The relationship between HRM practice and employee work-related attitudes
The concept of human resource management (HRM) has received focussed attention for around 20 years, with the catalyst being that “many US companies found they were being rivalled and in some instances overtaken, in markets they had dominated” (Ehrlich, 1994, p. 492). As Lodge (1985, p. 319) observes:
By the early 1980s there was still little disagreement that US corporate managers, employees and trade unions would have to change their ways in order to compete successfully for markets in America and abroad.
Harvard university academics introduced a new compulsory component of HRM into their MBA syllabus and reinforced this so-called “Harvard Model” with influential books and articles (Beer et al., 1984; Walton, 1985b; Walton and Lawrence, 1985). While there would surely have been a genuine desire to help US business, US society, and even US employees, there was also “a long-term effort to ensure that the Harvard Business School faculty provided leadership in human resource management” (Walton and Lawrence, 1985, p. xx).
The Harvard concept stresses that HRM should lead to employee commitment – not simply as a means to employer objectives of improved productivity and profits, but because “the fulfilment of many employee needs is taken as a goal rather than merely a means to an end” (Walton, 1985a, p. 49).
At about the same time as the Harvard concept was being developed another viewpoint was being promulgated by academics who supported a “strategic” concept of HRM, with the major work edited by Fombrun et al. (1984). This work emphasises that the “four generic human resource activities of all organisations:
- selection/promotion/placement process;
- reward process;
- development process; and
- appraisal process” (Tichy et al., 1984, p. 26)
need to be strategically aligned with the organisation’s overall strategic objectives.
Hard and soft HRM
British writers have focused on the differences between the Harvard “commitment” concept of HRM and the strategic HRM concept. Keenoy (1990, p. 368) sees the Harvard concept as “philosophically grounded in the recognition of multiple stakeholders and the belief that the practice and benefits of HRM can be achieved through neo-pluralist mechanisms”, while the strategic concept is ” almost uniformly unitarist in orientation and displays a quite singular endorsement of managerial values”. The former is frequently referred to as “soft” HRM, while the latter is “hard” HRM.
In theory, soft HRM fulfils employee needs as an end in itself, and the favourable attitudes generated from the use of “appropriate” HRM practices (Guest, 1997) together with “communication, motivation and leadership” (Storey, 1987, p. 6), result in commitment to the organisation and improved performance. Hard HRM is only concerned with the effective utilisation of employees (Guest, 2002) and emphasises the quantitative, calculative and business strategic aspects of managing the head count resource in as “natural” a way as for any other economic factor (Storey, 1987, p. 6).
If the reality of soft HRM practice was that it produced the benefits referred to in the rhetoric of academic evangelists such as Richard Walton then it should be highly favoured by both employees and employers. Questions have been raised by a number of researchers, however, about the ability of soft HRM to achieve these benefits. There are two concerns. The first is that the real motive behind its introduction is to undermine unions (Sisson, 1994). Indeed, going back to the introduction of the Harvard concept of soft HRM, Lawrence (1985, p. 362) writes about a seminal colloquium on HRM between 35 senior executives and Harvard academics. He claims that the question as to whether HRM was designed to keep the union out was “emphatically denied”, and he protested that all but three of the firms represented were at least partially unionised, with several having had a significant number of units of both kinds, with their “older plants unionised and the newer ones not”. An unreasonable interpretation therefore would be that these leaders of commitment HRM will happily dispense with unions given the opportunity. Indeed, Lawrence (1985, p. 362) admits that there was a general view “that in a well-managed unit with decision-sharing, a union was not needed to represent employee interests”. Guest (1990, p. 389) concludes, “… the main impact of HRM in the United States may have been to provide a smokescreen behind which management can introduce non-unionism or obtain significant concessions from trade unions”.
The second concern is that meeting the needs of employees has never been an objective in itself, and has simply been the normative view of what Harvard academics would like to see as the employment relationship. Truss et al. (1997, p. 70) in a study of soft and hard models of HRM, concluded that “even if the rhetoric of HRM is ‘soft’, the reality is almost always ‘hard’, with the interests of the organisation prevailing over those of the individual”. Similar conclusions have been reached by Keenoy (1990), Poole and Mansfield (1992), Guest (1995) and Legge (1995a, b, 1998).
If the soft model of HRM has validity, then there should be a clear relationship between the experience of soft HRM practice and positive employee attitudes (reflecting their needs are being met) and increasing employee commitment – as well as improved productivity. Surprisingly, however, most of the research and reporting on HRM has ignored the views of employees. Legge (1998, p. 14) points out: “when reading accounts of HRM practice in the UK and North America it is noticeable the extent to which the data are (literally) the voices of management”. Guest (1999, p. 5) agrees, claiming that “from its conception human resource management reflected a management agenda to the neglect of workers’ concerns”. Some studies have considered employee reactions to HRM (Gibb, 2001; Appelbaum and Berg, 2000; Mabey et al., 1998) – but Guest (2002, p. 335) is perceptive in his criticism that “a feature of both advocates and critics of HRM is their neglect of direct evidence about the role and reactions of workers”. Even the exceptions noted which gave employee reactions to HRM did not relate the employee experience of HRM practices to their reactions, and Guest (1997) argues for a research agenda that addresses this gap.
This paper aims to contribute to this gap by fulfilling two major objectives. The first is to add to the limited number of studies which have tested to see if there is a relationship between HRM practice and positive employee attitudes. The second objective is to see if there is a significantly stronger relationship between HRM practice and employee attitudes if employee perceptions of HRM practice are taken as the measure of HRM rather than employer perceptions, which has been the approach taken in general by previous researchers.
HRM practice and employee work-related attitudes
The soft model of HRM, as stated previously, suggests a relationship exists between the use of “appropriate” HRM practices and positive employee attitudes, and while theoretically these relationships remain poorly developed (Guest, 1997, 2001), a number of attitudes are nonetheless widely considered to be an outcome of soft HRM. For example, levels of job satisfaction, which is the affective perception that results from the achievement of desired outcomes (Harber et al., 1997), are found to be related to levels of HRM practice (Guest, 2002; Ting, 1997). High levels of employee commitment have also been found to be related to the use of “appropriate” HRM practice (Guest, 2002), and results from investing in HRM practices which benefit employees. For example, the provision of opportunities for training and skill development benefits the employee by equipping them with the necessary knowledge, skills and attitudes to function autonomously and responsibly (Guest, 2002). Furthermore, it improves retention and enables them to cope with change in the work environment (Guest, 2002).
Organisational fairness is the term used to describe “the role of fairness as it directly relates to the workplace” and is concerned with “the ways employees determine if they have been treated fairly in their jobs and the ways in which those determinations influence other work-related variables” (Moorman, 1991, p. 845). The HRM systems, policies, and procedures that operate in an organisation have been identified as impacting on an individual’s perceptions of bias and fairness (Kurland and Egan, 1999; Greenberg, 1990). Studies show that where employees believe they are treated fairly in the workplace then they hold positive attitudes towards the organisation (Moorman, 1991), whereas HRM practices that are perceived to be unfair have been found to result in the employee feeling bewildered and betrayed and thus less committed (Schappe, 1996).
As far as the relationship between HRM practice and employee attitudes is concerned, the HRM practice dimension has been measured by collecting data from employers in one of the two ways – using additive measures of HRM practice or self-reports about the extent to which particular HRM practices have been operationalised. A possible third approach would be to collect employee views about the operationalisation of HRM practice and then relate these reactions to their attitudes. These approaches are described below.
The additive approach
To date the most common approach for assessing the relationship between HRM practice and employee work-related attitudes is the additive approach. This involves employers indicating, using a “yes/no” response format, which HRM practices from a predetermined list currently operate in their organisations. The “yes” responses are then added together, and the assumption is that a higher sum indicates better HRM. Researchers using the additive approach usually correlate the total number of practices with the HRM outcomes being examined. Research exploring this relationship show a strong correlation between high numbers of HRM practice and positive employee attitudes (Fiorito, 2002; Guest, 1999; Guest and Conway, 2002; Appelbaum and Berg, 2000).
While this approach is frequently used, it has attracted criticism (Fiorito, 2002; Guest, 1997, 2002). The additive approach is very simplistic. If an employer claims a practice is utilised there is no differentiation on the basis on how well, or to what extent the practice exists. For example, there could be token training and development, or a high level of commitment to a professional and effective training and development programme. Both examples are credited as a “yes”. This perceived weakness has led some researchers to use employer self-reports about the strength of HRM practice.
Employer self-reports about the strength of HRM practice
Guest and Peccei (1994) suggest that an improvement on the additive approach is to assess the extent or strength of practice using response bands such as those contained in Likert scales. Such an approach sees employers self-reporting, either on the extent they consider particular HRM practices to be effective (Delaney and Huselid, 1996), or alternatively on the extent to which HRM practices have been operationalised (i.e. to what extent practices are practised) in their workplace (Kane et al., 1999).
Guest (2001, p. 1099) claims further research is needed to test the extent that employer and employee views on HRM practices coincide or differ. Specifically he suggests:
There is now a need for more research comparing the responses of managers responsible for developing and overseeing HRM practice and employees to identify levels of agreement about the operation of practices … such evidence as it is possible to glean from reports about levels of autonomy and consultation suggest that levels of agreement might be quite low.
Support for this assertion has already been found. For example, Appelbaum and Berg’s (2000) study comparing supervisor and employee perceptions about participative initiatives in US organisations found wide differences. Similarly, Kane et al. (1999) also found evidence of perceptual differences about HRM between different organisational stakeholders.
While managers and employers are important stakeholders, so too are employees. This has led to a strong call for employee voice to be heard in HRM research (Guest, 2001).
Employee self-reports about the strength of HRM practice
Given the apparent perceptual differences between employers’ and employees’ perceptions of HRM then a third approach that could be used would see employees’ reports on the strength of operationalised HRM practice analysed in relation to their levels of work-related attitudes. There is now growing support for assessing HRM from the employee perspective emerging in the literature (Fiorito, 2002; Gibb, 2001; Guest, 2001), and given that where the concern is employee attitudes, then it is intuitively logical to relate that to employee rather than employer perceptions of the effectiveness of HRM policies and practices.
The most common approach employed in studies assessing the relationship between HRM practice and employee work-related attitudes is the survey (Cully et al., 2000; Guest, 1999), and this is the method of data collection used in this study. Four areas of HRM practice are examined – good and safe working conditions, training and development, equal employment opportunities (EEO), and recruitment and selection. These areas of HRM have been selected because they have previously been identified as those likely to have the greatest impact on employee behaviour and attitudes (Guest, 2001). Three employee work-related attitudes are examined – organisational commitment, job satisfaction and organisational fairness. These attitudes are generally considered to be desirable outcomes that result from the use of soft HRM (Meyer and Smith, 2000; Guest, 1997).
Three approaches are used to measure HRM practice – an additive measure completed by the employer, and two perceptual measures, one completed by employers and one completed by employees. All three measures relate to practices characterised in the literature as soft HRM or best practice HRM (Guest, 1999; Guest and Peccei, 1992; Johnson, 2000; Kane et al., 1999). They are employee-centred (i.e. they should produce benefits for employees when operationalised), and thus should impact on employee work-related attitudes.
The additive measure, as stated, is the one most commonly used in research of this type. The additive measure used in this study asks employers to indicate, using a yes/no response format, which practices, from a comprehensive list of 80 practices (20 for each functional area covered), currently operate in their organisation (for example, “Do you consult employees on their training needs?”).
A 20-item scale consisting of statements about HRM practice is used to assess the strength of HRM practices (five items for each of the four functional areas). Respondents are asked to indicate, using a five-point Likert scale ranging from 1=strongly disagree to 5=strongly agree, the extent they consider each practice occurs in their organisation, for example, “EEO is promoted within this organisation” (see Appendix 1 for a full list of statements). The same scale is used for both the employer and the employee groups, with the wording amended to reflect their different perspectives. Employer and employee assessments of HRM using this measure are, therefore, concerned with the judgements these groups make about the extent a particular HRM practice has been operationalised in the workplace. The α coefficients for the five item measures relating to each of the four areas of HRM practice examined ranged from 0.83 to 0.88, suggesting high internal consistency exists (Browne, 2000). The statements about HRM practice on the perceptual measure closely reflect the ones used in the additive measure, but there is not a direct correspondence as that would have resulted in an excessively long measure.
Three widely accepted measures of employee attitudes are used. Guest (1997) suggests organisational commitment should be measured using the standard measure developed by Mowday et al. (1979) – the organisational commitment questionnaire (OCQ) – because this scale captures the extent the employee identifies with the organisation, their desire to remain in the organisation and their willingness to exert effort on behalf of the organisation. An adapted and shortened version of the OCQ is used in this study to measure organisational commitment. Again, respondents are asked to indicate the extent to which they agree with each of the statements using a five-point Likert scale ranging from 1=strongly disagree to 5=strongly agree (α=0.8535).
To measure job satisfaction, an adapted and shortened form of the Minnesota satisfaction questionnaire (Weiss et al., 1967) is used. The scale has six items and respondents are asked to indicate how satisfied they are with respect to each of the statements, using a five-point Likert scale ranging from 1=very dissatisfied to 5=very satisfied (α=0.8664).
While organisational fairness has two dimensions – distributive and procedural – this study is concerned only with procedural fairness and HRM policy and practice. The measure used is an adapted and shortened six-item version of the procedural justice scale used by Moorman (1991). This measures employees’ perceptions about the way formal procedures are carried out. Respondents are asked to indicate the extent to which they agree with each of the statements using a five-point Likert scale ranging from 1=strongly disagree to 5=strongly agree (α=0.9160).
Employer and employee respondents were also asked to provide a range of demographic information.
Data are analysed using SPSS, Version 11, using matched employer and aggregated employee data from 37 organisations. Correlations, using Pearson’s r, are completed to explore the relationships between HRM practice and employee attitudes. The multiple linear regression technique is used to explore the differential impacts of the three approaches being tested.
In total, 234 organisations were originally contacted to participate in this study. A total of 40 organisations agreed to participate. The person most responsible for operationalising HRM practice was requested to complete the employer survey. Employee participation was voluntary, and the company fully endorsed their participation. The employer participant was requested to distribute the surveys to a representative sample of their workforce, in terms of occupational classification, ethnicity and gender. The targeted population of employees consisted of a total of 1,075 full- and part-time employees from the 40 participating organisations. A total of 626 employees responded (a response rate of 58 per cent) by completing the survey and returning it in the reply-paid envelope provided. In total, 37 usable matched sets of data were obtained. These 37 matched data sets comprise one completed employer questionnaire (total usable employer responses=37) and on average 15 employee questionnaires for each organisation (total usable employee responses=572). The number of actual employee responses received for each organisation varied with workplace size, and ranged from eight to 31.
The demographics for the employer and employee samples are reported in Table I. New Zealand is a relatively small country and organisations employing in excess of 100 employees are usually classified as “large”, with those having between 50 and 99 employees classified as “medium” and organisations with under 50 employees as “small”. Three-fifths of the organisations participating in this study can be classified as large; one-fifth are medium, and the remaining five small. Of the 37 participating organisations, 19 are located in the private sector, 18 in the public sector. Around 35 per cent of the participating organisations are located in the service industry, with this group being evenly represented across both sectors. All other organisations are classified in the manufacturing, transport, retail, sales, education and health industries.
Both the employer and the employee samples are reasonably a representative of organisations and the composition of the workforce in New Zealand, with the only notable exception being the employee sample is skewed in terms of occupation towards professional and semi-professional groups.
The first analysis examines the relationship between additive measures of HRM practice and attitudes. The number of practices for each functional area varied considerably between organisations. Generally it is found that there are relatively high levels of practice for the areas of good and safe working conditions (M=15.95, SD=3.22), training and development (M=16.77, SD=2.76), along with recruitment and selection (M=16.52, SD=2.50). EEO is the one area, on the other hand, where a considerably lower mean level of practice exists (M=12.17, SD=5.96).
According to prior research, high numbers of HRM practice are associated with positive employee attitudes (Guest, 1999). The results of the analysis for this relationship are reported in Table II.
These findings differ from those previously reported. Using aggregated employee means, when levels of organisational commitment (M=3.72, SD=0.78), job satisfaction (M=3.72, SD=0.79) and organisational fairness (M=3.70, SD=0.86) are correlated with the combined additive measure of HRM practice, the results show that higher index scores in the number of HRM practices are not related to higher levels of organisational commitment (r=0.155, df=35, p=>0.05), job satisfaction (r=0.085, df=35, p=>0.05) or organisational fairness (r=0.146, df=35, p=>0.05). When this relationship is examined by functional area the correlations are also found not to be significant.
Next the relationship between HRM practice and employee attitudes is examined using first employer self-reports, and second employee self-reports. These results are also reported in Table II.
Statistically significant correlations are found to exist between the three attitudes measured and employee assessments of HRM practice for all the functional areas examined. However, no significant correlations are found between employer self-reports about the strength of HRM practice and employee attitudes.
Data are then analysed using a regression technique, enabling the predictive qualities of each of the three approaches for assessing HRM practice to be explored. The model used for regressions is shown in Figure 1.
All the measures of HRM practices were entered into a multiple regression equation so that the character of the relationship between different measures of HRM practice and employee attitudes could be explored. It is found that nearly 51 per cent of the variance in organisational commitment (R 2=0.506, p<;0.000), 58 per cent of the variance in job satisfaction (R 2=0.578, p<0.000) and 61 per cent of the variance in organisational fairness (R 2=0.610, p<0.000) can be explained by one or more of the measures of HRM practice.
The results of the regressions for the three measures of HRM practice, along with the three measures of employee attitudes, are presented in Table III.
When the results of the regression for the three different measures of HRM practice are analysed it becomes evident that employee assessments have the greatest level of predictive power for attitudes.
The regression reveals that numbers of HRM practice and employer assessments about the strength of HRM practice account for very little of the variance in attitudes. On the other hand, employee assessments of HRM practice appear to be very influential. Around 47 per cent of the variance in organisational commitment is explained by employee assessments of practice across the four HRM areas examined, with even larger and equally significant levels of variance being explained by this for job satisfaction and organisational fairness.
If we exclude employee assessments related to EEO from the analysis, as it uniformly records the lowest β across all three attitudes, the β statistic for the remaining three functional areas of HRM practice implies that for every increment in employee assessments, levels of attitudes increase by between 0.17 and 0.34. The most statistically significant and consistent increments are in relation to employee assessments of good and safe working conditions, along with training and development, and these are found to apply to all aspects of work-related attitudes measured. Also worth noting is that, not surprisingly, assessments of recruitment and selection practices accounted for the largest variance in levels of organisational fairness.
Discussion and concluding remarks
Consistent with the work of Appelbaum and Berg (2000), Cully et al. (2000) and Guest (1999), this study also finds that a significant relationship exists between HRM practice and employee work-related attitudes. There is a strong statistically significant relationship when HRM is researched from the employee perspective. However, there was no significant relationship when HRM practice was measured using the additive approach or employer self-reports. These results have implications for current approaches used to assess the HRM practice-employee work-related attitudes relationship.
This study demonstrates that by far the greatest correlate of employee attitudes are employees’ own assessments of the strength of HRM practice. Interestingly, higher levels of HRM practice, using an additive measure, are not found to be significantly associated with attitudes across any of the four functional areas of HRM examined. In fact, it is revealed that higher levels of HRM practice are in some cases related to decreased levels of job satisfaction and organisational fairness. For example, looking at the data for each functional HRM area, it is possible that higher levels of EEO practice may actually have no, or even a negative, effect on employee attitudes, with a strong inverse relationship being found for job satisfaction.
Similarly, weak relationships are found when the relationship between employer assessments about operationalised practice and attitudes are examined. So while Guest and Peccei (1994) suggest these areas are important aspects of best practice HRM, this study finds them to be unrelated to attitudes when data are collected from employers.
Put together these results suggest that if managers want to maximise employee attitudes, then implementing lots of practices is not sufficient. Practitioners need to be aware that the way they implement their HRM practices may be a more important determinant of employee attitudes than the number of practices they put in place. This builds on the research findings of other recent studies (Armstrong, 2001) that suggest it is possible that it is the quality of the practice that counts and not the quantity. It may be that practitioners may need to undertake regular attitudinal surveys to assess employee reactions to current HRM practice and assess these on a longitudinal basis to identify what is working and what is not.
It was beyond the parameters of this study to examine the complex relationship between HRM practice, employee attitudes and business performance; it has, however, explored the narrower relationship between HRM practice (for the areas of good and safe working conditions, training and development, EEO and recruitment and selection) and three employee attitudes (organisational commitment, job satisfaction and organisational fairness). While regressions tell us nothing about the cause-effect nature of relationships between assessments of HRM practice and employee attitudes, these findings strongly suggest that incorporating the employees’ experience of HRM into analyses exploring the HRM practice, employee attitudes relationship will contribute further to our understanding about what aspects of HRM have the potential to yield the greatest performance outcomes. Arguably, more so, it would appear, than exploring this particular relationship solely using additive measures or employer self-reports.
This study has some limitations. This study found the standard deviations for assessments of HRM practice and work-related attitudes amongst employee groups within each organisation to be reasonably low, showing close agreement. Nonetheless the aggregated data has been developed from a relatively small sample, thus it may not be representative of the bulk of employee views. The evaluative nature of the questions used introduces the possibility of common method variance (i.e. “an artificial correlation across questions due to mood or other contaminants” (Fiorito, 2002)). Also, the HRM practices used in this study are supposedly best practice; however, there is still insufficient evidence available to confirm that they are indeed best practice.
Finally, Gibb (2001) suggests effective HRM is about implementing best practice. While this may very well be true, it is clear from this study that simply introducing best practice, or even assessing the effectiveness of practice from the employers’ viewpoint, is no guarantee that they have been successfully implemented. If the organisation’s objective is to improve employees’ workplace attitudes, then possibly a better way to confirm which HRM practices are having the greatest impact, and hence are effective, is to correlate employees’ assessment of them with employees’ attitudes. Further studies in this area are now required.
Appelbaum, E., Berg, P. (2000), “High performance work systems: giving workers a stake”, in Blair, M., Kochan, T. (Eds),The New Relationship: Human Capital in the American Corporation, Brookings Institution Press, Washington, DC, .
[Manual request] [Infotrieve]
Armstrong, G. (2001), The Change Agenda: Performance through People, The New People Management, CODE, London, .
[Manual request] [Infotrieve]
Beer, M., Spector, B., Lawrence, P.R., Mills, D.Q., Walton, R.E. (1984), Managing Human Assets, Free Press, New York, NY, .
[Manual request] [Infotrieve]
Browne, J.H. (2000), “Benchmarking HRM practices in healthy work organizations”, American Business Review, Vol. 18 No.2, pp.54-61.
[Manual request] [Infotrieve]
Cully, M., Woodland, S., O’Reilly, A., Dix, G. (2000), Britain at Work: As Depicted by the 1998 Workplace Employee Relations Survey, Routledge, London, .
[Manual request] [Infotrieve]
Delaney, J.T., Huselid, M.A. (1996), “The impact of human resource management practices on perceptions of organizational performance”, Academy of Management Journal, Vol. 39 No.4, pp.949-69.
[Manual request] [Infotrieve]
Ehrlich, C.J. (1994), “Creating an employer-employee relationship for the future”, Human Resource Management, Vol. 33 No.3, pp.491-501.
[Manual request] [Infotrieve]
Fiorito, J. (2002), “Human resource management practices and worker desires for union representation”, in Bennett, J., Kaufman, B. (Eds),The Future of Private Sector Unionism in the United States, M.E. Sharpe, Armonk, NY, .
[Manual request] [Infotrieve]
(1984), in Fombrun, C., Tichy, N., Devanna, M. (Eds),Strategic Human Resource Management, Wiley, New York, NY, .
[Manual request] [Infotrieve]
Gibb, S. (2001), “The state of human resource management: evidence from employees’ view of HRM systems and staff”, Employee Relations, Vol. 23 No.4, pp.318-36.
[Manual request] [Infotrieve]
Greenberg, J. (1990), “Organisational justice: yesterday, today and tomorrow”, Journal of Management, Vol. 16 pp.399-43