Introduction

HIS IS A BOOK about corporate political design, and, in particular, about the differentiation and integration of the roles played in executive committees, CEO offices, and boards of directors. Our principal argument is that productive interpersonal relationships based on personal and professional trust are the key to the integration of these structures. Such role integration is particularly visible in what we call small numbers at the top as a mixture of role separation, role combination, and role-sharing among a reduced number of executives – usually between two and four.

   There are few incumbents at the corporate apex in the majority of organizations (Mintzberg, 1980). Corporate power, like social power, is always a phenomenon of small numbers. Sometimes the pinnacle of the organization takes the form of dominant coalitions, as noted by Cyert and March in their A Behavioral Theory of the Firm (1963) and by Thompson in Organizations in Action (1967), or of upper echelons, as in Hambrick and Mason’s (1984) influential piece; sometimes it is an inner core group, as in Kleiner’s (2003) text for practitioners.

   Although, as Khurana (2002) reminds us, individualism is assumed in most concepts of corporate power, rarely does one individual exercise great power in complex organizations. Yet performance is often attributed to the individual at the top. There are, however, numerous practices that do not fit that individualist assumption. Increasingly, two executives are sharing corporate governance roles traditionally occupied by one person (e.g. co-CEOs), and duos linked by trust are occupying two vertically contiguous roles (e.g. a CEO and a COO), in what Kirkpatrick (2004) calls “buddy acts” (p. 44). Sometimes circles of intertwined executives with the same or different status or job category, and with personal or professional trust, or both (what Hodgson, Levinson, and Zaleznik [1965] called “role constellations”), occupy the roles that constitute corporate governance.

   There are at least two reasons for us to focus the central chapters of this book on these duos, trios, and quartets and other executive constellations based on role separation, role-sharing and other role combinations, and on the trust that jells them: (1) for the inherent interest of the phenomenon of small numbers at the top, and (2) for the lessons it can contain for our knowledge of corporate governance.

   First, the occurrence of small numbers is an interesting and important social phenomenon in its own right, one that had been largely neglected for over a hundred years since Simmel (1902a, 1902b). There were some attempts to study small numbers at the top in the mid-1960s (Whisler, 1960; Daniel, 1965; Etzioni, 1965; Hodgson, Levinson, and Zaleznick, 1965), primarily in order to account for the growing complexity of the chief executive office in the multi-divisional organization. These works were followed some thirty years later by Stewart’s (1991) lucid piece on the chair–CEO relationship and, in recent years, by the work of organizational sociologists interested in the role of third parties in networks – Burt (1992), Krackhardt (1992), and Gargiulo and Benassi (2000), among others; and, even more recently, by authors interested in roles at the top (O’Toole, Galbraith, and Lawler, 2002; Sally, 2002; Hambrick and Cannella, 2004). The use of small numbers such as dyads, triads, and quartets in executive committees, CEO offices, and boards of directors¹ is the specific focus of this book.

   Second, it is hoped that the study of interpersonal ties as integrating mechanisms in the creation and maintenance of small numbers at the top will improve our knowledge of corporate governance, thereby providing us with a stronger base on which to construct effective corporate political designs. For that is the ultimate aim of this book.

   Knowledge has accumulated on the topic of horizontal structure designs – designs in which the differentiation in groups of specialized people and the corresponding integration or coordination of those units reflects the value chain of the company. However, as Tushman (1977), Lewin and Stephens (1994), and others have noted, vertical structures have been much less studied from a design perspective. Although the field of strategic leadership, based primarily on Hambrick’s upper-echelons theory, has generated an astonishing amount of high-quality work on top management teams, CEOs, and boards of directors, most of its works have, for methodological reasons, been primarily limited to the demographics of executive teams rather than their deliberate design characteristics and processes, as one of its academic champions has recognized (Finkelstein, 1992).

   There are at least four reasons why so little research has been undertaken on the design of vertical structures.

   (1) There are methodological challenges in the study of decision-making and decision makers at the top. It is difficult to secure access to the corporate inner sanctum (e.g. board deliberations are confidential), forcing scholars to resort to indirect research strategies, using such proxies as demographics, functional experience, and social networks. As a result, the measurable composition of corporate governance bodies has attracted the most research, but the core of decision-making at the top – in which interpersonal relationships, and trust and affection among executives play critical roles – still eludes academics.

   (2) When it is not demographics-based, the study of the incumbents at the apex of organizations has been diluted in the wide field of leadership studies, which mostly focuses on the personal characteristics and psychology of executives rather than on their actual behavior and their activities in performing the tasks prescribed by their roles. Studies of leadership are too often lacking in context, and do not account for the activities reserved for directors, executives, and their specific contingencies – a gap we try to fill with this book.

   (3) Corporate governance is, to a significant degree, an institutionalized function. A series of highly publicized business affaires have prompted the generation of a variety of norms, either social (e.g. professional codes of good governance) or legal (e.g. criminal law reforms such as jail penalties for CEOs convicted of misrepresenting the accounts of their companies), which have undoubtedly reduced the latitude for the design of distinctive corporate power structures.

   (4) Not only are the assumptions, values, beliefs, and rules about corporate governance more institutionalized in today’s investor capitalism than in the old days of managerial capitalism, but they are also more widespread across nations and business systems (Aguilera and Jackson, 2003). The demands for similar standards are increasing with the growth in global capital flows and are resulting in a higher degree of convergence in corporate governance.

   Despite this stronger institutional logic (Thornton and Ocasio, 1999) and wider convergence, this book argues that current corporate governance reforms are mostly based on weak norms and that there remains a wide range of possible variations in corporate political structures. This diversity (e.g. small numbers at the top) is not widely recognized in the business press² and even less so in the academic literature. It has long been acknowledged that executives have latitude in deciding the domains of their jobs and the manner in which they should assume their responsibilities in the face of constraints (see Stewart, 1982; Alvarez, 2000). The integration of roles at the level of corporate governance must be based upon weak structures such as interpersonal relationships because there are no other formal integrating mechanisms or superior hierarchy. Our position is in contrast to the normative environment of today’s businesses – an environment in which personal relationships are regarded with great suspicion in all corporate governance reforms. For instance, Blas Calzada, former president of Spain’s Securities and Exchange Commission, bluntly expressed these suspicions, declaring that the chair of a board and the CEO cannot be friends.

   There is an element of irony in today’s normative thinking, in that friendship and personal trust in corporate governance are mistrusted, whereas when considered as components of the social capital that provides the glue for modern societies, their value is positively recognized and even demanded by most analysts of the contemporary social scene (see, for example, Piore, 1995; Putnam, 2000).

   In the remaining pages of this Introduction, we first present a string of examples of what we call the “small numbers at the top,” in order to illustrate clearly what we mean by that expression. Then, in a section on the academic aim, we refer to our main theoretical approaches – contingency and role theory – and place them within current academic debates on the design of corporate power structures. We are of the view that organizational theorists have not paid sufficient attention to small-numbers structures, and that, even in recent research on shared leadership, the phenomenon of small numbers is treated only tangentially. In a section on social opportunity, we argue that this topic is especially appropriate and relevant in a period when changes in corporate governance are occurring throughout the world, based primarily upon well-known agency theory postulates, which we believe to be ideologically loaded and limiting for the range of possible governance practices. We conclude this Introduction with a description of the work process that led to this book and a summary of its chapters.

The small numbers at the top phenomenon

Despite a lack of academic acknowledgement, as noted by Stewart (1982, 1991), executive role-sharing and role constellations abound. Tandems, trios, quartets, and variants of these options are seen in a variety of contexts. They proliferate in creative enterprises such as entertainment, advertising, and fashion, where the gap between management and creativity is often bridged by teams comprising artists and managers. The luxury fashion brands of the Louis Vuitton Moët Hennessy Group have been co-run by a designer, in charge of the aesthetic concept, and an entrepreneur, responsible for the product’s commercialization. Music recording has long been a favorable place for enduring career couplings: Pairs such as Gloria and Emilio Estefan, the most powerful Latin music duo in the USA, have developed a personal and professional relationship over the past three decades, nurtured by their mutual loyalty and by their roots, their friends, and their troupe (Townsend, 2000). As in other tight partnerships, their joint trajectory is also embedded in their company, Estefan Enterprises, which includes the Crescent Moon Studios and video, TV, and cinema production companies, among other businesses. Film directors often couple their careers with producers who are close to them and who provide them with both autonomy and resources (Alvarez, Mazza, Strandgaard, and Svejenova, 2005). A case in point is the famous Danish film director Lars von Trier, at the core of the Dogme movement, who has teamed up with his producer and friend Peter Albæk Jensen (Brorsen and Strandgaard, 2002). Italian independent filmmaker Nanni Moretti works closely with his producer and friend Angelo Barbagallo. Asked how the two prepare a film, Moretti explains:

Rather than a team, we are two people – Angelo Barbagallo and me  . . . having our own production company, and such a good partner, enabled me to suddenly decide to make films, . . . to make documentaries or shorts, whatever we felt like  . . . when you have a conventional relationship with the film producer, a film will have very clear and distinct stages – the writing of the film, pre-production, the actual shooting of the film, post-production, the editing and all the rest. Whereas in some of these films that I have made in recent years, these stages have become much more blurred. (Wootton, 2001)

   Some of these complementary tandems play such an influential role in their firms that the company’s stock value falls when they leave, as occurred in the case of the departure in November 2003 of Gucci–Yves Saint Laurent’s chief designer, Tom Ford, and its CEO, Domenico de Sole. In creative industries, tandems and career couplings are observed not only between artists and business professionals but between two creative people, such as the filmmaking Coen brothers (De Felipe, 1999). Advertising pairs recurrently team up as well, often for campaigns, or they may move in tandem from one agency to another (Vagnoni, 1997). Similarly, in Nissan Design International, vehicles are designed by some twenty-five couples, each consisting of an intuitive creative and an analytic creative, who have been hired in pairs by the unit’s founder, Jerry Hirshberg, in the belief that the odd coupling leads to friction that, in turn, enhances creativity (Cubeiro, 1998). Musicians in string quartets also tend to develop long-term collaborations (Murnighan and Colon, 1991), as do scientists (Zuckerman, 1967). Pierre and Marie Curie, and Irène Joliot-Curie and Frédéric Joliot, are examples of Nobel Prize winners with both scientific and affective ties. Their accomplishments as couples were probably far greater and more lasting than they would have been had each worked individually (Pycior, Slack, and Abir-Am, 1996).

   United professional trajectories could also manifest themselves in some hierarchical partnerships at the executive level of creative organizations (e.g. co-Presidents, CEO–COO, co-CEOs). World-class media empires may need more than one person at the top (Grover, 1999). Teaming up in 1984, the creative Michael Eisner as CEO and the financially adept Frank Wells as COO, put the troubled Walt Disney Productions back on its feet. A Business Week article portrays their partnership in the following way:

When he first came to Disney back in 1984, Eisner’s comrade in arms was Disney president Frank Wells. An accomplished Hollywood lawyer and one-time head of Warner Brothers, Wells was the yin to Eisner’s yang. When Eisner got too excited about something, it was Wells who would bring him back to earth. Eisner would want to build a Mickey-shaped hotel, Wells would remind him of how much it would cost. (Grover, 1999)

   Michael Barker, Tom Bernard, and Marcie Bloom are co-presidents of Sony Pictures Classics. Before founding the company in February 1992, Barker and Bernard founded Orion Classics in the early 1980s. Marcie Bloom has been their partner since 1989.

   Small numbers at the top can be observed not only where internal differentiation is greatest, as in the firms engaged in artistic activities, but also in mainstream businesses. Small numbers at the top abound in the broad domain of family businesses, often with kin sharing management responsibilities persistently over time. In the 1997 Arthur Andersen–MassMutual American Family Business Survey, more than 11 percent of the business respondents reported that they had two or more CEOs and more than 40 percent believed that that would be their situation in the succeeding generation. In the 2002 American Family Business Survey almost 9 percent reported having two co-CEOs, and 3.5 percent having more than two CEOs; more than 35 percent said that co-CEOs are possible in the next generation, and two out of five of those likely to adopt co-CEOs responded that one of the co-CEOs may be a woman (Mass Mutual Financial Group and Raymond Institute, 2003). One common path to such couplings occurs when parents name their children as equal successors to the business. In June 2004, for example, Marc and Manuel Puig became co-CEOs of Puig Beauty and Fashion Group, the family-owned corporation carrying the well-known perfume brands of Carolina Herrera, Paco Rabanne, and Nina Ricci among others.

   Pairs of CEOs are also found in such steady and conservative sectors as investment banking. Goldman Sachs has hired co-CEOs for more than two decades: Weinberg and Whitehead, Rubin and Friedman, and Corzine and Paulson. But Goldman Sachs, a case we often refer to in this book, is not an isolated example in the financial sector. Winthrop H. Smith and Charles E. Merrill jointly ran the brokerage house Merrill Lynch for decades (Heenan and Bennis, 1999). In a similar arrangement, Charles R. Schwab served as Chair and David S. Pottruck as President, sharing the CEOship at Charles Schwab Corporation.

   There are pairs at the top in other traditional sectors. In the catalog retail company Spiegel, for example, the CEO role is shared by Mike Moran, James Sievers, and Harold Dahlstrand, who together constitute the Office of the President, each contributing different skills and responsibilities.³ At Ralston Purina, Patrick McGinnis headed the Pet Products Group, while sharing the CEO role with Patrick Mulcahy, CEO of the Eveready Battery Company between 1997 and 1999. The US superstore chain Bed, Bath & Beyond, was founded and led by co-CEOs Leonard Feinstein and Warren Eisenberg. In January 2001, the international news and technology information giant Reuters America Inc. announced the appointment of Alex Hungate and Phil Lynch as co-CEOs – not their first such partnership. And co-presidents Mark G. Parker and Charles D. Denson have served Nike since 2001.

   The high-tech industry also provides some pertinent examples, like the case of the co-founders of the Hewlett-Packard Company – Stanford buddies William R. Hewlett and David Packard (Kaplan, 2000). Another example is the introverted Bill Gates and the socially adept Steve Ballmer, who met at Harvard in the early 1970s (Heenan and Bennis, 1999). In the Microsoft duo, Gates was the strategist and Ballmer the tactician. While at the University of California in Berkeley, the technically creative Steve Wozniak hooked up with the marketing whiz Steve Jobs, five years his junior, forming the partnership behind Apple (Kaplan, 2000). This partnership broke up in the early 1980s when Wozniak left Apple.

   The leadership of political organizations also provides well-documented examples of tandems at the top, for reasons similar to those found in business: the impossibility of a single individual simultaneously conducting inspiring and disciplining activities, performing external roles (e.g. relations with the electorate) and internal functions (e.g. control of the party). Role transitions between such disparate roles are difficult to conduct and can endanger the coherence of public image demanded of politicians (Miller, 2001). In Spain in the 1980s, the long-standing government of the socialist party was headed by a tandem consisting of the Prime Minister and First Secretary of the party, Felipe González, who focused on institutional tasks and securing the support of the voters; and the Vice Prime Minister and Second Secretary of the party, Alfonso Guerra, who had the explicit role of coordinating the efforts of top officials in the administration and of harnessing the party bosses. Similarly, US President Bill Clinton and Vice-President Albert Gore demonstrated that complementary tasks are consistent with differences of character throughout their first presidential campaign, when Clinton picked Gore as his running mate to the end of their second mandate (Williams, 2001).

   Tandem leadership is not then a negligible phenomenon that occurs only in small economic sectors and marginal organizations. These and many more examples are scrutinized in Chapters 4 to 6, which focus specifically on duos and other small numbers. For an alphabetically ordered list of selected examples of small numbers at the top, see the Appendix.

The academic aim of the book

This book uses two academic platforms to build its arguments: contingency theory and role theory. We believe that both of these perspectives represent an opportunity to understand the phenomenon of small numbers at the top. Furthermore, neither theory has been sufficiently applied to contemporary vertical structures.

   The critical role of interpersonal relationships as an integrating mechanism was present in the initial versions of contingency theory, but was not pursued in its later developments. Yet the examples of small numbers at the top that have been presented in this Introduction could not have been effectively maintained without high levels of trust. In this book, we try to recover and update contingency theory, originally intended for horizontal structures, and apply it to the design of vertical structures – to the system of roles and processes commonly known today as corporate governance. This is the main theoretical strategy of the book, deployed in Chapter 1 on contingency theory and in Chapter 3 on role theory.

   Our thesis is that the key to the challenges of designing vertical structures is similar to the one that Tom Burns, pioneer of contingency theory, describes in the preface to the third edition (1994) of his book The Management of Innovation (1961), co-authored with G. M. Stalker. As Burns says: “My own preoccupation was with the structure and dynamics of interpersonal relationships, of the various interests pursued by individuals and of the alliances they formed to further them and the social sub-systems observably present in organizations” (p. ⅹⅲ).

   As we argue in Chapter 3, there is a characteristic of roles at the top that explains why interpersonal relationships are such an important integrating mechanism: roles at the top, although increasingly bound by social and institutional regulations, are still subject to the enactment – choices and actions – of their incumbents. There is wide latitude for designing the behaviors, activities, and tasks that roles at the top encompass, and, as a consequence, for the accommodation of interpersonal relationships among the role holders. To give one example, Stewart (1991) claims that the variations available for the division of tasks between a CEO and a chair of the board are numerous (e.g. the latter could be partner, executive, mentor, consultant), and the particular roles chosen depend on the degree of mutual trust and joint agreement about the content of the two roles between the two incumbents. It is for vertical structures, therefore, that the following assertion by Eccles and Nohria (1992), intended for structures in general, is most true:

the best structure often must be compromised in order to adapt to the particular individuals who are available . . . A sub-optimal structure that takes account of people’s strengths and weaknesses is almost always better than an optimal structure that makes overly heroic assumptions about what people are capable of doing . . . structure must be designed around and for individuals. (p. 141)

We might add that when designing top structures, both individuals and their relationships should be taken into account.

   Small numbers at the top, although existing in the practice of corporate political structures, are not well recognized or even welcomed by some management scholars and managerial traditions.

   In a revision of his classic article about the folklore and the reality of managerial work, and even in the midst of an argument emphasizing the vast difference in managerial roles and the enormous challenges of role transitions, Mintzberg (1989) is explicit in his skepticism toward role-sharing: