Introduction

Despite a growing body of research underpinning the field of electronic Human Resource Management (e-HRM), a distinct paucity of studies founded on macro theories emerges when this body of research is examined [89]. This paper argues that e-HRM diffusion in the subsidiaries of a Multinational Corporation (MNC) is, similar to other HRM practices, subject to a broad range of institutional factors, even though the nature of an e-HRM would necessitate a high level of integration and standardisation across the MNC to attain expected effectiveness and efficiency gains. It has frequently been argued that particularly US MNCs" are characterised by standardised, centralised and formalised HR policy-making processes and the introduction of US style HRM practices in their host countries [e.g. 1, 25, 50]. The level of diffusion of HRM practices in general across MNCs appears to diverge considerably and various attempts have been made to understand these differences. For a number of years, the international HRM (IHRM) literature has provided a forum for an ongoing debate, which has given rise to a host of institutional factors that may arbitrate the transfer of employment practices among multinational corporations" subsidiaries. These factors comprise home and host country effects [1,34,70], sector effects [13, 78,79], the institutional context and national business system [e.g. 10, 13, 13, 14, 24, 25, 26], dominance effects [35,62,74], organisational culture [2,7,48,86], or the impact of micro-political relationships between the subsidiaries and the head quarter [32,47]. A detailed discussion of these factors extends beyond the scope of this paper. Much of the debate surrounding these factors is founded on institutional theory. This paper will draw on both the business systems approach (or European institutionalism) and new or neo institutionalism. Indeed, Tempel and Walgenbach [90] opine that both theoretical stances can learn from each other. European institutionalism places greater emphasis on the regulative context [51, 93,95] than does the US-based new institutionalism, which focuses more on the socio-political background [17,83].

A review of the extent e-HRM literature underscores a dearth of research and discussion on the factors mediating the diffusion of e-HRM and e-HRM practices across the subsidiaries of MNCs. Consequently, the aim of this paper will be to explore whether the institutional factors that mediate the transfusion of standard HRM practices may also effectuate the transmission of e-HRM practices within the subsidiaries of an MNC or whether indeed other forces can be discerned. Data for this paper emanate from a review of the international HRM (IHRM) literature as well as an analysis of primary data derived from a series of in-depth interviews with key decision-makers and stakeholders in the areas of IHRM and e-HRM in the German and Irish subsidiaries of a US MNC. Findings from this research intimate a number of pertinent issues regarding the transmission of e-HRM practices throughout this MNC. Subsidiary variations in e-HRM utilisation may be explained by a multiplicity of factors including the strength of national business systems, the strategic salience of subsidiaries and micro-political power relationships between the subsidiaries and the head quarter. While the findings affirm both home and host country effects in the diffusion of e-HRM practices, the dominance effect of the MNC shapes employment practices in general [79,80] and not just e-HRM practices in the subsidiaries in distinctive ways. Due to the information intensive and dependent nature of an e-HRM system, e-HRM practices are subject to stronger forces for standardisation than standard HRM practices. This paper evinces that variability in e-HRM practices in the subsidiaries of an MNC arises, to a large degree, from the same type of factors governing the transfer of HRM practices across subsidiaries, albeit e-HRM practices are impacted in different ways. These findings also support evidence from the e-HRM literature, which illuminates other factors such as user acceptance as auxiliary key determinants of e-HRM success [40,82]. This paper is structured as follows: the succeeding section will illuminate the pressures of institutional duality on the transmission of organisational practices in general and e-HRM practices in particular across the subsidiaries. Then, the lack of research into the diffusion of e-HRM is highlighted. This is followed by an outline of the methodological approach for this research. Next, the research evidence will be discussed before the main conclusions are offered.

Institutional Duality and the Diffusion of e-HRM Practices

Internal integration and centralised decision making is of paramount importance in the operation of global e-HRM system [81]. Any deviation from the standard system would arguably compromise the quality of the data collected and ultimately impair the informative value of any subsequent analyses of this data. However, organisations are continuously faced with what has been described as "institutional duality", that is, different layers of institutional contexts that simultaneously impact the configuration of HRM (and thus also e-HRM) practices [60,76]. In other words, MNCs strive to attain internal consistency of policies and procedures to develop and sustain their corporate identity, while, on the other hand, MNCs are forced to tailor their policies and practices to suit the cultural, societal, and legislative environment of their host nation in order to achieve local efficiency [29,50,53]. Morgan and Kristensen [67] contend that the countervailing nature of these institutional contexts will ultimately lead to micropolitical conflict between the head quarter (HQ) and the subsidiaries and the subsidiaries themselves. A number of authors have argued this point [8,18,19,31]. Institutional isomorphic pressures may be categorised as normative or cognitive [83], high or low context-specific [11], or coercive, mimetic, and normative [29]. DiMaggio and Powell define isomorphism as "a constraining process that forces one unit in a population to resemble other units that face the same set of environmental conditions" [16]. Coercive isomorphism is the product of both formal and informal pressures of the host society in which the subsidiary resides, including government, employment legislation, trade unions, works councils, etc. [16]. Mimetic isomorphism focuses on organisational modelling (in benchmarking and imitating strategies and practices of key competitors) in response to uncertainty in the firm"s environment [16]. Normative isomorphism relates to the adoption of accepted work practices, standards and modus operandi of a specific institutional (sectoral) environment. Institutional isomorphism may ultimately result in organisations that are "virtually indistinguishable" and "interchangeable" [16,29]. These pressures can arise in the global or national (even regional), internal (the relational context) or external (the institutional context) environment of the firm [60]. Child [11,12] distinguishes between high context and low context dimensions to assess the level of influence different institutional contexts have on MNC practices. A high context dimension refers to factors that lead to a high level of embeddedness in the national and social institutional context, whereas a low context perspective is associated with factors such as the economy, market and technology, which are less dominant in moulding a company"s HR policies and practices. To a large degree, the level of differentiation and adaptation of HRM (and e-HRM) practices required by the MNC seem to hinge on the strength of the national business system (NBS) of the host country [13] and the magnitude of differences between the NBS of the home and host country [89]. MNCs pursuing a transnational or geocentric globalisation strategy appear to favour what has been described as a "cherrypicking approach", whereby the MNC selectively adopts HRM (and perhaps also e-HRM) practices from the respective home and host country context of their subsidiaries [43,49]. Furthermore, some employment practices that originated in host countries may be adopted by subsidiaries in other countries and even in the country of origin of the MNC this process is termed "reverse diffusion" [1,24]. The level of transfer of HR practices has been theorised by a number of authors. Morgan and Kristensen [67], for instance, argue that the larger the institutional distance was the greater the difficulty in transferring practices successfully would be. Kostova [59] differentiates between implementation and internalisation. She suggests that successful implementation and internalisation hinge on three sets of factors under the headings social context (regulatory, cognitive and normative), organisational context (culture) and relational context (commitment, identity and trust relationship with parent organisation). Building on this theme, Björkman and Lervik [6] put forward three dimensions of ascending levels of transfer successimplementation, internalisation and integration of diffused HR practices. Oliver [73] identifies a range of strategic responses to institutional pressures, which will ultimately affect the success of any transmission. These responses extend from manipulation, defiance, avoidance, and compromise to acquiescence. In comparison, the ERP and IS bodies of literature generally appear to consider large corporations in which these systems are rolled out as a homogenous mass rather than a heterogeneous system of political subsystems, that is, the subsidiaries. Few articles of the ERP implementation literature allude to culture and the organisational environment [75,85]. However, established IS success models strike one as focusing predominantly on user acceptance issues [15,71]. While this section has illuminated a range of institutional factors that may compel a MNC to adapt its (e-)HRM practices locally, the next section reveals the dearth of literature and debate on the transmission and local adaptation of e-HRM in subsidiaries of an MNC, which this paper is ultimately aiming to address.

Scarcity of Research into Diffusion of e-HRM

Based upon the arguments presented above it seems reasonable to assume that e-HRM practices in different subsidiaries of the same MNC ought to diverge in some way. To date however, the e-HRM literature has accepted and has somehow assumed that unilaterally imposed e-HRM practices will be adopted by subsidiaries in the same manner in which they were intended by an MNCs HQ, even thought the above discussion has highlighted that this is in fact not the case. The few quantitative studies that examine the diffusion of e-HRM and / or Human Resource Information Systems (HRIS) centre on the adoption of HR technology by a wider population of organisations rather than the adoption and subsequent diffusion of e-HRM in a single company [42,58,63,72,92]. It seems that only Smale and Heikkilä"s [84] study acknowledges that the introduction of an e-HRM system may give rise to conflict and micro-political behaviour, which necessitate negotiation and local adaptation of e-HRM practices to resolve these issues. Evidently, additional research is required to address this imbalance.

Methodological Approach

This research employs a single case study but multiple units of analysis approach to assess whether or not the diffusion of e-HRM practices in the subsidiaries of a US MNC is mediated by the same institutional factors that govern the transmission of standard HRM practices. This paper also aims to explore any differences and similarities that may exist between the transfusion of traditional HRM and e-HRM practices. Accordingly, this investigation is both instrumental [87] and exploratory [97] in nature. Due to the multifaceted nature of institutional factors a single case was selected to focus on the phenomenon under investigation and to avoid the moderation of results by additional extraneous variables, which would have been introduced by a multiple case analysis [4,14]. German and Irish subsidiaries of this MNC were chosen for this analysis due to the distinct cultural, economical, business and employment systems backgrounds in these countries [57]. Primary data for this ongoing study derive from fourteen in-depth face-to-face interviews with key decision-makers and stakeholders in the areas of IHRM and e-HRM utilisation in this multinational. Interviews were recorded and transcribed. Interviewees were chosen based on their level of involvement and decision-making power regarding the use of e-HRM in the subsidiaries. The semistructured interviews lasted between one and four hours and were carried out in the Irish manufacturing facility (with the HR director, HRIS specialist, and two line managers), the European HRIS headquarter (Head of Shared Services Centre Project Team, European HRIS Manager, Information Systems Analyst, Payroll Processing Manager) in The Netherlands, the International (European) Headquarter (Senior Director for HR Systems) in Switzerland, the Central European Sales Headquarter in Germany (HR Director Central Region and HRIS specialist), and the German manufacturing plant (Plant Director, Finance /HR Manager, Plant Manager / Head of the Works Council). The US MNC under investigation (Meddevco in the following) employs approximately 38,000 people in 120 countries. This medical devices company operates in the region of 270 manufacturing facilities, sales offices, research centres, education centres and administration facilities across the globe. This sector is highly regulated [27] and may thus be subject to strong institutional isomorphic pressures. Coercive pressures arise from the highly unionised nature of this sector in Ireland and Germany and the highly regulated nature of the employment regulation system in Germany [68,36]. Normative pressures emerge from the regulated nature of the sector itself, for instance quality and regulatory standards. According to the Irish HR Director in this study, key competitors in this industry frequently benchmark one another in terms of HR practices. This is not surprising, as many of the 140 medical device companies located in Ireland are clustered in the West of Ireland, which delivers additional mimetic pressures for institutional isomorphism among these companies. The majority of medical devices companies appear to be headquartered in the US [28].The e-HRM system in Meddevco is part of an enterprise resource planning system (ERP) named PeopleSoft (owned by Oracle). An ERP is a management information system that integrates information from all functional areas such as finance, production, marketing and HRM into a central data bank. The US HQ utilises practically every PeopleSoft module available to support the HR function. Online HR activities supported by the system include talent management, performance appraisals and e-recruitment, online training, and HR administration. The US-part of the corporation also maintains a HR shared services centre (HRSSC), which does not serve centres outside of the USA. The ERP and e-HRM systems were adopted at different stages in the life cycle of the different subsidiaries. In the Irish manufacturing site, PeopleSoft was adopted when the site was taken over by Meddevco. The Sales HQ for the Central Region in Germany (set up in 1970) commenced the introduction of PeopleSoft in 1999 (the same year the Irish site was acquired). It took almost two years to implement, according to the German HRIS Super-User. The German manufacturing site was acquired in 2000 and PeopleSoft introduced in 2004 and is, with reference to the German Plant Director and Finance Manager still not fully implemented, even though the European HRIS Centre considers the rollout actualised. This dichotomy accentuates the differences between three ascending levels of transfer of HR practices identified by Bjorkman and Lervik [6] -implementation, internalisation and integration. The evidence suggests that e-HRM transfer may only be at the initial level. In other words, managers and staff in the German manufacturing plant have not accepted the system nor do they see value in using the system. The next section will focus more closely on the key factors of e-HRM diffusion in Meddevco.

Key Factors in e-HRM Diffusion

The introduction to this paper has already drawn attention to the premise that internal consistency ought to be the key to maintaining a global e-HRM system. Thus, the data and types of information collected throughout the MNC"s subsidiaries ought to be uniform. It therefore follows that MNCs ought to control and keep isomorphic pressures to a minimum. However, this is not the case in this MNC. A number of examples exist where the corporation had to make concessions to individual subsidiaries and countries with regard to data entry. In the German manufacturing site for example, modules such as e-recruitment or talent management are not utilised due to the lack of manpower. In addition, this subsidiary is allowed to leave some employee information on the system unpopulated, as the pay scales of the employees covered by collective bargaining arrangement do not fit into the system. In some respects, the German Plant Director stated, the corporation simply turns a blind eye when qualifications of (German) employees are entered incorrectly, as the system of vocational training does not exist in the US. Moreover, the system recognises all but few German universities. German payment systems and the entry of German qualifications into the system also present barriers to utilisation in the German Sales HQ, according to the German Super-User. Managers in the German manufacturing plant feel a great deal of disenchantment and even disengagement with the system, especially since the corporation has now shifted its main focus of attention on the introduction of a new ERP system (by SAP), even though PeopleSoft will run in parallel for the foreseeable future, a decision which raised palpable concern with the HR Director for the Central Region. Similar issues arose in the international head office in Switzerland and limited adjustments had to be made to the system for compliance reasons. Since the international HQ in Switzerland was considered a show-piece for the organisation and an extension of the HQ in the US, additional customisations or divergence from e-HRM policies were not an option according to the Senior HR Systems Director located in the International HQ. At a European level, the MNC has to date been unable to introduce a HRSSC, which centralises all HR administration in a central location. According to some of the key stakeholders interviewed (German Plant Manager, Irish HR Director), European managers are fundamentally opposed to this idea and their combined resistance has already led to the failure of the first attempt to introduce such a HRSSC. In response, the MNC has set up a project team, which includes some European key stakeholders in order to pre-empt possible resistance in the next introduction attempt. Other issues arising from the implementation of HR self-service are of a logistic nature. Since HR self-service is provided via an intranet, employees will need access to a PC. However, not every employee can be guaranteed permanent access, particularly in the manufacturing sites in Galway (Ireland) and Heerlen (The Netherlands), according to the Irish HR Director and European HRIS Centre Manager, even if the organisation provides a number of computer kiosks on the shop floor. In addition to access to a PC, employees require the basic computer skills necessary to take advantage of the selfservice features of such a system. These skills cannot be taken for granted as the composition of staff differs in the subsidiaries. For instance, the majority of staff in the German manufacturing plant possesses a tertiary qualification, while most of the workers in the Irish Plant do not. In addition, people may have privacy and security concerns regarding online access to personal information [21,55]. Mimetic pressure to implement a HRSSC arises from the MNC"s key competitor, which has already introduced such a HRSSC.

Auxiliary key drivers for the global standardisation of HRM practices include organisation structure and culture [20]. However, the standardisation of e-HRM practices is just as subject to the "drivers for localisation" [20] as other HR practices such as recruitment or training are. At any given time, these localisation drivers provide a counter force to the drivers for standardisation and include, inter alia, national culture, national institutions and national business systems as well as the subunits themselves [39]. Meddevco consists of six distinct product divisions, each with its own support functions. Complex reporting and organisational structures, centralised control and decision-making in this MNC all act as strong drivers for standardisation. A so-called Human Resources Council (HRC), consisting of Senior Vice Presidents (SVP) of particular functions, for instance the SVP for Compensation and Benefits Systems, drive the multinationals" HR strategy. Nine of the ten members of this council are permanent constituents; only the European representative rotates on a yearly basis, which has some marked repercussions for the (lack of) representation of European interests in the HR decision-making process. While the HR Council develops corporate HR strategy, it does not have Board of Director status. The actual Board of Directors of the MNC ratifies any proposals before these are disseminated to the divisions. Asked about her influence on e-decision making processes the Senior Director HR Systems replied: None. All decisions are made by the HRC. If for instance a decision would be made to introduce SAP by the CIO [Chief Information Officer] and the HRC, I would not be involved in the decision making.

Other key stakeholders in this research (Irish HR Director, German Plant Manager, HR Director Central Region, Head of HRSSC Project Team), independently from one another, emphasised that one of their key roles was to interpret company policies and transpose these into the local business system. The data suggests that local managers apply some discretion in adapting some of these policies locally, for instance the Irish manufacturing plant refused to implement a salary modelling tool, which confirms evidence from other research in MNC subsidiaries [18,19,26,32].

National institutions and business systems in Germany and Ireland diverge considerably. It has frequently been argued that the employment relations (ER) system in Germany is highly regulated, whereas Irish ER appear to be a lot more deregulated by comparison [1,56]. The German system of co-determination is characterised by indirect worker participation through elected worker representatives and a myriad of formalised institutions [77]. It has also been argued that the key labour market institutions, multi-employer bargaining, co-determination, and initial vocational training, curtail managerial prerogative [69]. As one might suspect, legal compliance issues represent the key reasons for adapting e-HRM practices. However, the evidence also shows that these may be circumvented by using supplementary systems, as was the case with payroll systems, which are unique in each subsidiary. The localisation drivers, or high context specific drivers, in the unionised German manufacturing plant outweigh the drivers for standardisation to some extent. For instance, the introduction and subsequent changes or amendments of an e-HRM system would have had to go through a formal consultation process with the works council. Any veto by a works council effectively would have put a halt on the usage of the system [68]. However, the works council in the German manufacturing plant did not seem to object to the introduction of e-HRM, partially perhaps because the works council chairman and his deputy belonged to the management team of the German manufacturing plant. In the also unionised Irish manufacturing plant, the force of localisation drivers is low compared to the drivers for standardisation. In other words, trade unions were neither consulted about nor do they have any influence on the use of the system, as stated by the Irish stakeholders interviewed. It comes therefore as no surprise that the Irish subsidiary is rather more willing to adopt US e-HRM practices than the German plant, although the former has blocked a number of e-HRM practices and was able to do so but for reasons other than the drivers in question. The picture in the sales HQ for the central region mirrors that of the Irish manufacturing plant. Evidence from this research suggests that this is largely due to the absence of a union in the sales HQ. In the opinion of the HR Director for the Central Region, customisation of the e-HRM system could have been a more contentious issue had the central office in Germany been unionised or had there been a works council. The absence of a works council affords some advantages in the daily HR operations of the Sales Head Office.

Because we don"t have a works council, the recruitment process is simplified immensely for us. Because if you have a works council, then you have to first announce all positions internally for two weeks before you can go external. This for instance would be an issue that we would have to incorporate [in the system] if we had a works council, but since we have none … (German HRIS specialist)

The picture of e-HRM practices that manifests itself in the evidence provided is rather fragmented. Is it possible, therefore, to consider the e-HRM practices employed across these subsidiaries to be convergent (toward US model of e-HRM practices), divergent or in a form of stasis [66]?

The MNC operates a strict "no customisation unless legally required" policy regarding the rollout of e-HRM practices in its subsidiaries. The Finance Manager of the German manufacturing plant explained the customisation strategy.

The global aspect is always checked and if [a customisation] can be implemented globally and if it is advantageous not only for [us] then it is highly likely that it will be implemented quickly. If it is specific to our location and if you don"t have a sufficient rationale as to how important it is, then nothing will happen. This policy is an indicator of the high levels of control in the corporation, which is a view that is echoed by the German and Irish HRIS Super-Users. For the most part, the strategic subunits (SBUs) in this research lean towards the full adoption of e-HRM practices. As far as the MNC is concerned, this makes business sense. For instance, the above section on the role of e-HRM has stressed the significance of internal consistency regarding the collection, management and analysis of employment related information. Overall, the European HRIS centre and European subsidiaries appear to have little leverage concerning decisions made by the US parent"s HRC. The SBUs were neither informed nor consulted during the implementation phase. This seems to endorse Burbach and Dundon"s [9] findings on HRIS and e-HRM utilisation in Ireland, which stressed that nine out of ten organisations did neither consult with nor inform employees of the e-HRM implementation. This is rather surprising, since it is well established that employee involvement can be correlated with system acceptance and ultimately system success [40, 87,94]. The HQ drove system implementation in Europe and individual subsidiaries had no choice but to implement the system under the aegis of the European HRIS Centre. Despite HRIS Centre Manager"s suggestion that the introduction process went smoothly, interview data indicates that resistance to the initial implementation was extensive, as, according to the Head of the European HRSSC Project Team, the corporation appeared to assume a "sink or swim" approach to implementation. Evidence intimates that some form of resistance to the e-HRM system in general appears to persist in the subsidiaries in that some line and sales managers seem to maintain what has been referred to by interviewees as "shadow administration", in the form of Excel or paper-based files, by some of the interviewees to circumvent the use of the global system. While officials know of their existence and their inappropriateness, they appear to have resigned to the fact that they continue to be used. I am almost certain that there are still managers that still have these. One has to concede that managers do not work daily with PeopleSoft. One can be almost sure that one or the other personnel file still exists. Fine. You can"t do anything about it. It also won"t change in the near future. (HR Director for the Central Region)

The more recent introduction of a talent management system that requires employees to complete an extended online Curriculum Vitae, which is available to superiors, appears to be the cause of continued resistance in this MNC, whereby employees look to be reluctant to fill in these online profiles of themselves. With respect to the adoption success models presented above, the evidence furnished here evinces that e-HRM, while it may be implemented as far as the European HRIS Centre and US HQ are concerned, has not been internalised by the subsidiaries. That is, individual subsidiaries have not displayed wholesale commitment, satisfaction or psychological ownership [59] of various e-HRM practices. However, the MNC has introduced a simple but effective way to ensure greater levels of compliance with the e-HRM systemwithout completing their online talent profiles staff will not be promoted. A similar policy ensures that line managers use the online appraisal mechanismemployees will not receive pay rises or bonus payments if the appraisal has not been conducted via the online system. A different usage pattern emerges when the e-recruitment facility is taken account of. While the Irish manufacturing plant fully capitalises on the system"s features, for example, the system is even linked to an external job search website to increase exposure, the German manufacturing plant does not post any vacancies on this system, as stakeholders prefer to attract only local staff. This evidence suggests high levels of corporate control and high levels of embeddedness of the MNC in its home context [1,30]. For instance, the Irish Manufacturing plant was also able to prevent the implementation of a new system module.

They wanted us to introduce a Salary Modelling tool, which we thought was too complicated. The system we use is Excel based, simple, and very user friendlydifferent salaries can be determined straight away. So here we have been able to resist the introduction of new practices. (Irish HRIS Super User)

The Irish HR Director remarked on the issue of system implementation

We have now reached a critical mass of 2300 employees where we could say that no new systems would be introduced in Europe without our okthe economies of scale just would work -there wouldn"t be enough people in the rest of Europe to make it work.

As the Irish manufacturing plant also operates a sizeable research and development unit, their impact on system implementation may equally be attributable to sources of micropolitical power within the organisation as it may be related to institutional influences [8]. It is apparent that the Irish subsidiary has gained considerable strategic importance and resource power, which it is able to leverage in exchange relations with the HQ and other SBUs [5,18,33,61]. Unlike the Irish Plant, which has far reaching influence on the introduction or non-introduction of some practices, e.g. self-service HR, the much smaller German Plant has virtually no sway. The German manufacturing plant appeared to be somewhat disillusioned after the implementation of the HR system, which the following quote by the German Plant Director highlights.

You introduce such a system, because you want to benefit from its rationalising effects, because you want to introduce a global system that can communicate with each other in entire holding company. For us this means 75% more administration, because nothing is like it used to be, because nothing works the way we would like it to work. And now there is somebody who says, PeopleSoft, there you"ve got it and he doesn"t realise how could they actually manage it? How should they handle it? How much personnel will they need to derive any value from using the system? A [CEO] presses a button and sees his 100 best employees. [...]. He has a staff of 100 people that present everything that they generate out of the system on a silver platter-brilliant. But what use is it to me. I am not [the CEO]. May main priority is that my employees receive their correct wages at the right time. PeopleSoft can"t do that. There, I don"t care about PeopleSoft.

PeopleSoft is at the very back of my priorities. […]

Not all HRIS practices used in Europe originated in the US. In some cases, "Europe" has been able to successfully develop and implement unique e-HRM practices, which were adopted ex post facto by the US in a process of backward integration [22, 23, 24, 37]. For instance, internet recruitment and manager reports were two initiatives that were developed in Europe and after successful implementation rolled out in the US. The notion of a global uniform e-HRM system in this MNC is somewhat contradictory, as above evidence emphasised. Considerable differences in e-HRM applications exist among the subsidiaries even within the same country. Moreover, each subsidiary appears to rely on a number of parallel systems. In Europe, every subsidiary uses its own payroll and time and attendance system, which is possibly due to differences in national legislation, which governs payroll administration. Other third party systems include training administration, salary modelling tools or quality assurance systems. As the MNC largely expanded through an acquisition strategy, many of these systems were part of legacy systems still used in the corporation. According to some of the key stakeholders interviewed (HRIS specialists and German Plant Manager), the proliferation and incompatibility of these sub-systems with the global system presented a major barrier to the operation of e-HRM in the corporation and as far as they were concerned further evidence of the lack of commitment to the overall system by the HQ. Above discussion has stressed the multifarious nature of institutional factors and isomorphic pressures that impact on the diffusion of e-HRM across the subsidiaries of an MNC. A number of conclusions can be drawn from this analysis.

Conclusions

The basis for this paper was the assumption that e-HRM practices in the subsidiaries of an MNC were subject to the same or similar institutional factors as standard HRM practices were. An analysis and discussion of the interview data revealed distinct differences in the e-HRM practices employed in the subsidiaries. Moreover, a range of institutional factors that may account for differences in diffusion of e-HRM practices could be discerned. The evidence presented above highlights the complex nature of the relationship between home and host country effects, pressures for standardisation and resource capabilities of subsidiaries. This phenomenon is frequently referred to as "institutional duality" [60,67]. A number of authors have argued that the dynamic nature of national business systems further complicates an accurate assessment of the factors shaping the constellation of HR practices (and thus e-HRM practices) in the MNC [45]. Björkman and Lervik [6] suggest that the transfer of employment practices is first and foremost a social process that is influenced by corporate governance, subsidiary and HQ relationships, the nature of existing HR systems and the strategy used by the HQ to introduce practices. Moreover, the balance of decision making power in the MNC is the result of negotiation and micro-political activities between organisational actors and business units, which is mediated by contextual and structural constraints that the organisation finds itself in [31,65]. Smale and Heikkilä"s [83] study has evidenced that negotiation and micro-political activities are key factors in e-HRM implementation. Evidence from Ireland in this research seems to intimate the impact of micro-political influences in the form of strategic capabilities [38] in the manner in which the subsidiary could influence the introduction of certain e-HRM practices. Despite some evidence for convergence of e-HRM practices in the subsidiaries towards the e-HRM practices promoted by the MNC"s US HQ, discernible variations exist between the e-HRM practices used by the multinational abroad and those employed in its home country. This finding is reflective of other studies focusing on the convergence / divergence of HR practices in MNCs [3,10,54]. One of the key findings of this paper, which underlines the applicability of institutionalist theories to e-HRM research, is the dichotomy that exists between what the US and international HQ consider implementation and the conflicting reality in the subsidiaries. Commensurate with Kostova"s [59] and Björkman and Lervik"s [6] conceptual models of organisational (HR) practice transfer, some e-HRM practices could be considered integrated, for instance basic employee record administration has become part of the organisational routine, while other e-HRM practices such as talent management are merely implemented, that is, enacted. Differences in transfer of e-HRM exist owing to dissimilarities in the institutional, organisational and relational contexts of the subsidiaries in this research [59,60]. Applying Oliver"s [73] strategic response model, the range of responses to the introduction of e-HRM practices in the subsidiaries ranged from acquiescence, compromise, and avoidance to defiance. However, the level of adoption and type of strategic response vary with each practice from one subsidiary to the next. This paper poses that e-HRM research will benefit from the application of these models to ascertain the adoption success of e-HRM practices. While this single case study research is limited in focus, the evidence presented here moots that e-HRM diffusion across the subsidiaries is contingent on an intricate mélange of a variety of institutional factors. Furthermore, this paper has evinced the validity of institutionalist theory as a new research paradigm and macro theoretical foundation for future e-HRM research. This paper also advances the view that aspects of both neo-institutionalism and European institutionalism may advance the field of e-HRM research. However, as this paper is of an exploratory nature, additional research taking into account a broader range and number of MNCs and their subsidiaries will be required to further test these hypotheses.