation, distributed leadership, working time, employee engagement and teamwork. Besides topics in work and organizational psychology, he also teaches courses in field methods. Thomas Jønsson is assistant professor in work and organizational psychology at Aarhus University and member of the LINOR research unit. He is interested in research about

86


participation and shared influence in organizations, employee attitudes and social processes at work. He teaches courses about work and organizational psychology and research methods. Mark Shevlin is a professor of psychology at the University of Ulster and an honorary professor at the University of Aarhus. His research interests are in the areas of employee influence and organizational participation, clinical psychology and structural equation modelling. He has co-authored a popular book on statistics, Regression and Correlation: A Guide for Researchers and Students (Sage, 2000).