Introduction
In the past decade an increasing number of international managers from multinational organizations have participated in development programs at the Center for Creative Leadership. These managers work across the borders of multiple countries simultaneously. Some of them are expatriates. Most are not. And although many of these managers are not wrestling with the issues of relocating and adjusting to living in a different culture, they all find themselves dealing with cultural issues—defined in the broadest context—every time they pick up the phone, log onto their e-mail, or disembark from an airplane.
Working with these managers led us to ask fundamental questions about our current understanding of managerial effectiveness and whether or not it applied to managers who work in an increasingly complex global world. We asked ourselves: What do these managers do? Is it different from the work they did when they managed in their own countries, and if it is different, how so? What does it take for them to be effective when they manage across so many countries simultaneously? What do these managers need to know in order to be effective? What do organizations need to know and do in order to select and develop people who will manage and lead effectively in the global economy? This report is our attempt to address those questions. Although it is written for scholars, the practical implications of our work have been developed and published elsewhere (Dalton, Ernst, Deal, & Leslie, 2002).
Other researchers have explored the characteristics or competencies of global managers. Gregersen, Morrison, and Black (1998), for example, conducted interviews and gathered survey data from international managers in identifying five characteristics of successful global leaders: (1) context specific knowledge and skills, (2) inquisitiveness, (3) personal character (connection and integrity), (4) duality (the capacity for managing uncertainty and the ability to balance tensions), and (5) savvy (business savvy and organizational savvy). Associates of the Hay-McBer Group conducted critical incident interviews with 55 CEOs from a variety of industries located in 15 countries to determine the critical factors predicting global managerial effectiveness (Martin, 1997). They named competencies they believed are universal regardless of context (four competencies under each of three headings labeled Sharpening the Focus, Building Commitment, and Driving for Success) and identified three kinds of competencies that vary as a function of a given cultural context (business relationships, the role of action, and the style of authority).
These two major studies reach similar conclusions regarding a proposed taxonomy of effective global managers. Our aim in this report is not to offer yet another taxonomy. Our goal is to integrate a number of theories of leadership and managerial effectiveness as they relate to management in the global role, to test the utility of these theories, and to explore the dynamics that lie beneath their taxonomies. Using a more theoretical, integrative, and quantitative methodology, we want to better understand and link some of the precursors of global managerial effectiveness.
This study investigated a large number of variables from a variety of perspectives—far too many for a single research article and in far too much detail to be of great interest to practitioners. The purpose of this report is to provide a comprehensive review of all of the hypotheses and analyses conducted as part of our investigation—those that proved fruitful and those that did not.
Because of the complexity of this report, we offer the following road map. Following this introduction, we present and discuss our conceptual model, which we designed to help identify, appreciate, and explain the relationships among the skills, capacities, traits, and experiences managers need to be effective when their work is global in scope. The methods section relates to all subsequent chapters. We have organized the chapters following the methodology to match with our conceptual model. Each chapter presents background information on specific variables by reviewing key and relevant literature and by subsequently offering hypotheses regarding those variables, and concludes with results and a brief discussion.
Chapter 1 sets the stage by introducing the work of global managers—what they do and how it is different from managerial work in a domestic context. Chapter 2 investigates the relationship of personality to effectiveness in a global role and considers personality as a precursor to the presence of the skills and capacities necessary for effectiveness. Chapter 3 explores the relationship between learning capabilities (self-development, perspective taking, and cultural adaptability) and managerial effectiveness. Chapter 4 explores how being a cosmopolitan—an individual who has lived and worked in many countries and speaks a number of languages—may increase the likelihood for effectiveness in a global role. Chapter 5 focuses on the influence of workgroup heterogeneity and homogeneity on effectiveness—does experience in managing a diverse workgroup in a domestic role increase the likelihood that an individual will be effective in a global role? The final chapter concludes with a discussion of what global managers do, what it takes for a manager to be effective when the work is global in scope, and how global managers can be selected and developed.
This report is written for our academic colleagues and for research-oriented practitioners. It is our hope that other researchers will find our conceptual model useful and intriguing enough to continue the exploration.
Model
To create our model we turned to the literature and identified those variables that have demonstrated links to managerial and leadership effectiveness, including: role behaviors (Mintzberg, 1973), coping with pressure and adversity, integrity (Kaplan, 1997), knowledge of the job (Kotter, 1988), and personality (Barrick & Mount, 1991; Salgado, 1997).
We added other variables to our model that the literature and our experience led us to suppose might be particularly salient for international jobs. These included learning capacities (Spreitzer, McCall, & Mahoney, 1997) and three additional experience-based variables. One, cosmopolitanism, depicts an individual’s exposure to other languages and cultures during childhood and adolescence; the other two variables we called cultural heterogeneity and organizational cohort homogeneity. We defined cultural heterogeneity as an individual’s experience with managing diverse groups while managing within his or her own country. Organizational cohort homogeneity explores the influence of workgroup similarity on perceptions of effectiveness. In other words, we explored the effects of managers’ demographic similarity (for example, number of years with company, national culture) to their cohort on how effective they were perceived to be.
Our model (Figure 1) is conceptual, not statistical. Personality stands for an individual’s enduring traits that might help explain the kinds of experiences to which he or she is drawn and the kinds of capabilities and role behaviors he or she is most likely to have acquired.
Experience refers to those experiences and demographic variables that individuals bring with them to the job. Experience may be critical in understanding why one manager is comfortable with the unfamiliar factors inherent in global work but another manager is not. Experience may also influence the skills and capacities a manager has acquired over time.
Managerial capabilities includes three major categories of skills: learning behaviors, resilience, and business knowledge. Learning behaviors include the motivation and skill to work and learn across cultural differences, the willingness to take the perspective of others, and the capacity to learn from workplace experiences. These variables have held a tacitly strong position in the management-development literature and some of the global-management literature. Resilience refers to the ability to manage time and stress, factors that might be more salient when the management task is global in scope. The third skill group, business knowledge, represents knowledge of the business and business practices. We did not write specific chapters about business knowledge and resilience. Also not discussed here in any detail is the variable cognitive ability. We believe cognitive ability to be important to global managerial effectiveness, but because of the plethora of existing literature on the subject, and because of the difficulty in measuring the construct in a survey design, we excluded it from this paper.
Managerial roles stands for those behaviors that managers employ to carry out the basic functions of their work: managing relationships, managing information, and managing action. Current thought suggests that although all of the roles are important, the need for a manager to enact a particular role shifts as a function of context. Managerial work in an international business represents a particular type of context.
Global complexity represents the context of interest. We maintain that when the manager’s work is global in scope, the relationships of these variables to measures of effectiveness and to one another will be different from what they would be if the manager’s work were local in scope. We have operationalized global complexity as the additive function of having to manage across distance, country, and culture. The greater the time and geographical distances and the more countries that fall under a manager’s scope of responsibility, the greater the global complexity of the work. Thus, temporal, geographical, and cultural complexity separate low global complexity (domestic work) from high global complexity (global work).
Methods
The methodology described in this section pertains to all following chapters, which address specific parts of the conceptual model. Two hundred eleven managers from four organizations participated in our study. All managers included were approximately at the same organizational level. Ninety-eight of the managers were from a Swiss pharmaceutical company. Twenty-five worked for a U.S. high-tech manufacturing firm. Forty-eight worked for a Swiss hospitality and service organization, and 40 worked for a Swedish truck-manufacturing organization.
Group Assignment
We used two items from our biographical measure to classify managers into either a low- or high-global-complexity group. The first item (“In how many countries are you a manager?”) allowed the following responses: (a) one country—I am not an expatriate, (b) one country—I am an expatriate, (c) more than one country on the same continent, and (d) more than one country on different continents. The second item (“In how many time zones do you work?”) allowed six responses: (a) 1, (b) 2, (c) 3, (d) 4, (e) 5, and (f) 6 or more. In tandem, these two items formed a proxy measure we used to assess the level of global complexity inherent in a manager’s role.
To form the low- and high-global-complexity groups, we took the following steps. First, we put both items on the same 4-point metric to give them equal weight in an additive function. Specifically, the item addressing number of time zones was collapsed from 6 to 4 points. Responses (a) and (f) were kept as unique categories (representing the low and high extremes of the value, respectively), and responses (b) and (c) and (d) and (e) were collapsed to form two middle-range values. Second, the two items were summed (range = 2–8). Third, the median for the sample was calculated (median = 3). This procedure resulted in the formulation of two groups, a low-global-complexity group with values from 2 to 3 (n = 110) and a high-global-complexity group with values from 4 to 8 (n = 101).
Sample Characteristics
Both samples were predominantly comprised of well-educated white males with a mean age of 44 in the low-global-complexity group and 45 in the high-global-complexity group. The majority of managers in each group were educated in only one country. Members of the high-global-complexity group had been in their current jobs for less than a year. Managers in the high-global-complexity group were also more likely to have been expatriates in the past than were those in the low-global-complexity group. Although 41 countries were represented in the total sample, 43% of the group were Northern European by birth (German, Swedish, Swiss) and 18% were U.S. citizens by birth. Those percentages were reflective of the corporate headquarters’ locations of the four participating organizations. Participants lived in 30 countries at the time of the study, with 66.9% living in Switzerland, Germany, Sweden, or the United States. Considering the split between low and high global complexity, individuals in the high-global-complexity sample were proportionately more likely than the low-global-complexity sample to speak English (39%) or French (10%) as their native language. The individuals in the low-global-complexity sample were proportionately more likely than those in the high-global-complexity group to speak German (33%) or Swedish (18%) as their native language.
Standardization of Data
We investigated the influence of native country of target manager (culture) and organization type on the criterion measures to determine the need to standardize the data from the four organizations before merging it.
Using the native culture of participants, we created a measure to represent cultural regions (Ronen & Shenkar, 1985). The regions included Anglo (n = 63), Germanic (n = 68), and Nordic (n = 37). This accounted for 167 of the participants. Other regions—Arab (n = 3), Far East (n = 5), Latin European (n = 9), Latin American Spanish (n = 6), Near Eastern (n = 2), and Independent (n = 9)—were not used due to the small sample size. We then conducted a one-way ANOVA between groups to compare means of the three regions on the managerial effectiveness ratings. No differences were found in criterion scores as a result of the cultural region of the target manager.
We repeated this analysis to assess the impact of organizational type on ratings (location of headquarters was partly confounded with organizational type because the pharmaceutical and service industry companies had corporate headquarters in the same country). Organizational type included pharmaceutical (Switzerland), high-tech manufacturing (United States), service (Switzerland), and truck manufacturing (Sweden). Again, we found no differences from the boss’s perspective. Nonetheless, these data were standardized within the four organizations before merging the four data sets. Standardizing the data within organizations helped to control for variation due to organizational culture, business and economic context, and industry type.
Measures
Independent variables.
Personality measure. The NEO PI-R was used to represent the personality conceptualized as the five-factor model (FFM), which groups personality traits into five domains (Costa & McCrae, 1992). Each domain is made up of six facets or subscales and these six facets define each of the factor domains. The NEO PI-R was chosen because of its psychometric integrity and because extensive research has demonstrated that these five factors do appear to be universal, if not all-inclusive, across cultures (McCrae & Costa, 1997). (See Chapter 5 for further details. Alphas for the five factors are provided in Appendix B.)
Demographics. Each manager filled out a biographical form indicating gender, age, native language, country of birth, and race or ethnic origin.
Experiences. Each manager filled out a biographical form indicating tenure, years in current role, expatriate experience, languages spoken in the course of doing work, languages spoken before age 13, number of countries lived in, country currently living in, years of formal education, number of countries educated in, and major field of study. Managers were also asked to indicate for their most recent domestic job the relationship, sex, race or ethnic origin, age, native country, country of current residence, and functional area for ten members of their workgroup.
Roles. We used a measure of 75 items to represent role behavior. We derived these items from an existing instrument, SKILLSCOPE (Kaplan, 1997), and from our review of the literature. Of the 75 items, 9 were dropped during preliminary analyses due to poor item-total correlations. For data reduction purposes, the remaining 66 items were subjected to a principal component analysis (principal axis method, with orthogonal rotation). When requesting a 7-factor structure, 50 items loaded cleanly on their expected factor. We used inferences taken from the data and our conceptual understanding of the items on the remaining 16 items to incorporate an additional 6 items into the 7 scales. The remaining 10 items were dropped from subsequent analyses. (See Chapter 1 for a complete discussion of role importance and effectiveness.)
Capabilities. Capabilities comprised three sets of scales: learning, knowledge, and resilience. We used 37 items to represent the learning constructs. We derived these items from existing instruments (SKILLSCOPE: Kaplan, 1997; Prospector: McCall, Spreitzer, & Mahoney, 1997) and our review of the literature. Following the same analytic strategy previously described for roles, a request for a 3-factor solution resulted in 28 items loading on their expected factor. The remaining 9 items, which did not load in a meaningful way on any given factor, were dropped. A further investigation of the items representing the intercorrelations of the learning scales cultural adaptability and perspective taking, and the knowledge scale international business knowledge, suggested that further data reduction might be appropriate. A factor analysis of the scales for perspective taking, cultural adaptability, and international business knowledge produced a more conceptually satisfying and parsimonious solution. In the final analysis we used 10 items to represent the construct international business knowledge. We used 5 items to represent the construct cultural adaptability. Four items were used to represent the construct perspective taking. The remaining items were dropped from further analysis.
Eight items were used to represent the knowledge construct. Four items were used to represent the insightful construct. These items were derived from an existing instrument (Prospector: McCall, Spreitzer, & Mahoney, 1997) and our review of the literature.
Four items were used to represent the resilience construct coping. Seven items were used to represent time management and three items were used to represent integrity. We derived these items from SKILLSCOPE (Kaplan, 1997) and our review of the literature. (Scales, items, and alphas can be found in Appendix B.)
Dependent variables. Managerial effectiveness as depicted in our model (Figure 1) represents the observable things that people do related to stated goals. Additionally, job effectiveness is held to be multidimensional. A criterion taxonomy—a listing of the multiple facets of effectiveness—is a more precise means of organizing links between the specific predictors and specific effectiveness criteria rather than between predictors and overall effectiveness (Campbell, McCloy, Oppler, & Sage, 1993). To represent managerial effectiveness, 27 items were gathered from the literature, then supplemented and revised in consultation with one of the companies that sponsored the research project. The 27 items were written to address three dimensions of managerial effectiveness: business practices and outcomes, managerial and leadership qualities, and relationships.
Further analysis and discussion by the research team suggested that these dimensions were better represented as five factors rather than three. We derived the final five dimensions of managerial effectiveness using three steps:
- We conducted a principal components analysis at the individual-rater level with boss and direct report ratings combined (747 observations). Fifteen items loaded cleanly on one of five factors.
- After a series of discussions related to the data and/or our conceptual understanding, we incorporated from the remaining 12 items an additional 9 items into the 5 scales. Three items were dropped.
- The five derived scales and corresponding items were e-mailed to a group of CCL faculty members, who were asked to provide a name for each scale.
These steps resulted in the final five dimensions of managerial effectiveness. The first scale was called managing and leading. It represented the traditional leadership behaviors of setting direction, inspiring, and motivating. It also included items that reference an internal focus and traditional manager-to-direct report activities, such as selection, development, coaching, and managing conflict. The second scale was called interpersonal relationships. This scale represented relationships with peers and senior managers inside the organization. The third scale was called knowledge and initiative. These items combined the characteristics of broad knowledge and professional competence with the personal attributes of confidence, independence, and initiative. The fourth scale was called success orientation. This scale represented an orientation toward goal achievement and attainment of desired organizational outcomes. It also included an item related to the potential to reach the most senior job in the company. The fifth scale, contextually adept, had an external focus and included items related to the ability to manage external relationships. (Scales and alphas can be seen in Appendix B.)
Data Collection
Participating managers completed the 240-item personality measure (NEO PI-R), the biographical form, and the measure of role skills and capabilities. The surveys were all in English. Managers were assured that their individual results would be available only to the research team. On the personality measure managers rated themselves on a 5-point scale ranging from strongly disagree to strongly agree.
On the measure of role skills and capabilities managers first rated each item on a 3-point scale ranging from (1) this is not important to my current job to (3) this is extremely important to my current job. They then returned to the items and rated their own skill at performing each of the behaviors on a 5-point scale ranging from (5) this skill or perspective is one of my greatest strengths to (1) this skill or perspective is something I am not able to do.
Bosses responded to the 27-item effectiveness statements on a 5-point scale ranging from strongly agree to strongly disagree. A not applicable response category was provided. This measure was in English. Responses were returned directly to CCL. Respondents were assured that their individual responses would remain confidential to the research team.
Although boss and direct report ratings were obtained, direct report ratings are not discussed in this report. Early examination of the direct report ratings in the high-global-complexity condition offered little to enhance our understanding because, like most multiperspective ratings, these respondents see things differently. Our focus turned to the perspective of the boss because it provided the best position for commenting on effectiveness under different conditions. The direct report ratings were dropped from the remaining report; however, the correlation matrices can be found in Appendices E and F. (See Appendices C-F for alphas and intercorrelations of all the measures by rater source.)
Conclusion
The methodology discussed in this section relates to the results and discussion that make up the remainder of this publication (excluding appendices). As described earlier, Chapter 1 introduces the work of global managers—what they do and how it is different from managerial work in a domestic context. Chapter 2 focuses on the relationship of personality to effectiveness in a global role. Chapter 3 explores the relationship between learning capabilities (self-development, perspective taking, and cultural adaptability) and managerial effectiveness. Chapters 4 and 5 introduce the variables of experience into the model—life experiences related to the number of countries in which a manager has lived or worked, the number of languages a manager speaks or reads, and work experiences such as the influence of diversity on a global manager’s effectiveness. Chapter 6 offers a general discussion of the findings as they relate to the conceptual models, and draws some conclusions about what those findings have to say about managerial effectiveness in a global context.