Abstract
The article compares the importance for the quality of work of three forms of direct participation – individual task discretion, semi-autonomous teamwork and consultative participation – drawing on a representative national survey of British employees. It assesses their implications for employee welfare, specifically their subjective importance, their implications for learning new skills and their effects on psychological well-being. It finds that individual task discretion is the most effective form of direct participation, followed by consultative participation. Although there are also positive effects of semi-autonomous teamwork, these are weaker and less consistently significant. The article finds that the strength of the implications of direct participation vary between employees, particularly by occupational class and the importance employees attach to the use of initiative at work. However, with the exception of those with weak initiative orientation, its effects are generally positive across different categories of employee.
Keywords
Introduction
There has been a striking consensus across quite diverse theoretical schools that employee participation is important for the quality of work. However this broad convergence conceals some important differences in how participation is conceptualized. Perspectives rooted in neo-Marxian theory have tended to stress the importance of employees’ direct control of their work task, while an important strand of management theory has emphasized the benefits of semi-autonomous teamwork and stronger consultative communications between management and employees. There has been considerable controversy in the literature about whether teamwork and consultative communication represent real or pseudo forms of participation. An empirical assessment of this requires comparison of the implications of different types of direct participation for employee welfare, but at present there is a marked dearth of relevant comparative evidence using good representative data.
Moreover, little is known about whether the implications of direct participation differ between different types of employee. There may be differences between employees in the relative importance attached to the ability to exercise influence over the work process. Men and women may have significantly different orientations to work, leading to gender differences in the effects of participative mechanisms. It is also possible that occupational classes differ in what is regarded as important in work, which again might affect the consequences of participation. Arguably, there may be a tension between management-initiated forms of direct participation and traditional forms of union regulation of work, so the effects of direct participation on employees will vary between employees depending on the strength of unions in their workplace.
This article seeks to advance understanding of these issues by drawing on a nationally representative sample of British employees, interviewed in 2006. Following the sociological tradition, the quality of work is taken to refer to the intrinsic quality of work, in contrast to the quality of employment conditions, which relates to extrinsic factors such as pay and security. It compares empirically the different forms of direct participation using three criteria of employee welfare: subjective importance, implications for skill use and skill development, and finally consequences for psychological well-being. It both looks at the general effects of the different types of participation and examines whether they vary between employees as a result of differences in sex, class, work values and union workplace strength.
Forms of direct participation: Debate and issues
The concept of participation has come to cover a very wide range of different practices for involving employees in decision making about work (Gonzalez, 2009; Heller et al., 1998; Kalleberg, 2011; Wilkinson et al., 2010; Zhou, 2009). Potentially it covers both a variety of forms of direct participation, in which employees are personally involved in influencing decisions about work, as well as diverse forms of indirect participation, in which their influence is mediated by representatives (Wilkinson et al., 2010). The focus here is restricted to different forms of direct participation – forms of influence on decision making that involve employees as individuals rather than influence exercised through representatives. The issue of how one specific form of indirect participation – trade union power – affects the efficacy of direct participation, however, will also be addressed. The analysis also excludes another commonly cited form of participation – financial participation, particularly through diverse profit sharing schemes, for which we have no data. As Marchington (2005) has pointed out, however, there must be real doubts about the extent to which such schemes involve any effective involvement of employees in decision making.
Direct participation has been the centre of the most important controversies in the recent literature, with writers from different theoretical perspectives disagreeing sharply about the virtues of its different forms. The discussion has focused on three main types of direct participation – individual task discretion (or autonomy), semi-autonomous teamwork and consultative participation.
Individual task discretion
The earliest arguments for the importance of job task control focused on the extent to which individual workers could exercise control over their immediate work task. For neo-Marxian writers, such as Friedmann (1946) and Braverman (1974), this control was the critical factor allowing for the possibility of self-development through work. It required that individual workers were responsible for the design as well as the execution of the work. Sociological research confirmed empirically the strong impact of task discretion on job involvement (Gallie et al., 1998). More generally, Kalleberg and Vaisey (2005) found that ‘autonomy’ was a necessary condition for the workers they studied to consider their jobs to be of high quality. From the 1980s, the argument for the importance of individual control was reinforced by the results of research by social psychologists and epidemiologists (Johnson and Johansson, 1991; Karasek and Theorell, 1990; Marmot, 2004). These results emphasized the importance of the individual’s ‘decision-latitude’ not only for psychological well-being but also for physical health, in particular the risk of cardiovascular disease.
Semi-autonomous teamwork
The focus on individual job task control was challenged from the 1920s by a growing emphasis on the role of the workgroup for employees’ attitudes and behaviour. Evidence of the spontaneous power of workgroup decision making (Lupton, 1963; Roethlisberger and Dickson, 1939; Roy, 1954) led to new theories of management that argued that the incorporation of teamwork into the formal organizational structure could channel employees’ capacity to exercise control over their work in a direction that contributed to overall organizational objectives. They advocated the delegation of decision-making over the immediate work process to semi-autonomous teams that would have formally defined spheres of responsibility (Trist and Bamforth, 1951; Trist et al., 1963).
For some writers, the importance of semi-autonomous teamwork was reinforced by the changing functional requirements of advanced technologies (Blauner, 1964; Woodward, 1970). With the disappearance of individual work tasks, the restoration of the type of individual discretion that lay at the heart of neo-Marxian analysis necessarily would be taken off the historical agenda. But it could be replaced, it was argued, by the capacity to exercise control at a higher team level. For Blauner, the development of collective team control was a central factor underlying the long-term trend from ‘alienation’ to ‘freedom’ at work. The belief in the potential of semi-autonomous teamwork to resolve conflict between the needs of employees and organizational productivity has also come subsequently to inform a wide spectrum of theories of management – such as ‘high performance’, ‘high involvement’ or ‘high commitment’ management (Butler et al., 2004; Lawler,1986; Lawler et al., 1995; Walton, 1985).
Consultative participation
From the 1980s, with the sharp decline of trade union density and weakening of the bargaining power of the unions, a third form of direct participation became increasingly central to debate – the involvement of employees through management-created consultative channels. This shift also became a central theme of high-performance management theories (Boxall and Purcell, 2010). Consultation through wider workplace meetings, more localized briefing groups, problem-solving groups or quality circles would empower employees by allowing them to influence organizational issues through direct communication with management.
The emergence of these different forms of direct participation raises two principal questions, neither of which has been well addressed yet by empirical research. The first is whether managerially-initiated forms of participation (semi-autonomous teamwork and consultative participation) contribute as effectively to employees’ welfare as more traditional forms of task discretion or are better viewed as a form of pseudo participation. The second is whether their implications are relatively general among employees or vary between different types of employee.
Research questions
The relative importance of forms of direct participation?
There is broad consensus across different theoretical perspectives that individual employee control over the work task is a factor that is likely to enhance employee well-being. However, the view that semi-autonomous teamwork and consultative participation empower employees has been a source of sharp debate. Naville (1963) was one of the earliest to suggest that teamwork could be repressive for workers rather than facilitating their self-development or well-being. A later critical literature depicted teamwork as primarily a vehicle for managerial ideological control and cast doubt on the effective decision-making powers it offered to employees (Barker, 1993). It has also been argued that consultative participation is a form of pseudo-participation, where information and consultation are used to persuade employees to accept decisions that are already taken (Pateman, 1970; Ramsay, 1977). From this perspective, such management-initiated forms of participation were unlikely to make a significant contribution to job quality. However, others have argued that non-union channels of involvement reduce workplace problems and help satisfy employee needs (Bryson and Freeman, 2007).
If the argument is correct that semi-autonomous teamwork and consultative participation are effective forms of participation, it could be expected that they would have similar benefits for employee welfare as individual task discretion (Hypothesis 1); if they are forms of pseudo-participation their effects should be neutral or even negative (Hypothesis 2).
Variations among employees?
Arguments about the benefits of employee participation generally assume that its effects will be very similar across different types of employee. However, an older literature on the impact of job characteristics cautions against this assumption and argues for the need to take account of different expectations among employees (Hackman and Lawler, 1971; Hackman and Oldham, 1976). The range of potential sources of variation is large, but the focus here will be upon the possible importance of gender, occupational class, work preferences and trade union strength in the workplace.
Gender
An influential, if controversial, argument has been that women are characterized by greater diversity in types of work orientation than men, particularly with respect to the relative centrality of work and the family (Hakim, 1991, 1996). A significant proportion of women, it is argued, attach a relatively low importance to either the quality of their work or the opportunities it gives for advancement. If this argument is correct, it can be expected that women would attach less importance to direct participation and that its implications for their welfare would be less marked (Hypothesis 3).
Occupational class
The argument that more recent types of direct participation are mechanisms of control rather than empowerment could be expected to be particularly relevant to those in lower occupational class positions, where control is usually seen as particularly problematic. Further, since intrinsic work preferences (including the importance attached to the use of initiative) are also closely related to education, employees in lower class positions may attach less importance to direct participation and its positive implications for work quality may be correspondingly weaker. If this argument is correct, the effects of direct participation for indicators of work quality should be weaker in lower than in higher class positions (Hypothesis 4).
Job preferences
Part of the argument for differential effects by gender and class involve assumptions about variations in the importance that employees attach to influence and the use of initiative at work. Such differences in job preferences, however, may be more widespread in the workforce, reflecting differences in individual disposition and choice. Following the expectations of ‘job characteristics’ theory (Hackman and Lawler, 1971; Hackman and Oldham, 1976), those who attach less importance to involvement in decisions should experience fewer benefits from all forms of direct participation than those for whom it is an important need or value (Hypothesis 5).
Trade union strength
British trade unions were historically cautious about the development of non-union forms of employee involvement. To the extent that this perspective survives, it could be expected that they would seek to undermine their effectiveness as potential rivals to their own influence. They might also reduce their effects by raising aspirations about work and employment conditions in the workforce (Bryson, 2004). It could be argued then, at least in the British context, that the presence of strong trade unions would undermine the efficacy of direct forms of participation (Hypothesis 6). Alternatively, it has been argued (Gonzalez, 2009; Knudsen et al., 2011) that a strong trade union presence would strengthen the effectiveness of direct participation by ensuring that views expressed in consultation were translated into concrete changes in working practices to the benefit of employees (Hypothesis 7).
Assessing the importance of participation for the quality of work
How is importance for the quality of work to be judged? There is little consensus in the literature on the criteria for the inclusion or exclusion of specific aspect of the work situation as important for the quality of work. Three principal arguments can be identified. The first emphasizes the primacy of employees’ own views on what matters to them in a job (Clark, 2005). The second emphasizes ‘objective’ grounds in terms of the implications of job characteristics for an individual’s ability to use and extend their skills. This approach underlies the neo-Marxist concern with self-development. The third position focuses on the implications of work characteristics for psychological well-being and health. This can be seen as the effective position underlying the research agenda of psychosocial theorists of work.
Each of these arguments appeals to a powerful principle and there is no evident way of determining their relative worth. Rather, the strength of argument behind each suggests that relevance for the quality of work needs to be assessed in terms of each of the three principles. The objective here is to assess the relative importance of the three forms of participation both overall and for different categories of employee in terms of: a) the extent to which they are valued by employees; b) their implications for skill use and skill development; and c) their implications for psychological well-being. The measures for these variables are introduced in each of the relevant sections.
Data, measures of participation and analytic procedure
Data
The analyses are based upon the extensive source of data on work experience and work attitudes provided by the ‘British Skills Survey’ carried out in 2006. This survey includes questions on the control employees have over their work task, their involvement in teamwork, the presence of broader consultative mechanisms and the influence of trade unions, together with a wide range of questions on the individual characteristics of employees and other aspects of their work situation. In addition, the questionnaire included questions on job preferences and a measure of psychological well-being. As ever there are trade-offs in terms of types of data. National surveys rarely provide extensive questionnaire batteries for the measurement of specific concepts, but they have the great benefit of enhanced generalizability.
The survey consisted of a nationally representative sample of employees aged 20 to 65, with booster samples in Wales, Scotland, the East Midlands and Northern Ireland. Interviews were conducted at home, with a response rate of 62 percent. A total of 7787 respondents participated in the survey, 6829 of whom were employees. Sample weights were computed to take into account the differential probabilities of sample selection. All of the analyses that follow have been weighted accordingly (for further details see Felstead et al., 2007), with weights scaled to reflect initial effective sample sizes.
Measures of direct participation
The indicators of the different modes of direct participation have been taken to reflect the effective capacity to influence rather than just the formal existence of a participative scheme.
Individual task discretion
The measure is derived from four questions designed to assess how much personal influence people had over specific aspects of their work. They were asked: ‘How much influence do you personally have on … how hard you work; deciding what tasks you are to do; deciding how you are to do the task; deciding the quality standards to which you work?’ There was a four point response scale: ‘a great deal (of influence)’, ‘a fair amount’, ‘not much’ or ‘none at all’. Response points were scored from 1 for ‘none at all’ to 4 for ‘a great deal’. As the four items had a Cronbach alpha of .77 a summary index was created by averaging across the four items.
Semi-autonomous teamwork
A set of questions about team involvement enable us to distinguish different forms of teamwork. To begin with, employees were asked: ‘Do you usually work on your own or does your work involve working together as a group with one of more other employees in a similar position to yours?’ If people did work in a team, they were asked how much influence the work group had over how hard they worked, deciding what tasks they could do, deciding how they are to do the task and deciding the quality standards to which they worked. Scores on the items were averaged and those in teams with an average score of 3 to 4 (equivalent to a great deal or a fair amount of influence) were classified as working in ‘semi-autonomous teams’. A dummy measure of semi-autonomous teamwork was constructed that distinguishes, on the one hand, between those who were in teams that had a great deal or a fair amount of influence and, on the other, those who were either in teams that had low (little or no) influence over decision making or did not work in teams.
Consultative participation
The measure of consultative participation is constructed from a series of questions. Respondents were asked initially: ‘At your workplace, does management hold meetings in which you can express your views about what is happening in the organization?’ For those who could express their views, additional questions were asked about the issues on which consultation took place. These questions ranged from relatively everyday issues of health and safety and training, to operational issues of planned changes in work organization or in product or services, to strategic issues concerning the financial position of the organization and investment plans. A four point scale of consultative participation was constructed from these categories running from 0 for no participation to 3 for consultation on strategic issues.
Moderating variables
The moderating variables are sex, class, union influence and the importance attached to the use of initiative at work. The major groupings (first digit) of the Standard Occupational Classification (SOC) are taken for Occupational Class. The measure of trade union influence was constructed from two questions. People were first asked whether there were unions or staff associations at their place of work. Those who replied that there were unions were then asked ‘How much influence do the trade unions in your establishment have over the way work is organized?’, with response options of ‘A great deal’, ‘A fair amount’, ‘Not much’ and ‘None at all’. In the analyses, ‘union influence’ is a dichotomous variable distinguishing, on the one hand, those who were in workplaces where the unions had a great deal or a fair amount of influence and, on the other hand, those who were either in workplaces without unions or in workplaces with unions with not much or no influence.
The importance of the ability to use personal initiative at work can be seen as a proxy for the importance of participation, since the ability to use personal initiative entails the capacity to influence one’s work environment. The survey asked people the importance they personally attached to a range of job features that people may look for in a job, including the item ‘A job where you can use your initiative’. Overall 85 percent of employees regarded it as ‘essential’ or ‘very important’, while 15 percent had a weak initiative orientation – regarding it as ‘fairly’ or ‘not very important’.
Analytic procedure
The implications of individual task discretion, involvement in semi-autonomous teams and consultative participation are compared with respect to relevant criteria of employee welfare using a representative sample of employees. The survey is cross-sectional and therefore cannot provide direct information on causal sequences. The term ‘effect’ is used in the sense of statistical, not causal, effect. The approach is to assess the consistency of the pattern of the data with the expectations of different theoretical arguments. In a few instances, however, the survey provides retrospective questions on change over time, which can be used to tentatively explore further the plausibility of causal assumptions.
The procedure, with respect to each of the criteria of employee welfare, is first to compare the effects of the different types of direct participation, by taking their standardized beta coefficients when introduced together in a regression analysis without other variables. The net effects are then examined, taking account of a range of control variables to examine whether the initial results can be accounted for by variations in individual characteristics or other aspects of the work situation. The compositional controls used are: age, occupational class (one-digit SOC Classification), industry (one-digit NACE classification), size of establishment (dummy variables for 1–24, 25–99, 100–499, 500+), and dummies for temporary work and employment in the public sector.
Some indication of sequence can be found through recall data. This data is only available with respect to task discretion. The survey asks people, ‘compared to your job five years ago, has the amount of choice you have in the way you do your job increased, decreased or stayed about the same?’. The largest proportion (48%) had seen no change, 12 percent had experienced a decrease in their decision-making scope on the job and 40 percent had experienced an increase. In each case, then, an additional model is presented in which past change in task discretion is added to the full models (including both the measures of the current levels of direct participation and controls).
The next step is to test whether the implications of the three modes of direct participation differ by gender, class and the strength of unions in the workplace, by adding interaction terms (type of direct participation × type of employee) into the full model with controls. Finally, in the analyses for the effects of direct participation for skills and psychological well-being, a test of whether the effects of the different forms of direct participation vary depending on the subjective importance attached to participation is carried out by introducing an interaction term between the different types of participation and the importance of the ability to exercise initiative in people’s job preferences. 1
Results: The implications of direct participation for the quality of work
The subjective importance of direct participation
There are no directly comparable measures on the value employees attached to specific types of direct participation. However, an indirect approach is to examine their impact on employees’ job satisfaction (Kalleberg, 1977, 2011). Forms of participation can influence job satisfaction in different ways. Their effects may be due to the fact that participation leads to improvements in the work environment rather than because participation is felt to be important in itself. The strongest measure, then, of the subjective importance of the different types of direct participation is their effects on job satisfaction once other aspects of the quality of work and employment are controlled.
As can be seen in Table 1, taking the beta coefficients as the measure of comparison, individual task discretion is the form of direct participation most strongly associated with higher job satisfaction, followed by consultative participation. Influence over the job through participation in semi-autonomous teams is significant in the model that controls for other individual and work factors (Model 2) but the strength of the association is much weaker.
Forms of participation and job satisfaction (beta coefficients).
Note: OLS Regression, weighted. Beta coefficients. Model 2 controls: age, occupational class, temporary work, public sector, industry, and size of establishment. Overall model also included sex. Sig: *** = p <.001; ** = p <.01; * = p <.05. Model 3: controls as in Model 2 + skill use, learning requirements, hard work, speed of work, safety, working hours and risk of job loss. Model 4: controls as in Model 3 + change in task discretion. Model 5: controls in Model 4 + interactions. UN = Union strength; F = Female; CL = Class; D = Individual task discretion; T = Team influence; C = Consultative participation; DC = Discretion change.
Were the effects of the different forms of direct participation on job satisfaction due to their consequences for nature of work and employment conditions rather than to the intrinsic importance attached to them? Model 3 introduces controls for a wide range of work factors – skill use, learning opportunities, work pressure, health and safety, working hours and job insecurity. It shows that these factors lead to little change in either the strength or the relative importance of the coefficients. This finding suggests that the differing effects on job satisfaction reflect primarily the subjective importance of the specific forms of direct participation.
It might also be argued that the results may show, not an effect of participation on job satisfaction but, rather, that more satisfied employees are more likely to engage in participative practices. To explore this possibility, at least in the case of task discretion, an indicator of past change in ability to exercise choice over the job is added to the full model. It can be seen that past change in task discretion was highly associated with increased job satisfaction, even when current levels of task discretion and compositional controls are included – consistent with the view that task discretion does effect job satisfaction.
Finally, the issue of whether or not there were differences between types of employee in the subjective importance of different forms of direct participation is addressed by introducing interaction terms (Model 4). There was no difference between male and female employees in the effect of task discretion on job satisfaction but the positive effects of both team influence and consultative participation were less great among female employees. It is important to note that these were relative differences and did not imply that any of the forms of direct participation had a negative effect for women. Additional analyses by sex showed that consultative participation also had positive effects, albeit weaker than for men, for women’s job satisfaction (0.08, p = <0.001), while semi-autonomous teamwork had no significant effect for women.
There were no significant class differences in the importance of any of the forms of direct participation for job satisfaction. Similarly, trade union strength did not influence the effects of either semi-autonomous teamwork or consultative participation. Strong trade unions did, however, weaken the effect of task discretion. Again these differences were relative. Even where unions were strong, additional analyses revealed that task discretion still had a positive effect on the job satisfaction of employees (0.16, p = <0.001).
Direct participation, skill use and skill development
The second criterion for the importance of direct participation for job quality is whether it enhances opportunities for skill use and skill enhancement. The survey questions provide three relevant measures. The first asks how strongly they agree or disagree with the statement ‘In my current job I have enough opportunity to use the knowledge and skills that I have’. The second asks how far they agree or disagree that ‘My job requires that I keep learning new things’. Both were measured on a four point scale from ‘strongly agree’ to ‘strongly disagree’. The third captures experiences of training in the last 12 months. It distinguishes between those who had had no training in the last 12 months, and those with training experiences of less than a week, less than a month, less than six months and more than six months. The question was designed to capture training in a broad sense, providing a list of forms of training that included both on-the-job and off-the-job training.
The standardized Beta coefficients from OLS regressions are taken as the basis for the comparison of relative effects. The results are presented in Table 2. Model 1 gives the coefficients for the effects of the different forms of participation without controls. Model 2 shows the effects taking account of a range of other individual and work factors: age, occupational class, temporary work, public versus private sector, industry and size of establishment. Model 3 adds into the analysis the measure of past change in task discretion.
Forms of participation and skill use and development (beta coefficients).
Note: OLS Regressions, weighted, controls as in Table 1, Model 2. Regression for change in task discretion including controls for current participation. *** = p <.001; ** = p <.01; * = p <.05.
The relative effects of the various types of direct participation differ depending on the specific aspect of skill development under consideration. The strongest coefficients with respect to the opportunity to use knowledge and skills are those for individual task discretion. Although less marked, consultative participation would also appear to have an important effect. In contrast, the effects for semi-autonomous teamwork are relatively weak and non-significant. Turning to new learning on the job, all three types of participation have relatively strong effects. Finally, with regard to training experience, the effect of semi-autonomous teamwork is weakest, followed by individual task discretion. In contrast there is a strong effect of consultative participation on training opportunities.
Given that it could be the case that jobs requiring more skill development are conducive to greater direct participation, the retrospective question on change in individual task discretion as an indicator of sequence (Model 3) is again examined. It is notable that a past increase in task discretion has positive effects on skill use, learning on the job and training, even controlling for current levels of task discretion and a wide range of other individual and work characteristics. While there may be reciprocal effects, this supports the view that higher task discretion has a positive effect on skill opportunities.
Do the implications of the different types of direct participation vary depending on the type of employee? Taking the interaction terms (Table 3), there are only two instances of gender differences and they are not consistent in direction. Teamwork influence is less important for female than for male employees with respect to skill use – indeed, further analysis (not shown) indicates that it has no effect for women. However, consultative participation is more important for female than for male employees with respect to training opportunities.
Forms of participation and skill use and development: Interaction terms (beta coefficients).
Note: OLS Regressions, weighted. Interactions added to Table 2, 2nd model with compositional controls. *** = p <.001; ** = p <.01; * = p <.05. UN = Union strength; F = Female; CL = Class; D = Individual task discretion; T = Team influence; C = Consultative participation; DC = Discretion change. Adjusted R2 refers to full model + interactions.
The effects of participation also differ depending on the importance the individual attached to initiative. Stronger preferences for the use of initiative increase the impact of task discretion on opportunities for skill use (although task discretion continues to have a positive effect even among those with weaker preferences). 2 They also increase the effect of team influence on training. In contrast, union influence in the workplace makes no difference to the effects of any of the forms of direct participation for either skill use or skill development.
It is differences with respect to class that are most striking, especially for individual task discretion, and they are different from those predicted. Individual task discretion has significantly stronger effects at lower levels of the class hierarchy, both for skill use and for the two measures of skill development. More detailed analysis (not shown) indicates that, compared to managers and professionals, the effect of individual task discretion is stronger with respect to skill use for those in skilled manual, sales and operative work. It is also stronger for new job learning for both operatives and elementary workers. Compared to managers and professionals, its effect for training experience is significantly higher for all other classes with the exception of personal service employees. There are fewer significant class interaction effects for other forms of participation. However, both teamwork influence and consultative participation were more strongly related to learning opportunities for those in lower class positions.
Overall, the pattern confirms the importance of direct participation for opportunities for skill use and development in work; but the type of participation that appears to be most effective differs depending on the specific aspect of skill experience. Individual task discretion is most closely associated with the opportunity to make good use of one’s knowledge and skills in work, while consultative participation appears to be more closely related to training opportunities.
Direct participation and psychological well-being
What were the implications of different forms of direct participation for employees’ psychological well-being? Warr (1990) has argued for the importance of distinguishing two axes of psychological well-being – ‘contentment-anxiety’ on the one hand and ‘enthusiasm-depression’ on the other. To capture this well-being, respondents were asked a range of questions about their emotional responses to the job. They were initially asked, ‘Thinking of the past few weeks, how much of the time has your job made you feel each of the following … ?’ There followed a series of adjectives, some positive some negative. To tap enthusiasm-depression, the adjectives were ‘depressed’, ‘gloomy’, ‘miserable’, ‘cheerful’, ‘enthusiastic’ and ‘optimistic’. To tap contentment-anxiety they were ‘tense’, ‘uneasy’, ‘worried’, ‘calm’, ‘contented’ and ‘relaxed’. Responses were made against a standard 5-point frequency scale ranging from ‘never’ to ‘all of the time’. A reliability analysis showed that the reliability (Cronbach alpha) of the overall 12-item scale of psychological well-being is 0.55, whereas this rises to 0.87 for the six item enthusiasm-depression sub-scale, and to 0.84 for the six item contentment-anxiety sub-scale. For each axis, a scale was constructed by averaging responses to the items, with the negative items reversed. Higher scores and therefore positive coefficients indicate better psychological well-being.
Table 4 examines the effect of the different modes of participation first for overall psychological well-being and then separately for the two sub-scales for ‘contentment-anxiety’ and ‘enthusiasm-depression’. The first column gives the net coefficients for each mode of participation without taking account of other individual and workplace characteristics. The second column shows them with other factors taken into account, while the third shows the effects of the measure of past change in task discretion.
Forms of participation and psychological well-being (beta coefficients).
Note: OLS Regression, weighted, controls as in Table 1, Model 2. Separate regression for change in discretion including controls for current participation. Sig: *** = p <.001; ** = p <.01; * = p <.05.
There are significant positive effects on employees’ psychological well-being from all three of the modes of direct participation. With respect to both overall well-being and both of the two subscales, the beta coefficients suggest that it is the effect of individual task discretion that is strongest. Moreover, separate analysis confirms that this effect is the case for both men and women. Consultative participation also has a clear positive effect on each of the scales and for both sexes. The impact of teamwork influence is positive but more limited, with relatively small coefficients. It is only significant for contentment-anxiety in the model without controls. It is only significant for the ‘enthusiasm-depression’ scale once controls are taken into account.
As with skill development, the direction of causation cannot simply be assumed. It could be the case that those with lower psychological well-being are less likely to be involved in participative mechanisms either by choice or exclusion. Again, the retrospective measure of change in task discretion is examined to provide some insight into sequence. As can be seen from Model 3, there is some support for the view that participation has an influence on psychological well-being from the retrospective measure of change in task control. Past increases in task control were associated with higher psychological well-being on all measures but particularly with respect to the enthusiasm-depression dimension of mental well-being, even taking account of current task discretion and a wide range of other controls.
Did the effects differ significantly between types of employee? Interaction tests (Table 5) showed that there were no differences in the implications of the different forms of direct participation for the psychological well-being of men and women. In contrast, there were significant interaction effects for class. Those in lower class positions experienced fewer psychological benefits from either individual task discretion or consultative participation than those in higher class positions. With respect to task discretion, this difference could be partly accounted for by differences in the strength of preferences about the importance of the use of initiative in work. Among employees who placed a similar value on the use of initiative in work, there was no evidence of a class difference with respect to the effect of individual task discretion for either the overall psychological well-being or the depression scale. Again such relative differences did not imply that those in less skilled class positions were negatively affected by any of the forms of direct participation. Additional analyses for separate class groupings indicate that task discretion was still positive for the low skilled (elementary, operative and sales employees) for all aspects of psychological well-being, while consultative participation was positive with respect to both overall psychological well-being and enthusiasm-depression. 3 In the other cases, the effects were not negative but non-significant. Semi-autonomous teamwork had the weakest effects for the low-skilled and was non-significant for all three measures.
Forms of participation and psychological well-being: Interaction terms (beta coefficients).
Note: OLS Regressions, weighted. Interactions added to Table 4, 2nd model with compositional controls.
Sig: *** = p <.001; ** = p <.01; * = p <.05. Definitions, see Table 4. Adjusted R2 are for full models + interactions.
Trade union strength made no consistent difference to the effects of direct participation for psychological well-being. Employees in workplaces where trade unions were strong were more likely to experience beneficial effects from semi-autonomous teamwork. The initial effects for union strength on task discretion and consultative participation were non-significant, although it reduced the positive effects of task discretion once preferences about the exercise of initiative in work were controlled.
It is also possible that participation raised psychological well-being because it responded directly to people’s values of self-direction. If so, one could expect to find important variations in the effect depending on the importance they attached to being able to use initiative in their jobs. As can be seen in Table 5, such preferences have different implications for various modes of participation. Preferences with respect to the use of initiative make little difference to the impact of team influence on psychological well-being. There is, however, a marked difference between those with different initiative preferences in the effects of individual task discretion and consultative participation – with stronger preferences associated with substantially more positive effects on psychological well-being. Additional analyses of those with low initiative orientation indicated that none of the types of direct participation had significant benefits with respect to psychological well-being and, in one case, consultative participation, it was associated with higher anxiety (−.09, P = <0.01).
Conclusions
In the light of the controversies in the literature about the implications of different types of direct participation for the quality of work, the article has compared the importance for employee welfare of three types of direct participation – individual task discretion, semi-autonomous teamwork and consultative participation. For the comparison, it took three criteria of welfare – subjective importance, effects on opportunities for skill use and skill development and implications for psychological well-being. It examined both their overall effects and the extent to which these varied between different categories of employee.
The first issue is whether the different forms of direct participation provide broadly comparable benefits (Hypothesis 1) or whether semi-autonomous teamwork and consultative participation are better viewed as forms of pseudo-participation that either do not contribute to, or contribute negatively to, employee welfare (Hypothesis 2). It was certainly the case that individual task discretion (or autonomy) emerged most consistently as critical for the intrinsic quality of work. It has the strongest effect on job satisfaction, net of a wide range of variables reflecting work and employment conditions. It is important for both skill use and skill development – indeed, it is the form of participation that has the greatest importance for the use of skills and knowledge at work. Finally, it is the form of direct participation that has the strongest positive effect on psychological well-being.
However the other types of direct participation had generally positive, if rather weaker effects. These effects were particularly the case for consultative participation. It was highly significant for all of the three criteria of work quality – subjective satisfaction, skill use and development and psychological well-being. Moreover, it had the strongest effect of any of the forms of direct participation with respect to opportunities for training. Overall, the evidence points to the positive effects for the quality of work of consultative participation. It would require rather strong assumptions about ‘false consciousness’ to dismiss such consistent evidence of its benefits.
Working in semi-autonomous teams had the weakest effects of the three forms of direct participation. It has had the weakest effect on job satisfaction, it was unrelated to opportunities for the use of skill at work, and it had weak, albeit significant, positive effects for psychological well-being. Nevertheless, it did positively contribute to new learning and training experiences and in no instance did it have negative implications for employee’s work quality. Overall, while there was some evidence for its positive effects, collective team influence would not appear to be an adequate substitute for designing jobs in a way that ensure that employees have scope to use their individual discretion.
The second main issue was whether there were marked variations between different types of employee in the implications of direct participation for work quality. To begin with, there were relatively few differences found between men and women. There was no general support for the view (Hypothesis 3) that direct participation would be less important for the quality of work of women, due to differences in the relative centrality of work and family values. Semi-autonomous teamwork and consultative participation did appear to have less subjective importance for women, but individual task discretion was just as important for their job satisfaction. In most cases, the effects of direct participation for skill opportunities were no different for men and women. Where there were differences, they did not consistently point to weaker effects for women. Semi-autonomous teamwork did have weaker effects for women’s ability to use their knowledge and skills in the job but consultative participation made a bigger difference to the training opportunities provided by the job than was the case for men. There were no differences at all between men and women in the effects of any of the types of direct participation on psychological well-being.
There was also no consistent evidence that strong trade unionism either reduces the effects of direct participation (Hypothesis 6) or that it reinforces it (Hypothesis 7). In most cases union strength made no difference to the effects of the different types of direct participation. There is some evidence that it reduced the positive impact of task discretion for job satisfaction, although task discretion continued to have a positive effect even in workplaces where the unions were strong. It had ambivalent effects for psychological well-being, increasing the positive effects of semi-autonomous teamwork but reducing those of task discretion (but only once initiative preferences were controlled). Overall, it does not seem that there is a significant structural tension between a strong trade union presence and the effectiveness of direct participation in enhancing employee welfare.
There were more marked differences in the effects by occupational class. However these differences did not provide clear support for the expectation (Hypothesis 4) that direct participation would have less positive implications for employee welfare for those in lower compared to those in higher class positions. Rather there were respects in which the less skilled received greater benefits than the higher skilled from direct participation and respects in which the effects were weaker. Direct participation appeared to be particularly beneficial for skill use and skill development for those in less skilled class positions. This was especially the case for task discretion, which had consistently stronger effects across the skill measures the lower the class position. Moreover, both semi-autonomous teamwork and consultative participation had more positive effects for the less skilled with respect to learning opportunities in the job. At the same time, direct participation had less positive effects for the psychological well-being of those in less skilled than for those in higher skilled class positions. However, it is important to note that, despite this relative difference, individual task discretion still had positive effects for the lower skilled on all of the psychological outcome measures and consultative participation for two of the three.
Finally, it is also clear that individual work preferences do mediate the effects of participative arrangements (Hypothesis 5). Among those who attach higher importance to the use of initiative, task discretion and team influence have particularly strong effects with respect to skill use and development, and task discretion and consultative participation have stronger positive effects for psychological well-being. Among those for whom use of initiative was not important, task discretion and consultative participation were still beneficial for skill use and skill development but direct participation was mainly neutral with respect to psychological well-being.
Overall, the evidence points to three broad conclusions with respect to quality of work policies. First, it confirms the importance of direct participation for the quality of work and suggests that the agenda for extending it should remain a broad one. In particular, both individual task discretion and consultative participation have consistent positive effects, independently of each other. Second, although there are clear variations in the strength of the implications of direct participation for different types of employees, it had a generally positive impact for all categories of employee, other than for the relatively small sector of employees for whom use of initiative was unimportant. The importance of improvements in direct participation for the quality of work is then relevant very widely across the workforce – for women as well as men, for the low skilled as well as the high skilled. Finally, of the different forms of direct participation, it was the scope given for individual discretion over the way in which work is planned and carried out that had the strongest influence on employees’ welfare. This finding implies that the long- standing concern about creating opportunities for employees to take decisions about their everyday work tasks remains the most pressing issue on the agenda for improvements in direct participation.
There is clearly still a rich agenda for future research. It will be important to address more closely the evidence about causal direction and the possible reciprocity of effects than has been possible here. Although the analysis has sought to cast light on the plausibility of causal assumptions through the use of retrospective information, only longitudinal research can establish this in a rigorous way and longitudinal data about work experiences is currently very inadequate. More research is needed too into what accounts for the differential benefits of the different types of direct participation. Finally, the analysis has not been able to examine one of the propositions of the radical critique of direct participation – namely that such relatively local forms of employee involvement (however important for the quality of work) may reduce the importance attached to alterations in the nature of strategic decision-making in organizations, which has even more fateful consequences for employees’ long-term futures. The relationship between direct forms of participation and attitudes to democracy at work merits careful investigation.
Footnotes
Acknowledgements
The research team that designed the survey included, in addition to the author, Professor Alan Felstead of Cardiff University and Professor Francis Green, Institute of Education, London. The fieldwork was conducted by BMRB Social Research.
Funding
The 2006 Skills Survey on which this article is based was supported by grants from the Economic and Social Research Council, the Department for Education and Skills, the Department of Trade and Industry, the Learning and Skills Council, the Sector Skills Development Agency, Scottish Enterprise, Futureskills Wales, Highlands and Islands Enterprise, East Midlands Development Agency, and the Department for Employment and Learning, Northern Ireland. However, the author takes sole responsibility for the analysis.