Multinationals´ Policies and Local Responses: findings from cross-national case studies in Germany, France and the USA

This paper examines the complex interaction between globalisation practices of multinational corporations and the governing potential of response by local systems of employment regulation in light of rationalisation processes linked to financial market driven strategies. It investigates this issue by looking at optimisation programmes conceived and developed at the central headquarters of two German multinational companies and then traces their implementation at local sites in three different international settings: Germany, the USA and France. A central focus was the role that industrial relations systems play in the transformation process and whether (or how) the response varies according to the strength and coordination of the institutionalised system of industrial relations. The study shows that differences still remain in the form and strength of the industrial relations systems in these three countries. However, in all three cases, a new development in industrial relations is taking place in which areas potentially relevant for bargaining are left ‘unoccupied’. That is, they are missed completely in the traditional negotiating agendas of industrial relations actors.


Introduction
This paper examines the complex interaction between the globalisation practices of multinational corporations and the governing potential of local responses in light of rationalisation processes linked to financial market driven strategies.It investigates this issue by looking at optimisation programmes conceived and developed at the central headquarters of two German multinational companies and then traces their implementation at local sites in three different international settings: Germany, the US and France.It first looks at the programme's mission, and then how the programme gets translated into concrete measures that are undertaken in selected sites.In a second step, it considers the potential for local response to the demands coming from central headquarters, either through formal systems of employment regulation or through informal, but concerted efforts (Ram et al.,2001).In this way it offers lessons for two research traditions: one focussing on the dissemination practices of large companies in different national settings (Ferner and Quantanilla,2002), and the other on attempts to understand the modifying potential of different systems of employment regulation (Hall and Soskice,2001;Ebbinghaus and Manow,2001).
Initially, as the general streams of convergence vs. local determination in the discussions on the effects of globalisation on workplace practice and relations crystallised, the two different lines of argumentation became more and more polarised.Convergence proponents posited an increasing standardisation of company policy, divorced from considerations of national specificity and traditions (Reich,1993;Berger, Dore,1996).The internationalisation strategies of multinational companies were attributed with an economic logic that washed nationally specific differences away, moving them toward a single world economy and world society (Narr, Schubert,1994).The reasons offered for the ongoing trend toward convergence, as well as the inclination to view it as either a positive or negative trend differed greatly.Some views on convergence saw it as deriving from the dissemination of best practice (Ohmae,1996;Womack et al.,1991).Others were 'negative' scenarios, in which multinationals force their subsidiaries to implement particular work forms or structures without consideration of local specificities or problems.
Local diversity arguments tended to counter the convergence positions by looking at the enterprise level and pointing out how forms of best practice tend to adapt to local conditions or how attempts to force unitary structures simply fail.Thus, in contrast to convergence theses, local diversity perspectives questioned to what extent international companies can pursue their strategies and be unhindered by the need to adapt to local contingencies.Research in this area led to a decisive revision of the approaches that assumed the existence of 'footloose' global enterprises (Dörre,1996;Marginson et al.,1995).For instance, it was shown that the ability of local institutions and actors to exercise influence in global companies is linked to the type of internationalisation strategy the company is pursuing and how that meshes with the factors that the local site offers, including infrastructure, access to markets, qualifications, unemployment rate, wage levels and taxes, etc. (Flecker, 2000).In this type of argumentation, for companies which are pursuing strategies in which the local development of products and the use of differing production conditions are important (i.e.transnational companies), a comparatively strong regional reference and the development of differentiated product and production strategies will lead to high levels of autonomy (Hirsch-Kreinsen,1998).In this scenario, the power that individual sites can wield depends in large part on their resources and attractiveness as production sites.Studies in this genre argue that product type plays a large role in how autonomous local sites can remain.If product complexity is high, use of local conditions is more necessary and therefore divergent strategies are endured by the global company (v. Behr,1999;v. Behr, Hirsch-Kreinsen,1998).
Nevertheless, there are indications that even when so-called autonomy for local sites is permitted, it comes under conditions of high risk to the local site.Goals are set or negotiated and the path to reaching them is left to the discretion of the local site.However, failure to achieve these goals leads to consequences such as a return to centralised controls or, quite possibly, closure.Thus, local autonomy, or even the recognition of necessary variations in product markets and processes by the international company headquarters, does not automatically mean that local companies will remain untouched by strategies of 'coercive comparison' (Flecker, 2000).Goal setting, benchmarking and competition for contracts and investment funds are control mechanisms that can be implemented regardless of the level of standardisation or integration of production.Nonetheless, some researchers take the position that exactly how such control mechanisms get played out at the local level is not necessarily as straightforward or predictable as perhaps intended by the companies.They argue that, behind a surface of standardisation and aggregate information systems very disparate processes and structures actually exist in the majority of multinational companies (Dörrenbacher,1999).Another line of discussion suggests that management and workers at a local site may have similar interests in their response to unpopular directives from the central corporate level.This alliance further facilitates the maintenance of their difference.

Point of Departure
As can be seen in the analyses and studies that have been carried out, there continues to be a persisting diversity in the findings and also growing evidence that a polarisation of positions does not fully capture the picture of what is occurring in global companies at the local level.This paper does not presume to offer a synthesis of the various theoretical and conceptual directions offered up to now.It does, however, recognise the difficulty of presenting a well-balanced and accurate picture of the role of MNCs in the globalisation process and their effects on local sites given that the necessary objects of study are different actors with a variety of interests, diverse institutional settings and national contexts, and changes over time.
More recent studies of the role of MNCs in globalisation have attempted to deal with some of these problems by integrating more than one level of analysis -macro, meso and micro -in the research process to get a more complete explanatory picture.(Meil et al., 2003;Ferner et al., 2004) One outcome is a more contradictory and complex view of multinationals and globalisation in which simultaneous tendencies of centralisation and decentralisation, control and autonomy, convergence and local determination are apparent.These trends are not arbitrary, but also have a certain strategic logic behind them.
This study follows in this path.It first takes a close look at the practices of MNCs and then attempts to link micro level processes with meso level institutional systems.In taking different levels of analysis into account, it tries to explain not a single trajectory of development but a complex and partly contradictory one.We posited that large multinationals will in fact try to impose convergent structures or workplace conditions on their subsidiaries.This is the case for one reason, because the need for central headquarters to control their subsidiaries and engage in benchmarking leads to principles of uniformity and, for another, because central headquarters generally view the methods that they have developed as being the 'best' ones for the strategic orientation of the company.However, we argued that these central convergent strategies will meet with resistance as well as difficulty and will be transformed by local contingencies in their implementation process at the local site.In addition to local effects through varying institutional systems and labour traditions, we also expected host country effects: The thesis was that the content of the programmes and the way the MNCs approach implementation will be affected by the home country in which the programmes were developed.Both the MNCs in this study were of German origin with strong traditions in their home country and a strong German centre.Given that in their home country a strong labour voice and high levels of participation were common operating principles, we expected greater attention to local concerns.Furthermore, German companies, with traditionally less orientation to purely financially driven goals and more long-term perspectives (Jürgens et al. 2000), might be expected to pursue less purely financially motivated policies and should exhibit less centralised and unitary control of local sites than for instance US MNCs (Ferner et al 2004).
The main aim of our study was, however, not to determine whether local conditions matter, but rather to examine exactly how local actors, either formally (through existing institutional structures) or informally, contribute to changes in central company practices when these are implemented at the local level, or when applicable, what forms of resistance arise in response to policies that are centrally mandated from headquarters.Therefore a second aspect of the study was to see and understand how local actors respond to the measures coming from the MNC headquarters.Our assumption was that the institutional framework, for instance in the form of industrial relations systems, would be relevant in the transformation process and that the response would vary according to the strength and co-ordination of the institutionalised system of industrial relations (Albert,1993;Ebbinghaus, Manow, 2001).

Methods
Our empirical research centred on an examination of a comprehensive policy initiative being launched by a multinational electronics company (Company A), which was intended to be implemented at all of its (nearly 200) international sites.Company A seemed a logical choice as a case study.It is a traditional German company with a long history, a strong centre and a reputation for hierarchical organisational structures.However, it also has long term experience with international production sites, which have existed over a long time period at their individual national sites and therefore have established practices and traditions of their own.Another incentive for choosing Company A was that it was in the process of disseminating a comprehensive company-wide optimisation programme at all of its sites.Such 'explicit' (see Marginson et al.,1995) programmes are not that common.Many companies pursue 'implicit' programmes -best practice, employee involvement, personnel strategies, etc.These programmes are often vague or are not comprehensive -that is, they are geared to one department, one assessment measure, or one procedure.Global programmes with centralised methods, facilitators, training, backing by top management, a special department to carry them out, and lots of publicity are much less common.
For comparative purposes we also decided to examine the implementation of a programme in a company that was not an industrial production enterprise, but rather in the increasingly dominant service sector.We chose an insurance company (Company B) because of its economic importance and its German headquarters: it had sites in all countries relevant to our project and like Company A was in the process of launching a programme to be disseminated to all of its sites.We examined the direct implementation process of the two programmes at the local company level in three different countries and, for the highly diversified Company A, at different business units.In this case, we were able to keep one division constant for all three countries, and another for two of the countries.The national contexts -Germany, France and the USA -were chosen for their variations in institutionalised arrangements of industrial relations.
We conducted intensive interviews based on a pre-formulated list of questions with company managers, works councillors, and when possible, regular employees.We also spoke with union and works council representatives in France and Germany.Altogether 26 in-depth interviews were carried out in the three countries with individuals and groups in the time period between 2000 and 2002.Additionally literature from or about the companies, both from the intra/internet and published documents was analysed, and an extensive review of the secondary literature was carried out.

Description of the cases
This section summarises the study's findings regarding the dissemination of the MNCs' programmes, looking at what kinds of projects or practices were actually implemented at local sites.At the level of strategic programme goals, the attempt to achieve standardisation across the entire company versus the extent to which local specificity was permitted or even systematically encouraged was examined.At the level of practice, for instance in the use of tools and instruments, it was asked to what extent local and nationally specific flexibility is permitted or whether directives from the central headquarters have to be followed and are monitored accordingly.At a third level, the adaptation processes at the local or national levels were described, revealing potential regional differences across the multinational.Also considered were changes in orientation from previous company programmes, the relationship to the centre in the diffusion process, the general attitudes to the programme and implementation strategies by the individual sites, the effects for the site (for instance in terms of organisational changes, pay, job security), the reaction at the local site to the central HQ initiatives, and finally the effects on autonomy and control for the local sites.From these findings more general observations in terms of the interplay between convergence and local diversity were made.

Company A -Traditional manufacturing industry
According to those interviewed in Company A, including the CEO of the US holding, the main and explicit goal of the programme is the clear orientation of all the units toward company success (economic success) and an increase in competitiveness.The programme is geared to change what is seen as the insufficient profit orientation of unit managers and make the company results more transparent (and thus more comparable).At the individual site level, the main goal is the increase in the contribution to value added and 'an orientation to economic results in all areas, down to the smallest unit.' (German works councilor) The responsibility for the planning and implementation of the programme lies at the German headquarters.There, a programme manager, the head of the programme office, is responsible for all of the sites, world-wide.The service function of the programme office is the development of a toolkit containing instruments that should help the world-wide sites develop programmes to reduce costs and waste and improve quality.These include a range of the current state of the art tools for management and financial methods to optimise, improve, and monitor company processes in a number of different categories (for example, cost reduction, innovation, boosting sales, etc.)In each category there are a number of modules offered to help define and direct projects that are appropriate for a particular unit or problem.So with cost reduction, for example, the emphasis may be on structural change or on lowering purchasing costs or on optimising processes.Within these subcategories are other sub-categories of more refined modules.The tools are thus a means of information, a form of training and an impetus to get projects off the ground.Since the programme´s inception in,1998, the number of individual tools has swelled to about 2000.
The control function of the programme office is reflected in the motto of the programme: 'clear goals, concrete measures, and rigorous consequences' (Company A 2001).'Clear goals' means reaching EVA (economic value added) targets.'Concrete measures' means using the catalogue of tools geared toward portfolio optimisation, earnings-oriented sales growth, asset management, cost reduction, quality and innovation as defined by the company headquarters.Finally, 'rigorous consequences' means reviews every three months to monitor results.When there is 'persistent' deviation from targets the Corporate Executive Committee decides between four options: buy, co-operate, sell or close.
This type of programme heralds a clear break with those introduced in the 1980s and early 1990s whose goals were tapping productivity gains through employee participation (Dörre 2001;Applebaum & Batt,1994).The difference is clearly evident in the change in orientation between an earlier programme launched by Company A and the current one.The programme's forerunner at Company A had originally been launched at the beginning of the 1990s at the initiative of the corporation´s board of directors.The goals were to identify and improve the company culture and to create a vision for the future development of the enterprise.The central focus of the programme was a re-engineering process, intended to lead to an optimisation of time and process use in response to what was seen as an overly bureaucratic organisation and long development cycles.The goal was to disseminate the programme worldwide, in the 193 countries in which the company was active.The programme was geared toward an increased participation of employees in continuous improvement and more efficient design of work and production processes.In Germany, 105 sites participated in the first programme.The works councils, from the central HQ to each of the sites, were involved in the implementation and strongly supported the programme.The general tenor of the interview responses at the local German sites was a strong emphasis on participation and improvement of company work processes in the first programme.In France, the first programme was apparently also received positively, although it was admitted that a number of existing initiatives were placed under the programme label to fulfill programme requirements; also, the programme was not as widespread in France as in Germany.Interestingly, in the USA, the first programme never really got off the ground.Moreover, for those familiar with it, it had an extremely negative connotation of being completely oriented toward cost-cutting measures, which is exactly opposite to the impression it made in Germany and France.
At some point in the mid 1990s, the first programme began to fizzle out and a new programme generation was re-launched (beginning in 1998).It was linked in name to the first, but with a new orientation geared to designing projects that achieved concrete and measurable economic results, and an overall increase in the transparency of economic performance.Individual practices at the company level in the new programme launched in Europe in late 1998 and worldwide (after some initial problems) in 2001 may still evidence the participatory practices found in the initial programme.However, increasingly the message from the centre being transmitted to the companies, and in particular their management, is that financial results are the ultimate goal; and these must be visible immediately.The consequences of not heeding this message are looming in the form of sell-off or closure.The pressure to meet these demands immediately naturally encourages short-term cost-cutting measures to achieve the desired results without thought for the long term consequences of such measures.
At the German sites that were investigated, there was knowledge of the programmes and tools at the management and works council level, but less actively at the employee level.The introduction of the programme was usually announced through regional presentations in the form of large formal assemblies at which high ranking officials from the central HQ -the chairman of the board in some cases -presented the programme and its goals.In Germany, this meeting was received quite critically, the impression being that the programme was too abstract with no attempt to link the programme to specific problems or the actual situation of the individual site.(In fact, the standardised presentation was a strategy of the centre, which intended to give the same presentation to all of their sites around the world, thereby transmitting a signal of company cohesion.In all three countries, this strategy met with a negative response.)Additional information on the programme was provided on the company's intranetthe main means of dissemination.The relationship with the centre was otherwise very distant.It took the form of reporting of economic results.Intervention from the centre only came about if the economic value-added targets were not met or in cases in which the sites asked for help for programme implementation (which happened rarely).
Although the information for the programme in Company A came from the centre and it was central HQ goals that were being communicated, the implementation process itself proceeded at the local level.Tools and instruments were selected that met the requirements of the projects being launched and were perceived as being appropriate for the particular needs of the site.The programmes tended to be directed toward cost reduction, quality improvement and other internal company processes.The practice of placing already running programmes under the banner of the programme to meet the demands set by the centre was also used.
At the US sites of Company A which were investigated, there was also an emphasis on quality improvement and waste reduction, but there was an additional focus on customer integration and meeting customer needs (to be visible in improved sales figures).As in Germany, the programme was launched by a media presentation at which the US CEO (a German) and members of the German board of directors presenting the programme goals and tools.This assembly was aimed at management, so the knowledge of the programme at the employee level was quite marginal.The programme was officially launched in 1998-99 but was mainly ignored by the US sites, so it got underway very slowly.The actual introduction (under greater pressure to comply) took place in 2001.For most of the sites, the response to the programme was negative; the tools were perceived as being overly complex, thus requiring a large staff to interpret and implement them, which the US sites did not have available.This led to the impression that the German sites, and of course the central HQ in particular, had much more manpower available for such activities, leading to further bad feelings.Other criticisms were that the programme was too geared to internal company processes and incremental change rather than the larger picture, which would in the end, not help the sites to perform well in a difficult economic climate.Most US sites felt that they had already introduced programmes with similar goals and orientations (to improve business performance) before this particular programme was imposed upon them by the Germans.Thus the centralised programme implementation also created negative feelings in that the existing programmes were not valued or recognised unless they were put under the rubric and label of the programme initiated by central HQ.
Other than the kick-off meeting and pressure from their senior management to develop projects and the economic results to go with them, communication with HQ was rare and reduced to a reporting function.Information about the programme was available on the intranet, but it appeared in English rather late, and except in rare cases 1 was not actually used.The practice of using existing programmes and 'selling' them under the programme banner was widespread.
In general, the programme, although centrally launched, did assume that the actual measures would be designed locally with local contingencies in mind.However, the 1 In one case, an individual engineer used the information from the Intranet to launch a highly successful project in his company.The group involved described the process as almost accidental.They were looking for information to start a waste reduction project and stumbled across the information which was partly in German.Through individual initiative and getting their local management to agree (at first only haltingly), the group designed one of the most successful projects in the USA.
pressure to report results in terms of economic value added in shorter and shorter intervals has intensified over the last couple of years for the US sites.This has led to concerns at the sites that the orientation to short term financial results will affect their overall ability to grow and innovate.The strict reporting schedules and procedures and the lingering awareness of the consequences of failing to meet economic targets has clearly impinged on the previously very high levels of autonomy enjoyed at the US sites.
In France, the programme in Company A led to practices mainly centred around quality improvement, although there had been a shift away from a focus on the production area toward service departments and customer service oriented projects.As at the other sites, the programme specifics were launched at a large meeting by representatives of the central HQ.Once again this method was criticised as being too abstract and the message as having little relevance since it came from board members with whom the employees were unfamiliar.At one unit that was performing poorly, the presentation was experienced as a type of threat which immediately gave the programme negative connotations.Also, the very financially technical presentation gave the impression that the programme was geared mainly to management and was not relevant for the working situation of individual employees.As in the other international sites, the other information channel was the intranet.Actual implementation in practice was, as in Germany and the US, left up to the individual sites.In the French cases under study, a great deal of emphasis was placed on group development and using employee suggestions to develop measures to achieve the desired economic effects.This worked relatively well and offered a high level of continuity with the forerunner programme.France was in fact one of the few countries in which the personnel from the first programme were often allowed to remain in their positions for the re-launched programme.One problem mentioned, however, was whether funding to develop the projects from the suggestions made by the employees would be sufficient in light of the tight economic framework.Evaluation was also seen as problematic since the financial impact of many of the measures would occur later than the requirements for reporting results.In fact, the French sites stated that a major difference from earlier programmes was the evaluation of the practices undertaken in conjunction with the programme according to their contribution to short-term economic goals.The means of reporting differed between plants: at one, productivity increases of 7-8% were targeted; at another savings of 15 Mio € in 1-2 years, the time intervals of reporting were getting shorter everywhere.Best practice sharing between sites of a unit -one of the expressed goals in the programme -occurred through individual initiatives, but not on an organised basis.Help from the centre in the form of training or aiding communication between sites did not exist.Communication was reduced to reporting on economic results.

Company B: Insurance and Investments
Company B headquarters in Germany also initiated a programme that was meant to be disseminated world-wide and was unmistakable in its orientation through it name: EVA (Economic Value Added).Generally, however, the programme was much less structured and also less directed from the centre than the programme in Company A. There was no central department formed solely for the purpose of developing or controlling EVA.
An international task force was formed to design appropriate measures for financial reporting: an interesting difference from Company A in which guidelines and procedures were developed at the central German headquarters.It was expected that the various sites would use the reporting system so that each site's contribution could be monitored.If a site exhibited several poor results in succession and became a problem site, a trouble shooter from the centre might be sent in.Continued poor performance could result in a change of management, a merger, a different arrangement of responsibility areas, or closure.Besides reporting and its possible effects in negative cases, there was little interference from the centre.Exceptions were the co-ordination of projects that it made sense to disseminate and standardise internationally, for example IT systems.In these cases international teams were usually formed.The head of controlling in Germany pointed out that going onto the US stock market demanded a stricter reporting for financial targets and increased disclosure of financial assets, one of the motivations to monitor results more closely in a standardised manner.
At the German site of Company B, the works council supported the EVA programme since it led to transparency in areas that were costly to the company without an accompanying contribution.The background of this support was that at the time of the interviews ( 2002) there appeared to be little risk of personnel reduction.Moreover, the local sites enjoyed a very high level of autonomy in day to day operations, largely due to the business structure in which knowledge of the local market and local regulations was extremely important to the HQ.The direct effects of the programme for local sites were limited: organisationally there was virtually no effect and the consequences for pay were felt mainly at the middle management level.At the German sites there was an increase in performance-based premiums, but this did not affect base pay which was negotiated with the union.In the USA, performance pay systems and bonus packages which were negotiated directly between the employees and the personnel department did affect employees by creating 'winners'and 'losers.'As a result of profit sharing, the employees at sites with good economic developments earned more at the expense of employees at sites with poor economic performance.Such differences could even occur between different departments on the same site.Referring to the more severe effects of EVA reporting, the US chief of finance remarked that 'company B is slow at giving consequences to underperformers' .

Summary of the Findings
In the programme implementation in both companies, the two aspects of centralisation and decentralisation are clearly recognisable.The programmes and their goals are directed toward convergence to the extent that they are targeted at every company unit and subsidiary worldwide with fixed and centrally-set goals.They also serve to strengthen and especially to systematise company restructuring and optimisation measures.This means that all participation and rationalisation measures that exist at the sites are made more transparent, and those which do not conform to the central programme´s goals and orientations are targeted for possible elimination.Thus, the company´s control programme does serve to bring all of the sites of the MNC into line.On the other hand, each site has autonomy to decide how to meet the centrally-set goals.In fact, it is more or less accepted that existing rationalisation measures, as long as they are successful, can be defined into the programme, even though they were not specifically designed for, and may even be precursors to it.
There were positive responses to the programmes of the MNCs in terms of using the opportunity in Company A to initiate a change process or to increase transparency in Company B. However, most of the responses tended to be negative, especially in Company A. Expressed were a feeling of increased pressure from the centre without accompanying support, feeling threatened from HQ, and the impression that the programme from Company A might not necessarily hurt them, but it would not necessarily help them to improve performance either.Yet, because of the accountability tied to the programme, which was marketed as a comprehensive company strategy for optimisation, the sites would be held personally responsible for the results of economic downturns over which they might have little control.
There was no evidence of any organised reactions, either through formal institutional mechanisms, or informal means.In Germany, the works council could intervene in the case of personnel reduction with agreements on settlement packages if the reductions were due to benchmarking or best practice activities.However, when there was no direct outsourcing or displacement, it was easy for the company to get around these rules, which were in any case not preventative.By introducing measures for increased flexibility, the company could use temporary or contract workers, thus making permanent jobs obsolete.Collective agreements at the union and employer level protect wage levels.However, personal evaluation systems and performance agreements bypass the regular wage negotiations.For middle level management, pay and career opportunities are increasingly determined by their site's performance.
In France, many of the projects undertaken in conjunction with the programme in Company A were seen as participatory measures that had been generally on the increase in French companies after the implementation of the Auroux laws in the early 1980s.There was no organised response either in company level representation or the union level.Many of the effects on remuneration for individual employees and managers were similar to those in Germany.
In the USA only one site of Company A had union representation which limited its activities to formal issues.Company B had no union in the USA.At the local level the US sites seemed to follow a strategy of ignoring the demands from the centre for as long as possible.In any case, any organised response was bound to be limited because knowledge of the programme at the employee level was limited.Most information remained at management level and was only operationalised in the form of concrete and local optimisation projects.Among regular employees there was little awareness of the central HQ´s initiatives, except in very abstract terms.

Shift in orientation towards a market-based regime
As we proceeded with our research, we discovered that some of our initial premises were being called into question.The reason for this was that for the multinationals in our study, in particular Company A, a major shift in orientation for company policy had taken place.The direction of change was not so surprising (DiMaggio 2001;Hall, Soskice 2001).A crucial development with severe consequences for companies listed on the stock exchanges today is that the market is no longer made up of lots of small nameless stockholders.At present, large fund managers in investment, money market or pension funds, with extremely large numbers of shares invested in particular companies have restructured the balance of power between those who run the companies and those who 'own' them.Such 'investor capitalism' (Useem,1996) is characterised by institutional stockholders who are identifiable actors and can exercise power in influencing the activities of multinationals, especially when financial performance is not leading to the desired stock price performance.Investor capitalism creates a situation in which top managers -under pressure themselves -put extreme pressure on their individual site managements to achieve positive financial results on a regular basis (Dörrenbacher, Plehwe 2000;Kühl 2002).
Thus management decisions are now influenced or even controlled by shareholder considerations in a way that did not exist previously.'Work was organised within corporations in ways that ensured that managers were the key insiders... ' (O´Sullivan 2000, p. 71) Having no clearly definable owners to answer to, managers found themselves with a lot of autonomy and power to run multinationals.This did not mean a complete absence of accountability.There was a separation between beneficial ownership of corporate stock and strategic control over the allocation of corporate resources.(O`Sullivan 2000).This meant that daily decisions and even long term strategies were much less coupled to instant financial results than seems to be the case today.If managers argued for a certain strategic course in the era of managerial capitalism, they were credited with being most familiar with the needs of the company until proven otherwise.Some authors argue that local management is sometimes pressured into making decisions that are against their own inclinations or will.One example is being induced to raise dividend shares after a good performance (i.e.distributing profits) rather than saving the money for a 'rainy day' fund, which many managers in industrial enterprises will be inclined to do in order to weather the almost certain downturns in business cycles.
That this shift could be so visible even in a traditional German company was unexpected.In Germany, there was a strong historical tradition of managerial control of enterprises even before the war (O`Sullivan 2000).In post-war Germany, the industrial landscape was characterised by a complex system of interlocking shareholding networks including a strong formal presence of bank control in corporations.Labour, represented on the boards of corporations, also has a voice through co-determination.The absence of small shareholders and the role that financial institutions were willing to play in company processes led to a system in which industry was largely insulated from strict market forces.Additionally, the high level managers of industrial enterprises tended to have engineering backgrounds, giving them a different orientation from enterprises run by financial specialists with a background in accountancy.Together, this constellation created a system characterised by the term 'patient capital' -a financial orientation supporting Germany´s high quality niche product strategies with a long term perspective (Deutschmann 2002;Jürgens et al 2000).In fact, the German literature expressly argued that contemporary German companies were not characterised by the strong shareholder value strategies found in other industrialised countries, because of specific national bank and financial and goals, as well as low levels of private stock ownership (Jürgens et al. 2000;Vitols 2001).
For the multinational Company A, however, this reorientation was evident, and represents a fundamental shift in ideology: away from an 'engineering' productdriven company to one driven by financial results.Thus the central identity of the company is no longer based on its technological superiority, but rather on its financial performance, the same standard applied to any other multinational in any other industrial branch.The other major change, also deriving from the strong orientation to financial results, represents a shift in philosophy concerning the ties between the various business units, or in other words, the view of what the company as a whole means.Does the corporation as such, which is divided into a number of units by product type, have significance for the individual companies?In the past, the practice of this very solvent company, which had large capital reserves, was to move capital between units to temporarily help out in periods of financial difficulty.This policy has changed so that individual business units are now held responsible for their individual results.Poor performance on the part of a company or business unit is seen as a strain on the corporation as a whole: a kind of malignant tumor that is eating away at the good health of the corporate entity.These shareholder value orientations are apparent in both MNCs in terms of the newly introduced central reporting, transparency and hard consequences for poor performance, but are most obviously visible in the case of Company A.
It also became apparent in the course of our research that the literature on the effects of globalisation in company practice did not adequately recognise the importance of the shifts in company policy for local sites.It became clear, for instance, that the dichotomy between local autonomy and centralisation as clear alternatives for company strategy that is often portrayed in the literature on globalisation is misleading.The programmes at the two multinationals in our study whose implementation we were investigating were the perfect manifestation of this change.Such programmes are geared to utilise two strategies that were previously thought to be in contradiction to one another.The benefits of local autonomy are reaped by utilising the positive attributes of the local setting -skills, costs, expertise, experience, market, resources.Therefore companies are encouraged to make use of the general, often diffuse, tools and methods defined by the company headquarters to custommake and implement a project or measure at the local company level: autonomy is in this regard welcomed.However, the reporting and benchmarking tools available to measure 'success' are determined using centralised financial monitoring systems and the accountability (or control) linked to this process is very high.This is not merely a formalistic process; it has very concrete effects on work and work practice2 .
As we know, the logic of the market dictates that poor financial performance has to be met with a fairly immediate response.The response is not pre-determined, but 'restructuring' is one of the common answers.Simply cutting personnel won't necessarily have an effect, but rather has to be part of a larger plan which might involve some form of reorganisation (for instance, outsourcing to a supplier, resettlement to a cheaper site, i.e. regime shopping (Streeck,1998;2001), creating a subsidiary, etc.).Although it would be unfair to every restructuring measure brought about by financial performance pressure as only being short term and cost motivated, it is true that the extreme orientation to financial outcomes can result in measures that are either cosmetic and not really structural improvements, or are even destructive because of their limited horizon.One example is the use of external contractors for formerly in-house tasks, which takes the costs out of the internal accounts even though the costs still exist.Another is selling off segments that are doing badly temporarily, but have a vital importance for product innovation.It has even been implied that one of the strategies of investor capitalism is to pay extremely high settlements or financial packages to top management as an incentive to side with shareholders rather than internal company interests when a conflict between the two could arise (O´Sullivan 2000).
Obviously the reach of financial orientations embodied in the programmes described here affects the working relationships of employees as well as management.The regular employee possibly feels the direct effects of the programmes only marginally in the form of a variety of measures introduced at the company to improve quality, improve productivity, streamline company processes, and potentially in cost cutting measures and job redundancy.But the ultimate consequence is of course massive job cuts, sell-off or closure.Given companies' new orientation to financial results even at the cost of long term considerations, it is not that surprising that employees at local sites are willing to participate in measures that are designed to achieve economic value added targets to save their jobs.

Conclusion: Market based regimes and local response
Our study revealed that the programmes being implemented by the companies we investigated are geared very much toward principles compatible with the financial orientations of what has been called the 'era of investor (or shareholder) capitalism' in Germany.These programmes place the various sites of the multinational companies under severe pressure to achieve set financial targets within fixed time frames.Especially for the German multinationals in this study, this policy represents a break with past practices in which a productdriven strategy was pursued and in which individual divisions were supported through periods of difficulty.What has changed is a much more pronounced short-term orientation to achieve results and a harder exposure of individual units to the harsh economic consequences of failing to meet targets.A loss of autonomy has taken place to the extent that units at different national sites are more controlled from the central headquarters both with regard to transparency in financial outcomes and with regard to meeting established targets.Autonomy remains, however, to the extent that the means to improve performance is not prescribed, at least up to now.
One stream of literature identifies these financial orientations with a new post-Fordist form of company organisation, characterised by the emergence of new control strategies which consequently transform the previously Fordist mode of control.The new form makes control anonymous, with the use of market-based mechanisms.(Dörre and Röttger 2003).'Market based control, decentralised firm organisation, flexible work, and management dominated participation are elements of the new production model' (Dörre 2003 p.23).For these authors, it is precisely these aspects of making power and control anonymous that signal a new regulatory regime.
As has been shown here, the subtle connection between the strategies being pursued by global companies in the form of optimisation programmes and the effects on employees makes it difficult for systems of employment regulation to exercise a governance function.Moreover, most of the effects occur within individual companies, so that only internal company interest representatives have direct knowledge of the programmes.However, they do not have the capacity or the means to respond at the level of regulation.In all of our cases, the potential for active response by local industrial relations actors to the programme implementation, or even to the consequences of this implementation, was revealed as being very low.Other studies have also observed that, 'management uses the "diffuse power of the market" (Mann,1993) to keep employees and industrial relations actors in line' (Dörre p.,(19)(20).
From the point of view of the systems of employment regulation, it is extremely difficult to respond conflictually to a programme that appears so rational and objective.The rules of interaction being laid down by the company central headquarters are to improve your products and processes as you see fit, so that your company makes a profit and contributes to good market results.This sounds very reasonable.The measures of performance seem extremely rational and apply to everyone.Thus every unit and every individual becomes a financial optimiser.The implicit agreement between labour and capital shifts from promises of security and a piece of the pie in return for productivity and labour peace to openness to risk from market fluctuations in return for a piece of a faster growing pie, and possibly greater autonomy, flexibility and participation in the direct work process.
These operating principles of investor capitalism and market-regulated controls weaken the governance potential of existing systems of employment regulation whose negotiating agendas were determined in what regulation theory calls the 'Fordist/Taylorist' production era (Bergmann 2001).Whether or not one subscribes to the theory of production regimes, or agrees that a departure from the Fordist model has occurred, it is clear that the political and regulation forms upon which unions built their negotiation strategies, and which were mostly geared to traditional industries and traditional industrial workers, cannot be transferred to new market based systems.The content of negotiation falls completely outside the range of issues regulated in these systems.The next section looks more closely at the shifts that have taken place in the systems of regulation in the three countries in our study.

Mapping models of employment regulation: from Fordism to 'investor capitalism'
Industrial relations develop in close relationship to the structures in which they are embedded at a national level which shape them as a result of the particular constituents, regulatory supports and historical context in which they were formed.As central instruments in the regulation of work, industrial relations systems are tied to other regulation systems which are, in turn, supported by national labour and laws, education and training systems, and state instruments for labour market policy.The relative immobility of labour and the industrial relations system which represents it has reinforced the national specificity of these systems.However, the specific forms that industrial relations systems take do not only derive from the state regulatory systems in which they are embedded, but also are shaped by the historical form that conflicts between capital and labour have taken and the political ideologies that have inspired them.This becomes especially clear if we look at the history of union development as one of the main pillars of industrial relations systems.In Germany, for instance, unions see themselves as strongly linked to the politics of the German labour movement which was historically social-democratic in orientation -the goal was to achieve a viable democratic state in which labour had co-determination rights.In France, Italy and Spain, by contrast, labour movements were tied to syndicalist, anarchist political movements which led to a splintering of these movements into unions tied to particular political parties.In Great Britain, the shop steward system was based on structures of representation defined along 'craft' or occupational lines of demarcation and tactics of negotiation tailored to the work process.These nationally specific forms of industrial relations determine what the arenas of negotiation between capital and labour will be, the form that conflicts take and the levels and styles of negotiation between collective actors.
In the classic view of industrial relations research the goal of the regulation of work relations is to achieve a balance between the conflicting interests of an autonomous utilisation of capital on the one side, and the protection or improvement of labour's means to reproduce itself on the other.The research carried out from the post-war era to the present shows that a characteristic of industrial relations in Western industrialised economies was the strong link that arose between the orientations of the system of interest representation and the Taylorist/Fordist production model that dominated in the period of economic growth following the second World War.(Bechtle, Lutz,1989) The arenas of conflict and negotiation that emerged in the Taylorist-Fordist production system centre around issues deriving from the work relationship that crystallised in this era, such as wages, working conditions, working time, employment security, etc.This link is particularly strong in German industrial relations which can be clearly seen in the way wage forms are fixed in industry-wide negotiated settlements and in company level negotiated settlements, and also in the concepts underlying the labour process and work organisation that were promoted in this period.(Müller-Jentsch,1991) The wage forms that developed from the 1950s onward were oriented toward an attempt to (objectively) measure and calculate performance, leaving some room for compromise in the form of complaints against standard times (Vorgabezeiten).Because of the central role that the determination of performance levels and the wage setting tied to them has had in German industrial relations, work councils as well as leading union functionaries traditionally concentrated a good deal of their negotiating expertise on wage issues.Conditions of work organisation were thus developed which supported pay based on performance levels -generally in the form of the piece rate wage (Akkordlohn).The control of the system of wage determination by performance levels led to a range of institutionalised forms of regulation such as procedures, work content, routines and norms (REFA 3 ) in which the actors in the industrial relations system played a major role.Union representatives received certification in REFA industrial engineering procedures in order to better exercise their codetermination rights in setting performance standards.
Our initial assumption in examining the relationship between global company strategies and industrial relations was that the response at the local level would be affected by the characteristics of the specific systems of industrial relations that are found in the three countries in our study.We expected an effect on the implementation process as well as on the specific character that each programme would take at the different sites.We set out to investigate these assumptions by asking if the particular architecture of the various systems of industrial relations could predict the responses of industrial relations actors to the programmes as they were implemented at the local level.And, if there was no clear institutional response from traditional organisations of interest representation, were there nevertheless identifiable reactions at the company level which could be classified as 'controls' and could be considered as constituting a form of industrial relations, although not in the sense of national level systems of work regulation?Finally, we wanted to determine whether identifiable alterations to the industrial relations practices at the local level resulted from interaction with the programmes implemented by the global companies, which were meant to be instituted universally at all of their sites.
One of the first steps that we had to undertake was to select a way to compare the architecture of the various systems of industrial relations that we would be examining (Germany, France and the USA) in order to identify the reactions and control potential for the globalisation process.Our goal was to steer away from predetermined conceptions of the outcomes based on institutional contexts, such as the German system of industrial relations, that were familiar to us and to avoid abstract functionalist explanations of the demands of globalisation.
We began with the schema developed by Düll and Bechtle (1988) in which systems of industrial relations are classified in a space defined by the four dimensions of negotiation, norm building, conflict, and control (see Figure 1).Between the areas of conflict and control, one finds the poles between which industrial relations have moved historically, while between negotiation and norm building, the forms of institutionalisation are laid out.These latter two make possible a continuous, 3 REFA, the Association for Working Time Studies and Work Organisation, founded in 1924 as the 'Reichausschuss für Arbeitszeitermittlung' or the 'Government association for the determination of working time" carries out studies on work design, industrial organisation and company development.After the war, it became one of the most important instruments of the unions for categorising jobs using complicated time and ergonomic strudies.albeit limited, balance between conflicting interests, and thus allow the potential explosiveness in the naturally adversarial relationship between labour and capital to be overcome.Norm building represents the concrete form that regulation of the employment relationship takes, for example as rules, procedures, topics and arenas of negotiation.
In the early first attempts at 'mapping' historically formed architectures of industrial relations systems, Germany is typified by a configuration dominated by the link between negotiation and norm building (the normiertes Verhandlungssystem). Open forms of conflict at the company level are rare in such a system and broader forms of conflict are strongly regulated.The 'control' element is anchored in the 'Mitbestimmung' (co-determination) concept of the union.In France, the system of industrial relations originally developed mainly along the poles of conflict negotiation -conflict beyond the individual company level; at the company level the dimension of negotiation was traditionally very weak.Control was reinforced through state agencies (such as the 'inspecteur du travail').By the mid 1990s, a shift in the system of industrial relations was recognisable toward the poles of conflict and norm building, in which state regulation began to play a more important role (Lallement 2000).
The original model did not include the USA.If one tries to apply the classification system to the traditional US industrial relations system, one could characterise the post-war era's configuration of industrial relations as moving mainly between the fields of conflict and negotiation.Norm building played a role to the extent that the entire system was based on the collective bargaining contract, which often covered the same issues of wages, seniority rules, health and safety, pension funds, and job delineation.The little US legislation that exists supports the regulation process by protecting collective bargaining contracts, prohibiting union busting, penalising discrimination, and protecting the general right to unionise, etc.
The great disadvantage of the classification system presented here is that different levels of interest representation are all missing from it, making it impossible to see the balance between national and local dimensions, the level of institutionalisation and the level at which bargaining takes place.The advantage is that, although it sets the given conditions for action by collective actors in a particular context, it also theoretically allows for some measure of shift if the actors involved in the industrial relations system adopt new strategies.The classification demonstrates links or potential links between dimensions as they have crystallised in specific historical configurations of industrial relations systems in different countries, but existing links can break up, thereby opening up new possibilities for change.Thus it allows a dynamic evaluation of industrial relations response and provides a means to obtain a visual representation of how the response appears with regard to the changes in company strategy and modes of control.
In fact, a reading of the literature on the development of industrial relations in these three countries in the last decade has indeed revealed that a shift has occurred when mapping the system architecture of the German, French and US industrial relations systems using the classification scheme from Düll/Bechtle and this is supported by the empirical research we conducted (see Figure 2).As can be seen, the main dimensions of negotiation and norm building characterising the German system have not changed.However, two aspects have changed: one is that the internal company arena has gained in importance, which has been graphically represented with a larger area covered by negotiation.The other is that the area regulated by the 'normierten Verhandlungssystem' has been narrowed to represent the fact that a number of issues that affect the employment relationship today are simply not considered or covered by the existing processes of negotiation and norm building.In France, the graphic representation of the system of work regulation has shifted considerably from the original portrayal.In a move away from the poles of conflict and negotiation, the French system is now more and more characterised by the state´s regulatory influence in the form of broad-based laws affecting the work relationship (for instance on working time and training).The company-based areas of negotiation have also increased in importance, often in conjunction with implementation of state laws.Conflict is less characteristic in the private sector and the unions seem to have a diminishing role and dwindling awareness of company issues and problems.In the USA, the graphic representation also shows a considerable shift away fromthe post war system in the current situation.Previously moving along the dimensions of conflict and negotiation, with control in the form of rigidly defined job classifications and wage contracts, the USA was characterised by narrowness in the areas in which formal systems of interest representation could negotiate.Increasing marginalisation has resulted in a situation in which formal systems of interest representation exist only in isolated industrial and public sector pockets.Negotiation on formal issues still occurs in these pockets, but conflict plays less and less of a role.(Tilly and Tilly,1998) The mapping models reveal that a shift in the configuration of industrial relations has taken place in all three countries, although to different degrees, and also that there are still differences between systems of industrial relations.Although this is the case, the mapping models are not good predictive tools for the governance of systems of employment regulation under conditions of 'investor capitalism' because the content of negotiation falls outside the range of issues covered in the rules of regulation.Figure 3 shows the differences in the response potential of the formal systems of industrial relations in the three countries.
In Germany, the strength of the works councils at the company level provides a potential for an organised response, although in our cases they were mainly bypassed in the implementation process of the new generation of programmes geared toward financial results.The works councils, with their high levels of competence regarding company policy, know that new company strategies and financial priorities disempower them to act with the set of issues at their disposal to negotiate, but do not have the capacity or expertise to demand a participatory role.
In France, the state takes an active role in framing labour policy, but this is in relation to 'old,' traditional topics, such as working time, which are increasingly legislated for.The implementation level is the company, where the comité d´entreprise is becoming more involved in negotiations.Nevertheless, the role of the comité d´entreprise is a reactive one.The unions operate beyond the company level, attempting to exercise influence at the political level, but not on issues that are of immediate relevance at the company level because their relationship to internal company practices has become very diffuse.(Lallement 2001) In the USA, unions have a marginal presence in industry and their issues for negotiation are very traditional.Some unions in the public sector and in traditionally strong unionised sectors had been involved in the participatory movements of the 1980s (Wheeler,1990).However, in our cases the unions played no role in the implementation and negotiation of programmes aimed at optimising financial performance.In fact, it seems that the most visible level of protest or protection is as citizens and not as employees: issues in which employees lose their jobs as a result of financial practices or regime shopping become voting issues.In this way, protest departs from the realm of the workplace.
In sum, although the systems of employment regulation vary, there was little possibility for local actors to carry out a governance function in response to the pressure imposed on them by the central headquarters in the form of restructuring programmes oriented to financial performance.Informal types of response could be defined as taking control of the restructuring process at a local level to as great an extent as possible, which did indeed occur.However, an organised informal response to counter the pressure from the centre was not recognisable.This does not mean that the sites do not have autonomy, nor does it mean that they do not enact participatory measures which give individual employees a voice in the restructuring process.However, the level of autonomy, the ability to respond actively, not to mention the actual survival of the local unit, depends on an economic situation over which they do not have control.
The current challenge for formal systems of employment regulation, if they do not want to see their co-determination reduced only to the development of social plans for mass dismissals or the protection of seniority, is to reorient their negotiating strategy to deal with issues that do not presume a stable employment relationship in a productiondominated economy.A shift in employment relations has taken place toward a market-mediated work relationship which requires a corresponding shift away from employment regulation based on the Fordist production model.Of course, there are union initiatives at the EU level, in Germany, and the USA as well as workforce initiatives at the company level in new knowledge-intensive sectors which are trying to adapt new strategies to meet company demands for flexibility, and minimise precariousness and market-based controls (Waddington and Hoffman, 2000;Heidling et al, 2004;Hurd et al, 2003;Frege and Kelly, 2003)).At the same time, given the state of the stock market, the present period also has to be seen as a time of flux for the multinationals in terms of defining their strategies for the future.Former strategies of extreme rationality led by financial considerations do not seem to hold up to the arbitrariness of today's financial markets.Although the old social contract based on principles deriving from the Fordist era will most likely continue to disappear, it is still unclear what will replace it.

Figure 1 Figure 2 Figure 3
Figure 1 Classical model of industrial relations in Taylorist-Fordist production systems in Germany, France and the USA (consolidated negotiation)