Managing the Multinationals: An International Study of Control Mechanisms by Anne-Wil KatheHarzing, Cheltenham: Edward Elgar, 1999, 425 pp., ISBN 1-84064-052-9.

There is now a well-established literature on control mechanisms in multinational companies (MNCs). This book adds to this literature by attempting to prove empirically the role of international management transfers in control. It is drawn from Harzing’s doctoral research, based on a survey of MNCs headquartered in nine countries and with subsidiaries in twenty-two countries. Whilst not being a particularly “student friendly” book, and certainly not an easy read, it makes a substantive and useful contribution to the field.

The book addresses a number of key questions: (1) what are the characteristics of headquarters (HQs) and subsidiaries that explain the composition of control mechanisms? (2) what is the role of international transfers in controlling subsidiaries? (3) are alternative control methods used? (4) can different MNC configurations be distinguished? and (5) do MNCs that conform to a particular configuration outperform others?

After a brief introduction, Chapter 1 presents an exhaustive and thorough review of literature from three often disparate areas, which are set up as the “theoretical building blocks” of the study: (1) organization control mechanisms (drawn from organization studies); (2) MNC strategy and structure (drawn from the international management literature); and (3) MNC international transfers (drawn from the less developed but rapidly expanding literature on expatriate management). Chapter 2 then seeks to combine these three “building blocks,” examining: (1) the application of control mechanisms in MNCs; (2) the role of international transfers as a control mechanism; and (3) various ideal-type configurations of headquarters and subsidiary characteristics and control mechanisms. This involves some carefully built arguments to establish the key relationships and indicates an impressive grasp of a very wide range of literature.

Chapter 3, which deals with methods, tackles the issue of how these areas might best be operationalized. It begins with a sophisticated discussion of the links between contingency theory, strategic choice, national and organizational cultures, and MNC control mechanisms, and posits neo-contingency theory as the foundation of the study; that is, the argument that organizational actors have significant degrees of strategic choice, and that although organizations are influenced by their contextual contingencies, the relationship is not deterministic and outcomes are unpredictable.

So far all of this is reasonable enough, but where the methodology becomes more shaky is in the operationalization of particular concepts and their subsequent measurement through statistical analysis. Some rather sweeping decisions are taken here which underpin the whole project. First it is stated that “strategic choice is very difficult to operationalise and measure … especially when using a survey method of data collection,” and so the decision is made to “not pay specific attention to the concept of choice in the remainder of this thesis,” but merely to “keep in mind… that any relationship between culture and outcomes and contingencies and outcomes is mediated by strategic choice” (p. 166). This indicates the limitations of the methods used as against more qualitative and in-depth case study work, a weakness the author is happy to acknowledge later in the book. Second, it is confidently asserted that national cultures can be measured, although in doing so the author “had to make a number of simplifications,” and hence “for the sake of simplicity we equate nation with culture… we assume nations culturally homogenous” (p. 166). It is also assumed that what the author calls “vaguer” concepts such as strategy and control, although “perceived differently in different countries,” will nevertheless be “broadly identical in all countries” (p. 172). At least the author is prepared to accept that, in her own words, the assumption of cultures as homogenous “might stretch reality a bit” (p. 171).

The key aim here is to explain why there are differences in the application of control mechanisms between firms. It is argued that “multivariate statistical analysis should be able to sort out … [the] different influences, although much depends on the sample sizes” (p. 167). Indeed it does, and here they are not very large. The overall response rate was just under 20%. Although this may well “compare favourably with most of the other studies” (p. 211), and although responses are analysed in detail by industry, country cluster, size and age of headquarters, function of subsidiaries, etc., it remains the case that the actual number of respondents in many cases is small. Given that broad conclusions are drawn in the book concerning different HQ/subsidiary relationships and the influence of a variety of national systems on control mechanisms, it needs to be remembered that the whole project relies essentially on individual survey respondents from a fairly small number of organizations (there were, for example, a total of 25 respondents respectively from UK and Dutch companies, 16 from German, 14 from French, 14 from Swiss, 13 from American and 11 from Swedish).

The author does acknowledge many of these shortcomings in the final few pages of the book, describing as a limitation the fact that “we used a key-informant approach, [so] our results are based on the opinions of a single respondent in each organisation” (p. 367), and furthermore that because the study “used perceptual measures to operationalise some of the constructs, the answers to our questions might contain an element of perception… [which] might reduce the validity of our findings” (p. 367).

I have taken some time to get to these findings, but then so does the book! Even the two “results chapters” (chapters 4 and 5) continue the detailed discussion of methodological issues—e.g., the use of factor analysis to reduce questionnaire items to constructs, the reliability and cross-cultural equivalence of the scales used, etc.—and it is only in the relatively short (26-page) final chapter of the book (Chapter 6) that a clear statement and discussion of the major findings is presented.

Here, one conclusion stands out above all others, and that is the explanatory power of the country-of-origin concerning the type of control mechanisms used by MNCs. Harzing concludes that “if we look at the main subject of this thesis: control portfolios, we find strong differences between MNCs headquartered in different countries in the application of the various control mechanisms…. As our study shows, the country-of-origin effect has anything but lost its significance” (pp. 356–7). In particular, the survey indicates that direct personal control mechanisms (“personal centralised control”) are used to a larger extent in respect of subsidiaries of British and German MNCs, while their application is rather low for subsidiaries of Swiss and Swedish MNCs. Indirect personal control mechanisms (control by “socialisation and networks”) are most prominent in respect of subsidiaries of Swiss and Swedish MNCs and least prominent for subsidiaries of both French and Japanese MNCs. Both direct and indirect impersonal control mechanisms (“bureaucratic formalised control” and “output control”) are used to a high extent in respect of subsidiaries of Anglo-Saxon and German MNCs and to a relatively low extent for subsidiaries of Japanese and Finnish MNCs.

The specific role and/or function of subsidiaries is also found to explain a large amount of the variance in the type and level of control used in respect of them. Thus so-called “autonomous” subsidiaries experience a significantly lower level of control than subsidiaries with either “active” or “receptive” roles. Sales subsidiaries experience lower total levels of control, mainly because the level of personal centralized control is much less pronounced for these subsidiaries, whilst R&D subsidiaries experience a significantly lower level of bureaucratic formalized control.

As regards the role of expatriates, these are shown to form both a direct and an indirect means of control, i.e., directly supervising decisions taken at headquarters whilst also influencing the level of shared values between headquarters and subsidiary managers. The total level of control exercised by the HQ is particularly high in “global” companies (as opposed to “multidomestic” or “transnational” companies), mainly caused by a high level of the two direct control mechanisms—personal centralized control and bureaucratic formalized control. Expatriate presence in subsidiaries of this type of company is high, and centres on exercising direct control through the supervision of decisions taken at headquarters. The global configuration is found to be most typical of German and Japanese MNCs. Expatriates are thus “powerful links between the various units of an MNC … [and] are crucial in the realisation of the organisational effect” (p. 363).

As for the control/performance link, here too “the highest explanatory power could be attributed to the country-of-origin of headquarters and the industry in which the MNC operates” (p. 364). However, the author is rightly cautious here, pointing out that “explaining performance differences between firms is a hazardous enterprise” (p. 159), and making the important observation that “companies in different countries… do not necessarily attach equal importance to [the same]… performance indicators,” and as such “it would seem much more useful to investigate which performance indicators are considered important in various countries and to relate this to the country’s societal/legal/cultural background” (p. 364).

This statement indicates again how the author remains cognizant of the limitations of the study throughout, and it is refreshing to see an openness towards other methods as a means of building upon the findings. Most notably it is acknowledged that “since our study is cross-sectional rather than longitudinal, statistical correlations cannot unambiguously be interpreted as causal relationships…. A true test of the causality of the relationships tested in this book can only be offered by qualitative and/ or longitudinal research” (p. 368). Relatedly, “because of our emphasis on generalisability and the method of data collection—questionnaires with closed-ended questions—our results mainly focus on outcomes. The actual process underlying much of the relationships has remained a black box” (p. 368).

Other recent work has attempted to uncover and map these processes, and what this more qualitative research is able to show is the complex interplay between formal and informal types of control. In Ferner’s words, “different mechanisms of control, and their use in various combinations, reflect the realities of political processes within the company and the struggles for control and autonomy between different interest groups that underlie them”; and, crucially, “MNCs rely on complex combinations of formal bureaucratic control and other control mechanisms that not only coexist but also are mutually dependent” (“The underpinnings of bureaucratic control systems: HRM in European multinationals,” Journal of Management Studies, vol. 37 (2000), 521, 537, emphasis added). It is this mutual dependency which survey work of the kind reported in this book is unable to illuminate, and hence there is a tendency to distinguish too clearly between “bureaucratic/formal” control systems on the one hand and “personal/social” mechanisms of control on the other, and moreover to assume that managers face a clear strategic choice between the use of one or the other.

I have said that this is not an easy book to read. Partly this is due to the straightforward reproduction of the doctoral thesis structure. Indeed, not only does it read like a PhD thesis, but frequent reference is made in the text to “this thesis.” The style is also a little dry and dense in places, and whilst I am always imploring post-graduate students to ensure they include adequate “signposts” in their dissertations to indicate the flow of the sections, this book takes that principle to an unnecessary extreme. Far too many summary and conclusion sections are included, both at the ends of chapters and within sub-sections of chapters. A shorter book—which flagged key findings far earlier—would have made for a more accessible text and a more coherent sense of the central arguments.

Aside from these stylistic and methodological reservations, this is unquestionably a thorough and authoritative assessment of the field (the bibliography numbers something in the region of 700 references), and the study makes a substantial contribution on a number of fronts, not the least of which—set against the currently fashionable “globalization thesis”—is to highlight the enduring significance of the country-of-origin in multinational company activity.