Life on the Line: Exploring High‐Performance Practices from an Employee Perspective

Through a case&#8208;based approach, this study addresses the call for employee&#8208;focused research to help us understand whether the high&#8208;performance causal chain is the result of discretionary effort or management control and work intensification. Findings highlight the role of the manager and workplace relationships and emphasise the need for &#8216;good practice&#8217;.

Empirically, a 'common method bias' (Wood et al., 2012: 438) has been created, with a focus on cross-sectional data, and an (over)reliance on single (management) respondent data (Appelbaum et al., 2000). On the rare occasions that employee interests have been considered, they are usually questioned on issues of interest to management and through a managerial discourse (e.g. employees' motivation and commitment to the organisation and their willingness to participate in teams or work on a more flexible basis).
The following study addresses these empirical limitations and the lack of employeefocused research. As a result, the focus is on employees working in a 'highperforming' organisation, rather than just the presence (or absence) of particular practices and their associated performance indicators. To overcome epistemological limitations of previous research, and to achieve the objectives of the study, triangulation of both quantitative and qualitative data gathered from employees was collected at a single case study. Findings highlighted perceived variations in the implementation of practices, and the importance of work-based relationships. Consequently, findings support the focus for 'good practice' (Godard, 2004(Godard, , 2010, rather than 'high-performance' practice.

THE LACK OF EMPLOYEE-FOCUSED RESEARCH
Initial research on the high-performance paradigm placed an emphasis on theories of motivation, commitment and discretionary effort (Appelbaum et al., 2000;Delaney and Huselid, 1996;Guest, 1997;Purcell et al., 2003;Truss, 2001), claiming that they provided a more coherent account of the links between HRM and organisational performance (Delaney and Huselid, 1996;Guest, 1997;Huselid, 1995;Truss, 2001). The focus on motivation, commitment and discretionary effort was argued to 'present a step forward in theorising about linkages' (Truss, 2001(Truss, : 1127. However, this approach has its limitations. The most obvious concern focuses on issues of causality with the causal chain remaining untested. Instead, it is based on assumptions with little empirical evidence to support these mechanisms (Takeuchi et al., 2009). These assumptions are largely the result of a lack of employee-focused research. As a result, a two-dimensional perspective on employees has developed. Instead, research is usually from the viewpoint of the manager and rarely from an employee perspective (Harley et al., 2007).
The lack of employee-focused research is concerning, especially as it is established that management and employees having different perspectives on work (Liao et al., 2009). An argument that is further supported by data from the Workplace Employment Relations Survey (Cully et al., 1999;Millward et al., 2000) and the British Social Attitudes Surveys (Bryson, 1999), which indicate that management and employees hold very different views on a wide range of HR policies. Whether following a unitarist or a pluralist frame of reference (Fox, 1974;Heery, 2016), there is a need for employee-focused research. From a pluralist approach, the need for employeefocused research is clear, especially as it considers the employment relationship to be 'exploitative, dehumanising and conflicted' (Heery, 2016: 4). Without employeefocused research, we cannot truly understand the degree of subsequent resistance. Although a unitaritist approach argues that employer and employee interests can be integrated (Guest and Peccei, 2001: 209), there is still a need to understand both sides of the relationship in order to integrate these interests. It is clear therefore that the reliance on a single management respondent clearly results in bias (Appelbaum et al., 2000). Consequently, it is simply assumed, rather than proven, that satisfied workers (as designated by managers) exert discretionary effort and display higher productivity. This assumption has become so embedded in research that is now taken for granted; it is rarely analysed by consulting employees.
In addition, it is not just the presence of HR practices, but how these practices are managed (Guest and Conway, 2001). As research has shown, it is not just the presence of practices, but the perceived intentions behind them (Khilji and Wang, 2006) as there is a disconnection between what management say they are implementing and what employees experience (Liao et al., 2009). To understand how employees perceive high-performance practices requires moving from a focus on macro-level research and focusing on the micro-level (Chang, 2015;Lepak et al., 2007). With the majority of organisations providing the minimum set of HR practices due to legislation (Paauwe, 2004), it is important to consider how these practices are being implemented at the micro-level. In particular, variations may exist in large organisations in how practices are perceived by specific groups of employees (Guest, 2011). It is likely that different employee groups may experience identical HR practices differently (Liao et al., 2009). Variations may be intentional and match the requirements of particular employee groups (Lepak and Snell, 2002;Lepak et al., 2007), or the result of variations in line management implementation (Guest, 2011).
Further analysis of the effects of high-performance practices has suggested that instead of having positive benefits it actually results in increased levels of stress. Any benefits in performance are argued to be the result of the exploitation of employees, not the result of them willingly exerting discretionary effort (Keenoy, 1999;Legge, 1995;Godard, 2001;Appelbaum 2002;Delbridge and Turnbull, 1992;Ramsay et al., 2000). As a result, the 'question arises as to whether or not HRM practices might positively affect firm performance through exploiting workers in a way that has detrimental effects on their well-being' (Paauwe et al., 2013: 10). Based upon concepts of labour process theory (Ramsay et al., 2000), HPWS is seen as a management 'style' in order to coerce employees into working harder, with control being sought through the rhetoric of commitment. The theory argues that if management follows a 'high-road' or 'soft' approach it is under the veil of rhetoric and that in reality these approaches are simply a 'wolf in sheep's clothing' to derive control (Keenoy, 1990(Keenoy, , 1997Legge, 1995). However, claims of control and exploitation are unable to be either accepted or rejected, because of a lack of empirical evidence. Studies exploring issues of increased stress have provided mixed results (Godard, 2010;Ramsay et al., 2000). There is therefore a need to examine whether increases in productivity are the 'outcome' of commitment and motivation, or control and work intensification. This can only be achieved through employee-centred research, as employees are the only ones that can stipulate whether they are experiencing increased control and work intensification or expressing greater motivation. Without data from employees on their experiences of work, we only learn of the rhetoric and lose sight of the reality (Legge, 1995;Truss et al., 1997). There is a need therefore to test these assumptions (Guest, 2011).
Although it has been argued that the neglect in employee-focused research has been remedied through the inclusion of employees in the Workplace Employment Relations Survey data (Guest, 2011), empirically, the focus is still on quantitative data. Even the strengths of quantitative research, such as systematic measurement, would appear to be limiting factors in this area of research, as there is very little agreement on what should be measured (let alone how). In addition, quantitative data fail to give us the depth that is required to fully explain the relationship between highperformance practices and organisational performance. To look inside the 'black box' demands the triangulation of quantitative data and a collection of qualitative data from a variety of sources, as many authors have acknowledged but few have seriously pursued (Becker and Gerhart, 1996;Boxall, 1991;Gerhart, 1999 andGuest, 1997). This could be taken a stage further by actually stepping inside the black box and conducting participant observational research, but there are even fewer studies of this ilk (Delbridge, 1998). To do so, it requires in-depth case study research.

METHODOLOGY
Following the research procedure set out by Truss (2001Truss ( :1127Truss ( -1129, the question of HPWS and organisational performance was 'inverted' by selecting a highperformance organisation and then looking at what HR practices the company employed, and whether these could be constituted as a bundle of high-performance practices or an HPWS. Access was secured based on anonymity; hence, the company is given the pseudonym of Healthcare Limited. Classified by the FTSE Global Classification System as a 'multi-department retailer', it is the UK's leading provider of health and beauty products. In line with research on the high-performance paradigm, the focus of this study was on the manufacturing division. The manufacturing division develops and produces a variety of products and is the integrated supplier to the main company and to third parties. Because of access and location, the study examined four of the manufacturing divisions of five departments (referred to as D1, D2, D3 and D4).
According to company information and data gathered from semi-structured interviews with members of the HR team, Healthcare Limited employed HR practices that are typically included in various bundles of 'best practice' (listed in Table 1). Thus, if Healthcare Limited was included in a survey of HPWS/organisational performance, the (single) management respondent could confidently tick the relevant boxes, leading to the conclusion that the company is a 'best practice' organisation and a high performer.  Arthur (1992), Pfeffer (1994Pfeffer ( , 1998, Delaney et al. (1989), Huselid (1995), MacDuffie (1995) and Delery and Doty (1996)

Data collection
Research was conducted in the manufacturing division of Healthcare Limited in 2002. The timings of the employee attitude survey, employee focus groups and participant observations were at the discretion of management gatekeepers. Prior to undertaking employee-focused data collection, interviews were conducted with key stakeholders.

Semi-structured interviews
Semi-structured interviews were conducted with the Director and Head of HR for the manufacturing division, the Head Shop Steward, the General Manager and departmental managers from the four manufacturing departments. In total, eight semistructured interviews were conducted.

Participant observations
Participant observation was conducted in each of the four main production departments in January 2002. An evening shift (from 13:30 until 22:00 hours) was worked in each department to gather practical experiences of what life was really like. Employees were informed of the reasons for the research and were assured that it was independent from management. Whilst working on the line, no preferential treatment was given or received. Notes were taken when opportunity arose (largely during break times) and were handwritten on a small notepad out of sight of employees to avoid alienation. To ensure comprehensive data collection, all observations were recorded. This included descriptions of the environment, work practices and processes, conversations and my feelings whilst working on the line.

Employee focus groups
Five employee focus groups were conducted over a two-day period in May 2002. In total, 39 employees participated (one employee was selected to attend but did not attend). Each group involved two employees from each department, therefore ensuring that there was representation from each department in every focus group. This facilitated and widened the discussion between employees as they compared and contrasted their experiences from different departments, although at the expense of direct departmental comparisons. However, because of production constraints, it was not possible to take eight employees off the line per department at the same time.
The focus groups lasted for approximately an hour and were conducted in a predesignated room just off the production floor. Line managers randomly selected employees on the basis of their workload at the time of the focus group sessions. It is important to note that this could have created a potential bias in participant responses. However, responses from participants suggested that this was not a significant issue. Questions were structured around opinions of specific HR practices (e.g. training, teamworking, suggestion schemes, payment systems and management relationships).

Employee attitude survey
The survey consisted of a number of attitudinal measures using previously validated scales (Bacon and Blyton, 2001;Cully et al., 1999;Poole and Jenkins, 1998). Included in the survey were questions concerning specific HR practices (e.g. the amount of training received, consultation, job redesign, teamworking and financial benefits) and questions on the degree of work-related stress over the past year. A 10-item measure was used to measure the level of commitment and motivation felt by employees (Mowday et al., 1979). In addition, a 21-item measure of job satisfaction was used (Bacon and Blyton, 1999. The employee attitude questionnaire was distributed to all available full-time permanent shop floor workers in the four main departments in September 2002. Because of the nature of the factory work, an article-based survey was utilised. With the support of management, a shift was spent in all four departments, and employees were given the opportunity between line changes and break times to use an allocated room to complete the survey. The employee attitude questionnaire was administered to all available full-time permanent shop floor workers in the four main departments. Considering respondents who were off sick/on holiday as 'unreachable' then an 'active response rate' of 99 per cent was achieved as only one respondent refused to complete the survey. The sample was representative with respect to age, gender and other demographic characteristics. Because of the survey being 'vetted' by management 'gatekeepers', questions regarding control, stress and work intensification are limited.

Data analysis
Qualitative analysis on both participant observations and focus groups followed an iterative approach and manually coded to generate broad themes. There were then categorised and grouped. Further rounds of analysis proceeded to develop meaningful codes. Survey data were manually inputted into SPSS for statistical analysis between variables. Analysis of participant observations, focus groups and survey data was undertaken independently of each other, before being triangulated.

Quantitative analysis
Demographic information was gathered to ascertain whether gender, age and the number of years worked at the company has an effect on responses to employee opinions of high-performance practices and mediating variables. Respondents were predominantly male (59.9 per cent) and aged between 40 and 49 years (34.3 per cent). The majority of employees (76.6 per cent) were members of a trade union.

Motivation, commitment and antipathy
Employees were asked 10 statements on a seven-point scale that focused on levels of commitment and motivation. Factor analysis identified three components of this scale that are both statistically significant and logically grouped. The components have been categorised as motivation, commitment and antipathy (Table 2). It is evident from the 'motivational' component that the statements are a reflection of employees' motivational levels, for example, 'I'm willing to put myself out just to help the company' can be classified under Bailey's (1993) description of discretionary effort. The 'commitment' component reflects an employee's desire to remain at the organisation for both financial and personal reasons. The final component, 'antipathy', groups together statements that pertain to an employee's desire to either leave the organisation, or reluctance to encourage a friend or offspring to join the company. Regression analysis was conducted on the three components against age, gender, tenure, the department that the employee worked in and their grade level (Table 3).
Although cross-sector analysis prohibited further understanding of the causal mechanisms behind relationships, regression analysis on levels of motivation has identified interesting relationships regarding levels of motivation and the age (regression co-efficient = À0.146, p < 0.05) and tenure of a respondent (regression coefficient = 0.212, p < 0.05). Levels of motivation were negatively related to the age of the respondent, with respondents classified in the lower age categories expressing higher levels of motivation. In addition, a positive relationship existed between the  tenure of a respondent and their levels of motivation (regression co-efficient = 0.212, p < 0.05), with higher levels of motivation expressed by long-term employees. Focusing on specific statements, respondents were asked whether they were willing to 'put themselves out for the company'. The majority of respondents (76.6 per cent) claimed be willing to go that 'extra mile'. There was a relationship between a respondent's willingness to exert extra effort and their gender (Spearman's correlation = 0.136, p = 0.027). Although the relationship was weak, men were more likely to agree to the statement that they are willing to exert themselves. In addition, willingness was associated with the grade level of the respondent (Spearman's correlation = À0.131, p = 0.033), with the grade level having a negative relationship. This suggests that the higher the grade level the more the respondent is likely to agree with the statement concerning their willingness to exert extra effort. Respondents also claimed that they like to make extra effort not just for themselves, but also for the company (89.9 per cent) and that to be able to contribute towards the organisation made the majority of respondents feel good (86.2 per cent). It is evident therefore that respondents were willing to and wanted to exert discretionary effort in their work, both for themselves and for the company. Regression analysis was conducted on levels of commitment with independent variables of age, gender, tenure, the department and grade of a respondent. Relationships occurred between levels of commitment and the age and tenure of a respondent. Long-term employees expressed higher levels of commitment (regression co-efficient = 0.248, p < 0.01). Focusing on specific statements regarding commitment, the majority of respondents agreed with the statement that they were proud to work at the company (82.5 per cent), with respondents that had previously been employed more likely to agree that they are proud to tell people where they now worked (Spearman's correlation = À0.126, p = 0.041).
Statements regarding an employee's desire to leave the company and their reluctance to recommend the organisation to a friend or offspring, classified as a component of antipathy (Table 2), shows relationships with age, tenure and grade. Older respondents were less likely to think about leaving the organisation and more likely to not recommend the company to a friend or offspring compared with younger aged employees (regression co-efficient = À0.169, p < 0.05). Long-term respondents were more inclined to express antipathy towards the company than short-term respondents (regression co-efficient = 0.431, p < 0.01). In addition, antipathy was more likely to be expressed by respondents on higher-grade levels (regression co-efficient = 0.157, p < 0.05). There were no relationships between antipathy and age, gender or department.

Control and work intensification
Because of management 'vetting' of the survey, limited questions were allowed regarding issues on control and work intensification. To gauge employee opinions concerning control and intensification, respondents were asked questions regarding their levels of satisfaction (Table 4) over the amount of control they had (low levels would signal high degrees of management control) and satisfaction over their current work load (assuming that respondents would be dissatisfied if they felt their work load was too intense). These questions were logically grouped together and had mathematical significance. Respondents were also asked whether they were experiencing workrelated stress (Table 4). Management was apprehensive about questions related to stress being included in the survey, believing employees would not be able to differentiate between stress from work and from their personal lives. To resolve this issue, a definition of work-related stress was included in the question.
Components of control/work intensification and stress were correlated to variables of age, gender, tenure, department and grade level (Table 5).
Satisfaction regarding issues of control/work intensification were related to the gender of the respondent (regression co-efficient = 0.163, p < 0.05), with female respondents more inclined to state that they were dissatisfied with statements pertaining to control/work intensification. This suggests therefore that female respondents were dissatisfied with the amount of control they had within their work and the level of effort and workload that they had. However, over half of respondents (55.5 per cent) were satisfied with their workload ( Table 6), suggesting that work intensification, for the majority of respondents, was not significantly high. However, satisfaction over workload had a relationship with the number of years a respondent had worked at the organisation (Spearman = 0.140, p = 0.021); the longer a respondent had worked at the organisation, the less satisfied they were with their current workload. Satisfaction was also linked to the grade level of a respondent (Spearman = 0.153, p = 0.013), with higher-grade level respondents being more dissatisfied with their workload.
Satisfaction was associated with the number of years an employee had worked at the company (chi-square = 64.106, p = 0.003) and with the department that the  Question 21-3: Your current work load level Question 12-3: Have your levels of workrelated stress increased over the past year? Question 21-4: The effort required to do your job Question 12-1: Do you suffer from workrelated stress Question 21-11: The amount of variety in your job Question 21-10: The freedom to choose your own method of working employee worked in (chi-square = 56.051, p = 0.003). Employees from D4 were more inclined to claim they were dissatisfied with the amount of effort required to do their job, with employees from D3 being more likely to state that they were satisfied. Responses suggest that over time workload had been intensifying, but that the effort required doing the job was declining.

Variations across departments
Employee opinions on the level of support are associated with the department that an employee worked in (chi-square = 29.879, p = 0.000, 0.012 level of significance) and had a negative relationship with the number of years that the employee had worked at the company (Spearman's rho = À0.261, p = 0.000, 0.01 level of significance). This suggested, therefore, that the longer an employee worked at the company the less support they felt they receive from management.
An employee who felt management provided sufficient support was more likely to agree that they felt 'part of the company' (Spearman's rho = 0.238, p = 0.000, 0.01 level of significance) and agree that they were 'quite proud to be able to tell people who it is they work for' (Spearman's rho = 0.214, p = 0.000, 0.01 level of significance). They were also likely to agree to the statement that they would recommend the job to a close friend (Spearman's rho = 0.234, p = 0.000, 0.01 level of significance) and/or to a son/daughter (Spearman's rho = 0.216, p = 0.000, 0.01 level of significance). An employee was also more likely to disagree with the statement that they thought about leaving the company (Spearman's rho = 0.287, p = 0.000, 0.01 level of significance) if they felt supported by management. Finally, if an employee believed that they received sufficient support from management then they were more likely to be satisfied overall, with their job (Spearman's rho = 0.440, p = 0.000, 0.01 level of significance).
Respondents were equally split when asked about the relationship between management and employees. An employee's opinion of the quality of this relationship was dependent on the department that the employee worked in (chi-square = 52.541, p = 0.000) with employees from D1, on average, stating the relationship was 'good' (65.4 per cent), with only 18.5 per cent of employees from D2 claiming it was 'good'. Employee opinion had an impact on commitment levels, with an association between opinions of the relationship with management and an employee's reluctance to change employer if the company was not doing well financially (chi-square = 43.450, p = 0.019). The relationship between management and employee had a profound effect on an employee's commitment levels. A positive relationship occurring between an employee's opinion of the relationship between them and management and being less likely to think about leaving the company (Spearman's rho = 0.342, p = 0.000, 0.01 level of significance), and being less likely to leave for economic reasons, such as increased pay elsewhere (Spearman's rho = 0.216, p = 0.000, 0.01 level of significance). A possible explanation for the impact on commitment is the effect the relationship has on an employee's opinion of the company. A positive relationship occurred between an employee's opinion of the relationship with management and an employee agreeing that they feel a part of the company (Spearman's rho = 0.394, p = 0.000, 0.01 level of significance), and agreeing to feeling 'proud' to tell people where they worked (Spearman's rho = 0.323, p = 0.000, 0.01 level of significance).
As an additional sign of commitment, an employee who believed that the relationship between management and employee's to be 'good' was also more likely to agree that they would recommend the company to a close friend (Spearman's rho = 0.347, p = 0.000, 0.01 level of significance) and/or a son/daughter (Spearman's rho = 0.313, p = 0.000, 0.01 level of significance). In addition to the impact the relationship between management and employee can have on employee commitment levels, there was also an impact on employee motivation. Although the relationship was relatively weak, results do suggest that an employee was more likely want to exert discretionary effort if they feel that the relationship between management and employees was 'good' (Spearman's rho = 0.212, p = 0.000, 0.01 level of significance). Levels of job satisfaction are also affected by the relationship between management and employees, with employees who believe the relationship was 'good' being more inclined to be satisfied overall in their job (Spearman's rho = 0.477, p = 0.000, 0.01 level of significance).

Control and increased work-related stress
Employee responses during focus group interviews and participant observation show that employees were experiencing work intensification, expressed by a worker during the focus groups as there being 'more and more pressure on you, targets are getting higher. They are cutting back on costs, the machines are going quicker, [and] your work is harder' (participant 18, focus group 3).
It was evident that there was sufficient insecurity amongst workers to ensure compliance. The paternalistic management style towards permanent workers emphasised the 'privileged' position that they were in, with both visual and verbal differences expressed between 'temps' and permanent employees (e.g. with signs stopping temporary workers from entering certain facilities, but allowing permanent employees in). They were constantly reminded of the insecurities outside of their position, with referrals of increased competition, cut backs and the use of temporary contracts. As highlighted by a worker during focus group interviews, the 'business has changed', with 'very few jobs that come permanent now-a-days' (participant 3, focus group 1) and that this was now the 'nature of it' (participant 10, focus group 2). Although they have a permanent contract they are reminded of the rarity of them, emphasising the need to 'hold on to' the position that they have with the company and therefore comply with management wants. With management stating that there is a shortage of permanent contracts, temporary workers are likely to comply with management in order to achieve one.
Opinions regarding work-related stress from this study showed increased stress amongst employees. During focus groups, interviewees commented on the growing number of employees off sick due to stress, explaining that this was due to having 'more on you now' (participant 17, focus group 3) with increased pressure from management to maintain levels of production and continuity of output during shift changeovers. It is evident that employees were also under increasing stress due to the greater automation of the line, with workers expressing difficulty keeping up with the machine with bottles flying down the line like cars on a formula one track. As a result, the task was physically stressful, and when coupled with the line manager's constant criticisms for not working hard enough, emotionally stressful.
It is evident from the findings of this study that management within the manufacturing divisions had control over shop floor workers. Management exerted simple control over employees through the enforcement of discipline and formal rules. For example, this is evident in the 'petty' rule enforced by management regarding the use of 'communal' gardens built for employee use during lunchtime breaks. Management forbid shop-floor employees from wearing their white coats in the gardens and its canteen facilities. As a result, shop floor workers were unable to use them as they did not have sufficient time during their lunch break to go to their lockers and change prior to going to the gardens, and then changing again before returning to the line. Technical control of employees was demonstrated in participant observations, with the first-hand experience of how the automation of the machine sets the pace at which employees' work, regardless of whether employees could keep up with the machine. The ad hoc manner of promotion argued by employees to be dependent on whether your face fitted emphasised the bureaucratic control management possess over shop floor workers: 'if your face don't fit, you don't get anything' (participant 26, focus group 4).
In addition to these three types of control, management at Healthcare Limited also has economic control over employees working in the manufacturing departments. The company had established an oligopoly over the area. As such, employees were aware that they were unlikely to acquire the same level of pay, pension and reasonable working conditions elsewhere. In essence, employees become economically 'trapped' (participant 25, focus group 4), thereby decreasing possibilities of resistance.

Motivation and discretionary effort
During focus group interviews and participant observation, the motivational importance of relationships was emphasised, specifically the impact of positive working relationships. When given the opportunity to talk about issues that were relevant to them, employees did not discuss high-performance practices, but focused on relationships with their colleagues, line managers and senior management.

Management relationships
From an employee perspective, one of the most influential variables in the degree of satisfaction they experience in their jobs was the 'relationships' that they have in the workplace. More specifically, it was the relationship that employees had with their line/department managers that had the greatest impact on their motivation; and their relationship with their co-workers that had the greatest impact on their commitment and loyalty levels.
According to interviewees, in addition to recognition from management, showing employees 'respect' (participant 7, focus group 1) had a profound motivational effect on their reported experience of work. In interviewees' opinion, not only do management no longer show respect by saying 'good morning' (participant 19, focus group 3) they do not 'even look you in the eye and acknowledge you' (participant 21, focus group 3). Although interviewees generally supported this opinion, it was noted that the degree of 'disrespectful behaviour', as interviewees' saw it, was dependent on the individual manager and therefore frequently resulted in departmental differences. The level of respect given affected an employee's desire to exert discretionary effort as they regarded 'doing extra' as a favour to the line manager rather than the company. As a consequence, interviewees were more inclined to do more for some managers rather than others.
The impact of poor management and departmental relationships is evident in relation to work-related stress. A situation that was argued to have 'very much increased over the past few years' (participant 1, focus group 2), but that every department was different. The extent to which an employee feels and is able to deal with work-related stress was acknowledged by interviewees to be dependent on 'where you work and the people you work with' (participant 6, focus group 1) and 'on who you've got on your back' (participant 5, focus group 2). Interviewees felt that there was very little empathy and that they were 'not getting support from the management to help us' (participant 3, focus group 2). As a result, of the increasing pressure and lack of empathy from management, interviewees claimed an increase in work-related stress, and also an increasing number of mistakes in the production process.
Responses highlight the importance of middle managers in achieving the desired 'outcome' of HR practices. It was evident from data collected from focus groups that there were discrepancies in the implementation of practices. The level of management support received was variable, depending on the individual manager and the individual department. A significant number of employees in D3 and D4 complained about insufficient support from management, specifically when they were experiencing personal or family problems. Employees working in the D1, who had praised their line manager and management team for the personal support that they had received, contrasted this. Whilst working in the different departments, the relationship (or lack of) was noticeable. In D1, employees could be frequently seen talking to a member of the management team, in what was clearly a relaxed manner. In other departments, however, this was not the case. Instead, an air of 'parent and child', or 'teacher and pupil', was tangible, producing an oppressive environment.

Co-worker relationships
What was strongly emphasised by employees is the obvious fact that shop floor workers do not work on the line to fulfil some childhood dream or ambition; they are there essentially for the money. It is a job that pays relatively well for what is required and the skills that they have. Life on the line is not a voyage of self-discovery to employees, money is a necessity, it is 'what you work for' (participant 29, focus group 4), whilst friendship is the 'tolerator'. '[I don't] mind coming to work because I like the people I work with. It's not the job itself; it's the people that make it bearable' (Participant 31, Focus Group 4).
Friendships at work were seen as an important factor in breaking up the boredom and monotony of the job. Developing friendships with both colleagues and management encouraged employees to come into work and to remain with the company. Friendships, or certainly good working relationships, with colleagues were important, not only in making the job bearable but also for making it possible.

DISCUSSION
Reviewing the company's HR policies and practices, there is a stark contrast between those for shop floor workers and for management and graduates. A classic example of this can be seen in the recruitment strategy of the two groups. Shop floor workers are recruited by an external agency with very little involvement from Healthcare Limited; in contrast, graduates and management are recruited by the company's extensive HR department using very progressive methods (including in-depth application forms, assessment centres and psychometric testing). In addition, general behaviour towards shop floor workers and management is openly different. It is evident therefore that when required, management do follow best practice, including effective management of the 'processes'. However, the company does not need to invest either time or money in shop floor workers; they are not regarded as an investment, but simply as a resource. It is evident from employee responses that any variations are attributable to management style.
When reviewing the survey, specifically variables of commitment, motivation and antipathy, results are more indicative of fealty than the picture of 'active commitment' painted by advocates of best practice HRM/HPWS. Respondents that had worked for the company in the long term were more likely to express antipathy, yet expressed higher levels of commitment towards the organisation. As stated by employees, they remained because of 'economic' control, with employees acknowledging that they would not be able to achieve the same amount of pay for their level of skills elsewhere, claiming that they would prefer to be unhappy then to take a drop in pay. Employees had resigned themselves to the realities of life on the line and found ways of living with the system.
All employees, in all departments, had access to the same HRM practices, incentives and opportunities to participate, yet not all employees' exercised discretionary effort in their work. Where positive relationships were evident, and respect and recognition reciprocated, then the impact and subsequent 'outcomes' of HPWS practices were apparent, emphasising the importance of maintaining a positive psychological contract (Guest and Conway, 1997) and 'good management practice' (Godard, 2010). This indicates, therefore, that positive relationships are important variables. However, the question of whether these variables are the 'outcome' of HPWS practices or are the precedent to HPWS practices remains.
In addition, survey results demonstrated the importance of processes and highlighted the positive impact of relationships at work, both with management and fellow workers. Focus group discussions and observations emphasised the significance of relationships and the motivating effects of respect and recognition. Data demonstrated that what matters to employees is not necessarily 'best practice', but simply good practice (Godard, 2010). Godard suggested that moderate adoption of HPWS practices may be better than low levels of adoption, but only slightly less advantageous than high levels (Delaney and Godard, 2001;Godard and Delaney, 2000). This is a suggestion that is upheld by the present study.

CONCLUSION
Although Healthcare Limited employed progressive HRM/HPWS practices, it was apparent that, on the factory floor, they were not having the impact advocated by the 'high-performance' paradigm. Instead, the 'process' or causal chain between the HR practices and the companies organisational performance occurred not via discretionary effort, but through control. Theoretically, this study has shown that the reality of working on the line is one of management control, work intensification and stress, supporting previous research by the labour process school (e.g. Braverman, 1974;Friedman, 1977;Burawoy, 1979Burawoy, , 1985Edwards, 1979;Keenoy, 1990Keenoy, , 1997Legge, 1995;and Ramsay et al., 2000). However, it is unknown whether this applied to just the manufacturing division and may not have applied to employees in the 'core' business where considerable time and investment in recruitment and selection has been made (Liao et al., 2009).
Although it would appear through quantitative data collection that because of poor 'management' or a lack of effective devolution of HR practices, the organisation was failing to fully achieve these desired 'outcomes' within the manufacturing divisions. Yet closer examination through qualitative research shows that the company was 'managing' the situation rather well. It was evident that through rigorous control of the labour process, the organisation does not require discretionary effort from employees. The manufacturing division had an oligopoly power of the employment relationship by focusing on high wages and restrictions of permanent contracts. Resistance is low with employees acknowledging that they are unlikely to acquire equal pay and job security with the reasonable working conditions of the organisation elsewhere. With respondents asserting that they would want better for their offspring, it is apparent that employees accept the realities of working on the line, and although express disgruntlement, are unlikely (or unable) to leave.
This study shows that when employees are given the opportunity to speak, as opposed to tick a box on a management-focused questionnaire, a very different picture of their motivation to work and their commitment to the organisation emerges. When given the opportunity, the issues that come to the fore focus on the work-based relationships, an issue that is (often) overlooked in the literature promoting highperformance practices (excluding Purcell et al., 2003); give employees a voice, and they do not talk through a 'managerial perspective' about performance or refer to high-performance practices. Instead, they refer to the motivational (or demotivational) effsects of managerial relationship and the long-term commitment effects of friendships at work. This suggests that discretionary effort is the result of managerial relationship and teamworking, rather than high-performance practices. Instead, of striving for 'best practice' HRM, management should pay more attention to 'good personnel management' (Godard, 2010).
Findings from this study have shown that what is required is triangulation of methods to provide an in-depth understanding, and a subsequent need for future research to do so, specifically, the need to focus on employee perspectives at the micro-level, and examining variations in the implementation of practices. This study has contributed to the limited number of employee-focused research; however, to enable generalisations to be made, there is still a need for further exploration in this area, specifically looking at the issues that are of importance to employees.