Abstract
Firms can be conceived as political entities in which various stakeholders struggle against each other to reach goals. From this standpoint, the organizational structure of a firm, together with the power relations and strategies deployed by stakeholders, constitute a governance regime. It is understood that the economic performance of a firm is impacted by the type of governance regime which is at play among its different stakeholders, but little is known on the effects of governance regimes on organizational creative performance. Adopting a center/periphery perspective, we look at business groups, a well-known type of business firm in creative settings. A salient question is thus to know how firms affiliated to various types of business groups perform vis-a-vis unaffiliated firms, and what are the underlying political processes. We define three different regimes of governance—empires, federated arrangements, and kingdoms—each characterized by specific regulation and integration features. Regulation refers to the type of rule (from the center to the periphery), to the autonomy of peripheral entities, and to the regime’s level of flexibility; integration captures the extent to which projects of the center and the periphery are consonant as well as the local elites’ type of mobility. We devise empirical tests for a realm of the creative economy by employing organizational and network analysis of 293 fashion houses over seven seasons. Overall we find that, although both empires and federated arrangements perform better than kingdoms, federated arrangements perform best in creative contexts because they are hybrids of empires and kingdoms.
Introduction
A firm can be conceived as a “political coalition” (March, 1962) in which various stakeholders (or actors) seek to gain power in order to reach their goals (Pfeffer, 1981). However, the goals of the firm itself “are not given; they are bargained” (March, 1962, p. 672) and the way they are implemented depends on the political processes through which organizational conflicts are solved. In turn, these processes will be determined by the structure of the firm under consideration. Together with the distribution of power among actors, the organizational structure forms a regime of governance. As Clegg, Courpasson, and Phillips (2006, p. 331) explain, “When we connect power to the social fabric of organizational forms we think of organizations as regimes.” Regimes of governance, because they guide and integrate power struggles among actors, have been explored as far as their impact on economic performance is concerned, notably via their organizational structure angle (Clegg et al., 2006, Chapter 2). However, little has been said about the effects of such regimes on organizational creativity. Thus, this article aims at understanding the impact of different regimes of governance on organizational creative performance.
What types of regimes of governance can be considered when it comes to organizational creativity? We focus on the way the center of the organization is related to its periphery. The center/periphery question, that was notably developed by Shils (1982), is central in the social sciences. In organizational settings, as Chandler (1991) explains, it is to be found in the relationship between the corporate center (“HQ” or headquarters) and the periphery of its affiliated businesses which it coordinates. A broad distinction is usually made between creative organizations that are affiliated with “business groups” and those which are unaffiliated, that is to say, independent. The “business group,” defined as a set of “legally separate firms bound together in persistent formal and/or informal ways” (Granovetter, 2005, p. 429), is a significant type of business organization in contemporary capitalism at large (Granovetter, 2005; Guillen, 2000; Khanna & Yafeh, 2005).
This type of organizational context constitutes an excellent site for studying the impact of governance regimes on creativity because they are highly relevant in the creative industries—the industries in which creative and aesthetic values are central to their production process (e.g., Caves, 2000)—where the impact of business group affiliation on organization-level innovation and creative performance remains a debated question (Peterson & Anand, 2004; Thornton, 2002). For example, Lopes (1992) explains that in the American music industry multilabel music companies protect music producers against the vagaries of an uncertain market, allowing them to be more creative and mitigate the risks of failing. However, in the French fashion industry, Crane (1997) suggests that, when costs of entry are low, smaller, newer, and unaffiliated firms are more innovative and creative than their bigger, older, and affiliated counterparts.
This leads to an additional question about the type of affiliations among the principal components of a business group—its headquarters (or center), and its various affiliated firms—which needs to be considered together with the affiliated vs. non-affiliated distinction. Adopting such an approach to regimes of governance is a way to expand their understanding through a more granular and subtle lens, for example by considering “hybrid” regimes of governance, as it has been suggested in research at the crossroads of organizational and political theories (Clegg et al., 2006; Courpasson & Dany, 2003; Courpasson & Thoenig, 2008; Dahl, 1956). More specifically, we look at how regimes that mix and subsume the features of traditional regimes based on business groups and unaffiliated firms emerged to tackle evolving socioeconomic conditions. Thus, our research question can now be refined: what is the impact of different types of governance regimes—related to the various ways of being affiliated with a business group vs. not being affiliated, including hybrid regimes—on the creative performance of a firm?
The focus on emerging hybrid regimes can also help us move the center/periphery question beyond the centralization/decentralization debate by suggesting other and novel arrangements of power distribution between the center and the periphery. More specifically, adopting a political perspective, we define three different regimes of governance: empire, kingdom, and their hybrid form, federated arrangement. Each is characterized by specific regulation and integration features which we detail. Regulation refers to the type of rule (from the center to the periphery), to the autonomy of peripheral entities, and to the regime’s level of flexibility; integration captures the extent to which projects of the center and the periphery are consonant as well as the local elites’ type of mobility. In this sense we also contribute to a renewed body of research in regime theory (Wijen & Ansari, 2007) and about the role of elites in center/periphery settings (Zald & Lounsbury, 2010), in particular as far as the dynamics of hybrid forms is concerned (Reed, 2012). Finally, we pay special attention to the development of entrepreneurial endeavors in relation to the center/periphery question in the three governance regimes, since it is a process of high relevance in creative contexts.
The fashion industry provides a case study to address the aforementioned questions. This is a relevant empirical setting in which to explore the features and effects of regimes of governance because it is characterized by significant organizational and political structural diversity (Djelic & Ainamo, 1999). Fashion is also an industry in which creativity plays a major role, and we believe that it is an excellent context to study the impact of governance regimes on creative performance.
Theory
Regimes of governance in organizations
The concept of a “regime of governance”—sometimes referred to as a “mode of governance” (Batterbury & Fernando, 2006)—stems from a long tradition in philosophy and political science (see Grawitz & Leca, 1985, for a thorough review) and is used in a variety of contexts; recent examples include the influence of policy networks and political institutions on the public life of various French and British cities (John & Cole, 2000), the state’s management of the poor in India (Corbridge, Williams, Srivastava & Veron, 2005), and the study of organizational dynamics in post-bureaucratic contexts (Courpasson & Dany, 2003).
In organizations, regimes of governance emerge at the crossroads of power relations and organizational structure, with the latter being “the major ingredient of a system of governance” (Clegg et al., 2006, p. 330). Various regimes of governance exist, and the type of relations that materialize in a given regime engenders particular organizational arrangements as well as methods of coordination that occur between a center and its periphery. Understanding how a center is related to a periphery is an essential question, not only in sociology and social sciences in general (see for example Shils 1982), but also in organizational analysis and in strategy, where the relation between the corporate level and the affiliated components in multibusiness firms and business groups is key (see Porter, 1985), as well as the distinction between business group-affiliated and unaffiliated firms.
Three regimes: understanding the center/periphery question through regulation and integration
We distinguish between three different regimes of governance taken from political sociology where center/periphery relations have been the focus of research geared to understand the rise and decline of political systems and various levels of political performance and longevity: the empire, the federated arrangement, and the kingdom. With these three types of regime, we believe we introduce the governance regimes that are most relevant to our research question focused on business group affiliation and creative performance. We discuss in our concluding remarks the existence of other regimes under conditions that are not covered in our paper.
The first two types—empire and federated arrangement—are found in business groups but not the third (kingdom). However, all the regimes must address similar center/periphery coordination issues. Because they emulate the structures of the aforementioned types of political regime, we refer to business group-based regimes of governance which must deal with legally independent firms as empires and federated arrangements, and to unaffiliated organizations which do not depend on business groups’ headquarters for some of their functions as kingdoms. All are sites of numerous power relations—from the interweaving of relations among managers to the power struggles across departments for funding—and these relations are inscribed in whichever regime of governance drives the relationship between center and periphery.
Two underlying dimensions organize our typology of governance regimes: regulation and integration. These two dimensions are sociological properties that characterize relations among social entities in general (Durkheim, [1897] 1952). They have also been used together to analyze the structure of “complex organized systems” whether they are an “organization or an interorganizational networks” (Crozier & Thoenig, 1976, p. 561). Here, we use these two concepts in order to specify the power relations between central and peripheral entities. First, regulation—“the basic mechanism in organizations” (Crozier & Thoenig 1976, p. 561)—refers to three interrelated components. The first is the type of rule from the center to the periphery, direct or indirect, a central factor when characterizing governance and organizational power relations. As Clegg et al. (2006, p. 330) explain, a governance regime can be conceived “as a space and occasion in which concrete modes of rule will be in struggle with each other.” To this, we add two associated elements: the degree of peripheral autonomy, and the level of flexibility of the organizational structure. Second, by integration we refer to the ways through which the center and the periphery constitute a whole (Crozier & Thoenig, 1976; Holzner, 1967), and more specifically we focus on the extent to which peripheral projects reflect the center’s mission, as well as on the features of peripheral elites’ mobility, related to what Clegg et al. (2006, p. 330) term “patterns of elite production.”
Business group affiliated firms and unaffiliated firms as context
Business groups constitute an excellent context to study the relationship between center and periphery from a governance and regime theory point of view because they are characterized by a key distinction between corporate center and peripheral affiliated firms (Chandler, 1991). A second key distinction is between affiliated and unaffiliated firms. As an illustration of this difference, consider the movie industry where “big studios” such as Sony are business groups to which “distribution divisions” such as Columbia or Tristar are affiliated (Cattani & Ferriani, 2008, p. 829). Alternatively, Lionsgate or the Weinstein Company are independent, unaffiliated studios. Similar distinctions can be found in the music industry between multilabel organizations and independent labels (Lopes, 1992), or in fashion between houses that are affiliated to big conglomerates and those which are not (Crane, 1997).
Research on business groups can be divided into two streams. First, researchers have looked at the varying characteristics of business groups, and how they are to be distinguished from other organizational types (Chung, 2006; Smångs, 2006). For example, Granovetter (2005, pp. 433–4) lists six dimensions along which business groups vary, including forms of cohesiveness (sources of solidarity and the extent of “moral economy”), structural features (ownership and authority), and types of relationships with external actors (financial institutions and the state). Taking an international perspective, Khanna and Yafeh (2005) classify business groups’ features according to their risk-sharing profile; they find that Japanese, Korean, and Thai business groups provide their affiliated firms with insurance mechanisms, although this is not the case in other emerging markets. Focusing on Singapore and Korea, Tsui-Auch and Lee (2003) distinguish between family-controlled business groups and business groups employing external managers, and analyze how these two types of business groups fared in the context of the Asian currency and financial crisis of the late 1990s.
A second stream of research focuses on assessing the performance of business groups. Researchers adopting this perspective either compared the performance of firms affiliated with business groups with the performance of independent or unaffiliated firms, or they analyzed the performance of different types of business groups. According to Granovetter (2005, p. 438), business group affiliated firms generally, but not invariably, perform better than unaffiliated firms. Social and economic contexts also play an important role in the formation and impact of business groups. For example, Chang, Chung, and Mahmood (2006) find that the impact of business group affiliation on innovation depends on the institutional context in which the affiliated firms operate. In Korea, where institutional infrastructures for innovation are more developed than in Taiwan, business groups foster more innovativeness than in Taiwan.
A central insight of these two interrelated research streams is that the impact of business groups on performance depends on their organizational and institutional characteristics. Vissa, Greve, and Chen (2009) emphasize how business groups in developed economies differ from those in emerging markets; they show that affiliated firms display behavioral patterns that help them tackle market uncertainties more effectively than do independent firms. This characteristic is especially useful in the context of developing economies, where markets are typically not supported by efficient institutions (Guillen, 2000; Keister, 1998). We further explore these insights by looking at how different types of business group affiliation, or lack thereof, are related to specific types of governance regimes that impact the creativity of organizations in the fashion industry, notably as far as the center/periphery question is concerned.
Applying regimes of governance to creativity in the fashion business
The fashion industry, a creative industry akin to the music and film industries, is replete with political metaphors such as “empire” or “imperial” which are routinely used to designate specific organizations (Hass, 2007) or to characterize the field itself (Lipovetsky, [1987] 1994). Similarly, the power of fashion designers is an object of popular fascination and they are often referred to as “emperors” or “kings.” Examples include the Paris-based German couturier Karl Lagerfeld, who is nicknamed “Kaiser Karl” because of his influence on the field (Orth, 1992); the Italian designer Valentino, referred to as “the last emperor” of fashion in a 2008 documentary film by Matt Tyrnauer; Aldo Gucci, the son of the founder of the Italian house Gucci who was known as L’Imperatore (the Emperor) (Petriglieri and Stein, 2012); and the early 20th-century French designer Paul Poiret, who called himself the “king of fashion.”
Beyond these illustrative metaphors that capture attention, however, the relevant analytic point is the degree to which the political structures of empires, federated arrangements, and kingdoms are replicated in the center/periphery relations that we observe in a specific and relevant instance of a creative economy, the fashion industry. Fashion houses are the main competitors in the fashion industry—they are the brand names customers recognize, and they attract all the attention of the media (Breward, 2003). A fashion house is an organization in which a creative team (usually led by a creative director) design collections twice a year. The collections are generally displayed at fashion shows to the press, buyers, and selected clients.
It is important to emphasize at this stage that a fashion house is either affiliated to a business group—for example, Marc Jacobs is affiliated to conglomerate LVMH and Yves Saint Laurent to conglomerate PPR—or unaffiliated, as it is the case with Chanel. In unaffiliated houses, all the functions that would be taken care of by the business group are internalized and carried out by the fashion house’s management team. In the case of empires, business groups do not compete as such in the market, but through their affiliated brands, or fashion houses. In the case of federated arrangements, the corporate center is a house that competes in the market alongside its affiliated houses. Kingdoms are houses and as such compete directly. Thus, performance comparisons can be made at the level of competing units, so that organizations affiliated with either of the two types of business group-based regimes of governance are compared directly to the unaffiliated organizations which are characterized by the governance regime we call kingdom.
We adopt the aforementioned metaphors but incorporate them as research-based analytic categories to contend with in our work. Table 1 summarizes the characteristics of the three governance regimes that we study and whose particulars are reviewed in this section.
Features of Three Regimes of Governance
Business group affiliated firms and the “empire” regime of governance
Many firms are affiliated to business groups in arrangements that can be analyzed as “imperial” governance regimes. From the viewpoint of regulation, empire refers first and foremost to a form of political dominance wherein one political entity (the center) dominates one or more other political entities (the units or components of the periphery). Crucial to the imperial regime of governance is the notion of “indirect rule.” Starting with Eisenstadt’s (1963, 1967) classification of historical empires and later with Doyle’s (1986) work, the study of empire was transformed from the older, “imperial” model—which focused on expansion and territorial exploitation (e.g., Hobson, [1902] 1965)—to the question of rule, or how the center exerts control over the periphery. Tilly (1997) introduced uncertainty into the analysis of empires; in this approach, the center dominates through indirect rule via peripheral elites who retain access to local resources.
The nature of indirect rule has recently come to be specified in social network terms (Barkey, 2008). An empire in this view is based on a “brokerage” (Burt, 2005) relation between its center and the different components of its periphery with a structure that emulates a hub-and-spoke pattern of relations. In this brokerage model of empire, peripheral components are given considerable leeway and autonomy while continuing to benefit from the resources of the center, especially centrally directed rewards and promotions. Finally, some recent work defines empire as a “negotiated enterprise” (Barkey, 2008; Nexon, 2009). In this sense, empire is a large, composite, multiethnic, and multireligious political formation in which relations between center and periphery are regulated through flexible and negotiated arrangements.
With regard to integration, an imperial regime of governance rests on the notion that peripheral projects are not suppressed by the imperial project (Barkey, 2008); here a project is defined as a sense of belonging, a sense of sharing the goals and benefits of an enterprise. Empires maintain a supranational mission of legitimacy—say, a religious identity with geopolitical claims or a mission civilisatrice (as claimed by European imperialists)—a necessary glue for the vertical incorporation of diverse elites. Even though imperial polities are geographically dispersed and culturally diverse, the sense of an overriding mission (shared at least by elites) maintains the integration of center–periphery relations.
In empires, “legibility” (Scott, 1998)—the ability of those who govern to read the territory they control and account for its resources and responsibilities in order to carry out such governmental functions as taxation and military recruitment—emerges from center to periphery regulation as well as an integrative play between the multinational identities of empire and a greater understanding of legitimacy built over time through governance as a shared project. As empires expand, they need to collect the facts on the ground, recognize the role of existing institutions, and decide about continuity and change. Empires manage resources and manpower while providing ample space for difference, variation, and strategic rotation of local elites across the empire. Legibility requires reading the territory and peoples of the empire and then integrating them within a structure flexible enough to allow for separate and negotiated compacts that empower the periphery without forsaking control of it. In this way, understanding empires requires focusing on the relations between the center and the periphery, where legibility, the control of resources, and contention are negotiated and resolved. In this context, basic behavioral models of power focusing on individuals’ decisions and outcomes—what Lukes (2005) calls the “one-dimensional” view of power—do not always satisfy, especially where complex organizations, interwoven networks of associations, and organizational fragmentation, as seen in empires, present a lack of clarity as to which actors wield power.
In the fashion industry, LVMH and PPR (see Hass, 2007) are the two most salient examples of empires that display the structure of political empires with their indirect rule, significant peripheral autonomy, structural flexibility, multiple projects integration, and strategic rotation. LVMH, Moët Hennessy–Louis Vuitton, which owns fashion houses such as Marc Jacobs, Céline, and Fendi, was founded in 1987 by Bernard Arnault; as of 2011, he remains the key shareholder of the group. Known as Pinault-Printemps-Redoute until 2005, PPR (which owns fashion houses such as Yves Saint Laurent, Gucci, and Alexander McQueen) was established in 1963 by François Pinault and is now headed by his son François-Henri Pinault. These two business groups compete in several markets and have organized their activities in different divisions, but fashion is the most significant activity for each (PPR, 2009; Solca & Wing, 2009). There are also other, smaller, fashion empires. For example, the Swiss conglomerate Richemont is a major player in the luxury industry and includes jewelers (such as Cartier) and watchmakers (such as Piaget); Richemont also owns some fashion brands, such as Chloé. Similarly, the Spanish group Puig owns several fashion houses, for example Nina Ricci. More generally, and consistent with our approach, fashion houses that belong to business groups competing outside the fashion industry as well as in fashion are coded as being affiliated with business groups. This is for example the case with French house Lanvin, which has been owned since 2002 by Harmonie SA, a group headed by Shaw-Lan Wang, who is also involved in the publishing business in Taiwan. Lanvin is coded as being affiliated to an empire.
With respect to regulation, fashion empires exert indirect rule on their affiliated fashion houses, leave them considerable autonomy, and allow for a negotiated structure. They do not micromanage their brands and instead allow them plenty of leeway to make both strategic and operational decisions, as long as they reach specific financial targets. In some instances, even those targets can be renegotiated, as was the case between French fashion house Christian Lacroix and conglomerate LVMH when Bernard Arnault decided to accept losses from this house for a while (Messarovitch & Arnault, 2000). But much like other empires in varied political and business realms, fashion empires exert a measure of indirect control by facilitating the sharing of resources among their different houses, which connect through the center and not much directly between themselves.
In terms of integration, fashion empires exhibit features similar to political empires or business empires in other industries. First and foremost, all of them own diversified portfolios of brands that each benefit from considerable autonomy; hence these fashion empires integrate multiple brand projects. Second, similarly to political empires (see Barkey, 1996), fashion empires organize the international professional mobility of top designers and managers among their different houses and divisions—for example, in 1997 LVMH offered New York designer Marc Jacobs, who was director of his own house acquired the same year by LVMH, the leading creative position at Louis Vuitton in Paris.
Unaffiliated firms as “kingdoms”
Many firms operate autonomously in markets and are not affiliated to business groups and can be characterized as “kingdom” governance regimes. From a historical vantage point, the continuum of regulation ranges from empire at one end to, at the other end, a form of unitary kingdom. With authority concentrated at the center of the absolutist kingdoms in Western Europe, the monarch was the only human source of law and ruled with the aid of a permanent professional army and bureaucracy. The monarch, army, and bureaucracy were all invested in centralized direct rule; having moved away from the feudal and imperial forms of decentralized and indirect rule, they left little or no autonomy to peripheral components that led to an inflexible center–periphery structure.
During the 19th century, the kingdoms (absolutist and constitutional regimes) of Western Europe gave way to modern nation-states in which the goals of center and periphery were consonant. Consolidating the unity and project coherence already under way in absolutist kingdoms, nation-states coupled a strong central leadership with tight control of local elites’ mobility. Thus, although both kingdoms and nation-states are strongly regulated (through direct rule from the center, minimal peripheral autonomy, and an inflexible structure), they are also strongly integrated. The nation-states that emerged with the rise of democratic institutions led to increased levels of integration. That is, the nation-state tends toward a homogeneous unitary society in which project coherence is high. The contemporary nation-state is quite forceful in its direct integration of society—both through democratic or authoritarian bureaucratic power and what Mann (1984) calls “infrastructural power,” by which elites are controlled and overseen by the state. In this context, where complexity is reduced and direct relations dominate, power relations are much less negotiated than in empires; they are about direct compacts with the state or domination and lack of autonomy.
In fashion, most houses are not affiliated to conglomerates and thus compete by themselves on the market. They are kingdoms. They present a single, unified face to the market, and do not devolve power to a corporate center that manages several affiliated units. French fashion house Chanel exhibits all the features of a kingdom; it is under the direct financial rule of its owners (the Wertheimer family), and under the direct creative rule of its artistic director (Karl Lagerfeld). Moreover, although Chanel manages several lines, these lines have no operational autonomy, and thus the overall organizational structure of the house is not the result of a negotiation between the center and the periphery. From an integration point of view, Chanel is characterized by a unitary brand project and mobility is tightly controlled by the house itself.
Hybrids governance regimes: federated arrangements
The political tension between center and periphery, between unity and peripheral autonomy, and the contest between pressures for integration versus fragmentation of large-scale diverse communities has led actors in modern times to devise political systems that differ from empire yet still retain some of its characteristics. The federated arrangement, which in the political realm can take the form of a federation or a confederation, is a solution that is situated in the political space between empire and kingdom, and it incorporates different layers of complexity in the dimensions of regulation and integration. It is often an alternative arrangement to the kingdom, developed to bring stability when increased complexity and diversity threaten cohesion. In business contexts, federated governance regimes are based on business groups as is the case with empires, but share some aspects with kingdoms.
As far as regulation is concerned, federated arrangements feature a mix of direct and indirect rule, vary in terms of peripheral autonomy, and deploy a flexible though institutionalized structure. In a sense, they are “hybrid” regimes of governance that have emerged as alternative ways (to those of empire and kingdom) for coping with tense relations between center and periphery. A federated arrangement can thus be understood as the combination of “a shared government (for specified common purposes) with autonomous action by constituent units of government that maintain their identity and distinctiveness” (Watts, 1998, p. 118). The existence of multiple localities is a key to the understanding of federated arrangements.
The main difference between empire and federated arrangement is that the latter relies on institutionalized flexibility: the federated center and the federated arrangement’s constituent units have areas of authority as provided by a constitution that empowers each component in turn. Whereas empire involves a dominant center that indirectly controls and influences its periphery via a negotiated and flexible structure, federation involves a “constitutionally structured dispersion of power” sometimes referred to as noncentralization (Elazar, 1997; Osaghae, 1990; Watts, 1998, p. 124). This dispersion of power defines prerogatives, and the center can exert direct rule in some matters and indirect rule in others. Legibility in the federated arrangement is achieved through the application of agreed-upon constitutional principles that have been organized throughout the polity. In comparing empires and federated arrangements, we can say that imperial flexibility stems from central dispositions and indirect rule whereas federated flexibility is strategically agreed upon by center and periphery. In federated arrangements, local constituencies can fiercely defend their identity and entitlements (de Tocqueville, [1835, 1840] 2003), leading to various levels of project integration, but the center can have widespread prerogatives as well that directly influence local politics. Similarly, elite mobility in federated arrangements is mostly organized at the highest level—that is, the center where peripheral projects converge, while local and peripheral entities retain some leeway. Historically, in the political realm, federated states have emerged later than empires. The tensions of hybridity are maintained in the constant negotiations, movements, symmetric and asymmetric power relations that are built into federated arrangements, but are refashioned by agreed-upon amendments.
In the fashion industry this structure has alternative configurations since companies may adapt to the market by growing in different ways. One model involves growth from within, whereby fashion houses develop brands that are related to their original brand—a process known in marketing as “brand extension” (e.g., Aaker & Keller, 1990). In such cases, houses (e.g., Armani) behave like federated arrangements, owning several sub-brands that compete as separate entities in the fashion industry. For example, Giorgio Armani and Emporio Armani, two lines of Armani, each have their dedicated fashion show and practically act in the market as separated fashion houses. Armani strongly regulates these sub-brands, yet they display relatively distinct projects and appeal to different customers (see White, Godart & Corona, 2007). Finally, some fashion houses, though they exercise only limited regulation of their portfolio of houses, still overshadow those subsidiaries. These houses represent a different form of federated growth: one that occurs via acquisition, as in the case of Diesel which owns houses such as Maison Martin Margiela. At this stage, it can be relevant to note that historically federated arrangements have emerged only recently, after empires had become accepted regimes of governance in the industry. This fits with Courpasson and Dany’s (2003) finding that hybridization is a recent phenomenon. However, the fact that hybrid business groups have emerged only recently does not preclude the definition of their characteristics. Hybrids, even though they emerge from previously existing socio-cultural formations, display their own idiosyncratic features, such as the famous “punk” style which according to Hebdige (1979) is a hybrid of reggae and rock influences, or the rocknroll music genre which is a hybrid of several other streams of popular American music, as explained by Ennis (1992). Punk and rocknroll are hybrids, but they nonetheless have their own features that can be summarized, defined, and incorporated in emerging socio-cultural formations.
Hypotheses
We can now think about how these comparative categories of governance regime can help us reconsider organizations and their creative outputs. We have defined three types of regime of governance in the fashion industry in relation to our central research question regarding the impact of business group affiliation, or the absence thereof, on creative performance; these three types cover the spectrum of possible governance regimes as far as the focus of the current paper is concerned. First, fashion organizations that compete on their own in the market, regardless of how many non-autonomous fashion lines they manage, are kingdoms. Second, fashion organizations characterized by the existence of a corporate center that operates a portfolio of fashion houses—but where the corporate center does not compete on its own in the fashion market—are empires. Third, we view some fashion organizations as federated adaptations to the market and conditions of competition. Some operate several lines that compete separately on the market as fashion houses, but are clearly associated through a common affiliation. Others are fashion houses that compete on their own in the market, but have acquired other fashion houses that they dominate from an operational standpoint. In the case of federated arrangements, the corporate center can compete on its own in the market. Adopting political terminology and concepts is useful because it helps identify types that exhibit “hybrid” characteristics, which are espoused as fashion business groups adapt to the market’s vagaries. Following Dahl’s (1956) use of the concept of a “polyarchy” that combines features of democracy and oligarchy, the notion of hybrid governance regime has more recently been applied to organizations (Clegg et al., 2006; Courpasson & Dany, 2003; Courpasson & Thoenig, 2008). The key idea to be found in this perspective, as far as the present article is concerned, is that hybrid regimes emerge from political struggles in concrete settings, not as top-down design solutions.
There is a stream in research on the fashion industry that tends to give more creative credit to fashion houses that are autonomous (Crane, 1997), which we call kingdoms, at least in contexts where the cost of entry is low. However, because empires and federated arrangements are governance regimes that foster some level of diversity and respect differences while sharing power, resources, and knowledge, we expect them to function especially well in creative industrial contexts. In the music industry, for example, Lopes (1992) shows that multilabel music companies (similar to empires or federated arrangements) enable the music industry to manage downstream uncertainty while allowing for stylistic innovation and diversity. Likewise, Dowd (2004) shows that, even though the music industry’s concentration in the hands of just a few major players can hinder diversity, this effect may be mitigated by a certain degree of decentralization of the kind found in fashion empires and federated arrangements. In describing the movie industry, Cattani and Ferriani (2008) report that the big, imperial studios and their multiple “distribution divisions” do not hinder the expression of individuals’ and teams’ creativity.
As pointed out by Chandler (1991, p. 33), corporate centers provide units in “multibusiness firms” with “loss-prevention” and “value-creating” support. While what he calls a “multibusiness firm” is not a business group per se—the former constitutes a single legal entity with several divisions while the latter is a set of legally separated entities belonging to a common organization (see Douma & Schreuder, 2008, pp. 351–2 for a full discussion)—they share similar organizational logics for example as far as the relationship between the center and the periphery is concerned. Thus, we conflate multibusiness groups with business groups since the highlighted mechanism can be found in both contexts. More specifically, in creative contexts, governance regimes based on a business group structure mitigate market uncertainty, giving more opportunities to affiliated firms to explore creative options. Additionally, as far as regulation is concerned, a certain degree of indirect rule, some peripheral autonomy, and structural flexibility can help affiliated firms to adapt quickly to ever-changing customer tastes. From an integration point of view, the existence of multiple projects, and the rotation of elites in affiliated firms, fosters creative diversity and innovation: each affiliated firm can focus on creative endeavors that fit its own history, and creative talents are given more opportunities to hone their skills through organized mobility across the business group units.
The resilience of empires and federated arrangements is also based on their ability to connect different communities and sensibilities through their affiliated firms, and thus to increase the “legibility” (Scott, 1998) these regimes of governance have on the various markets in which they compete. Cities and fashion styles play the role of nations and cultures in political empires. Political empires and federated arrangements are brokers among nations and cultures; likewise, fashion empires and federated arrangements are brokers among cities and styles. Fashion business groups connect fashion capitals that are otherwise disconnected, and they also occupy a central position in the “stylistic space” of fashion.
In sum, regimes of governance related to business groups are expected to have a positive impact on the creative performance of their affiliated firms. In other words, firms competing in creative contexts typically benefit from business groups which help their affiliates face uncertainty without compromising their creativity. For all these reasons, we define the following hypothesis, applied to our context.
However, it is not clear whether the empire or the federated regime should perform better. The distinction between the two types of regime is salient in fashion and its related industries, but needs to be further unpacked for theoretical developments. For example, Bernard Arnault, the CEO of fashion and luxury empire LVMH, elaborated in an interview about why houses affiliated to an empire such as LVMH are likely to perform better than other houses: “Success, in this type of business, derives from size which must remain relatively small, allowing to have direct, permanent, and personal contact with products, teams, and craftsmen. This can work only in a completely decentralized organization” (Messarovitch & Arnault, 2000, pp. 102–3). He goes on, detailing LVMH’s imperial structure vis-a-vis alternative, more centralized, governance regimes which we term “federated arrangements.” The crux of the governance difference between imperial and federated governance regimes is that, in the latter, the center itself is directly involved in market-specific competition. In other words, while federated arrangement centers such as Diesel compete directly as a brand in fashion, this is not the case for LVMH or PPR which compete only through their affiliated brands.
This has major regulation and integration implications: while in empires fashion houses are solely responsible for developing their collections, federated arrangement centers may decide to influence the creative output of their affiliates, thus limiting their autonomy and trying to create some level of unity project-wise. This is clear for Armani, for example, where the collections of the affiliated brands are closely controlled by the center, and specifically by Giorgio Armani himself, and also for Diesel where design outputs of the subsidiaries such as Maison Martin Margiela are watched over. The Martin Margiela case can help further illustrate the specific mechanisms at play in fashion federated arrangements. Renzo Rosso, the founder of Diesel, the Italian business group that bought Maison Martin Margiela in 2002, has a sound personal knowledge of fashion design, as he was for example called a “jeans genius” and “fashion innovator” by renowned fashion journalist Suzy Menkes (2003). This sound knowledge, developed at Diesel, informs the way he handles and manages the firms bought by Diesel. When it became clear that Martin Margiela would leave the house he founded in 2009, Renzo Rosso, along with Giovanni Pungetti (the chief executive officer of Maison Martin Margiela), decided that Maison Martin Margiela could survive as a creative powerhouse without a creative director, a very innovative and risky decision. However, this control can lead to political reactions from affiliated fashion houses, such as in the conflict between Prada and Helmut Lang in 2005. In federated arrangements, the sharing of creative insights from the center to the periphery is prevalent, but affiliated organizations may not hesitate to defend their identity. The existence of a positive impact on organizational creativity of a certain level of political conflict between the center and the periphery would fit with research on creativity in teams, where a reasonable amount of conflict is shown to have a positive impact on creativity as far as non-routine tasks are concerned (Jehn, 1995).
However, one could expect houses affiliated to empires, which are characterized by greater autonomy of affiliated houses, to perform better than those affiliated to federated arrangements because of this higher level of autonomy. Nevertheless, the federated arrangement regime of governance is the result of a hybridization between empire and kingdom, and hybrid organizational regimes have been shown to perform well in several contexts because they draw on the strength of different forms and thus are more flexible and better prepared to handle unexpected market evolutions (Shane, 1996). Similarly, Courpasson and Dany (2003) explain that hybrid forms of “bureaucratic” and “democratic” organizational governance are more effective because they balance adherence to extant organizational goals and entrepreneurial endeavors. Nonetheless, the impact of hybrid governance regimes on creativity remains to be fully understood. At this stage, it is important to emphasize that a hybrid governance regime is more than a simple mix of two other governance regimes. As Clegg et al. write: “hybrids are mostly conceived as more or less harmonious combinations between presumably opposite organizational characteristics,” but the fact is that “political hybrids are contested social productions” (2006, p. 332–3). Hybrid regimes such as federated arrangements are the result of conflicts and power struggles. They perform better for exactly this reason: they are the product of an organizational political history. We expect fashion houses affiliated to federated arrangements to perform better than houses under other regimes of governance. We believe that federated arrangements—as hybrids of empires and kingdoms—are especially well-suited for a creative industry because they actually exert some creative control on their affiliated units, and deal with some level of conflict. Hence we offer the following hypothesis, which like the previous one is applied to our empirical context of the fashion industry.
Data and Analysis
Data
Data were collected using several online and print sources on the fashion industry. A crucial first step was establishing the population of fashion houses to be included in our analysis. There is a wide array of producers competing in the fashion industry, from mass-market clothing retailers such as Japan’s Uniqlo or Spain’s Zara to elite Parisian haute couture maisons such as Stéphane Rolland or Anne Valérie Hash. However, the center of the industry is composed of the houses that organize fashion shows in the four main fashion capitals: New York, London, Milan, and Paris (Godart 2012, p. 51). These houses define the trends and attract most of the attention of the fashion buyers, press, and customers through their fashion shows. The influential fashion magazine Vogue, now available online via its international website Style.com, is a reference for the entire fashion industry (see Godart & Mears, 2009). Each season, Vogue editors and journalists review the fashion shows organized in the four major fashion capitals, providing pictures of all the designs presented by the creative directors for each show. In this sense, Vogue’s fashion show reviews provide an exhaustive picture of the collections of the fashion houses that constitute the center of the industry. This population consists mainly of well-established, upscale fashion houses with the financial means to organize a fashion show. Thus, mass-market retailers and smaller houses are excluded.
Since the Fall 2006 season, Vogue has offered a stylistic analysis of fashion houses’ collections. In particular, Vogue defines six main styles (or trends) that, according to its expert analysts, characterize the current season—including “hippie revival” and “boy meets girl” for Spring 2008. For each style, Vogue provides pictures of designs (and the name of the fashion house that produced them) that correspond to this style; fashion houses can then be classified in terms of those styles. A given house can display several styles in a given season, which is important to know because stylistic choices by creative directors in fashion design are likely to influence the creative performance of a fashion house. Also, knowing the stylistic choices of fashion houses allows us to build a network of seasonal stylistic affiliations for houses, a “stylistic network.” Thus, for example, two houses displaying the “hippie revival” style for Spring 2008 (Prada and Giorgio Armani) would be affiliated in that season. From this point of view, we follow a new trend in social network research that expands the notion of a network from a set of ties among actors to a set of ties among actors and concepts, ideas, or aesthetic features (see Carley, 1994; Mohr, 1998). We collected data on the seven fashion seasons for which information was publicly available when we gathered data for this paper, namely, Fall 2006, Spring 2007, Fall 2007, Spring 2008, Fall 2008, Spring 2009, and Fall 2009. A total of 293 fashion houses were reviewed during these seven seasons and are included in our dataset, but the unit of observation is the fashion house-year, of which there are 814 (there are many market exits in the fashion industry, as well as many newcomers, and these two facts account for this number).
Standard organizational data on fashion houses is scarce compared to what could be found in other industries. For example, financial data such as number of employees, revenue, and operating income are not available for privately held fashion houses or for individual brands belonging to fashion groups. This means that the sales figures of renowned fashion houses such as Chanel or Christian Dior are not available to outsiders. However, corporate websites and online news providers such as Factiva offer additional data, such as the founding year of a fashion house.
Variables and analysis
A salient issue faced by social scientists when dealing with creative performance is the operationalization of the concept of creativity as a dependent variable, which can be studied at the level of individuals, teams, or whole organizations (Drazin, Glynn & Kazanjian, 1999). Here we are interested in the impact of different governance regimes on organizational creativity. Measuring creative performance in an industrial context such as the fashion industry is not an easy task, notably because there is no obvious and objective way to measure such creativity. In a recent attempt to do so, Cattani and Ferriani (2008) assess creative performance in the movie industry in terms of awards and nominations of movie professionals and teams, and this approach could be adapted to fashion. However, the main fashion awards, such as the CFDA Fashion Awards in the United States and the British Fashion Council Awards in the United Kingdom, are mostly national and hence of limited use for studying the international population of fashion houses in our dataset. A better way to measure organizational creativity is then to rely on the opinion of selected industry experts, a perspective that fits with Amabile’s (1996) claim that something (for example, an idea or a product) is creative when it is novel and useful, the novelty and usefulness being best assessed by professionals familiar with a given social context.
A ranking of fashion houses corresponding to Amabile’s criteria for defining creativity is compiled, every season, by the leading French fashion trade magazine Journal du Textile. We used it to build our dependent variable. This ranking is international in scope because it is based on assessments by 70 international buyers working for leading retailers such as Neiman Marcus in the United States or Le Printemps in France. Each season these buyers award points to fashion houses and they are specifically asked to assess the creativity of the collections. The Journal du Textile compiles a ranking based on the points awarded. The ten highest-ranked fashion houses are distinguished as “Masters.” This Journal du Textile ranking is widely viewed as a worthwhile reference used by professionals, and it was employed by Crane (1997) to assess the creative performance of fashion houses. There seems not to be a strong national bias in the ranking procedure. Background interviews—conducted with fashion professionals in New York and Paris in order to understand better the “codes” of fashion (see Hsu & Hannan, 2005; Rao, Monin & Durand, 2005)—confirm that this ranking is used internationally by fashion recruiters to spot up-and-coming designers and by investment banks to assess the performance of fashion houses. The ranking has been described as positively impacting sales (in the sense that houses that are deemed creative tend to sell more over time), but creativity and commercial potential are two variables that needs to be conceptually distinguished. Creative houses such as Christian Lacroix can fail to command high volumes of sales, while less creative houses (in the mass market, for example) can sell a lot.
The independent variables of theoretical interest are categorical variables describing the type of regime (either Empire or Federated arrangement) under which a fashion house functions, where Kingdom is the reference. A value of 1 for the variable indicates that a fashion house is affiliated with an empire or a federated arrangement and a value of 0 indicates that it is not. Within our data set, a total of 29 houses are affiliated with one of the empires. Furthermore, it contains a total of 24 fashion houses that can be classified as belonging to the federated arrangements regime of governance. Additionally, we would like to highlight the fact that, given the way we have operationalized the governance regimes for the fashion industry, there is only very limited space for coding uncertainty. Affiliated houses are clearly identified in industry and firms’ reports, and the nature of the business groups is easily identifiable given the criteria we have defined.
Categorical control variables include Season dummies and City dummies. Including a seasonal categorical variable is a way to account for unobserved temporal effects. City location (with London as a reference) allows us to address the idea—preeminent in such disciplines as economic geography—that place plays a key role in market performance (e.g., Saxenian, 1996; Scott, 2007).
Fashion house age is an independent variable used to control for organizational market experience and knowledge. It is calculated as the difference between the current year of the fashion season under consideration and the year the fashion house was founded.
In order to control for fashion house-specific characteristics of creativity, we included two additional variables. Stylistic diversity is measured as the count of Vogue styles or trends to which a fashion house is affiliated for a given season; and Number of designs is a count of the designs that were assessed by Vogue. These two variables need not be correlated. A fashion house could be affiliated with many styles and present only a few designs for each style, or it could present many designs in only one style.
A measure of media coverage by selected outlets is used as a proxy for fashion house size because prior research (Fang & Peress, 2009) shows that media coverage and firm size are highly correlated; organizational size has been shown to impact organizational dynamics in many contexts (see Kalleberg & Mastekaasa, 1998). This measure, Media Coverage, is constructed using the number of articles mentioning the name of a given fashion house in the six months preceding the focal fashion season (see Godart & Mears, 2009). This number is determined after reviewing the newspapers, magazines, and websites classified by Factiva as belonging to the “clothing and textiles” industry; all languages available in Factiva are covered in order to avoid national or linguistic bias. Media coverage matters because magazines and newspapers are major outlets for the diffusion of information to customers (Moeran, 2006), and this is increasingly true of websites as well. This variable is divided by 100 so that the regression coefficients can be interpreted.
Because the market performance of firms has repeatedly been shown to depend on their position in socioeconomic networks (e.g., Burt, 1992; White, 2002), we introduce two network variables to control for structural effects in the performance of fashion houses, using UCInet (Borgatti, Everett & Freeman, 2002) to measure the variables. In this network, the nodes are fashion houses that are connected through the sharing of styles during the focal season. First, the Structural holes variable captures “constraint,” which is a continuous measure defined by Burt (1992) to account for the control by an actor of structural holes (in other words, this actor is connected to social networks that are not connected otherwise). Second, Coreness is a continuous measure of network centrality that accounts for the position of an actor within a core–periphery structure (Borgatti & Everett, 1999); it has been used, for example, by Cattani and Ferriani (2008) to explore creativity in the movie industry. In this paper we construct a network in terms of stylistic affiliations. For a given season, houses are “affiliated” if they share a style. Structural holes in this case would account for the ability of a given house to blend styles that are otherwise disconnected, and Coreness would indicate whether a given house is promoting mainstream or more idiosyncratic styles.
We use a logistic procedure with robust standard errors to test our hypotheses because the dependent variable is binary—namely, whether or not a fashion house was listed in the Journal du Textile’s “top 10”. This Top 10 variable is measured over the course of seven seasons. As a robustness check, which yielded the same results as the logistic regression for our main hypotheses, we also ran the models using a negative binomial procedure where the dependent variable was the count of Journal du Textile points rather than the fact of belonging to the Top 10 or not. For this robustness check, we used a panel data approach, but without fixed effects at the level of fashion houses because the fixed effects specification would lead to dropping most of the observations. Tables 2 and 3 display, respectively, the descriptive statistics of the variables of interest, and the Pearson correlation coefficients for these variables. Table 4 reports test results for H1 and H2.
Descriptive Statistics
Pearson Correlation Coefficients
Note: all coefficients with absolute value bigger than 0.07 are significant at 0.05.
Logistic Regression on Fashion Houses’ Creative Performance
Notes: Criterion for creative performance is membership in Journal du Textile’s ‘Top 10’ list (N = 814 house-seasons). Standard errors in italics.
p < .05; **p < .01; ***p < .001.
We focus on model 4 which is our fully specified model. Both of our hypotheses are supported. Empires and federated arrangements have a positive and significant impact on the likelihood of a fashion house to belong to the Journal du Textile’s Top 10 ranking. In the creative context of the fashion industry, governance regimes characterized by an affiliation with a business group are generally beneficial to an affiliated organization. This is because in both empires and federated arrangements, the existence of some level of indirect rule, peripheral autonomy, some flexibility, multiple projects, as well as rotation of elites mitigates uncertainty for affiliated organizations and allows them to gain legibility of the market. As a matter of illustration, a fashion industry case can help understand the mechanisms at play for example in empires. In 2002, when Shaw-Lan Wang bought Lanvin, and included the French fashion house in her empire, a clear understanding was that she would help revive the company by giving time to the newly appointed creative director, Alber Elbaz, to develop a coherent brand and not intervene in the day-to-day operations. This indirect rule exerted on the periphery, giving Lanvin autonomy and flexibility, was also facilitated by Shaw-Lan Wang selecting the right creative director at the right time since Alber Elbaz had a good understanding of the realities of the fashion business, given his previous experiences in other houses (Hirschberg, 2005). A couple of years later, Lanvin became one of the most profitable and creative fashion houses in the world, and indeed Lanvin is well-placed in the Journal du Textile ranking, being number one since Fall 2009. Being affiliated to an empire (or a federated arrangement) buys time for the affiliated fashion houses, and also provides them with resources that can be aptly used when the local elites (here the creative directors) are well selected. In this sense, governance regimes based on business groups foster innovation and creativity by protecting their affiliated firms from the vagaries of creative industries. As Greenwood and Hinings (1996, p. 1048) point out, change, and notably the action that leads to it, “is not disembodied” and thus understanding the dynamics of change, innovation, and creativity can be eased by the understanding of the impact and dynamics of various governance regimes, as suggested in this article.
However, the positive impact of federated arrangements on creativity is more important than the impact of empires, as seen in a larger coefficient (1.19 vs. 0.92). This difference stems from the hybrid nature of a federated arrangement, a governance regime that is the result of a political process and draws on the strengths of both empires and kingdoms. Again, this hybrid nature does not mean that federated arrangements are simply an intermediary state between empires and kingdoms. Hybridity emerges from political struggles. Some autonomy of affiliated unit is advantageous, but we also find—perhaps counter-intuitively—that this autonomy needs to be somehow bounded. In fashion federated arrangements, the center keeps some control of the creative process that takes place in its subsidiaries. This comes from the political struggles that led to the emergence of the federated arrangement, either the acquisition of a house by another house, or the creation of a spin-off. In both cases, the center—and the individuals who compose this central elite—have a personal and deeply ingrained knowledge of the industry. This is the case for example for Renzo Rosso at Diesel or Giorgio Armani. This knowledge not only gives them legitimacy to intervene in the creative process of the affiliated units, making them more creative, but also helps the center to increase its own creativity. For example, if Maison Martin Margiela benefited from its affiliation to Diesel, Diesel also benefited from acquiring Maison Martin Margiela, both financially and creatively. Additionally, federated arrangements can trigger some level of center/periphery tension and conflict that can be beneficial to creativity.
Fashion house age has no impact on creative performance. Stylistic diversity has a positive and significant impact in models 1 and 2, but is not stable, and importantly the effect does not appear in model 4. Thus we cannot conclude that this variable matters for a fashion house’s creative performance. Additionally, we find that the size of a fashion house’s creative output (measured with the Number of designs variable) does matter; it has a positive and significant impact in model 4. This indicates that houses that create more designs are likely to be considered to be more creative. We can conjecture that this is because in creative industries, which are characterized by high uncertainty, creating more designs is a way to increase the odds of producing something that will be positively received by market actors. We also find that Media coverage has a positive and significant impact, which suggests that a broad presence in the press and on the internet—an indication of the size of a fashion house—is positively correlated with the perception of this fashion house as being creative by relevant industry players. However, future research might be necessary to assess the causality between creative performance and organizational size as expressed in media presence. One could hypothesize that bigger houses are able to command better assessments, not only because of their inherent market power but also because of their more important resources giving them more leeway to be creative. Finally, we find that social network variables (Structural holes and Coreness) do not affect the performance of fashion houses.
Conclusion and Discussion
Summary and implications
In this paper we explored the impact of various regimes of governance on organizational creative performance. Along with Clegg et al. (2006), we theorized governance regimes as the correspondence between concrete power relations and organizational structures, with an emphasis on the latter. Political and power struggles in organizations within the creative economy depend on various regimes of governance that lead to differential creative outcomes. Conceptualizing governance regimes in organizations by focusing on the way the center is related to peripheral units, we defined the concepts of empire, kingdom, and federated arrangement by highlighting their features along two dimensions, regulation and integration. We then developed a rigorous comparison between political and business regimes of governance and argued that political and economic regimes often share important structural characteristics. Finally, we evaluated the empirical validity of these concepts by testing several hypotheses in the creative business context of high-end fashion. As a consequence, it appears that social scientists can draw from political regimes not only to decode business regimes but also to formulate the underlying strategic choices faced by an organization. Notably, the notion of hybrid regimes, stemming from Dahl’s (1956) concept of a “polyarchy” and aptly used, for example, by Courpasson and Thoenig (2008) in organizational analysis, has a lot of potential for empirical research, and we have tried to expand it to the creative realm.
Empirically, we found that hybrid regimes of governance—i.e., federated arrangements—generate the most creative organizations. They benefit from the advantages of the two regimes they stem from, empires and kingdoms. In the fashion industry, as in other creative industries, business group affiliation in general protects affiliated organizations against the vagaries of inherently uncertain creative processes, favoring change, creativity, and entrepreneurship in general. However, houses affiliated with federated arrangements proved to be more creative than those affiliated with empires. This is due to differences in the political structures of these regimes of governance. Empires are characterized by indirect rule that lets affiliated houses develop their own collections and creative approach—as long as they meet profitability criteria—because the center does not have specific creative competences. In federated arrangements, the center exerts its power on the creative process of its subsidiaries, transferring its knowledge and experience to them, and in turn benefiting from their skills. In federated arrangements, the subsidiaries are only partially autonomous, and their creativity is influenced by a mix of direct and indirect rule from the center.
More generally, we established a relationship between the regime of governance of an organization and its creativity. While the link between power structures and efficiency has long been an object of research (Clegg et al., 2006), the question of power structure and creative performance has been only marginally tackled, and our paper has shed new light on this question. More precisely, the existence of a decentralized power structure, as found in business groups, generates more organizational creativity in cultural industries. However, the centralization vs. decentralization dimension is not enough to fully understand the dynamics of organizations, as explained for example by Courpasson (2000) in his work on “soft bureaucracies.” The same goes for the understanding of how power influences creativity in organizations such as fashion houses. While centralization and decentralization are important and instrumental concepts, they need to be enriched via a more fine-grained study of political dynamics. The role of elites and how they relate to the different units that compose the organization they manage is a crucial point. Business groups where elites are actively involved in the creative process (i.e., federated arrangements) perform better than those where it is not the case. Power decentralization generates creativity, but its effects are most noticeable when it is balanced, that is to say when the central elites of an organization are legitimate and active enough to influence the creative process. These processes can generate tensions and conflicts, but as long as they do not degenerate into a standstill, they can be creatively beneficial. In this sense, we found that there is indeed a “social movement” of “ruling corporate elites to rejuvenate the pillars of their legitimacy” (Clegg et al., 2006, p. 389); in the creative industries it takes the form of an encroachment of the center on the creative endeavors of the periphery.
In this way, we also contribute to the general understanding of power in organizational settings, where the concept of power in strictly behavioral terms is not fully satisfactory. It is more fruitful to formulate power in terms of relations across center and periphery, of direct and indirect forms of control, and as emerging from the various types of relations that exist in concrete social and economic contexts, in situ. In this sense, we also go beyond what Lukes (2005) terms the two-dimensional and three-dimensional views of power which are about setting the agenda and manipulating the desires of individuals, respectively. As pointed out by Clegg et al. (2006, p. 217), the three approaches distinguished by Lukes all go back to the basic notion of power as convincing others to do what they do not want to do; we look at larger social structures and at how power is embedded in them. We also draw from both Weber ([1922] 1968) and Elias (1998). Weber’s focus was on the formal specification of the organizational structure, and power in impersonally defined organizational positions, rather than in the persons who happen to pass through them. Elias, in contrast, was taken not by formal delegation, but by the fact of interdependence in large organizations—that is, in organizations commanding the greatest quantity of resources. They both urged the incorporation of a more depersonalized understanding of power, but also one that is more embedded in context. We followed their heed and specified regimes of governance as social formations emerging at the crossroads of organizational power structures and concrete power struggles, and offering the intriguing characteristic of being replicable as “blueprints” for action, not only across organizations, but also across contexts (White, 2008, ch. 6).
The focus here has been mostly on firms in creative fields, but we believe that the political concepts of empire, federated arrangement, and kingdom can also be used in non-creative settings, notably because the political features we highlight are valid beyond the creative economy. We do not expect empires or federated arrangements to perform better in every context. For example, in commodity industries—where cost is a main driver of competitive advantage—a unitary project and an inflexible structure can yield important cost reduction advantages by allowing for more streamlined processes. In such a context, kingdoms might well perform better than empires or federated arrangements.
We believe our article belongs to a broader attempt at “rethinking the firm” by using a variety of social scientific approaches (see Koza & Thoenig, 2003). Political sociologists and political scientists have learned a tremendous amount from organizational studies. Many political analyses of state–society relations, much state-related research, and numerous studies of social and political movements have borrowed concepts from the fields of organizations and social networks (e.g., Davis, 2005; Diani & McAdam, 2003). The reverse does not hold, however, because most of the organizational literature in business studies has ignored important models of political association (Pfeffer, 1993)—this despite major contributions by political sociologists and scientists to the field of organizational behavior that involve game theory (Axelrod, 1997), rationality (Simon, 1997), and power (e.g., Dahl, 1957). These various fields should be brought together in comparative studies of political and business governance regimes. That would allow us to establish whether and how the nature and structure of governance can significantly improve our understanding of organizational and strategic choices by business actors.
We have also expanded the literature on business groups in two ways: first, by defining several types of business groups and unaffiliated firms in terms of political governance models (empire, kingdom, and federated arrangement); and second, by evaluating the impact of these business groups and unaffiliated firms on creative performance. In showing that different types of governance regime have differential impacts on creativity, we have demonstrated the fruitfulness of using models of political governance to help understand businesses in creative fields such as fashion. We thus contribute to a stream of research that aims to define different types of organizations and their impact (e.g., Hannan & Freeman, 1977; Hsu & Hannan, 2005; Powell & DiMaggio, 1991). We integrate two major research streams on business groups, on the one hand the identification and specification of various forms of business groups (e.g., Granovetter, 2005) and, on the other hand, the assessment of the performance of firms affiliated with those various forms of business groups as compared with independent or unaffiliated firms (e.g., Chang et al., 2006). Concepts defined in historical and political sociology can be advantageously used in organizational studies. In particular, the business group literature will be enriched if we can specify the different types of such groups that populate an economy. And all organizations (whether or not affiliated with groups) must deal with similar coordination issues, which can be put in sharper relief when viewed through the lens of political science and sociology concepts.
In the creative context of today’s high-fashion industry, affiliation with business groups generally lends greater market performance. This better performance follows from a sharing of resources and insights, which is facilitated by the business group structure. In the fashion industry explored here, several elements give hybrid federated arrangements an edge over empires: a mix of direct and indirect rule, the large but bounded autonomy of affiliates, the existence of a flexible but institutionalized structure, the integration of multiple projects, and mobility at the top. Uncertainty and ambiguity are pervasive in markets (Knight, 1921), leading actors to devise complex social mechanisms to cope with it (e.g., Corona & Godart, 2010; Huault & Rainelli-Le Montagner, 2009; White, 1981, 2002; White et al. 2007). In the context of fashion, hybrid regimes of governance are better equipped to handle the vagaries of creativity, and by allowing some level of tension and conflict they increase their creative process. In this sense, not all business groups are equal.
Another point worthy of more thorough discussion is ethical in nature. This paper suggests that an empire has certain structural and organizational features that allow it to perform well and to sustain the diverse identities of its components. Because our case study is in a business context, we do not address the question of democratic representation. Yet this question cannot be avoided in political contexts, where empires—in scholarly research as well as the media—have understandably garnered a negative reputation (especially with respect to imperialism and colonial empires). Similarly, the literature on hybrid political regimes stresses the undemocratic nature of polities that present themselves as democratic since they perform basic electoral functions (Diamond, 2002). The various regimes of governance defined in this article must however be distinguished from the political regimes that regulate the appointment of officials and the rights of citizens.
Future research
The three regimes of governance analyzed in this paper constitute only a subset of what political science and political sociology can bring to a theory of governance regimes in business contexts. Our focus on empires, federated arrangements, and kingdoms was justified by our goal of analyzing the impact on creativity of different types of governance regimes around the question of whether creative organizations are affiliated or not to business groups. However, other research questions could benefit from the introduction of additional concepts. For example, a transition from a given regime of governance to another takes time. Two transitory forms of governance, the hegemon and the warlord, could be analyzed in further research.
In hegemony, member states constitute a community that agrees on principles of international politics and on a basic set of values and shared interests. Hegemons have been present in history at different times; moreover, empires and kingdoms have periodically assumed a hegemonic role and then abandoned (and sometimes reclaimed) this position. In the field of fashion, one could hypothesize that some fashion houses played, at the height of their influence, the role of focal center of a hegemon. This would be the case, for example, of Christian Dior at the time of the “new look” in the late 1940s. Warlords throughout history have been independent and unruly local agents who were able to gather forces (soldiers, mercenaries) around them and then to establish zones of autonomy—pockets of self-reliance that thrived in the interstices of state–society relations. In fashion, Karl Lagerfeld can be viewed as a warlord who reinvented many facets of fashion design though his multiple experiences and affiliations.
More generally, the operationalization of the governance regimes we have defined in this article is context-dependent. For example, the question of what constitutes an empire vs. a federated arrangement in the movie or music industries would require additional developments. We suggest that the level of industry expertise of the center and its elite could be the main criterion that sets federated arrangements apart from empires. Additionally, organizations sharing a governance regime vary in managerial practices. For example, there is a major difference between the two main fashion empires in the type of high-level creative profile they favor: whereas LVMH bets on fashion superstars (Karl Lagerfeld, Marc Jacobs), PPR hires talented designers who are famous in the world of fashion but not necessarily well-known by customers (Hass, 2007). Understanding the antecedents and consequences of managerial differences within shared governance regimes would generate intriguing research. Similarly, there exists variations on (and other structures besides) center and periphery, such as structures with multiple centers, different levels of periphery, and so forth. These were not considered here but would make for interesting supplementary cases in the exploration of how various types of governance regimes affect firm performance.
Footnotes
Acknowledgements
We thank Andrea Baldo, Peter Bearman, David Courpasson, Felipe Csaszar, Mario Diani, Paul Evans, Mark Mortensen, Florence Rambaud, Jean-Claude Thoenig, and three anonymous Organization Studies reviewers for their insightful suggestions and comments.
Funding
This research received no specific grant from any funding agency in the public, commercial or not-for-profit sectors.
Notes
Author biographies
Karen Barkey is Professor of Sociology and History at Columbia University. She studies state centralization/decentralization, state control and social movements against states in the context of empires. Her research focuses primarily on the Ottoman Empire, and recently on comparisons between Ottoman, Habsburg, and Roman empires. Her recent book, Empire of Difference: The Ottomans in Comparative Perspective, is a comparative study of different forms and moments of imperial organization and diversity. It was awarded the 2009 Barrington Moore Award from the Comparative Historical Sociology section at ASA and the J. David Greenstone Book prize at the Politics and History section of APSA.
Frédéric C. Godart is an Assistant Professor of Organisational Behaviour at INSEAD. His research interests include the structure and dynamics of creative industries, the development of design as a significant economic activity, and the history of fashion. He received a PhD in sociology from Columbia University, New York. A former fellow of the École Normale Supérieure de Cachan in France, he also holds an MPhil in social and political sciences from the University of Cambridge and an MS in management from Sciences Po in Paris. He was previously a research and business analyst for McKinsey & Company’s strategy practice.