DS1

akem_k

article.pdf

Home >Business & Finance homework help >Management homework help >DS1

Distributing HRM responsibilities: a classification

of organisations Mireia Valverde and Gerard Ryan

Universitat Rovira i Virgili, Department of Business, Avinguda Universitat, Tarragona, Spain, and

Ceferı́ Soler ESADE, Avinguda Pedralbes, Barcelona, Spain

Abstract

Purpose – The purpose of the paper is to show that HRM is not the sole responsibility of HR departments, but also of other agents inside and outside the organisation, such as top and line managers, and external HRM service providers. This paper seeks to examine how organisations distribute HRM activities and responsibilities among these agents. The study attempts to classify organisations according to agent distribution and to explore whether a number of internal and external context characteristics affect this distribution.

Design/methodology/approach – The survey in the paper shows the methodology chosen in order to collect and analyse factual data about the participation of different agents in HRM activities and the characteristics of the organisations and their context. The questionnaire obtained a valid sample of 231 Spanish companies. A multiple correspondence analysis approach was taken in order to cluster the organisations.

Findings – The analysis of the data in this paper clearly produced seven groups, each involving organisations that allocated responsibilities to the various agents in a similar fashion. However, no common contextual characteristics were found among the companies in each of the seven categories.

Research limitations/implications – The main limitations in the paper are related to the length of the questionnaire, the adequacy of the respondents related to the unit of observation; and the relative newness of the techniques used.

Originality/value – The main contribution of the paper consists of the incorporation of existing partial areas of study in the field of HRM (roles of the HR department, devolution to line management, and outsourcing of the HR function) into an integrated study.

Keywords Human resource management, Responsibilities, Devolution, Outsourcing, Spain

Paper type Research paper

Introduction The management of people occurs in all organisations, whether in a formal, informal or even unconscious manner. Because of this, people management materialises in a wide range of forms and is performed by a number of different agents, depending on the organisation, its environment and other characteristics. However, the focus of HRM studies has traditionally centred on the HR department rather than on the HR function as a whole.

In this paper, the HR function is not understood simply as the set of activities performed by the HR department, but as all managerial actions carried out at any level regarding the organisation of work and the entry, development and exit of people in the

The current issue and full text archive of this journal is available at

www.emeraldinsight.com/0048-3486.htm

PR 35,6

618

Personnel Review Vol. 35 No. 6, 2006 pp. 618-636 q Emerald Group Publishing Limited 0048-3486 DOI 10.1108/00483480610702692

organisation so that their competencies are used at their best in order to achieve corporate objectives (Valverde, 2001a, p. 19).

In this sense, agents other than the HR department may be involved in the management of people in organisations, mainly:

. Top management, who make strategic decisions (including HRM strategic decisions), establish the organisation’s values and philosophy (Schuler and Jackson, 1999) and influence its whole approach to managing people (Sisson and Storey, 2000).

. Line management, who traditionally have been given responsibility for some operational aspects of managing people, but whose role in this function has continued to increase since the advent of HRM (Storey, 1992; Lowe, 1992).

. External HR agencies, usually contracted by organisations to provide administrative HRM services or professional, specialised HRM solutions (Cook, 1999). The outsourcing of these activities has also been found to be on the increase (Hall and Torrington, 1998b)[1].

The HR roles of these agents inside and outside the organisation have been thoroughly examined in the extant literature on an individual and separate basis. This is particularly true for the literature on the HR department, but also for studies on the devolution of HRM activities to line management and on outsourcing the HR function to external specialist agencies. However, the simultaneous examination of all agents that have an input in the HR function has not been previously undertaken.

If our goal is to understand new directions in the roles and responsibilities of the HRM function, a good starting point might be to redress some of the imbalance between focusing on specific agents in isolation instead of the HR function (its activities and responsibilities) as a whole.

Thus, the interest of the authors was precisely to investigate how HR activities and responsibilities are distributed among the agents responsible for HRM (the three outlined above and the HR department). The absence of previous theories, or models, of HR agents, in the literature, lead to an exploratory approach to this study. This would have implications for the methodological design but also for the revision of the literature. The approach taken to revise the relevant literature consisted of observing what had been written about each agent or actor and their responsibilities in HR. However, it is beyond the scope of this paper to summarise the extensive literature on the four agents. Rather, we summarise below the main characteristics of this literature.

The literature is more prescriptive or normative than descriptive. Thus, we find greater emphasis on what various agents should do to manage people, as opposed to establishing what they actually do (Gratton et al., 1999). Often this literature lacks a strong empirical evidence (Jackson and Schuler, 1999). Also, there is a tendency to concentrate on the wider concept of “role” (including activities, responsibilities, power relationships, influence and position, etc.) with little focus on specifying which particular responsibilities are allocated to each agent (Legge, 1995). Descriptions and classifications tend to concentrate on the study of just one of the agents (e.g. classifications of HR department roles, as in Ulrich et al., 1997; Tyson and Fell, 1986; or Labelle and Dyer, 1992) or, at most, on how two agents distribute responsibilities between themselves (e.g. the study of the devolution to the line of responsibilities assumed to pertain to the HR department as in Cunningham and Hyman, 1999; or Hall

Distributing HRM

responsibilities

619

and Torrington, 1998a; or the study of the use of external HR specialist agencies as in Redman and Allen, 1993). There is no developed literature incorporating the study of all four agents.

Even in the very few instances of studies that examine more than two HR agents), these have not attempted to integrate the distribution of all HR responsibilities. As a consequence, the non-explained aspects of these partial studies (e.g. what makes a HR department adopt a specific role) are often assumed to be caused, by the actions of other actors in the organisation. The lack of an integrated approach to the study of actors and their responsibilities in HR management is clear.

The idea of HRM as a partnership of multiple agents or stakeholders is beginning to be more widely employed. However, this trend appears mainly in textbooks rather than in research publications (for example, Jackson and Schuler, 2000). Additionally, partnership is seen as involving two or at most three HRM agents (as in Mohrman and Lawler, 1999).

The main reasons for the weaknesses that characterise the literatures dealing with the roles and responsibilities of different HRM agents lie, on one hand, in the complexity of the phenomena being studied. On the other hand, they are caused by the large amount and lack of clear identification of factors that intervene in producing variability in phenomena such as the degree of devolution to the line, strategic or administrative roles of the HR department, or the amount of HRM tasks being outsourced. However, previous research has identified factors, which influence HRM practice in organisations. These are largely related to organisational context, both external (social, economic, political, etc.) and internal (size, structure, culture, etc.) (Gunnigle and Moore, 1994; Truss et al., 2002). These must be taken in consideration when addressing the study of the agents involved in HRM.

Purposes of the research This paper presents the results of the second part of a larger research project investigating how HRM responsibilities are spread or distributed among agents and the extent to which this distribution differs among organisations in Spain. Put simply, we address the question of “who does what” in HRM.

The specific objective of this part of the project was to classify Spanish organisations’ behaviours according to the distribution of agents involved in different HRM activities at different levels (strategic, operational, administrative, technical, etc.). The grouping of these organisations could provide a map or taxonomy of HRM types or models. A secondary objective was to identify the influence of contextual factors on the adoption of a specific HRM type or model.

In order to achieve the general and specific objectives set by the research project and to operationalise the questions it poses, the study put forward two propositions to be explored empirically.

P1. The exact distribution of who does what, in other words, which agents carry out various HRM activities, is different for each organisation. In this sense, the configuration of the HR function is unique to each organisation. However, it may be possible to identify common patterns of behaviour among organisations, thus identifying different types, groups or models of HR functions according to the distribution of agents carrying out HRM activities.

PR 35,6

620

P2. As the internal and external contexts of organisations shape their management practices, similar types of HR functions will be found in organisations with similar contextual characteristics.

Methodology As stated in the introduction, the topic of research is exploratory as there are no previous theories to be tested. Therefore we are presented with a case of theory building, where the issue needs to be mapped or discovered (Layder, 1993). In many cases, exploratory research is carried out using qualitative research methods (Alverson and Skölberg, 2000). However, the fact that the phenomena under study involved the collection of factual, non-subjective information in order to obtain the classification and profiling posed by the research propositions, an objectivist epistemology was deemed more appropriate (Burrell and Morgan, 1979), and thus a quantitative methodology (the survey) was chosen. Furthermore, and given that the second proposition of the study is related to the impact of organisational context, a contingency approach was taken. Contingency theory, stemming from general systems theory addresses the existence of fit between organisational variables (mainly size, structure and strategy) and organisational outcomes (Donaldson, 2001). In this study, the contingency approach is relevant as the second proposition searches for the relationship between organisational contingencies and models of HRM.

Data collection and sample The primary research was based on a questionnaire that was included in a larger survey of HRM practices, namely the Spanish node of the Cranet-E project (Cranet Survey, 2000; Brewster et al., 1991). The Cranet-E project is an ongoing project involving a network of universities gathering survey data on HRM practice in Europe. The sample frame used in Spain comprises all organisations that employ more than 200 employees and which are members of the main Spanish employer association (from where the database was extracted).

In addition to company size, an additional requirement for accepting the questionnaire was that the organisation had a HR department. This was to ensure the comparability of data for analytical purposes. The results of the primary research reflect the responses of 231 organisations (10.5 percent of the 2,200 questionnaires posted) in the private sector employing more than 200 employees and having a HR department.

Tool and measures The data collection tool consisted of a questionnaire, which gathered information on two main aspects of the organisations sampled:

(1) Participation in HRM activities. The main part of the questionnaire outlined 52 typical HRM activities. This list included activities at the various levels of HRM decision-making (strategic, policy, operational, administrative) and across a wide range of HRM areas (resourcing, training and development, change management, etc.). Respondents were asked to identify which agent was responsible for each activity in their organisation. They were also asked to allocate a percentage representing the proportion of responsibilities that were shared by more than one agent, as illustrated in Appendix 1 (Figure A1). The

Distributing HRM

responsibilities

621

list of 52 HRM activities was generated from the revision of the literature as well as from the results of two focus groups comprising representatives of the four agents studied (HR managers, managing directors, line managers and owners-managers of HR service providers).

(2) Contextual variables. The contextual variables used in the questionnaire were necessary in order to explore P2. Information was obtained for a considerable number of these variables, including traditional contingent variables and other characteristics of the internal and external context of the organisation typically observed in HRM studies (e.g. Tayeb, 1987; Mayne et al., 1996; Jackson and Schuler, 1999, etc.). The operationalisation of these variables is summarised in Appendix 2 (Tables AI and AII). The main variables are also outlined in the results section. It is important to highlight that some of these variables were already included in the standard parts of the Cranet-E survey (mainly typical organisational details such as size, sector, ownership, etc. and some HR related measures such as employee turnover and absenteeism, degree of formalisation and centralisation of HR policies and strategies, etc.). Others were added in the specific part of the questionnaire, such as organisational structure, culture, and other organisational characteristics not included in the Cranet-E survey. Both parts of the questionnaire were pre-tested and subsequently minor changes were incorporated into the final version.

Previous classification of HRM activities A previous phase in this project (Valverde et al., 2003) allowed the grouping of the 52 HRM activities into seven meaningful clusters or types of HRM activities. These types of activities constituted the starting point for our analysis, and were labelled as follows:

. Strategic decision making and leadership, including activities such as deciding on a major staff reduction, taking final decisions in collective bargaining, instigating and promoting change, designing the organisational structure, etc.

. Operational decisions and daily people management, including activities such as deciding on hiring a new full-time employee, identifying training needs, carrying out performance appraisals, deciding salary changes based on performance appraisal, etc.

. Service delivery, with activities such as acting directly as trainers or interviewing candidates.

. Policy making and diagnostic, with activities such as establishing a flexible contracts policy, defining criteria for career development or analysing the organisation’s labour market.

. Monitoring and follow-up activities, with activities such as evaluating the results of training activities or monitoring the established procedures to ensure health and safety.

. High level specialist HRM, with activities such as aligning the HR strategy with corporate strategy, establishing the general framework for labour relations in the organisation or advising on courses of action in case of conflict.

PR 35,6

622

. Administrative and technical activities, with activities such as administering contracts, storing and handling employee related information or organising means and resources for training activities.

This classification of activities also provided a general profile of distribution of HRM responsibilities among the four agents considered, as shown in Figure 1. This graphical illustration depicts a profile of who is responsible for each of the seven types of HRM activities. For example, it is possible to observe in Figure 1 that strategic activities are mainly the responsibility of top management (first profile), while the responsibility for administrative and technical HRM tasks is given mainly to the HR department (last profile). Other types of activities, such as operational and daily people management activities or service delivery activities are more spread among the various agents. It is important to understand that the profiles presented in Figure 1 represent

Figure 1. Agents’ profiles in the

seven clusters of activities

Distributing HRM

responsibilities

623

the average behaviour of the 231 organisations of the sample, and the objective of this study is to identify behaviours that are significantly different from this average profile.

Analysis Further considerations were made with regard to the possibilities for data analysis. One of the reasons for the non-existence of similar studies in the past was limitations, both in computer power and relevant statistical techniques. The use of new data mining methods in the present project highlights promising possibilities for the simultaneous study of large numbers of variables in future HRM research. The specific tool that permits the study of this sort of relationships is multiple correspondence analysis, which is an extension of correspondence analysis to allow for the simultaneous analysis of large numbers of variables. It is an interdependence technique highly adequate for dimensional reduction and perceptual mapping (Greenacre, 1989), both of which are objectives of this study. It is based on the search for associations between objects (groups in our case), and their illustration through a set of descriptive characteristics, which will assist in profiling the relationships (Hair et al., 1998). The main advantage of this technique is the possibility of plotting rows and columns in one same two-dimensional space (data matrix). Also, it does not limit itself to a simple recount of occurrences but also takes into consideration the weights of each category. This permits the calculation of statistically significant relationships.

From the analysis of the data matrix made based on the 231 cases and the classification of seven types of HRM activities, the objective was to group or cluster the organisations. The general process was:

. A principal component analysis in order to work with non-orthogonal data; and

. The clustering of organisations.

Results This section outlines the results produced by the analysis of the data. It is structured around the two propositions outlined earlier. For each proposition, the results of the analysis are described and interpreted.

In order to carry out the analysis, the organisations were clustered according to the contribution of each agent to each of the seven types of HRM activities outlined in the methodology section. The clustering process aimed at identifying profiles of behaviour that were significantly different from the average profile for the whole sample (see Figure 1), and significantly different from each other.

The dendrogram obtained from the clustering process suggested that it would make statistical sense to cut the tree graph in three, five or seven groups. The final choice was the seven-group partition as this preserved the greatest amount of statistical information and provided a richer picture of the reality of HRM in the companies studied.

The seven resulting groups of organisations and their characteristics will now be described.

Group 1: HR as a shared function among internal agents In this group, the data point to an even distribution among the three internal agents, implying a sharing of most HR activities between these agents in this cluster of 18 organisations (7.8 percent of the sample). To reflect this we use the label HR as a shared

PR 35,6

624

function among internal agents. In this cluster, with the exception of strategic decision-making and leadership activities, where top management has a distinct input, our data reveal that responsibility for all other types of HR activities are rather evenly distributed between the other three internal agents.

Remarkably, in this group of organisations, our analysis suggests that top management is not only involved in HR activities that could be considered the responsibility of “every manager” (such as operational decision making and daily people management) but is also involved in other activities, such as delivering training, monitoring and follow-up activities, and to a lesser extent even in administrative and technical activities. Similarly, line managers are involved in day-to-day activities, as one would expect (such as operational decisions and day to day management of people). However, they are also involved in more specialist HR activities.

This group of organisations is also characterised by the very limited use of outsourcing. There is only a barely noticeable presence of external providers in the delivery of training and in administrative and technical activities.

Group 2: HR as a shared function lead by top management Turning to the profile of the second group of organisations, a slightly different picture emerges from our analysis. At one level this group is not very different from the first group. The similarity lies in the fact that the three internal agents (top management, line management and HR department) have a significant input in most HRM activities. However, group 2 presents one key difference, namely, the significantly stronger involvement of top management in almost all HRM activities – hence our label HR as a shared function led by top management.

Indeed, the top management in these 60 organisations (26 percent of the sample) is a significant actor in the majority of activities: First, it has a predominant input in strategic and leadership activities. Top management input in these types of activities is high on average (57 percent) across our whole sample, but much higher in this specific group of organisations (72.3 percent). This leaves little room for line management and the HR department to be involved in strategy making (10.4 and 11.6 percent respectively). Second, top management is also the differentiating agent in activities that would be generally considered the tasks of HRM specialists. The relatively high involvement (21 percent) of top management in administrative tasks, such as deciding the types of contracts to be used, organising means and resources for training activities, or developing processes to cater for the personal needs of employees, is also noteworthy.

Top management is also the differentiating agent in more generalist activities: operational decisions and daily people management, and Policy making and diagnostic. In both of these types of activities, the distribution of agents is quite evenly spread among internal agents, with top management having a slightly higher input in activities such as acting as change agents, implementing changes in work processes or generating employee commitment. This would suggest that top management is heavily involved in day-to-day people management. Finally, group 2 makes very little use of external agencies for most HR activities, with the exception of training delivery.

Distributing HRM

responsibilities

625

Group 3: HR function, an agent for each job, a job for each agent In organisations in group 3, there is no one single agent, which can be identified as the HRM protagonist. Instead, our data would suggest that a specific agent has a highly prominent input in each set of activities. This suggests a clear-cut differentiation of activities within HRM, and the subsequent allocation of specific agents to take care of separate sets of HRM responsibilities. Hence, the label HRM, an agent for each job, a job for each agent.

The activities in which an agent has a highly salient input in this group are: strategic decision-making and leadership, with top management’s input of over 70 percent; service delivery, an area that is highly outsourced (by 76.9 percent); high level specialist HRM; and administrative and technical activities, where the HR department plays a central role (60.8 and 77.5 percent respectively).

The input from line managers in group 3 is also worth noting (ranging between 10 and 35 percent), but they do not feature any more significantly in any specific activity. Finally, this group is also unique in its significant use of external providers (33.8 percent for monitoring and follow-up activities, which, in this case, is shared with the HR department). Group 3 is a relatively small group, made up of 19 organisations (8.2 percent of the sample).

Group 4: Partial outsourcing of a wide range of HR activities The most common characteristic linking organisations in group 4 is the fact that for almost all types of activities there is a distinct and manifest presence of external provision. This does not mean that the degree of use of outsourcing is very high (as it is never larger than 20 percent), but that it extends to a wide range of HR activities. Hence, the label Partial outsourcing of a wide range of HR activities. All the organisations contained in this small group (six in total) use external providers both for activities where it would normally be expected, such as training delivery (19.6 percent) and administrative and technical activities (5.4 percent), and for activities where it is less expected, such as strategic decision making and leadership (5.6 percent) or operational decisions and daily people management (10.9 percent). Some of these activities could be included in the category of outsourcing high level consulting as identified in the literature (Cook, 1999), such as the use of consultants to assess a change process, to re-design organisational structures (strategic activities), or to develop specific HR policies (high level specialist HR).

In any case, group 4 is not to be dismissed because of its size, as it appears early on in the statistical analysis and is very stable, meaning that the behaviour of these six companies is very similar and consistent. Thus, the tendency towards outsourcing a wide range of HR activities is clear but limited to this smaller number of organisations.

Group 5: Outsourcing specialised activities and sharing generalist responsibilities Group 5 is characterised by that fact that the engagement of external providers is directed at specific types of HR activities. They mainly outsource administrative and technical activities (12.2 percent), training delivery (19.6 percent), and many high level specialist HRM activities (16.8 percent). These are all specialised activities within the HR function.

The outsourcing of these activities is counteracted by a smaller intervention of the HR department in these areas with the consequence that for this group, it is possible to

PR 35,6

626

talk about the pursuit of a leaner HR department (Griffiths, 1993). On the other hand, the three internal agents (top management, line management, and HR department) share the activities related to managing people from a more generalist point of view (such as operational decisions and daily people management, leadership, etc.).

Group 6: HR function as the exclusive domain of the HR department Because of the strong predominance of the HR department in all types of activities, this group of 30 organisations has been labelled HR function as the exclusive domain of the HR department. Indeed, the values of participation of this department range from 70 to 90 percent in five of the seven types of activities explored in our analysis.

It is fair to question whether there is a relationship between the prominent role for the HR department and the fact that HR specialists were the respondents to the questionnaires. However, an erroneous self-attribution from respondents would have resulted in an indiscriminate allocation of high values, regardless of the activity. Instead, the two most generalist activities (strategic decision making, and operational decisions and daily people management) have attained smaller values for the HR department and larger values for top and line management. This supports the validity of the findings. In summary, this group of 30 organisations reflects a powerful and pervasive HR department, with little by way of a role for external providers among these.

Group 7: HR function as the domain of the HR department supported by the line The profile of the final group, comprising 86 organisations (over one third of respondents), is, in part, similar to the previous group with one major exception i.e. the prominent role of Line management in certain activities. Thus in this group HR department is the main agent in four types of activities and the differentiating agent in strategic activities, just as in the previous group. However, there are two types of activities where line management is the protagonist: operational decisions and daily people management, and service delivery.

In this regard, the HR department counts on the support of the line for the implementation and dissemination of personnel policies at operational level (such as carrying out performance appraisals or deciding salary changes based on the results of performance appraisal) and, particularly, for the delivery of training. The role of the line in this group of organisations is as executors of established policies. Hence the label, HR as the domain of the HR department, supported by the line.

The fact that it was possible to meaningfully segment and group the organisations in our data set leads us to suggest that there is support for P1. The multiple correspondence analysis identified seven clusters or groups of organisations according to the relative involvement of the four HR agents in the seven types of HRM activities. Thus, the groups obtained reflect different approaches to distributing the work involved in carrying out HRM activities among the HR department, top and line management and external agencies.

The second proposition essentially addresses the question of whether there are any contextual characteristics that are common to the organisations in each of the seven groups identified. The introductory section of this paper considered the importance of the organisational context in the study of management models in general and of HRM in particular. Whether from a “fit”, contingency, or configurational perspective, the

Distributing HRM

responsibilities

627

variables that make up the internal and external organisational context are assumed to play an important role on the determination of HRM models and practices (Duberley and Burns, 1993; Delery and Doty, 1996; Wright and Snell, 1998).

Thus, a number of variables indicating internal and external dimensions of the organisations studied were measured to examine whether each group had similar contextual characteristics.

The statistical identification of these possible common characteristics is normally referred to as characterisation or profiling of groups (Hair et al., 1998) and it usually represents the final step in a clustering process. The process simply involves the comparison of clusters (the seven groups of organisations obtained in the multiple correspondence analysis) on the basis of relevant contextual illustrative variables.

The comparison of average values of these illustrative variables across clusters can be obtained through:

. The observation of the values of the illustrative variables in the results of the multiple correspondence analysis (MCA).

. Carrying out basic cross tabulations.

The results of the MCA showed surprisingly few illustrative variables appearing as significantly different among the seven groups. In fact, no one single contextual variable (for example, sector) appeared in the outputs of significant differences for all groups. This was a clear indication that, if differences were found, they would probably be slight and not statistically significant.

Therefore, it was necessary to turn to the running of crosstabs, knowing from the outset that any differences that we might find would not be robust. With the crosstabs carried out, virtually no differences were found. Thus, the remainder of this section will simply outline the variables measured (more details in Appendix 2) in order to facilitate possible replication of the study:

. Sector. Even though the variable sector is often cited as a determinant of types of HRM (cf. Andrews, 1986; Brewster et al., 1996), it is not clear which models are more likely to predominate in a given sector (Labelle, 1992). In this study, sector was one of the contextual variables that did produce some differences, but only in a limited number of groups. In particular, there was a higher proportion of organisations in the service sector among the Shared among internal agents group, while the Agent for each job, job for each agent group were more likely to be in the industry sector.

. Organisational size. Even though the survey was directed to organisations of 200 or more employees (in order to maximise the possibility of finding organisations with a HRM department), differences were nevertheless sought on the basis of their size (medium: 200-499 employees; large: 500-1,999 employees; very large: 2,000 or more employees). No relationship was found between an organisation belonging to a particular group and its number of employees.

. Structure. Structure was operationalised as more than one variable: first, organisations were asked to choose which type of structure was most similar to theirs (simple, functional, matrix, etc.). Second, a number of measures were taken on a semantic scale regarding the degree of centralisation, formalisation of

PR 35,6

628

policies and procedures, communication channels and hierarchical levels. No relationship was found for any of the variables related to structure.

. Technological system. Also measured on a semantic scale, no relationship was found for either the degree of standardisation or the complexity of processes undertaken in the organisation.

. History. The history of the organisation was operationalised into two groups of variables: First, the age of the organisation. For this, no relationship at all was found for most groups, except in group 1 (HRM as a shared function), which contained younger organisations. Second, organisations were asked whether they had been involved in acquisitions, mergers or takeovers during the last three years. No relationships were found regarding these measures.

. Employee characteristics. No relationship was found with regard to the level of turnover, their degree of unionisation or the proportion of workers with university education. A very faint relationship was found regarding work intensity, with groups 1 and 2 being more work intensive.

. Environment. Four measures were taken on the characteristics of the environment. However, no relationship at all was found with the degree of uncertainty facing organisations, or for the speed of change (dynamic), complexity or hostility they are subject to.

. Culture. Although there are comprehensive and well tested questionnaires for the measurement of culture, the survey design meant that a shorter format was needed. Thus, the decision was made to produce a summarised description of cultural orientations (Valverde, 2001b) according to the classification by Harrison (1972). Respondents were then asked to read these descriptions and allocate 100 points distributed according to the similarity between the culture in their organisation and the descriptions provided. No differences were found for three types of culture (power, achievement, and role). Only people culture distinguished some groups, with Group 1 (HR as a shared function) including more organisations with people oriented cultures and Group 6 (Exclusive domain of the HR department) including many organisations with no people culture at all. Because these results on people oriented culture are not statistically significant, caution should be placed in making inferences about them.

. Characteristics of the HR function. A number of measures were taken to operationalise the characteristics of the HR function. Among them, no relationship was found regarding the size of the HR department, the presence of the HR manager on the board of directors, nor the existence of a HR strategy.

As seen in this summary of the profiling results, the analysis of the internal and external contextual variables was not successful in differentiating the groups with regards to their characterisation. This finding means that P2 is not confirmed. This is discussed in the conclusions section.

Discussion The analysis of the data produced seven groups, each involving a number of organisations. The distribution of HR activities was similar among the organisations within each of the seven groups (and thus different to the distribution of HR activities

Distributing HRM

responsibilities

629

of the organisations in the other groups). Therefore, each of the seven groups of organisations was labelled according to their main traits, as shown in Table I. This confirms P1.

The two larger groups in the sample reflect the two major tendencies in terms of the contribution of each agent to HRM. Group 2 (26 percent of the organisations in the sample) is an example of the tendency of shared involvement by different agents. Group 7 (37.2 percent of the organisations in the sample) represents the tendency towards a higher input of the HR department in all activities. It is also interesting to observe in detail the arrangements of the smaller groups. The occurrence of this type of distribution may be less common but more innovative than the groups with more representation. Indeed, this may be an interesting development in terms of revealing possible new directions in arranging the responsibilities of the agents in HRM.

In order to find potentially common characteristics among the organisations in each group (P2), information was obtained for a number of contextual factors as well as specific characteristics of the HR function. Unexpectedly, among the organisations in each of the seven groups, no common contextual characteristics of any significance were identified. Thus, P2 is not confirmed. The research did not identify any relationship between the contextual characteristics of the organisations and particular ways of managing HR. In each group it is possible to find examples of different sectors, sizes, degree of unionisation, structures, etc. for all contextual factors measured.

Little evidence of similar findings is available in the HRM literature. Rather, this literature suggests that HRM policies and practices are determined by contingent variables (e.g. Jackson and Schuler, 1999; Mayne et al., 1996). However, Budhwar (2000), in a study on strategic integration and devolution to the line, found that organisational contingencies were not related to specific levels of integration or devolution. Moreover, Poole and Jenkins (1997), when also studying devolution, found some sector and size effects, but acknowledged that “they do not generally indicate a highly contrasting pattern” (p. 353). Further investigation of these results is therefore necessary.

A number of limitations for this study must be outlined in order to understand its scope and guide future research.

With regard to the questionnaire: in general, the measurement instrument took too long to answer. At this stage of theory development this was a necessary evil, as it would not have been rigorous to reduce further the 52 items (which were the result of a previous reduction from a longer list) without having tested them first with an

Group Label No. org. Percent org.

1 HR as a shared function among internal agents 18 7.8 2 HR as a shared function led by top management 60 26 3 HR function, an agent for each job, a job for each agent 21 9.1 4 Partial outsourcing of a wide range of HR activities 6 2.6 5 Outsourcing specialised activities, sharing generalist

responsibilities 7 3

6 HR function as the exclusive domain of the HR department 33 14.3 7 HR function as the domain of the HR department supported by the

line 86 37.2

Total 231 100

Table I. Labels and sizes of the final groups of organisations

PR 35,6

630

extensive sample. However, it is important to highlight that the reduction of items should be possible from now on, given the satisfactory results of this first round.

With regard to the unit of observation: perhaps the most important limitation of this study is the fact that the HR manager was chosen as the representative of the organisation (unit of observation) to respond to the questionnaire. Indeed, asking HR managers about their own function involves measuring a partial perception of reality. It has long been identified that HR managers tend to overestimate their contribution to HR activities (Ritzer and Trice, 1969; Mintzberg, 1978). Thus, even though HR managers are adequate respondents because of their knowledge of the topic, a multiple respondent approach, including representatives of the four agents studied, could have added richness to the results of the study.

With regard to the analytical techniques: even though new data mining techniques offer significant opportunity for the exploration of large amounts of data and variables, it must be said that not all statisticians consider these techniques to be fully developed (Aluja Banet and Morineau, 1999).

The main contribution of the study consists of the incorporation of existing partial areas of study in the field of HRM (roles of the HR department, devolution to line management, outsourcing of the HR function) into an integrated study. This represents a new and more holistic approach to the study of HRM.

Conclusion The results of the study empirically confirm the first proposition that the authors originally set out to explore, producing a classification of ways of arranging the participation of agents in HRM. However, the second proposition with regard to common characteristics among the organisations of each group was not confirmed.

An alternative explanation may explain this result: that the model of HR function adopted by an organisation may not be contextually determined, but instead it is a matter of corporate choice. That is, companies actually choose what type or model of HRM they want and distribute the responsibilities of their various agents accordingly, rather than being determined by organisational contingencies. If this assertion is true, it poses a number of questions in regard to future investigation. For example, who makes this corporate choice, a specific person (and if so, who: director general, HR manager, etc.) or a dominant coalition? Does the presence or influence of a single person affect the choice (for example, a HR manager with certain competencies, experience, etc.) or does the configuration of agents in the HR function follow traditions within the organisation? And even, is there any relationship between the distribution of HR responsibilities and the distribution of responsibilities in other management areas?

It is important to highlight that the concept of choice has been dealt with in the literature as the opposite of situational or environmental determinism (Poole, 1990). However, the idea of HRM policy or strategic choice, based on the notion that managers can exert considerable discretion when making HRM choices (Flood, 1998; Gunnigle, 1991), always seems to be juxtaposed with a range of internal and external constraints and limitations to these choices (Kessler et al., 2000). Since the present study found none of the variables observed to be determinants of the distribution of HRM responsibilities between different agents, the idea of choice proposed here is even more open (more management discretion), and it challenges the notion that discretion must be juxtaposed with contingent constraints.

Distributing HRM

responsibilities

631

Finally, the closing of an exploratory research project must involve putting forward the next set of questions to be studied in order to prepare the ground for the continuation of the research (Gill and Johnson, 1991). Bearing in mind the results of this study, together with the limitations outlined above, a number of suggestions are made in order to indicate the focus that further research might adopt.

The study of a large number of organisations has proved very useful in identifying a wide range of behaviours in the distribution of HR responsibilities. More research along the same line (even straight replication) is needed in order to consolidate the results of this study. Further, since the study was carried out in only one country, it would be interesting to see whether the involvement of HR agents in the various types of activities is similar for different countries.

Another area of further development is related to the measurement instrument and collection procedures. In this sense, the questionnaire should be refined and shortened by using the results of the reduction of variables obtained in this project and by carrying out consecutive rounds of data collection in order to assess validity and reliability. Also, regarding data collection procedures, it would be valuable to incorporate the views and perceptions of HRM agents other than HR managers. This would be achieved by including top and line managers, and representatives of external agencies as respondents.

Perhaps the most interesting and potentially fruitful area of further research is that related to P2, that is, the effect of contextual variables on the adoption of one of the seven types of HRM according to their distribution of HR activities among agents. The alternative explanation put forward in this paper (that is, the arrangements of HRM agents being a matter of management choice) needs to be explored empirically (perhaps with more in-depth methodologies) and contrasted with alternative explanations such as environmental determinism and strategic HRM choice.

Note

1. For example, in Europe, Hall and Torrington (1998b) measured the role of the HR department, their involvement in strategy making, outsourcing and devolution to the line in one extensive study, and in the USA the Corporate Leadership Council (1995) studied the dissolution of traditional HR, the HR department as service to the line, and the use of consultants.

References

Aluja Banet, T. and Morineau, A. (1999), Aprender de los datos: el análisis de componentes principales. Una aproximación desde el data mining, Ediciones Universitarias de Barcelona (EUB), Barcelona.

Alverson, M. and Skölberg, K. (2000), Reflexive Methodology, Sage, London.

Andrews, J.R. (1986), “Is there a crisis in the personnel department’s identity?”, Personnel Journal, Vol. 65 No. 6, pp. 86-93.

Brewster, C., Hegewisch, A. and Lockhart, J.T. (1991), “Researching human resource management: methodology of the Price Waterhouse Cranfield Project on European Trends”, Personnel Review, Vol. 20 No. 6, pp. 36-40.

Brewster, C., Tregaskis, O., Hegewisch, A. and Mayne, L. (1996), “Comparative research in HRM: a review and an example”, International Journal of Human Resource Management, Vol. 7 No. 3, pp. 585-604.

PR 35,6

632

Budhwar, P.S. (2000), “Evaluating levels of strategic integration and devolvement of human resource management in the UK”, Personnel Review, Vol. 29 No. 2, pp. 141-61.

Burrell, G. and Morgan, G. (1979), Sociological Paradigms and Organisational Analysis, Heinemann, London.

Cook, M.F. (1999), Outsourcing the Human Resources Function: Strategies for Providing Enhanced HR Services at Lower Cost, AMACOM, New York, NY.

Corporate Leadership Council (1995), Vision of the Future. Role of Human Resources in the New Corporate Headquarters, The Advisory Board Company, Olympia, WA.

Cranet Survey (2000), Estudio Cranfield-ESADE: gestión estratégica de recursos humanos, una década de investigación. Informe de conclusiones 2000, ESADE, Barcelona.

Cunningham, I. and Hyman, J. (1999), “Devolving human resource responsibilities to the line; beginning of the end or a new beginning for personnel?”, Personnel Review, Vol. 28 Nos 1/2, pp. 9-27.

Delery, J.E. and Doty, D.H. (1996), “Modes of theorizing in strategic human resource management: tests of universalistic, contingency, and configurational performance predictions”, Academy of Management Journal, Vol. 39 No. 4, pp. 802-35.

Donaldson, L. (2001), The Contingency Theory of Organizations, Sage, London.

Duberley, J.P. and Burns, N.D. (1993), “Organizational configurations: implications for the human resource/personnel management debate”, Personnel Review, Vol. 22 No. 4, pp. 26-34.

Flood, P. (1998), “Is HRM dead? What will happen to HRM when traditional methods are gone?”, in Sparrow, P. and Marchington, M. (Eds), Human Resource Management: The New Agenda, FT Pitman, London.

Gill, J. and Johnson, P. (1991), Research Methods for Managers, P. Chapman, London.

Gunnigle, P. (1991), “Personnel policy choice: the context for human resource development”, Journal of European Industrial Training, Vol. 15 No. 3.

Gunnigle, P. and Moore, S. (1994), “Linking business strategy and human resource management: issues and implications”, Personnel Review, Vol. 23 No. 1, pp. 63-84.

Gratton, L., Hope-Hailey, V., Stiles, P. and Truss, C. (1999), “Linking individual performance to business strategy: the people process model”, in Schuler, R.S. and Jackson, S.E. (Eds), Strategic Human Resource Management, Blackwell, Oxford.

Greenacre, M.J. (1989), Theory and Applications of Correspondence Analysis, Academic Press, London.

Griffiths, W. (1993), “A leaner, fitter future for HR?”, Personnel Management, October, pp. 36-9.

Hair, J.F., Anderson, R.E., Tatham, R.L. and Black, W.C. (1998), Multivariate Data Analysis, 5th ed., Prentice-Hall, Englewood Cliffs, NJ.

Hall, L. and Torrington, D. (1998a), “Letting go or holding on – the devolution of operational personnel activities”, Human Resource Management Journal, Vol. 8 No. 1, pp. 41-55.

Hall, L. and Torrington, D. (1998b), The Human Resource Function. The Dynamics of Change and Development, FT-Pitman Publishing, London.

Harrison, R. (1972), “Understanding your organization character”, Harvard Business Review, May-June.

Jackson, S.E. and Schuler, R.S. (1999), “Understanding human resource management in the context of organizations and their environments”, in Schuler, R.S. and Jackson, S.E. (Eds), Strategic Human Resource Management, Blackwell, Oxford.

Jackson, S.E. and Schuler, R.S. (2000), Managing Human Resources: A Partnership Perspective, 7th ed., South-Western College Publishing (Thomson Learning), Cincinnati, OH.

Kessler, I., Purcell, J. and Coyle Shapiro, J. (2000), “New forms of employment relations in the public services: the limits of strategic choice”, Industrial Relations Journal, Vol. 31 No. 1, pp. 17-34.

Distributing HRM

responsibilities

633

Labelle, C. (1992), “A role-based taxonomy of human resource organizations”, Cornell University, PhD thesis, New York, NY.

Labelle, C. and Dyer, L. (1992), A Role-based Taxonomy of Human Resource Organizations, working paper, Center for Advanced Human Resource Studies, Cornell University, New York, NY, pp. 92-35.

Layder, D. (1993), New Strategies in Social Research, Polity Press, London.

Legge, K. (1995), Human Resource Management Rhetorics and Realities, Macmillan, Basingstoke.

Lowe, J. (1992), “Locating the line: the front-line supervisor and human resource management”, in Blyton, P. and Turnbull, P. (Eds), Reassessing Human Resource Management, Sage, London.

Mayne, L., Tregaskis, O. and Brewster, C. (1996), “A comparative analysis of the link between flexibility and HRM strategy”, Employee Relations, Vol. 18 No. 3, pp. 5-24.

Mintzberg, H. (1978), “Patterns in strategy formation”, Management Science, Vol. 24 No. 9, pp. 934-48.

Mohrman, S.A. and Lawler, E.E. III (1999), “The new human resources management: creating the strategic business partnership”, in Schuler, R.S. and Jackson, S.E. (Eds), Strategic Human Resource Management, Blackwell, Oxford.

Poole, M. (1990), “Editorial: human resource management in an international perspective”, International Journal of Human Resource Management, Vol. 1 No. 1, pp. 1-15.

Poole, M. and Jenkins, G. (1997), “Responsibilities for human resource management practices in the modern enterprise: evidence from Britain”, Personnel Review, Vol. 26 No. 5, pp. 333-56.

Redman, T. and Allen, P. (1993), “The use of HRM consultants: evidence from manufacturing companies in the north-east of England”, Personnel Review, Vol. 22 No. 2, pp. 39-55.

Ritzer, G. and Trice, H.M. (1969), An Occupation in Conflict: A Study of the Personnel Manager, Cornell University, New York, NY.

Schuler, R.S. and Jackson, S.E. (1999), Strategic Human Resource Management, Blackwell, Oxford.

Sisson, K. and Storey, J. (2000), The Realities of Human Resource Management. Managing the Employment Relationship, Open University Press, Buckingham.

Storey, J. (1992), Developments in the Management of Human Resources, Blackwell, Oxford.

Tayeb, M. (1987), “Contingency theory and culture: a study of matched English and the Indian manufacturing firms”, Organization Studies, Vol. 8 No. 3, pp. 241-61.

Truss, C., Gratton, L., Hope-Hailey, V., Stiles, P. and Zaleska, J. (2002), “Paying the piper: choice and constraint in changing HR functional roles”, Human Resource Management Journal, Vol. 12 No. 2, pp. 39-63.

Tyson, S. and Fell, A. (1986), Evaluating the Personnel Function, Hutchinson, London.

Ulrich, D., Losey, M.R. and Lake, G. (1997), Tomorrow’s HR Management, Wiley, New York, NY.

Valverde, M. (2001a), “Mapping the HRM function: an exploratory study of responsibilities and agents in the managing of people”, doctoral thesis, Universitat Rovira i Virgili, Reus.

Valverde, M. (2001b), Introducció a la gestió de recursos humans, Ediuoc, Barcelona.

Valverde, M., Quijano, S. and Soler, C. (2003), “HRM as a shared function: an exploratory study of responsibilities and agents in the managing of people”, paper presented at the IV International Workshop on Human Resources, Cadiz, May.

Wright, P.M. and Snell, S.A. (1998), “Toward a unifying framework for exploring fit and flexibility in strategic human resource management”, Academy of Management Review, Vol. 23 No. 4, pp. 756-73.

PR 35,6

634

Appendix 1

Appendix 2

Summary of context variables Tables AI and AII.

Figure A1. Fragment of questionnaire to illustrate its structure

Sector Sector Categories

Size Size as indicated by number of employees Data History Organisation involved in mergers, acquisitions and

takeovers Age (as indicated by the year when organisation was founded)

Y/N

Employee characteristics Employee turnover Employee absenteeism

Data Data

Proportion of labour costs as percentage of operating costs (as indicator of labour-intensive organisations) Proportion of trade union membership among employees

Data Data

HR function characteristics Degree of formalisation of HR policies Degree of formalisation of mission and strategy

SD SD

Degree of centralisation-decentralisation of HR policies

Presence of an HR department Y/N Size of HR department. Basic measure to be turned into a ratio of number of HR professionals for every 100 employees

Data

Presence of the HR manager in the board of directors Y/N Degree of involvement of the HR manager in strategy development

SD Table AI. From the Cranet-E survey

Distributing HRM

responsibilities

635

From the Cranet-E survey From the specific questionnaire More details of the questionnaire construction, variable operationalisation and final primary research design may be obtained from the authors by contacting Mireia Valverde at: [email protected]

Corresponding author Mireia Valverde can be contacted at: [email protected]

Structure Which of these organisational structures resembles more that of your organisation?