The strange case of the missing worker: Human resource management, employees and organisational performance

It would be fair to say that the conventional wisdom concerning HRM, employees and performance rests on a rather thin body of evidence

(pages 18-23 of printed journal)

By Bill Harley

HRM and performance

Since the 1980s, it has become increasingly common to hear business leaders say ‘our people are our most valuable asset’. Corporate mission statements are replete with references to the importance of employees in driving business success. Indeed, it would be fair to say that the view that employees are fundamental to organisational success has become more-or-less a given in the contemporary business world. This view has been used by many to argue for the importance of sound human resource management (HRM), on the assumption that if employees are managed well organisations will perform well.

There is now a large and growing body of empirical research which demonstrates positive associations between HRM and performance, and which has further bolstered the view that good HRM drives performance. There is, however, a problem. In spite of frequent claims that the effective management of employees is the key to business success, an examination of the empirical research shows that there is rather less evidence than one might suppose that HRM causes performance gains. Perhaps even more surprising, given the apparent centrality of employees to organisational success, there is a dearth of research which actually examines the role of employees as the link between management and performance. Indeed, it would be fair to say that the conventional wisdom concerning HRM, employees and performance rests on a rather thin body of evidence.

The quest for the Holy Grail

Although the management of people has existed as long as there have been organisations, the profession of HRM as we now know it is a creature of the 1980s. There are different accounts of its emergence, but one compelling version is that it came out of work done by business academics in the US in the early 1980s. These scholars, concerned with the question of why US industry performed poorly during the 1970s in comparison to many newly industrialised economies, looked to Japanese management practices. While critics have claimed that these researchers, seeing Japan through American eyes, completely misunderstood Japanese management, the fact remains that they built an influential account of the link between people and performance on the basis of their observation. Their central argument was that Japanese companies performed well in part because they treated employees as valued members of the company who needed to be nurtured, developed and rewarded in return for which they gave commitment, loyalty and effort. This view of people management, which might be termed ‘developmental humanism’, formed the basis for the emergence of HRM, with its emphasis on treating employees as valuable resources.

As the discipline of HRM grew, both among practitioners and academics, it spread from its birthplace in the US across the Anglophone world and then into Europe and Asia. A discourse of ‘strategic HRM’ – the linking of employee management to corporate strategy – gained currency and advocates of this approach began to argue that HRM was so important that it should have a board-level role in companies rather than just being an operational function. Of course, the best way to get a seat at the board table is to demonstrate that one is crucial to business performance. As practitioners sought an increasingly central role in organisations, and HRM academics sought to show their relevance, a demonstrated link between HRM and business success became what Karen Legge memorably called the ‘Holy Grail’. If only it were possible to demonstrate that HRM was a key driver of business success, this would validate both practitioners and academics. This imperative led to a flurry of research activities in the 1990s seeking to demonstrate the HRM-performance link.

What do we know about HRM and performance?

Much of the research on HRM and performance has been based on the concept of ‘high performance work systems’ (HPWS). HPWS involve sets of mutually-reinforcing HRM practices, which are said to drive improvements in performance. Conventionally, HPWS are regarded as having three key elements. First, participative forms of work organisation in the form of self-managing teams and high-discretion jobs encourage employees to work creatively and productively. Second, to work effectively in such forms of work organisation, employees must have the necessary skills, which can be achieved by rigorous recruitment and selection, and the provision of training. Finally, if skilled employees are to work effectively, they must be motivated, which is pursued via reward systems which encourage good performance, notably through performance-based pay systems. Advocates of the HPWS approach argue that organisations which put in place participative forms of work and ensure that employees are skilled and motivated, will reap rewards in terms of better business performance as a result. Since the mid-1990s, research on HRM and performance has largely focused on testing associations between HPWS practices on one hand and indicators of organisational performance on the other.

As a result of the growing body of empirical research, a consensus has emerged among many scholars and practitioners – that ‘good’ HRM in the form of HPWS causes improvements in business performance. This view has provided a justification for the proliferation of university courses in, and departments of, HRM as well as for a great deal of advice provided to organisations by consultants. At the risk of oversimplification, the bulk of this research has consisted of studies which have collected organisational-level data on the presence of HPWS practices and various indicators of business performance, for example labour productivity, turnover and product or service quality. Statistical analyses have been conducted to measure associations between HPWS on one hand and performance on the other. In many cases, the research has shown positive associations between the practices and performance. On the face of it, this would seem to provide the proof that practitioners and scholars have sought – that HRM improves performance. To put it another way, the seekers of the Grail would appear finally to have found it. There are, however, problems with much of the research.

I certainly am not suggesting that the research is in some way substandard or that results have been manipulated. To the contrary, the research published in high-quality international journals has been rigorous and well executed. Nonetheless, it has important limitations which are often overlooked by those who use the results to argue the case for the legitimacy of HRM. First, most studies are cross-sectional – that is, they involve data collected at a single point in time – and they measure associations between variables. However, it is not safe to claim a causal relationship on the basis of positive associations. Second, the measures of HPWS and performance are often limited and problematic. Third, the associations shown in the studies are often very weak indeed. If we take into account these limitations, it becomes clear that we should be very cautious about accepting as a ‘fact’ that HRM causes performance gains. Indeed, the strongest claim we can make is that most studies are not inconsistent with this apparent fact. If, however, we push our scepticism to one side for a moment and accept the claims of advocates of HPWS, an obvious question is: how do HPWS practices cause performance gains? It is at this point that the paradox of the ‘missing worker’ starts to become apparent.

Where are the workers? Mainstream and critical accounts

It is possible to discern at least two very different theoretical explanations of the apparent causal mechanisms which link HPWS and performance. While the two theoretical models are more-or-less diametrically opposed, what unites them is the assumption that HPWS practices influence performance precisely because of the effect that they have on employees and the way that employees experience their work. The dominant theoretical model, which I call here the ‘mainstream’ model, suggests that HPWS practices deliver performance gains precisely because they deliver positive outcomes to employees. That is, they deliver ‘win-win’ outcomes, in which both employees and organisations benefit. Very simply and briefly, advocates of this model argue that HPWS practices deliver increased discretion, satisfaction and commitment, as a result of which employees not only stay with organisations and increase their work effort, but also have the opportunity, skills and motivation to work effectively. To put it another way, the theory tells us that HPWS lead to autonomous, skilled and motivated employees, working ‘smarter not harder’ in pursuit of organisational goals.

This is a very appealing explanation, in the sense that it suggests that everyone can benefit from good HRM, but it has not been without its critics. The alternative ‘critical’ model presents a rather less appealing explanation for the impact of HPWS on performance. Often informed by Marxian understandings of the dynamics of management-worker relations, in particular the view that the defining feature of management is the drive to squeeze more effort from workers in pursuit of increased profit, critics have argued that HPWS is just a new way of extracting employee effort. The focus of critics is primarily on work organisation and motivation mechanisms. Participative work organisation, they argue, actually represents a shifting from management to workers of responsibility for decisions. Motivation mechanisms, notably performance-based pay, represent a way to put pressure on employees to increase output. From this perspective, HPWS can be seen as a way of squeezing effort from employees, so any performance gains arise at the expense of workers.

The two models provide starkly contrasting accounts of how HRM contributes to performance. Advocates of the mainstream model argue that performance gains arise through positive effects on employees, leading to a win-win situation. Critics on the other hand argue that performance gains reflect a win-lose situation in which gains to management come at the expense of employees. The views are united, however, by their common assumption that performance gains arise due to the effect of HPWS on employees. Crucially, they rely on assumption about the role of employees, with little reliable empirical evidence to support their claims.

What do we know about the role of employees in the HRM-performance chain?

When I first began to research in this area in the late 1990s I was very forcefully struck by the gaps in our knowledge. While advocates of the two opposing theoretical models argued long and hard for their positions, there was a remarkable lack of evidence which would allow any adjudication of their claims. This seemed quite remarkable given the centrality of employees to the main theoretical models. There were three specific gaps. First, there appeared to be no research which sought to test the competing claims of the models. Second, and even more remarkable, there was virtually no research on the HRM-performance link which analysed employee data. Third, there was not even any research which tested the impact of HPWS on employees, a central part of the models. It seemed extraordinary to me that a discipline called human resource management appeared to be leaving the human experience out.

Since that time, I have sought to fill some of these gaps, using survey-based data. My concern has been twofold. First, I have taken the impact of HRM on employees as a topic worthy of attention in its own right, regardless of its implications for performance. Work remains one of the fundamental human activities – the way we experience it is a critical determinant of our life experience. Hence, an understanding of the impact of managerial practice on employees’ experience of work is an important facet of social scientific research. Second, I have done some work on the links between HRM and performance which has involved employee data.

My research on the apparent impact of HPWS on employees has called into question much of the conventional wisdom. Employing secondary analysis of large, cross-industry surveys in Australia and the UK, it has shown that some HPWS practices appear to have no effect on employees, some to have positive effects and some to have negative effects. My research, exploring employee responses as the link between HPWS and performance, shows different practices having different impacts – some positive, some negative – on employees. It provides little evidence, however, that these employee responses influence organisational performance. These results, which rely on rigorous analysis of reliable data, provide little comfort to advocates of mainstream or critical models and suggest that theoretical accounts which present HPWS as ‘good’ or ‘bad’ for employees in a general sense are overly simplistic. Further, the results suggest that causal paths from HRM to performance are likely to be much more complex than the existing theoretical accounts argue.

Where to from here?

In the time that I have been researching this area, there has been important work done by others which has also begun to fill in some of the gaps in our knowledge. No doubt incremental gains will continue to be made. For the moment, however, we must recognise that we know much less than is commonly assumed. First, the ‘fact’ that good HRM causes performance gains continues to rest on rather limited evidence. Secondly, the ‘fact’ that HRM leads to such gains because of its effects on employees is even more questionable.

The lack of conclusive evidence about the impact of HPWS on employees suggests that we need to move beyond the prior debates between mainstream and critical theorists and ask new questions. If, as the data suggest, some HPWS practices sometimes appear to have positive impacts on employees, some negative impacts and some no apparent impact, we need to refocus inquiry. Specifically, we need to ask: in what circumstances do particular practices have particular outcomes for employees? In what circumstances might these outcomes have an impact on organisational performance? What are the apparent causal paths from HPWS to employees to performance? To answer these questions requires detailed research focusing on specific groups of employees, in different parts of different industries, which can tease out these complexities. Much more work is required before we can feel confident that we understand the HRM-employee-performance link.

Top ^

A condensed version of his Inaugural Lecture given at the University of Melbourne on 6 October 2009.

Professor Bill Harley is Professor of Management in the Department of Management and Marketing at the University of Melbourne.