SCR 8518
Diffusing HRM
An empirical stur!J
ef
the implementation
ef
Human Resource Management poliry in the
Dutch operations
ef
two 1\1NCs in the food industry.
Master Thesis
Hester Houwing
9840893
Department of sociology
Section labour and organisational sociology
University of Amsterdam
Diffusing HRM
An empirical study of the implementation of Human Resource Management
policy in the Dutch operations of two MN Cs in the food industry.
Contents
A c k n o w l e d g e m e n t s - - - 7
1. Introduction 9
2. Diffusion via structure, power, and culture 11
2.1 Structure follows strategy 12
2.2 Power 15
2.3 Corporate culture 15
2.4 National environment of subsidiaries 18
3. Defining the central concept and operationalisation 25
3.1 Identifying HRM 25
3.2 Research design 26
3.3 Making the theory operational 28
4. Unilever 31
4.1 Company profile 31
4.2 Developments in human resource management 33
4.3 Structure, power, and culture 37
5. Sara Lee 41
5.1 Company profile 41
5.2 Developments in human Resource Management 43
5.3 Structure, power, and culture 47
6. Diffusion 51
6.1 Differences in diffusion among HRM practices 51
6.2 Comparing the case studies 53
7. """'~L~UL'JLL-~~~--~---~---~---·---~59 Sources _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ 61 tera.turc=-~~~~~~~~~~~~~~~~~~~~~-61 Inte1-views - - - 62 Websites _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ 63 ~,. _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ 64
Acknowledgements
This thesis is an outcome of a research commissioned by the Dutch trade union federation FNV and preformed by AIAS, the Amsterdam Institute for Advanced labour Studies. I was very happy to be able to participate in this research on globalisation and industrial relations as an intern, and it enthused me to write this thesis on diffusion of HRM in multinational companies. Two people who took part in the research were also my supervisors, Jelle Visser and Marc van der Meer. I would like to thank them first and foremost as they are both engaged in so many projects but were always dedicated when discussing my thesis and encouraged me to stay focussed. Secondly the other researchers taking part in the project on globalisation: Rob de Boer, Marta Kahancova, and Damien Raess, who took time out of their busy schedules to provide me with comments and advice. The people working with me on the 'power floor' at AIAS, who were always the first to know when something went wrong during the writing of my thesis, and had to cope with my complaining. My sister Charlotte who took time out to read one of the first drafts of this thesis and corrected my English writing. Finally David whose sense of ambition is an important motivation for me, and has supported me throughout the writing of this thesis.
Hester Houwing,
Introduction 9
1. Introduction
In the globalising economy of today's world multinational corporations (hereafter MNCs) find themselves under increasing global competition with other MNCs. In their search for a strategy to improve their competitiveness MNCs experience pressures for local differentiation as well as for global integration. In addition to experiencing pressures as a result of globalisation MN Cs are proactive in shaping and determining processes associated with globalisation; MNCs are pivotal actors when it comes to rapidly transferring goods and capital across the globe. In deciding on a global strategy companies continuously weigh the costs and benefits of on the one hand adopting a strategy based on internal consistency or on the other hand typified by adaptiveness to the conditions in local operations. The same considerations have to be made with respect to the handling of personnel, currently usually referred to as human resource management (hereafter HRM). MNCs make decisions on whether they should adopt a top-down approach in setting out HRM policy, or if they should let operations in otl1er countries decide on how personnel can be dealt with in the most appropriate rnanner.
The aim of this thesis is to show to what extent HRM policy in the Dutch subsidiaries of two MNCs is imposed on local management from company headquarters and to what degree local management and workers determine the way HRM policy is implemented. Rubery and Grimshaw have noted that the diffusion of similar practices that have proven to work within the company (often termed 'best practices') among subsidiaries yields various benefits. 111ese include administrative efficiency and technical and organisational gains (2003: 199-200). On the other hand, top-level management often weighs the potential benefits that might accrue from a certain amount of freedom to local 1nanagement to adapt corporate directives to the local environment (Harzing and Van Ruysseveldt, 1995). In addition to the degree of freedom that MNCs accord to their subsidiaries, management as well as employees' representatives in local operations make use of the resources they have at their disposal to exert power to influence policy set out at corporate headquarters.
Local HRM policy is the result of a process of diffusion of HRM policy within a MNC. The idea underlying the concept of diffusion is tl1at HRM policy is formulated somewhere in the MNC and then spreads throughout its (international) operations. HRM policy as it takes shape in the local subsidiaries of a MNC can be expected to be the outcome of conflicting interests of on the one hand international company management and on the other hand of management and employees in foreign operations. Company structure is the first determinant of the degree of centrality in the setting out of corporate policy. The more centralised a MNC is structured, tl1e more decisions (tegarding HRM policy) are made in a top-down manner. Secondly, the amount of power that actors can wield over other actors determines the influence that local management and employees can exercise in developing policy. Power is based on the control over resources that are valued highly by international company management, such as the profitability or special expertise of a subsidiary. Power is closely bound up with a third central determinant of the diffusion process: corporate culture. The values that characterise the telationships between actors or groups of actors within a MNC determine to a large extent how HRM policy is diffused and as a result how local HRM policy takes form. These three variables are seen here as pivotal in the
10 Diffusing HRM
process of diffusion of HRM policy within :MN Cs. The provisional hypothesis is that changes in these variables matter in the process of diffusion of HRM policy.
The concept of power deriving from the possession of resources valued by others has been developed by Ferner and Edwards1 (1995), and used by Coller and Marginson (1998) in an
empirical study of a large :MNC in the food industry. The analysis conducted here draws on the research done by Coller and Marginson but extends it in two fundamental ways: first, instead of research based on one :MNC, this thesis will deal with an empirical comparison between two :MNCs. This might provide an impetus for generalisation. Secondly, Coller and Marginson point to the importance of trade unions and works councils and the power they can exert in the diffusion process, an issue that they did not focus on in their study (p.15). This thesis will also shed light on these actors in the process of transferring HRM policy.
This thesis explores how changes in HRM policy find their way from the point where new HRM strategies are formulated to the implementation in the various operations that make up the organisation. The national context in which the subsidiaries operate in terms of industrial relations and corporate governance will also be taken into account. The data is based on empirical research within the Dutch operations of two large :MNCs in the food sector, Unilever and Sara Lee, and focuses on the diffusion of HRM policy within the corporations. Both cases serve as an interesting object for study as both have experienced significant changes in their HRM policy over the last five years.
The analysis is guided by the following research questions:
Which mqjor changes have taken place with regard to HRM poliry in the Dutch operations qf Unilever and Sara Lee over the last five years?
Flow is HRM poliry df!fused· from the cotporate centre Jo the subsidiaries (top do2vn) or strongly influenced
ry
local conditions?
Wry is 1~TRM poliry d~ffused the way it is and what is the role qf cotporate management, local management and emplqyees' representatives in the diffusion process?
Whal is the role qf compaf!Y structure, power, cotporate culture, and the national context in the process qf diffusion?
In the next chapter the theoretical framework underlying the research is discussed. This is followed by a definition of the central concept, an operationalisation, and a review of the methodology in chapter three. Chapters four and five present a description of the two cases selected for study: Unilever NL BV and Sara Lee/DE. In addition the major changes in HRM policy of the Dutch subsidiaries of both :MNCs are presented here. The final two chapters analyse how changes in HRM policy have been diffused and which of the variables have been most essential in this process.
1 Power is understood in this thesis as the overcoming of the resistance of other actors to achieve a desired result (Ferner and Edwards, 1995: 231)
Diffusion via structure, power, and culture 11
2. Diffusion via structure, power, and culture
HRM policy in the subsidiaries of MN Cs is the outcome of company characteristics, especially how these characteristics determine the way conditions in the local environment of the subsidiary are dealt with. These company characteristics are here divided into company structure, power, and culture (see 2.1). These three variables determine the extent to which HRM policy is modelled after the strategy that the corporate office has set out, or follows local practices. MNCs have become ever more subject to conflicting forces that on the one hand lead companies to centralise substantive parts of their business, while on the other wield pressure to adapt to local
Structure
Structure
Figure 1. Variables influencing HRM policy in local operations
Company Management [Strategy/HRM policy] Power Local management Power Culture Culture
Industrial relations Corporate governance
structure International level National level LOCAL
HRM
POLICY12 Diffusing HRM
circumstances (Coller and Marginson, 1998). The choices that top management makes in this respect are reflected in the structure of the company, the corporate culture, and power relations between (actors in) the parts that make up the company.
HRM policy not only takes shape under the influence of company characteristics such as structure, culture, and leadership style. The environment in which subsidiaries operate allows local management and employees to influence the diffusion process by creating a situation in which: "[n]ext to adaptive behaviour there is room for independent and innovative (idiosyncratic) behaviour" (Paauwe, 1991: 111). The important factors in the environment and consequently in the allowing room for manoeuvre in local operations are the corporate governance structure of the host country and the system of industrial relations (see figure 1 ).
HRM policy is made up of a number of distinct practices, for example staffing and employee development. Rosenzweig and Nohria (1994) have argued that each of these practices vary in the extent to which they are to be carried out the way top management prescribes or left free to be carried out on the basis of local circumstances:
HRM practices for which there are well-defined local norms and which affect the rank-and-file of the affiliate organisation are likely to conform most closely to the practices of local competitors. Practices for which there are diffuse and poorly defined local norms, or which are seen as being critical to maintaining internal consistency or arriving at critical decisions, are less likely to conform to local norms (Rosenzweig and Nohria, 1994: 233).
Rosenzweig and Nohria identified six HRM practices for their study. They ranked these six practices according to the degree of conformance to local practices, and found that their data supports this ranking. Starting with the HRM practice that most closely resembles local practice, their ranking is as follows: 1. time off; 2. benefits; 3. gender composition; 4. training; 5. executive bonus; and 6. participation. In chapter three the splitting up of HRM policy in distinct practices is taken up. This categorization differs a little from the one by Rosenzweig and Nohria. In chapter six will be determined to what extent the manner of diffusion fluctuates over the different HRM practices.
2.1 Structui-e follows strategy
One of the major tasks of the top corporate management of MNCs is the development and execution of company strategy. The way HRM is applied is directed by the strategic goals of a company (Haveman, 2003b: 77). Purcell (1989) distinguishes between three types of company strategy: first-order, second-order and third-order strategies. In a first-order strategy long-term organisational goals are set forth. A company's second-order strategy is the way in which an organisation is structured in order to achieve the first-order strategy. Both strategies shape the context in which third-order, or functional, strategies are laid out. HRM practices are a type of functional strategy, and are thus affected by first and second order strategies (Purcell, 1989: 67). Harzing (1995: 28) refers to strategy as the (one or various) product/market combination(s) of a company. For each stratet,ric business unit, a company generally identifies a market entry and participation strategy (for example wholly- or partly-owned subsidiaries), and a competitive strategy. She depicts organizational structure as a key element in the implementing of company strategy and HRM as a functional strategy in the furthering of a company's objectives.
Diffusion via structure, power, and culture 13
Where these functional strategies are formulated is the outcome of the first-and second-order strategies that are on their part the result of the way top management decides how to handle contradictory environmental pressures. There are ER-managers in the highest right down to the lowest levels in an organisation. If a MNC adopts first- and second order strategies based on centralisation, then the ER-managers in the higher levels with have more authority in the setting out of HRM policy relative to the executives at a lower level. Conversely, if a company follows a strategy typified by a high degree of responsiveness to local conditions, then the corporate office will restrict its influence to the setting out of broader HRM directives that are left free to be filled in at lower ranks in the organisation.
Purcell terms first-order strategy 'portfolio planning'. He argues that as companies experience increasing pressures to become more diversified, the corporate office is required to assess what constitutes the appropriate mix out of a portfolio of businesses in terms of costs and possibilities to generate cash for the company. The existence of a portfolio of businesses creates opportunities for local managers to handle employee relations in the way they deem useful, and this creates possibilities to handle HRM issues in different ways in various operations. Because local managers experience more freedom, variations in broad ERM-guidelines set out at corporate headquarters are more likely to occur. Second-order strategies bear upon the extent to which a multi-divisional corporation is decentralised, how business units are managed, and what the nature is of the relations between business units (e.g. if they are dependent on each other, or if they are vertically integrated). As in the case of first-order strategies, here too the higher the level of responsibility assigned to lower levels such as subsidiaries, the higher the degree of freedom to shape the process of diffusion of HRM policy.
Co1npaf!y structure
In describing the link between company strategy and company structure Paauwe and Dewe (1995a) refer to Chandler's (1962) axiom 'structure follows strategy'. Paauwe and Dewe point out that with this adage Chandler argues that different growth strategies of companies call for different administrative structures (1995a: 54). Examining strategy-structure linkages without including managerial values and attitudes as well as other processes are termed first-generation models (ibid. 55). These first-generation models strongly emphasize the importance of a 'structural fit' between strategy and organisational structure. Without dismissing the 'structure follows strategy' adage realisation grew among scholars that the concept of structural fit ceased to capture the complexities that companies increasingly faced. In addition, first-generation approaches reduced the roles and responsibilities of individuals to static elements. As a result second-generation approaches started to develop in the 1980s (ibid. 56). The difference between the two schemes is that whereas the first is occupied with the structural fit between strategy and structure, the element of investigation in the second is the role of management as an instrument to handle the implementation of company strategy. In this process the leadership style is supporting, and management directs their efforts towards stimulating employee development and learning (Paauwe and Dewe, 1995b: 76). Employee development as well as other HRM practices gain relevance in these approaches.
14 Diffusing HRM
The structure of a MNC has an important influence on the way HRM policies are transferred (Rubery and Grimshaw 2003: 205, Purcell 1989). To analyse structure I will use the typology of MNCs by Ferner and Edwards (1995), who differentiate between:
1. Diversified conglomerates of firms relying primarily on financial controls; 2. Integrated multinationals relying on 'organic' growth;
3. Network-based firms growing through acquisition; and
4. Federal (business services) firms formed as associations of independent national partnerships.
It should however be noted that these types of firms are ideal types and it is possible that companies show some variations based on their particular company history. The company structure, dependency among units, and cultural relations that exist within these types of companies can be typified as follows.
Structure/ Authority relations Resource dependency/ exchange relations Culture relations Diversified financial control MNC Strong - Top-down Units highly dependent on centre Weak Source: Ferner and Edwards, 1995: 244
Type ofMNC Integrated MNC Strong - Top-down Units highly dependent on centre Strong Decentralised network-based MNC Weak Inter-unit exchange relations Very strong FederalMNC Weak Inter-unit exchange relations Strong
Coller and Marginson (1998:9) have made an attempt to identify in which type of MNC transnational influence over employment practices is most far-reaching. They find that a certain kind of structure highly determines the degree of influence from the top of the organisation to the lower ranks. In this structure the MNC is organised around international businesses rather than national subsidiaries. For the power that local management and employees can exert in influencing policy it is an essential difference whether a MNC is organised per country or along the lines of economic activities or brands/products. The reason for this is the way employees' representatives are organised, which is mainly nationally. Additionally, business strategies that attempt to secure international synergies or in other words try to integrate production also tend to exert more transnational influence. Drawing on Coller and Marginson the typology of Ferner and Edwards can be further elaborated. Therefore in this thesis a further distinction is made between integrated MNCs organised per country and integrated MNCs organised per economic activity.
Diffusion via structure, power, and culture 15
2.2Power
Evety large enterprise has some sort of authority structure in which hierarchical relations between the parts that constitute the organisation are laid down. In organisations authority comprises "formal control over resources and decision-making" (Ferner and Edwards, 1995: 232). However, there is also another form of power as MNC's subsidiaries "may have greater or lesser power resources depending on their strategic importance, their ability to generate profits, their control of special skills, their role as intermediaries with key decision-makers in the environment, and so on" (ibid. 237). In each organisation resources are dispersed over various (groups of) actors. Because people or groups of people within the company value resources that other groups have, horizontal and vertical exchange relations exist. The nature of these resources might be financial, based on advantages in knowledge, or otherwise.
A process of bargaining occurs as actors try to get hold of the resources other actors have. As groups draw on the resources at their disposal they exercise a degree of power running through the formal power based on authority as laid down in the company structure. Not only groups within the company, such as international or local management, but also external organisations such as trade unions can exert influence because of the resources they possess (for example the power to mobilise workers). Within local operations local HR management might have the ability to constrain corporate transnational influence, and mitigate the process of diffusion of HRM policy. Also workers (trough unions or works councils) might be able to deploy the power that accrues from their resources and shape I-IRM policy.
2.3 Corporate culture
People always act on the basis of a set of (unconscious) values, also in their workplace. Corporate top management often makes use of or tries to influence the value set of their employees in order to increase control over what happens in the enterprise. Although the value set of an organisation is not analytically distinct from any other value set, it is often referred to as corporate culture. A large array of scholars has developed ideas about and definitions of corporate culture. Yukl (1998: 329 /330)) has discussed corporate culture in relation to problems that organizations face:
"Internal problems include the criteria for determining membership in the organization, the basis for determining status and power, criteria and procedures for allocating rewards and punishments, an ideology to explain unpredictable and uncontrollable events, rules or customs about how to handle aggression and intimacy, and a shared consensus about the meaning of words and symbols. [ ... ] A major function of culture is to help us understand the environment and determine how to respond to it, thereby reducing anxiety, uncertainty, and confusion. [ ... ] As solutions are developed through experience, they become shared assumptions that are passed on to new members. Over time, the assumptions may become so familiar that members are no longer consciously aware of them''.
A variety of other definitions of the concept exists: Claver et al. have described corporate culture as "a set of values, symbols and rituals shared by the members of a certain firm, describing the way things are done within an organization when solving internal managerial problems together with those related to customers, suppliers and the environment" (2002: 152). The control that can be exerted on the basis of culture relations emanates from the fact that culture is for the larger part an unconscious matter; people are most of the time unaware of the set of values that
16 Diffusing HRM
directs their behaviour. A company culture is a form of indirect control, and corporate values are a manifestation of a structure of corporate power (Ferner and Edwards, 1995: 234). At the same time corporate culture also constitutes a central element of HRM; culture can be a powerful tool
in creating employee commitment as it impacts on the way workers perceive their own interests in relation to those of the company. Symbolic acts associated with the culture of a company are a pivotal element in the legitimating of actions.
The values that exist within a MNC reflect the attitude that corporate headquarters has towards the management of their international operations. To distinguish between various attitudes I draw on the typology constructed by Perlmutter (1969) of an 'ethnocentric', 'polycentric', or 'geocentric' attitude of company management. In an ethnocentric MNC company strategy is based on a notion of pre-eminence of corporate headquarters. Strategy in a polycentric MNC can be characterised as highly responsive to conditions in countries in which the MNC operates. A geocentric strategy is characterised by a wish to combine some degree of internal consistency and a measure of responsiveness to local conditions. A certain attitude reflects a particular type of organisational structure and leadership style. When the top-management of a company has an ethnocentric or home-country attitude it implicitly regards the parent company and the criteria that the parent company employs as superior. This stance entails the imposing of the parent company's methods and conditions for doing things through the use of home-country nationals to manage foreign subsidiaries. The organisation is highly centralised and communication of issues such as HRM is one-way and top-down: from company headquarters to the local operations. The room to manoeuvre for local management is minimal.
A polycentric or host-country oriented approach is based on the notion that "local people know what is best for them" (Perlmutter, 1969: 12). The result is a decentralised and loosely structured company with subsidiaries that are left free to shape the way they handle their operations according to their idiosyncratic preferences. Often the only thing local managers have to adhere to are certain financial objectives. These managers are essentially recruited in the country where the subsidiary is located. Management of subsidiaries is left free to be shaped by idiosyncratic preferences, and local management and employees (and/ or their representatives) have considerable power.
The third ideal-type is tl1c geocentric or world-oriented approach. This third attitude is essentially a combination of both other forms, and is characterized by a high degree of interdependency and collaboration between top management and the management of subsidiaries. A central element is that all managers are encouraged to pursue 'worldwide objectives' meaning that they should not aim to defend country objectives but always look where (in the world) production can be carried out most profitably. The international management team includes the heads of the foreign operations, and nationality is not an issue when recruiting staff.
In the geocentric approach HRM policy is for a large part elaborated at company headquarters and relatively standardised. Managers at all levels in the company arc required to internalise the corporate culture, but have a certain degree of freedom to attain company objectives. Because these objectives are mainly financial the co1-porate office can be viewed as limiting itself to financial control over its subsidiaries. Adaptation to local circumstances is of secondary importance to these 'worldwide' financial objectives. In this management style the interplay
Diffusion via structure, power, and culture 17
between pressures for global consistency and isomorphism with the local environment is best felt.
A corporate culture "can be seen as a channel of influence for diffusing practices from one part of the corporation to another, as a result of shared understandings as to what may or may not be done" (Ferner and Edwards, 1995: 240). The diffusion of HRM through corporate culture has a strong connection with a global leadership style and works through a process of socialisation. The recruitment of managers from the home countly or the international transferring of managers is an important element in this socialisation process: "the use of corporate training programs, visits to head office and inter-subsidiary networking are some of the means to acquaint local management with corporate culture" (Rubery and Grirnshaw, 2003: 213). Rube1y and Grimshaw distinguish between companies that rely primarily on controlling financial outcomes of the subsidiary, and companies that try to control behaviour and processes (2003: 211 ). Organisations that adopt the first type of strategy leave it up to local management to decide on how to achieve the targets set. Companies that follow the second type of strategy seek to influence behaviour through the appointment of international managers, the deliberate cultivation of a corporate culture, and "through the use of coercive comparisons between parts of the organization to encourage the adoption of best practice" (ibid. 212).
Because organisations are increasingly organised along the lines of international business divisions, possibilities arise for the corporate office to compare plants on a global scale. This process has been termed 'coercive comparison' by various authors and entails the diffusion of best practices across sites (Ferner and Edwards, 1995: 245). Means of diffusing practices include the use of expatriate managers to handle foreign operations and/ or international meetings of personnel managers from operations in different countries to exchange information on policy initiatives. The appointing of international managers is a certain leadership style and the fostering of a corporate culture is a power resource. The employing of coercive comparisons is an outcome of the use of this power resource and can be utilized in the diffusion of HRM policy. A leadership style in which international managers are appointed in addition to the development of a strong corporate culture are two major means through which MNCs seek to control behaviour. The third is the use of coercive comparisons between plants in order to encourage the adoption of best practice (Rubery and Grimshaw, 2003: 212). A sirnilar 'best practices strategy', is a mechanism in the diffusion of organisational practices (Ferner and Edwards, 1995: 230). Top-level ER-managers need information on practices from their operations and thereby rely on formal authority relations. As a result of the way a company is structured (internationally) local operations possess this information and therefore the exchange of valued resources also comes into play. In this step of the process of diffusion management and employees in local operations can exert countervailing power to some extent. Unions, works councils, and local management can exert a degree of power based on their resources. When carrying out a best practices strategy the performance of various operations is assessed according to a number of efficiency measures. Plants are compared and when a plant has a better score on these measures then pressure is put on other plants to adjust their HRM practices. The managerial network also plays an important role in this. This is underpinned by the way managers arc recruited and trained/ socialised.
18 Diffusing HRM
2.4 National environment of subsidiaries
The object of study here are the operations of two MNCs in the food business 111 the
Netherlands. The extent to which diffusion takes place in a top-down fashion depends on the constraints and possibilities that the host-country environment offers. In order to gain insight in the environment of these local operations I use Hall and Soskice's ideal-typical distinction between liberal market economies (LMEs), and coordinated market economies (CMEs). National economies can be labelled a LME or CME on the basis of the way firms solve coordination problems in five spheres: 1. vocational training and education; 2. relations with employees; 3. relations with other firms; 4. structure of corporate governance; and 5. the system of industrial relations. In LMEs, market institutions are the means by which coordination problems in these five spheres are resolved whereas in CMEs ventures with other actors are coordinated through non-market relationships (Hall and Soskice, 2001: 7-8). Hall and Soskice typify the Netherlands as a CME because the institutions in the Netherlands support non-market relationships between firms and other economic actors. To analyse the process of diffusion of HRM policy within MNCs I will focus on the spheres industrial relations and corporate governance. The sphere of vocational training and education forms part of the field of HRM. Relations with employees and inter-firm relations are here taken up in the sections on corporate governance and industrial relations.
Corporate governance structure
When regarding institutional features of countries systems of corporate governance play an essential role. Corporate governance is the term used to refer to the modes of corporate ownership and control. Mart,>inson and Sisson (1994: 17) define two types of corporate governance structures that have implications for management approaches to industrial relations. These are the 'insider' and 'outsider' model. 111ey term the 'outsider model' the Anglo-American model, because it is found in companies originating from Anglo-Saxon countries. The 'insider' system is also labelled the 'continental' mode because it is associated with French and German systems of corporate ownership (ibid. 30).
Two defining features of the outsider model are a high degree of institutional share ownership, and closely bound up with this, a focus on short-term financial performance. Companies in the outsider system attach substantial importance to shareholder-value. This translates in an emphasis on financial control, exercised through continuous internal monitoring and performance evaluation. Two salient differences in terms of exposure to hostile takeover and financial performance between the insider and outsider model have industrial relations consequences. Insider systems have institutions to protect firms from hostile takeovers and companies focus on long-run performance. In contrast, the outsider system is characterised by exposure to hostile takeover and concentration on shorter-run financial performance. As a result owners of firms in outsider systems are less committed to their assets. In terms of industrial relations, in the outsider model employees are "more likely to be regarded as disposable liabilities and to be associated with a cost minimization approach to labour management" (Marginson and Sisson, 1994: 31). In the 'insider' system of corporate governance there is a financial system based on long-term bank credit, less prominence of the stock market and therefore less exposure to hostile takeovers.
Diffusion via structure, power, and culture 19
Furthermore, "[i]nternal controls tend to be operational rather than financial in character" (ibid. 30). Companies operating in the insider system are characterized by an emphasis on longer-run investment strategy, which results in a stronger commitment to company assets. Because ernployees are also part of the assets of a company in the insider model employees are regarded as highly valuable in obtaining company objectives. Additionally stakeholders including banks and employees (and their representatives) enjoy numerous rights. In the insider model of corporate governance the company and its stakeholders are both encouraged to make long-term commitments to each other.
The Dutch system is an example of the insider system; major banks do not dominate Dutch capital markets and corporations. Furthermore, securities regulation protects listed firms from hostile takeovers. A central element in the Dutch system is the separation between management and supervision of companies, the so-called 'two-tier' system. This is laid down in the
Structuurregeling of 1970. This act mandates companies to appoint a board of supervisors to supenrise the board of directors. This system does not apply to MN Cs that employ over 50°/(i of their workforce abroad. Dutch subsidiaries of such exempted MNCs or of foreign MNCs are subject to mitigated application of the Structuurregeling. The period when the Structuurregeling was introduced was one in which the societal consensus was that there should be a balance of interests between shareholders and a firm's employees (Raaijmakers and McCahery, 2000: 35). An important aspect of corporate governance is the relationship between companies and external providers of finance, i.e. banks and shareholders (Hall and Soskice, 2001: 337). The two systems are therefore also often called the stakeholder (insider) model and the shareholder (outsider) model. In the stakeholder model the concerns of all groups that have an interest in a company (employees, suppliers and even entire communities) are taken into account in decision-making by company management. In the shareholder model only shareholders enjoy rights to influence company management and creating shareholder value is the principal goal of the company. The insider or stakeholder model and the outsider or shareholder model further correspond with the distinction between CMEs and LMEs; in a CME the insider model is most common whereas in a LME the outsider model is encountered most frequently. However, the concepts of CME and LME are broader because they cover not only the sphere of corporate governance, but also industrial relations, relations between firms and employees, inter-firm relations, and vocational training and education.
The defining features of both models have profound implications for HRM, especially for the various elements that constitute HRM (see 3.1 ). One might imagine that in the outsider or shareholder model emphasis on shareholder-value, short-term financial performance and profit maximisation will lead to minimal investments in employee training and development. Additionally, the accent on performance will entail the using of performance-related remuneration. In practice this would involve the setting of measurable targets and the regular assessing of company and/ or employee performance according to these targets.
It can be expected that in the insider system the longer-term strategies of companies will permeate employer-employee relations. This will firstly influence the way employees are evaluated and how they are correspondingly rewarded. The targets set to appraise performance are likely to be longer-term than in outsider systems and assessments will correspondingly be held less
20 Diffusing HRM
frequent. In the stakeholder model corporate management will be more inclined to make investments in employee training and development than in a model where cost-minimisation is an integral part of company strategy.
Industrial relations
The term industrial relations refers to the relationships between employers and employees (at the business level) or between employers' associations, trade unions (and works councils), and the government (at national level). The structure of industrial relations in a country is a central determinant of the possibilities or obstacles for company management in the managing of their personnel. Industrial relations in a CME can be characterised by a relatively high degree of employee involvement in managerial decision-making. In CMEs in Northern Europe, including the Netherlands, there is a considerable degree of cooperation and harmony between works councils and/ or unions on the one hand and company management on the other (Slomp, 1995: 305).
Industrial relations cannot be understood without reviewing how they have come into being. The Dutch system of industrial relations is the outcome of events and processes that have taken place in the history of the Netherlands. Drawing on Visser and Hemerijck (1997) I trace this history back to right after the Second World War when Dutch policy makers emphasized the need for industrialisation, an increase in exports and a favourable balance of payrnents. Because production had to be carried out cheaper than in neighbouring countries the general consensus was that wages had to be restrained. By 1970 the Netherlands had become a high wage economy. Since the late 1960s the unions had been able to incorporate automatic price escalators in annual collective agreements determining 75 percent of annual wage increases by 1980 (Visser and Hemerijck, 1997: 96).
The 1980s started with economic recession triggered by the oil crises of 1979. In the fist half of the 1980s unemployment rates doubled compared to the second half of the 1970s. According to policymakers profitable private enterprise was to be achieved through lowering (wage) costs and the improvement of the operation of the labour market through decentralized wage bargaining. In the context of these policy developments the position of the unions deteriorated; strikes were out of the question and membership figures plummeted. Slowly the unions moved away from an emphasis on income redistribution and towards the issue of working hours reduction. In 1982 the Wassenaar Accord was reached, which put mage moderation at the core of negotiations between social partners. In return for wage moderation, employers agreed to working time reduction. Since the Wassenaar Accord collective agreements have come under increasing pressure as employers have argued for more individualised labour conditions (Rojer and van Rij, 2004: 18). These developments can be regarded as decentralisation, a process that was taken a step further during following negotiations, especially in 1993.
In the face of another economic recession in the early 1990s employers' demanded wage freezes. In 1993 the social partners agreed on a policy of wage restraint to combat unemployment. The unions agreed to wage restraint in exchange for working time reduction. The result was a new accord, signed in 1993, 'A Nen; Course. Agenda for colledive har:gaining in 1994'. The major priority of the new accord is job creation, to be achieved through decentralisation and flexibility. Both the
Diffusion via structure, power, and culture 21
accord of 1982 and the one of 1993 marked cooperation between employers and unions. In
addition, both accords have led to the "controlled decentralisation of Dutch industrial relations" (Visser and Hemerijck, 1997: 111 ). In 1996 the "Flex Act" was the next step towards more flexibility. In the Act rights of "flexworkers" were strengthened whereas rights of permanent workers were reduced. After the turn of the millennium, the growth of the Dutch economy turned into decline. In the collective negotiations reached in the first years of the 21 st century further wage restraint was agreed on as the path to economic recovery.
In the Netherlands agreements are reached at the central level between employers' associations, trade unions, and the government. General attitude of unions in collective agreements is one of compromise due to the threat of exclusion (Visser and Hemerijck, 1997: 112). Most collective agreements are negotiated by the largest trade union FNV so that coordination in the negotiation process is high. Besides these central agreements there is a growing number of company collective agreements. The two cases selected here also have their own company agreement. However, the overriding majority of the Dutch labour force is covered by a sectoral collective agreement.
The Dutch government discusses labour issues with social partners in the "Social Economic Council", SER (Sociaal Economische Raad). Central agreements between employers' associations and trade unions are negotiated in the "Labour Foundation", STAR (Stichting van de Arbeid). The government and the STAR confer twice a year on the socio-economic state of affairs and the development of income and employment. The STAR does not negotiate the terms of labour agreements. The negotiations between unions and employers associations that take place in the STAR generally have the function of guidelines and serve as a framework for further negotiations at other levels.
The Dutch peak employers' association for large and medium-sized firms in industry and services is VNO-NCW (Vcrbond van Nederlandse Ondernemingen and Nederlandr Christeiijk Werk;geversverhond).
The association organises about 90% of the (larger) enterprises in the non-agricultural private sector. In addition, there is the Federation of Employers in Small and Medium-sized Enterprises called "MKB-Nederland" that represent 35-40cYo of the SMEs. Agri- and Horticultural enterprises are organised 1n "I;fO-N ederland" (Federatie van Land- en Tuinbouwot:ganisaties)
(reformmonitor.org). Unilever and Sara Lee/DE are both members of the Algemene Werk;gevers Vereniging Nederland (A WVN), the advisory bureau of VN 0-N CW. The A WVN is involved in the realisation of over half of all collective agreements in the Netherlands and acts as a coordinator within and between branches of industry (Buitelaar and van den Toren, 1999: 90).
In the Netherlands the Civil Code defines many basic norms with respect to the rights and duties of employers and employees. Employee representatives are endowed with a number of rights towards the employer/management, and thereby in the shaping of HRM policies: the 'WoR' (wet op de ondernemingsraden -law on works councils), stipulates that companies in the Netherlands with over 50 employees are required to install a works council. In the Netherlands, as well as in other countries where works councils operate, trade unions and works councils do not act in isolation from one another. Although large variations in cooperation between the two parties exist, they can be considered as 'communicating vessels' or 'twin institutions' (Slomp, 1995: 317).
22 Diffusing HRM
Works councils have the "right of initiative", the "right to advise", and "the right of consent" on issues such as working hours, holiday rules, pension plans and training. In collective negotiations works councils have an advisory role. Companies with 10 to 50 employees may have "PV"
(personeelsvertegenwoordiging -staff representatives) if desired by employees. PVs can reach agreements on working hours (www.reformmonitor.org). In the insider model the employees and managers in local operations will have more influence on corporate decision-making because their views are taken into account. This will provide them with a power resource. This influence will mainly derive from the way the company is structured, and will rest upon authority relations. A corporate culture works as a reinforcing mechanism in legitimising these structures of authority.
About 63 % of Dutch trade union members belong to the "Federation of Dutch Trade Unions", FNV (Federatie van Nederlandse Vakverenigingen). 19 % belongs to the Christian union, CNV
(Christe!!Jk National Vakverbond), and about 8 % to the association for high- and medium-educated personnel, Ml-IP (Vereniging voor Hoger en Midde!baar Personee~ (reformmonitor.org). Trade union density in the Netherlands is quite high in relation to worldwide density levels, but ranks among the lowest in Europe (ILO, 1997 /98); about a quarter of the Dutch non-agricultural labour force is a member of a trade union. Although collective agreements covering the overriding majority of Dutch workers are the outcome of collective agreements between unions and employers' federations Dutch trade unions have no official rights to negotiate labour conditions. As Visser and Hemerijck (1997) have noted: "In the absence of a legal right of recognition for unions and given the threat of exclusion, coalition building is the only remedy"
(p.
83). The rate of workers covered by a collective agreement is substantially higher than the rate of union members. This is due to an Act that states that the Dutch minister of social affairs can declare a collective agreement valid for all companies in a certain sector. Collective agreements are usually concluded at sectoral level, but Unilever and Sara Lee both have a company agreement for their Dutch operations. These company agreements also cover the greater part of Unilever and Sara Lee employees in the Netherlands.HRM and collective labour agreements are closely bound up. HRM can be regarded as broader than collective agreements because it bears upon more elements of the employer-employee relationship. An outcom~ of the increasing relevance of l-IRM is that the status of trade unions (and works councils) is being undermined. As will be discussed in the next chapter, within I-IRM the focus is on the individual employee. This stands in contrast to the collective negotiations that unions are involved in. In recent decades, employees are more often involved early in processes of change in the company and as a result works councils and unions no longer have an advantage over their fellow workers in terms of information. HRM often entails allowing employees more freedom to shape their working conditions, and they therefore no longer feel that they need the trade unions to negotiate their labour conditions for them (Haveman, 2003b). Haveman has argued that collective negotiations with trade unions are of secondary importance within a company's BRM policy (2003b: 77). According to him this is due to the fact that "modern J-IRM policy is not only about the distribution of value within a company, but also about the creating of added value''. He adds that works councils only have a modest task in this respect. However, a number of issues that are negotiated in collective agreements also form part of the reahn of HRM
Diffusion via structure, power, and culture 23
and in this way collective agreements can act as instruments that affect HRM policies. This creates tension between the two spheres (Van den Toren, 2003). It has been found that the central items of both institutions generally overlap, and that the collective agreement can be depicted as mitigating and codifying HRM policy (ibid. 57-58).
It is important to note that the two organisations in this study both have their own company collective agreement. It is reasonable to assume that a link between company strategy and HRM policy is most easily established in a company-level agreement: as management and employees have a degree of power to influence labour conditions, a company collective agreement becomes a tool in obtaining BRM-objectives. (Haveman, 2003b: 78/79).
Defining the central concept and operationalisation 25
3. Defining the central concept and operationalisation
3.1 Identifying HRM
Human Resource Mana._Jl,ement
HRM can be defined as "a collection of employment and personnel practices and activities and their implementation" (Paauwe, 1991: 104). HRM is essentially equivalent to personnel management but using the term human resources to refer to employees "[ ... ] has gained widespread acceptance over the last decade because it expresses the belief that workers are a valuable and sometimes irreplaceable resource" (Gomez Mejia et al., 1998: 2). Paauwe en Dewe have described HRM as a management philosophy that arised out of economic and technological developments that took place at the end of the 1970s (Paauwe en Dewe, 1995b: 77). These economic developments include the shift in focus in production from solely cost price to quality, flexibility, and innovation. Another development is the rising number of people working in the service sector. These people are often higher educated than workers in other sectors and they need to be dealt with in a different manner. Technological developments have changed the nature and content of many tasks employees perform. This has resulted in changes in the way production is organised; production processes could now be integrated to a higher degree and responsibility could be shifted to lower levels in the organization. The result is an increase in the need to promote commitment from employees. The notion of a corporate culture is fundamental in this respect (ibid. 78).
Storey (1989) has placed the emergence of HRM in the 1980s within a broader context of overall innovations in the strategies of MNCs, and links these to changes in the environment of organizations. General innovations include increased flexibility, the restmcturing of the organization in business units, and the use of performance-related pay (Storey, 1989: 1 ). As a consequence of these strategic changes, the management of the employment relationship has also been altered. The switch to HRM is closely linked to the different position that personnel management was accorded within the company, namely more in line with company strategy. lfaveman argues that as personnel managers in the Netherlands in the early nineties were increasingly renamed 'manager human resources' their goal became to link company strategy, the creating of added value, and HRM policy (Haveman, 2003b: 76).
In order to be able to identify what constitutes HRM it is helpful to examine the differences between HRM and previous forms of personnel management. A defining feature of HRM as compared to traditional personnel management is the idea that people are an important asset in gaining a competitive advantage and should therefore be utilized to their 'full capacity'. Furthermore, there is a focus on individual rather than collective relations between employees and management. Related issues are developments in the direction of integrated, performance-related reward schemes, a focus on flexibility in work organisation, and the advancement of commitment in contrast to securing compliance (Storey, 1989: 2-10). According to Legge (1989) HRM policies are integrated in the strategic business planning and are mutually consistent in promoting commitment and a willingness among employees to adopt an attitude of flexibility and adaptability. (Legge, 1989: 25).
26 Diffusing HRM
Legge has compared the model of HRM with that of personnel management and has made an attempt to determine the differences and similarities between them. In doing this, she distinguishes between the normative and the descriptive aspects of both models. Important similarities between the two approaches are that both attach importance to the integration of ERM/personnel principles with business objectives, and to the development of employees' abilities and talents, in order for them to make a valuable contribution to the goals of the organization. Legge stresses that differences between the models are mainly within the normative aspects. The first distinction is that HRM focuses on what is imposed on managers whereas personnel management mainly denotes things that managers impose on employees; "[ ... ] personnel management appears to be something performed on subordinates by managers rather than something that the latter experience themselves" (Legge, 1989: 27). Secondly, HRM places managers in a more proactive role, and they are often held accountable for the way they have utilized labour, measured in terms of bottom-line results. Finally, HRM emphasizes the managing of a corporate culture as a central task for senior management.
Legge as well as Storey broadly depict the same defining aspects of HRM. The fist is the recognition that managing people in a way that stimulates their own development can help companies in achieving their strategic objective. The second is the integration of HRM policies within strategies aimed at achieving corporate goals, for example performance-related remuneration. The third aspect is that corporate culture characterized by flexibility and commitment to the company's objectives is fostered.
A range of scholars has discussed the elements that constitute HRM. The building blocks discerned here are based on divisions made by Storey (1989: 7), Legge (1989: 28), Gomez Mejia et al. (1998), Paauwe and Dewe (1995b), and Rubery and Grimshaw (2003: 218)
o Recruitment and selection: staffing, and the socialization of employees, as well as managing reorganisations with regard to employees;
o Appraisal: evaluation, performance measurement; o Remuneration: wage policy, benefits;
o Employee development: performance management and measurement, training and development.
o Governance: employee relations and communication, industrial relations, worker participation.
3.2 Research design
The research is made up of an empirical and a theoretical component. The theoretical part consists of a review of the existing literature on the diffusion of practices in MNCs, and the mechanisms underlying this process. This has been discussed in chapter two. The empirical part of the research can be divided into two phases; the first phase consists of desk research on company reports such as annual and social reviews. As a timeframe the data covering the last 5 years has been selected. This research is largely a preparation for the second phase of the empirical part of the research, the conducting of six semi-structured interviews. For both
Defining the central concept and operationalisation 27
companies, three interviews are held; one with a functionary of the personnel department, one with the president of the central works council, and one with a trade union official. The interviews are held within the framework of research done by the Amsterdam institute for advanced labour studies (AIAS) on globalisation and industrial relations in MNCs2. The questionnaire used in the interviews covered a range of topics including changes in the position of employees' representatives and HRM as a result of processes associated with globalisation (see appendix). The analysis in this thesis draws on the answers that respondents gave to the questions on the coming into being of HRM policy and changes in its content, changes in collective agreements, changes in the role and position of trade unions, and changes in the relations between (inter)national company management, unions, and works councils3 (questions
6, 7, 9, 12 and 13).
MNCs are characterised by their organisation over a range of countries. Looking at diffusion is especially interesting in these internationally structured enterprises. This research attempts to show the processes and mechanisms whereby MNCs diffuse HRM policy within the operations in various parts in the world. Drawing on research in this area for a single MNC by Coller and Marginson (1998), the aim here is to make a comparison between two MNCs. For reasons of comparability, two MN Cs within the same sector were selected. The choice for Unilever and Sara Lee was partly given in by the fact that they were selected for study in the research by AIAS. Both Unilever and Sara Lee are manufacturers of foodstuffs, household/body care products, and in the case of Sara Lee also clothing. Both companies are global leaders in tl1eir sector and experience strong global competitive pressures.
The multi-national food industry provides a particularly interesting example as MNCs in this sector are because of the nature of the products located in a range of countries. Foodstuffs cannot be hauled around the globe unlimitedly; they are highly perishable and therefore need to be produced in relative close proxinUty to buyers. Unilever and Sara Lee have historically both grown through acquisitions. With the taking-over of other companies specific features of these companies in terms of for example management style are also brought within the scope of the
.M:NC. The result is that both Unilever and Sara Lee are characterised by far-reaching internal diversity. This diversity is interesting in the light of the research questions posed here: does international company management maintain local variations or does it adopt a strategy of high coordination throughout tl1eir operations? Through a comparison the findings might be generalised to other MNCs operating in the food industry, or MNCs with a comparable international structure in terms of product-market links4•
2 The research was commissioned by the Dutch trade union federation FNV.
3 As can be seen form the way the questions were constructed changes were an important issue. In the AIAS research the focus was on a timeframe of roughly 5-10 years. In this thesis the timeframe is the most recent 5 years.
28 Diffusing HRM
3.3 Making the theory operational
The issue that is explored in this thesis is how HRM policy is diffused within MN Cs. Where is it developed and how is it diffused/implemented throughout the worldwide operations of the MNC? I have argued that this happens either on the basis of one global HRM policy for all operations set out by corporate management in line with company strategy or on the basis of incorporating local variations and allowing room for local management to develop and implement HRM policy the way they deem the most efficient. This is the dependent variable to be explained. The independent variables that are considered as determining this process of diffusion are company structure, power relations, and corporate culture (see figure 2). Structure determines how the various parts of the MNC relate to each other and how much formal authority (management of) subsidiaries has in setting out HRM policy. Power refers to the influence that (groups of) can wield over each other and over policy making in general. Corporate culture determines how top management handles relations with local management (a ethno-, poly-, or geocentric attitude of top management), and how local management handles interaction with stakeholders such as trade unions. A corporate culture is an important tool in the production process. Every organisation has a set of values that determines how employees and management experience their mutual relationships and their role in the obtaining of corporate goals. By influencing this corporate culture, leaders can indirectly influence the behaviour of employees (Y uld, 1998: 329).
The variables structure, power, and culture determine the diffusion process and will be compared for the two MNCs. The way a company is structured determines the room to manoeuvre for local management. In addition, it detennines to what extent local management and employees' representatives can influence corporate policy. In order to establish how Unilever and Sara Lee are structured I use the slightly modified and extended typology of multinational enterprises by Ferner and Edwards (1995). Based on Ferner and Edwards, and Coller and Marginson (1998) five types of firms are taken up in the analysis. To discern power relations in the two companies the resources that international company rn.anagement, national company management, and employees' representatives have at their disposal will be considered. Recourses can include some form of specialized knowledge, access to certain materials needed in the production process, or the financial relevance of the operation for the company as a whole in terms of profitability.
In order to analyse the role of corporate culture it is important to pinpoint which are the company's values. A starting point is provided by Perlmutter's distinction between a ethnocentric, polycentric and geocentric attitude of top management. These attitudes with regard to hierarchical relations influence the character of the relations between higher-level- and lower-level management and between lower-lower-level (local) management and employees. Furthermore, how have these values come about? Is there one comprehensive corporate culture set out by international management to be incorporated by every employee or is the enterprise culture in subsidiaries left free to develop on the basis of local values? Discerning the values that guide peoples behaviour can be problematic as they can be unconscious to a large extent. A research method to use in this respect is observing people in their daily behaviour. In this thesis only inte1-views were held but the values that have (indirectly) come to the fore in these inte1-views will be discussed. In addition, company documents on the topic are analysed.
Defining the central concept and operationalisation 29
Values determine relations between actors in terms of status and power (Yukl, 1998: 329), and therefore also establish which parties can develop and implement HRM policy. In order to determine if manner of diffusion varies over different HRM practices, the building blocks of HRM will be discussed separately for both companies in the following two chapters. The results will be discussed in the light of the analysis by Rosenzweig and Nohria in chapter six.
Figure 2. Variables influencing HRM policy in local operations
Company Management [Strategy/HRM policy]
Structure - financial control MNC - integrated MNC organised per country - integrated MNC organised per activity - decentralized global network company - federal international firm Power - resources Culture - corporate values - leadership style: • ethnocentric • polycentric • geo-centric Local management Structure - financial control MNC - integrated MNC organised per country - integrated MNC organised per activity - decentralized global network company - federal international firm
Industrial relations Relations between unions and employers'
associations (and the government). Power - resources Culture - leadership style: • ethnocentric • polycentric • geo-centric Corporate governance structure - insider (stakeholder) model - outsider (shareholder) International level National level LOCAL
