Multidivisional vs metanational governance of the multinational enterprise

Abstract

The effective and efficient management of diversified business firms that supply multiple products and operate in multiple, dynamic markets, especially large multinational enterprises (MNEs), builds upon a number of specific governance principles. These governance principles allow the alignment of environmental characteristics, strategy and organization. Given the rising need to “learn from the world”, Doz et al., in their influential Harvard Business School Press book entitled From Global to Metanational, have proposed a new set of governance principles described under the “metanational” umbrella concept. This paper revisits the metanational, using a comparative institutional perspective; here we contrast multidivisional and metanational governance principles. A comparative institutional analysis suggests that the metanational's application potential in terms of actually improving the effectiveness and efficiency of MNE governance may be subject to more qualification than suggested by Doz et al. Senior MNE management must therefore reflect carefully before substituting metanational governance principles for the more conventional, multidivisional ones with established contributions to managerial effectiveness and efficiency.

APPENDIX

Three Foundations of TCI Teory Applied to the MNE

Bounded rationality reflects the scarcity of the human mind: “…human behavior is intendedly rational, but only limitedly so” (Simon, 1957: xxiv). Given a complex and uncertain environment, economic actors are unable to know, process and act on all current and future information. Hence they are unable to produce comprehensive contingent claims contracts (Arrow, 1974). Economizing on bounded rationality is widely accepted as critical to both value capture and value creation in the MNE, especially when exploiting the MNE's firm-specific advantages, that is, its proprietary knowledge, which typically exhibits public goods characteristics (Rugman, 1981; Rugman & Verbeke, 2005b). Economizing on bounded rationality is also important when contemplating alternative governance mechanisms for knowledge management; here, the ease of linking the knowledge exploration process with subsequent production and marketing activities, that is, the challenge of effective and efficient innovation in its entirety, is a key consideration.

Bounded reliability reflects the scarcity of making good on open-ended promises: good-faith contractual representations do not always result in the realization of the promised outcomes or performance milestones due to a variety of factors. Safeguards or enforcement mechanisms (Argyres & Liebeskind, 1999) to heighten detection of – and provide punishment for – reneging are symptomatic.

Importantly, the concept of opportunism, which is a key element in Williamsonian TCE, and has been criticized as an inappropriate foundation of management theory (Ghoshal & Moran, 1996), is only one expression of bounded reliability, as assumed implicitly in much of modern TCI theory and visible in numerous modern cases on MNE management. Opportunism implies ex ante false commitments and/or ex post malevolent reneging on commitments. In a Williamsonian world, safeguards need to be established to reduce the probability of opportunistic behavior, and to punish it when it occurs. This is important both when dealing with outside contracting parties in a situation of small numbers bargaining and asset specificity, so as to mitigate the risks of cheating, and to reduce shirking inside the MNE. However, precisely inside the MNE, with managers often committed to spend a substantial portion of their career within a single company (or at least within a single industry), with their professional mobility dependent on their reputation, and with their professional pride to do what is in their mind the best job possible, more common expressions of bounded reliability may prevail, especially benevolent preference reversal over time.

Benevolent preference reversal may result from both recurrent, unintentional over-commitment and recurrent “local” prioritization, with the word “local” referring to the main activities, dispersed in geographic space, for which specific individuals and groups are made responsible. These problems cannot be simply reduced to opportunism issues. Commitments requested by CHQ and intentions expressed to CHQ to achieve a particular outcome/performance level do not always result in the realization of the promised outcome/performance milestones due to a variety of factors, including the misalignment of incentives. In the absence of opportunism, the bounded reliability problem is not that reality turns out differently from prior expectations (which is a mere bounded rationality issue), but that individuals recurrently experience (benevolent) preference reversal over time, a well-known phenomenon in psychology (Steel & König, 2006). For example, at the level of a division, the divisional manager may typically promise the execution of specific investment projects in accordance with CHQ preferences, and commit to specific performance requirements. However, a combination of factors such as a substantial distance in time from any sanction in case of non-achievement, a substantial distance in space from the headquarters’ monitoring apparatus, and the relative proximity and intrinsic satisfaction derived from focusing on autonomous, locally driven investment opportunities with immediate local rewards such as an improvement of relationships with local stakeholders, etc. drives preference reversal. This occurs especially if ultimate performance cannot be appraised objectively and in full (a bounded rationality problem), as is the case when substantial reciprocal interdependencies exist among various subunits in the organization so that an individual's (or unit's) performance cannot be measured accurately.

Apart from describing behavioral reality more correctly and fully than the opportunism concept, there are two additional advantages of adopting the bounded reliability concept: first, opportunism is a concept reflecting abundance (in terms of propensity to cheat and shirk), in contrast to bounded rationality. Bounded reliability, as is the case with bounded rationality, reflects scarcity, in this case scarcity of making good on open-ended promises. Second, for decades there has been a debate between economists and institutional theory scholars on the drivers of change: in economics the main driver is increased efficiency; in institutional theory it is increased legitimacy. A number of empirical articles in business have attempted to “test” these two alternative explanations for changes in managerial settings. By building upon the bounded reliability concept, the need for alternative explanations may in many cases be reduced. For example, historically the relatively slow diffusion of multidivisional governance, in spite of its proven efficiency impacts (not taking into account the causal ambiguity problems associated with such a managerial innovation and the related risk perceptions, nor the possibility of technology-driven rigidities), has been caused in part by the presence of powerful functional groups, in favor of the status quo inside large diversified firms. From an institutional theory perspective, powerful stakeholders considered such a change in governance as illegitimate. From a TCI perspective, governance change would have increased bounded reliability problems, as the stakeholders resisting change would have become “unreliable” (reneging on their employment contract expectations by simply exiting the company or by sabotaging the new system's implementation, etc.), thereby reducing the effectiveness and efficiency of the firm's functioning. However, when a crisis situation unfolded, and a search process was undertaken by the stakeholders concerned to find a solution, the new governance form's legitimacy increased and the danger of reduced reliability of groups previously opposed to the new form faded away. In general terms: expected increases in legitimacy of a particular course of action, such as the choice of a governance form or pattern of behavior, can often reasonably be interpreted as the equivalent of reduced bounded reliability challenges originating from stakeholders able to affect the firm's functioning.

Favorable Organizational Context for Effective and Efficient Knowledge Management in its Entirety

When penetrating foreign markets, firms must first deploy existing, non-location-bound (or internationally transferable) knowledge in those foreign locations (Buckley & Casson, 1976; Rugman, 1981). Second, in order to be successful there, they must also engage in investments, permitting the combination of non-location-bound knowledge from the home country with new, location-bound knowledge in specific host countries or complementary non-location-bound knowledge required to operate profitably abroad. Since both the initial and the recombined knowledge bundles typically have public goods characteristics, the MNE faces a probability of serious loss if these bundles are unintentionally absorbed by external actors, or improperly combined inside the MNE, in host markets. Third, a more recent phenomenon is the need for selectivity (and resulting corporate coherence) in the face of easy access to multiple technologies in multiple foreign locations, meant to create reverse knowledge transfers. Here, the MNE acts as a knowledge network, with several home and host country operations involved as actors in knowledge creation and diffusion activities (Rugman & Verbeke, 2001).

Cantwell's (1995) incisive analysis of MNE technological diversification strategies has demonstrated that three stages of international knowledge recombination must be distinguished. The first stage, covering the post World War I era up to 1970, had MNE foreign subsidiaries engaged primarily in the adaptation of home country knowledge to host country requirements. Technological diversification (albeit largely incremental) thus went hand in hand with geographic diversification. The second stage, which lasted until the mid-1980s, was an era of increased technological interrelatedness, also at the international level. MNE knowledge accumulation from international sources increased, driven partly by a reduction in transport and communication costs, and in many cases by the changing nature of the technological knowledge itself, which permitted its easier diffusion across sectors and national borders. In the third stage, which is still ongoing, new capability creation requires the use of internal networks, with a strong need for selectivity in the choice of interconnected locations (e.g., “higher order” regions), that contribute to new knowledge combinations from various sources: see also Cantwell and Iammarino (2000) and Cantwell and Piscitello (1999). The modern MNE that strategically integrates complementary knowledge sources from different geographic locations (Cantwell, 1989) and harnesses multiple technologies (Cantwell, Dunning, & Janna, 2004a; Cantwell, Gambardella, & Granstrand, 2004b) is thus one that requires, even more than before, a focus on dynamic selectivity in governance. Even in the context of “asset seeking” international diversification (Cantwell et al., 2004a, 2004b), whereby the strong need to manage tacit knowledge and the high potential benefits of learning-by-doing may act as stimuli for internalization, it is important to note that bounded rationality and bounded reliability challenges, though mitigated through internalization as compared with the use of market mechanisms (such as technology licensing), are not eliminated completely (Coase, 1937). Senior managers must monitor and mitigate the effects of bounded rationality and bounded reliability inside the firm, sometimes making extensive use of price-like mechanisms (Hennart, 1991; Rugman & Verbeke, 2003). Importantly, most of the innovation activity described above in large MNEs occurs within global product divisions or regional divisions: see, for example, Rugman (2005).