Printer Friendly

Top management team communication networks, environmental uncertainty, and organizational performance: a contingency view.

All else equal, more open communication and more reliable information are usually better for an organization (Eisenberg and Witten, 1987). Researchers disagree, however, on how best to achieve effective communication and information reliability. At the group level, one school of thought aligns with Coleman's (1988) network closure argument that the dense communication ties common in more closed networks promote trust and cooperation among group members (e.g., Garguilo and Benassi, 2000). Moreover, the resultant tight coupling among team members creates an environment conducive to efficient knowledge sharing (Hansen, 1999).

Other researchers instead follow the weak ties argument of Granovetter (1973) and the structural holes argument of Burr (1992), asserting that closure-based group cohesion can be a source of rigidity that hinders adaptive coordination in groups performing complex organizational tasks (Garguilo and Benassi, 2000). These researchers argue that weak ties linking otherwise disconnected individuals in more open communication networks create opportunities for "brokerage" of valuable new information across the structural holes in otherwise closed networks (see Burt, 1992, for an extended discussion). Put simply, the more distant and infrequent communications common in more open, weak tie networks can provide new and useful information for groups facing complex tasks.

Each of these perspectives has merit. However, the lack of knowledge about how the properties of group communication patterns influence group performance for complex tasks is especially acute regarding one important and highly consequential organizational group--the top management team (TMT) that makes strategic decisions. This paper contributes to the literature on TMT communication patterns and firm performance by arguing that the communication-performance relationship is context-dependent. The systems approach to communication proposes that the extent of the "fit" between network communication patterns and the communication environment will be positively associated with group performance (Euske and Roberts, 1987). A central and elemental task for TMT members is communicating to share information about the firm's environment in order to clarify or reduce environmental uncertainty (e.g., Galbraith, 1973). This paper argues that organizational performance will depend in part on how well a TMT's communication pattern matches the level of environmental uncertainty facing a firm. Thus, this paper proposes an overall "fit" hypothesis between TMT network communication patterns and environmental uncertainty, which allows the context(s) within which each of the divergent views of team communication might be valid to be examined.

This study examines the complex relationship between the "fit" of a TMT's communication network density and centralization with its firm's environment, and the resulting organizational performance, using data obtained directly from 404 TMT members in 32 firms. While the results of this study provide general support for a "fit" perspective, for one hypothesis the fit relationship was not as anticipated. These results indicate that as environmental uncertainty increases, concomitant increases in TMT network communication density are necessary for high performance, but unexpectedly, concomitant increases in TMT network communication centralization are also necessary for high performance.

The next section reviews relevant literature and develops the network structure "fit" hypotheses. Then the sample, the methodology, analyses and results are presented. The paper concludes with a discussion of the implications of findings for future TMT communication research.

THEORETICAL REVIEW AND HYPOTHESES

TMT Communication Patterns

The TMT is arguably the most consequential group within an organization (Carpenter et al., 2004). Furthermore, communication factors have long been known to affect important outcomes in organizations (e.g., DeWine, 2001; Downs and Hazen, 1977; Falcione, 1974). Yet, although researchers have argued that understanding TMT working processes such as communication is crucial in building the understanding of how TMTs affect organizational performance (Smith el al., 1994), and although prior research has shown that communication-related TMT outcomes such as social integration (Smith et al., 1994), behavioral integration (Carmeli, 2008), and decision-making process rationality (Mueller et al., 2007) influence organizational performance, research that has examined the effects of intra-firm TMT communication patterns on organizational level performance has been missing in the literature. This study takes an initial step toward addressing this important and overlooked issue. The conceptual model reflected in the following discussion is presented in Figure 1. Solid lines in the figure represent relationships tested in this study, while dotted lines represent the theoretical mechanisms underlying the tested relationships.

Communication among TMT members is central to the group's ability to carry out its functions (Mintzberg, 1973). Some researchers have found that frequent communication among TMT members increases information exchange and team cohesion, thus leading to better performance (e.g., Barrick et al., 2007). Others, however, have found team communication frequency could be the result of high conflict within the team (Smith et al., 1994), thereby reducing decision-making speed and thus performance (Eisenhardt, 1989). These results suggest that there is an appropriate level of communication within TMTs which is sufficient for information exchange yet less than that level which would preclude rapid decision-making. It is argued next that this appropriate level depends upon the informational requirements imposed by the firm's environment. This argument is consistent with the recent calls by researchers for increased consideration of contextual factors, such as the environment, in examining the TMT-organizational performance relationship (e.g., Doz and Kosonen, 2007; Cannella et al., 2008).

TMT Communication Patterns and Environmental Uncertainty

Organizations must develop structures and processes appropriate for the imperatives of their external environment if they are to achieve success or even survive (Pfeffer and Salancik, 1978). Of various environmental attributes, uncertainty has been portrayed as one of the most important in the management literature (e.g., Duncan, 1972). Milliken defines uncertainty as a "[manager's] perceived inability to predict [an organization's environment] accurately" due to lack of "sufficient information" or "inability to differentiate between relevant and irrelevant data" (1987:136). Inherent in this definition is the idea that insufficient and ambiguous information characterizes the state of uncertainty. Organizations deal with environmental uncertainty in many ways, including by planning ahead, delegation of authority (Thompson, 1967), structural differentiation and specialization (Galbraith, 1973; Lawrence and Lorsch, 1967), environmental scanning systems (Yasai-Ardekani and Nystrom, 1996), or developing flexible, informal decision-making processes (Fredrickson, 1984; Mintzberg, 1973). All these activities involve the collection, processing, and comprehension of information. Environmental uncertainty is therefore a key contextual factor that may influence the relationships of TMT processes and organizational performance (Cannella et al., 2008).

Information captured by an organization is generally processed through both formal procedures and informal interactions within the management team. Formal procedures process information following a predetermined chain of command. Informal procedures, however, are spontaneous and cannot be pre-designed but are also an important part of information processing network. For example, in a study of how patient complaints were channeled to problem-solvers in a hospital, Stevenson and Gilly (1991) find that managers often avoid passing problems to formally designated problem-solvers and instead use personal ties to forward information to problem-solvers.

[FIGURE 1 OMITTED]

The use of informal systems for communication would be even more pronounced when environmental uncertainty is high. Miller (1992), for example, provides empirical evidence suggesting organizations that perform best in uncertain environments have weaker linkages between their structural and process variables. And as the use of formal structures and processes become less pronounced in an organization, informal structures and processes inevitably emerge as substitutes (Monge and Eisenberg, 1987). Social network theory identifies communication networks as an important informal organizational structure that influences management effectiveness (Krackhardt and Hanson, 1993).

When top managers of organizations regularly interact through both formal and informal channels, patterns of communication emerge and may evolve into communication networks. These networks become important informal organizational structures that govern information transmission and exchange among TMT members. Two key attributes characterize patterns in communication networks: communication density and communication centrality. Network communication density refers most often to the proportion of actual communication linkages in a network relative to the number of possible linkages among network members, but it sometimes includes in more advanced forms the addition of the frequency of communication across linkages as well, so that the most dense networks will be those with many, high frequency linkages (Paldam, 2001). Network communication centrality, on the other hand, refers to the prominence of members in a network based on the number of individual linkages each possesses, thereby capturing the degree of concentration of communication/information on particular individuals in the network (Ibarra, 1993).

Prior research has established the performance effects of network communication centrality and density (Guzzo and Shea, 1992; Shaw, 1964). Reagans and Zuckerman (2001), for example, report that organizational units with denser networks achieved greater levels of productivity than did those with sparse networks. In a recent meta-analysis, Balkundi and Harrison (2006) conclude that the empirical evidence supports the overall idea that network density and centrality influence team performance in organizations. Below presents the reasons why these effects of network characteristics on performance are influenced by the context in which the networks are embedded.

Communication Network Density

When environmental uncertainty is high, TMTs with more dense communication networks are likely to be more effective than those with lower density. High communication network density indicates more linkages with more frequent communication across linkages within the TMT. This increased communication density improves information processing and facilitates rapid decision-making and implementation (Eisenhardt, 1989). Organizations in environments of high uncertainty have a high level of ambiguous information emanating from the environment, especially pertaining to cause-effect relationships (Milliken, 1987). With these ambiguities, differences in managers' interpretations, if not well communicated, could lead to confusion and potential group conflict (Tekleab et al., 2009). High network density is more likely to bring sufficient communication channels that help "glue" the differentiated TMT roles together (Molm, 1994). Dense interpersonal communication channels are useful in transmitting highly complex subject matter, as well as in disseminating information that can both reduce the potential for conflict (Arendt et al., 2005) and promote team cohesion, which together can improve performance (Michalisin et al., 2004; Li and Hambrick, 2005). In uncertain environments, organizations especially need to strengthen the roles of boundary spanning positions, to break down department and position barriers so as to help make the organization more flexible (Ibarra, 1993). Dense communication among top managers in an organization strengthens the integration of various boundary spanning positions at the top level and helps break down interdepartmental barriers (Barrick et al., 2007).

TMT communication network density benefits organizations by reducing resistance to change and promoting mutual understanding (Johnson el al., 2001), attributes that are especially important to the success of organizations in highly uncertain environments. Hence, dense communication network improves speed in decision-making and decision implementation, which are critical to successful adaptation in uncertain environments. TMT communication network density also affects the effectiveness of decisions made by a TMT both individually and collectively. Frequent communications among team members spawn team creativity and contribute to organizational innovation (Johnson, 1990). Although a single manager's judgment depends on background and experience (Goll and Rasheed, 2005), more communication among TMT members may serve to share more information within teams, stimulate perspectives and increases the overall cognitive capabilities of the group, thereby increasing the probability of reaching more comprehensive and effective decisions for team level decision-making (Larson et al., 1998). At the individual level, when expediting decision-making in an uncertain environment, a top manager may have to sacrifice decision quality for speed. If team members maintain intense communication among each other, such a trade-off might not be necessary as the collective "brain power" of the team can by utilized to ensure high decision quality and speed.

Conversely, when the organization's external environment is relatively stable, less abnormal information is received from the environment and most organizational processes and decisions can be based on mostly formal procedures (Fredrickson, 1984). Because TMTs in stable environments have less need to communicate with one another, a relatively sparse TMT communication network may be all that is necessary, and may be most efficient, for high-performing organizations. Efficiency is especially important to success for organizations in stable environments (Frederickson, 1984).

In sum, TMTs lacing highly uncertain environments require denser communication networks in order to perform well because they must handle so many non-standardized tasks. On the other hand, with low to moderate environmental uncertainty, a relatively sparse communication network will be associated with higher performance, because "over-communication" would slow the coordination and implementation of programmable tasks, resulting in organizational inefficiencies. Thus,

Hypothesis 1: As environmental uncertainty increases, concomitant increases in TMT communication network density will be positively related to firm performance.

Communication Network Decentralization

While the absolute amount of information shared among a TMT's members is important, whether each individual manager gets the chance to access the information needed for decision-making is equally important. In some TMTs, certain executives may have access to more information than others, while in other TMTs information is more evenly distributed. Network centrality can be conceptualized as both a group-level concept and an individual-level concept. At the individual-level, it shows the relative importance of a single member within a network, based on number of communication ties, and is often called centrality. At the group-level, it is called network centralization and refers to the evenness of the communication tie distribution across network members. That is, the greater the network centralization, the more likely a single actor dominates communication in the network while the other actors are considerably less important (Wasserman and Faust, 1994). Low TMT communication network centralization, on the other hand, indicates that information access through communication ties is relatively evenly distributed among managers. In other words, the group's network is decentralized with respect to information sharing and communication.

Compared to centralization, network decentralization offers several advantages for adapting to highly uncertain environments (Burns and Stalker, 1961). Environmental uncertainty involves high complexity and rapid change in environmental issues (Duncan, 1972). A decentralized communication network is more capable of fulfilling TMTs' complex task requirements because it better promotes interdependence and cooperation among managers (Mohn, 1994). Moreover, decentralization can better match the speed of change in the environment through a broader distribution of decision-making, which provides more autonomy to TMT members by providing instant information process in response to external stimuli (Tushman, 1979). Decentralization also facilitates open communication within groups (McGrath, 1984). When ambiguous reformation from an uncertain environment is assimilated within a centralized organization, on the other hand, it first has to be channeled to the central authority, usually the CEO. This may result in an information overload on one individual while others are denied quick access to information by the network structure, which likely restricts the effectiveness of the group (Galbraith, 1973).

Higher communication network centralization tends to be associated with CEOs who want stronger control over the organization. This control preference of the CEO, when combined with limited inter-TMT communication, likely affects the relationship between the CEO and TMT subordinates by reducing the social and behavioral integration of the team (Finkelstein et al., 2009). Such circumscribed leader-follower relationships likely depress group morale and make members reluctant to express themselves, negatively affecting innovation and the free flow of creative ideas within the TMT and increasing group rigidity (Pfeffer and Salancik, 1978).

Thus, when a firm faces high environmental uncertainty, a centralized TMT communication network will be sub-optimal because of the loss of flexibility, innovativeness, and inability to quickly process reformation (Eisenhardt, 1989). Although some centralized CEOs might try to encourage risk-taking actions to induce creativity and innovation, such efforts are unlikely to be successful in centralized networks. Instead, organizations with appropriately decentralized TMT communication networks are more likely to achieve better performance in uncertain environments due to information transmission and sharing, high delegation of authority to functional managers, high interdependence and cooperation among members, and enhanced operational flexibility. Conversely, when perceived environmental uncertainty is low, strategic plans are more easily codified. There is less need for information sharing, interdependence, and cooperation because established formal procedures are sufficient for getting things done. Thus,

Hypothesis 2: As environmental uncertainty increases, concomitant increase in centrality is negatively related to performance.

METHOD

Sample

The CEOs or presidents of 32 organizations located in a major Southwestern metropolitan area were contacted and agreed to participate in this research. This sample size, although small, is comparable to those of other studies that have looked at TMT networks (e.g., Athanassiou and Nigh, 1999) because such studies require contacting all or most TMT members at each firm. The firms in the sample ranged in size from 40 to over 26,000 employees and were evenly split between the service and manufacturing sectors. These firms represent seventeen 3-digit SIC industry sectors, ranging from 1731 (electronic contracting service) to 8911 (energy consulting). Thirteen firms were public while nineteen were private. A total of 404 executives participated in this study. These TMT members had spent an average of nine years with their organizations (s.d. = 8.1) and 2.9 years in their current position (s.d. = 3.4) at the time of the study. Their titles included CEO, president, vice president, and general manager. Average size of teams is 12.5 with a range between seven and fifteen. The 32 organizations constituted a convenience sample selected because of formal or informal ties with the local university or the interviewers. Access was a primary consideration because the interview-intensive research design placed considerable demands on the time of each TMT.

Mapping TMT Communication Networks

TMT communication networks were constructed based on retrospective reports by each TMT member of recent formal and informal contacts with all other members of the TMT A study taking this approach faces several methodological challenges and limitations. The first challenge is in identifying members of the top management team. Following Amason (1996) and Simons and Peterson (2000), each CEO or president of a participating organization was asked to provide the names and titles of other executives who frequently participate in strategic decisions and are considered to be "top" managers who are involved in strategic decision-making at the top levels of their organizations.

Second, there has not been a universally adopted time span during which a researcher should document the actual paths and frequency of communication. This study chose to investigate the communications among TMT members during the two weeks immediately preceding the interview. The two-week period was chosen as a compromise because a busy manager should be able to recall contacts he or she had with other TMT members with a relatively high level of accuracy, especially when it comes to informal contacts. Nevertheless, the two-week period provides a reasonably long "snap shot" reflecting the typical communication pattern among TMT members. As such, a trade-off between potential limitations of retrospective recall and representativeness were sought in this study. Moreover, this approach has advantages over commonly used approaches that simply ask managers to rate the overall frequency of their communications with other managers (e.g., Smith et al., 1994) because this study documents specific instances of communication events among each dyad of TMT managers and each individual report can be checked for reliability with the report from the other member of the dyad. Generally, this study is more specific and detailed while being less intrusive in excavating embedded, intra-organizational communication networks.

Third, communication network researchers sometimes consider various communication media, such as face-to-face and telephone, in building communication networks of groups because different media convey information of different types and different degrees of richness. This study chose not to differentiate various media because empirical evidence in support of the influence of media on complex tasks has been largely inconsistent (see Sub, 1999; Dennis el al., 2008) and because the major interest of this study was in the overall number of communication ties and their frequency of use. Hence, each tie may encompass multiple communication media.

A multi-step process was followed to collect information for building the TMT communication networks. An initial interview was first conducted with the CEO or president of each participating organization. During this meeting the general purpose of the research was described and confidentiality was assured. Next, a team of MBA students, each of whom was enrolled in a course on the structure and internal processes of organizations, was assigned to each firm. Each team was trained in the use and coding of a structured interview form prepared specifically for this research. The MBA team then scheduled a one-hour interview with the top executive and one-hour interviews with each of the other CEO-identified TMT members. At least two team members attended each of the one-hour interviews. During the interviews, the MBA team members read questions and marked answers in a manner similar to that typical of telephone research interviews. The advantage of this approach was twofold. First, the interviewer could answer participants' questions about any scale on the questionnaire immediately. Second, and even more important, the interviewers could ensure that the mapping of communication patterns described below was performed and coded correctly for each TMT member.

Next, each TMT member in an organization was specifically asked to report the frequency of their contacts with each other TMT member through a variety of communication channels, including telephone, email, fax, individual meeting, group meeting, written documents, and others (such as video conference, shared database, lotus notes, etc.) The frequency of contacts between each pair of managers was coded into one of four levels: "4"--more than six times per week; "3"--three to five times per week; "2"--one to two times per week; "1"--less than one time per week; and "0" --never. When asymmetric contact frequency was reported by a pair of managers, the greater amount was taken as the actual communication frequency for the pair to guard against any under-reporting due to recall problems. Finally, for each organization, a communication network was constructed based on the managers' reports.

Measures

Like other social networks, a network of communication can be understood through two fundamental individual-level network characteristics: the number and the strength of ties (Granovetter, 1973). Theoretically meaningful network level features are derived from these individual elements, as described next.

TMT Communication Network Density. Wasserman and Faust (1994) suggest using an information-based measure to gauge overall network density. This study followed their recommendation by operationalizing TMT communication network density based on an information-based measure that takes into consideration both the strength and number of links contained in a network. The frequency of contacts made between each pair of managers reflects the strength of connection, while the presence of a one-to-one contact between two managers represents a tie in the communication network (Gargiulo and Benassi, 2000; Hinds and Motensen, 2005). A coraposite measure of overall network density was then calculated by dividing the sum of contacts actually made by the total contacts possible. That is, Network Density = [summation]Vk /g(g-1), where g is the number of nodes or managers and Vk is the value (recoded contact frequency between each paired managers) of kth link (communication channel). This network density measure captures the total amount of information exchange by capturing the frequency of communication that passed through the established direct communication channels.

TMT Communication Network Decentralization. Consistent with the network density measure, this study adopted an information-based approach also developed by Wasserman and Faust (1994) to measure network decentralization. The inverse of the Wasserman and Faust (1994) network centralization measure is used as an indication of network decentralization. The Wasserman and Faust (1994) network centralization measure is a group-level index derived from an individual-level information centrality index proposed by Stephenson and Zelen (1989), which measures the information centrality of an actor (i.e., communicator) in a network based on valued, non-directional links in the network. The Stephenson and Zelen (1989) index can be interpreted as the proportion of total "information" flow in a graph controlled by an individual actor. Accordingly, the group-level centralization index developed by Wasserman and Faust (1994) measures the concentration of group information- sharing, by examining the accumulated differences in the amount of information controlled by each individual actor when compared with that of the individual actor who controls the most information. In short, the greater the group-level centralization measure, the more ties and information are concentrated on a single actor in a network. The detailed mathematical procedure for calculating this index is outlined in Wasserman and Faust (1994: 195-197). Finally, the network density and network centrality indices discussed above were computed using the social network analysis software UCINET (Borgatti et al., 2002).

Environmental Uncertainty. Environmental uncertainty refers to the instability and unpredictability of environmental elements facing an organization (Duncan, 1972; Pfeffer and Salancik, 1978; Milliken, 1987). Because relationships between environmental characteristics and organizational responses are based in part on managers' decisions, which are based on their perceptions, perceptual measures of the environment should be used when studying these relationships (e.g., Boyd et al., 1993). This study, therefore, used a perceptual measure during the interviews to determine TMT members' perceptions of environmental uncertainty facing their firms. Previous empirical research on environmental uncertainty also has identified three relatively established dimensions of the construct: dynamism, complexity, and munificence (Duncan, 1972; Thompson, 1967; Dess and Beard, 1984). To save managers' time and following past research, this study adopted a simplified five-item measure of environmental uncertainty modified from Miles and Snow's Perceived Environment Uncertainty instrument (Miles and Snow, 1978). Specifically, managers were asked to rate their environmental munificence by evaluating their industry's growth rate compared with that of the general economy based on a Likert scale of 1 to 5, with 1 indicating the lowest comparable growth rate and 5 indicating the highest comparable growth. Environmental complexity was captured through the managers' evaluation of the extent of variability in the pattern of industry growth and expansion. Environmental dynamism, the most discussed component of environment uncertainty, was captured by three items asking for perceptions of industry change regarding the product life cycle, competitors, and customer preferences.

Exploratory factor analyses were conducted on the items. The results showed that a two-factor solution explained 71% of the total variance. However, the variables representing the first factor have a relatively low correlation of 0.35 (p = 0.055), while the variables representing the second factor are all highly correlated (r > = 0.50, p < 0.01). In addition, the Cronbach's alpha across all five items of environmental uncertainty was only 0.57, which is below the normal standard of 0.60 (Nunnally, 1978). Based on these results, the two items (industry growth and expansion) that composed the first factor was dropped and averaged the last three items as the final measure of environmental uncertainty. The Inter-Rater Reliabilities (IRR) for environmental uncertainty ratings among the managers for all 32 organizations ranged from 0.61 to 0.98, with an average of 0.86. The ratings of each manager in an organization were averaged in further calculations. The final Cronbach's alpha among three items (product life cycle, customer demand change, and number of competitors) was 0.75.

Organizational Performance. Organizational performance has two distinct, yet interrelated, components: efficiency and effectiveness. The measures of these two components can vary significantly for organizations. Given that the sample consisted of both service and manufacturing organizations from different industries and with different ownership types, general measures of efficiency and effectiveness were necessary for comparing performance across organizations. As a result, two commonly-used measures of performance, growth and profitability, were adopted to measure organizational performance.

Because archival performance data were unavailable for the private organizations in the sample, managers were surveyed and their assessment of the performance of their organizations was used as a subjective proxy for objective measures (Dess and Robinson, 1984). The subjective measure inevitably introduces judgmental bias. In order to minimize the possible bias, managers identified as TMT members were requested to evaluate their organization's performance from two angles: overall performance and performance compared with those of industry peers on both profitability and sales growth. Managers were asked to be as impartial and independent as possible. Also, a letter of confidentiality was signed to ensure their evaluations would not be shared with their colleagues and would not be disclosed to any third party without their permission.

Inter-rater reliability among managers in each firm on each of the performance items was very high: all were above 0.60, with most above 0.80. As a result, the ratings of all managers on each performance item were averaged for each firm. Factor analysis showed that the six performance variables loaded on three factors distinctly. Cronbach's alpha for the two growth items did not meet the normal standard of reliability (alpha = 0.575), while items of overall performance had an alpha of 0.712 and those of comparative performance had an alpha of 0.705. Based on these results, overall performance and comparative performance (versus other firms in the same industry) were used as the two distinct variables of performance. The two variables were calculated by taking the average of managers' ratings of their organizations' achievement of growth and profitability goals.

Control Variables

Besides the above variables, this study also controlled for the following variables that could confound the testing of the relationships hypothesized in this paper: CEO centrality, which is measured as the information centrality of a CEO using the Stephenson and Zelen (1989) index; TMT tenure heterogeneity, which is measured as the average combined tenure differences between the tenure in position of each of managers in a top management team and the tenures of each other managers in the group, that is, 1/n [n.summation over (j=1)] [n.summation over (i=1)] (Ti - Tj); TMT functional heterogeneity, which is measured using Blau's (1977) categorical dispersion measure. Each manager's functional background is classified into one of eight broad categories adapted from previous studies (Michel and Hambrick, 1992; Wiersema and Bantel, 1992), including CEO/President/General Manager, Consulting, Finance/Accounting, Marketing/Sales, Production/Manufacturing, R&D, Human Resources, and Planning.

ANALYSES AND RESULTS

Table 1 presents the means, standard deviations, and Pearson bivariate correlations for all variables used in this study. On average, the TMT members seem to be satisfied with their organizations' performance. The means of their ratings of both overall performance and comparative performance are above 3.0, the average performing point in the Likert scale. This suggests that the sample consists of mostly organizations with normal performance.

Following past research on contingency relationships (Alexander and Randolph, 1985; Bourgeois, 1985), and in order to conserve degrees of freedom, fit was measured as the difference between contingent variables. That is, consistent with the hypotheses, fit was measured as "matching" rather than "interaction" (Venkatraman, 1989). The two fit variables were calculated by deducting Environmental Uncertainty variable from Network Density and Network decentralization. In each case, the larger the fit variable, the better TMT network communication is in meeting the demands of environmental uncertainty.

Three sets of regression models were run to test the hypotheses. In the first set of regression models (see Table 2: Model la and Model 2a), the dependent variables were the two performance variables and the independent variables were three control variables only. In the second set of models (Model lb and Model 2b), two independent variables were added to Model la and Model 2a to test the main effects of TMT network density, decentralization, and environment uncertainty on organizational performance. Three observations were detected as outliers and were removed.

In the next set of regression models (Model 1c and Model 2c in Table 2), the fit variables were added as independent variables. Due to strong multicollinearity, the main effects were removed from the model testing the interaction effects. These regression models are intended to test the direct influence of fit on performance, or the interaction effects of different types of fit on organizational performance.

In the base models of la and 2a where only the control variables were included, CEO individual centrality had a positive and significant coefficient (p < 0.001), suggesting that greater information concentration on CEO leads to higher organizational performance. This result is consistent with correlation coefficient in Table 1.

In the second set of models, wherein the main variables were added as independent variables along with the control variables, the coefficients of TMT communication density on both performance measures are positive and significant (p < 0.01 on both performance measures). This indicates that TMT communication network decentralization and environmental uncertainty appear to have no direct influence on performance.

Hypothesis 1 predicts that the fit between TMT communication network density and perceived environmental uncertainty is positively related to performance. The hypothesis was tested by Models I c and 2c. As the results in Table 2 show, the coefficients of communication network density on both performance measures are significant (p < 0.05) and positive, consistent with what is predicted in the hypothesis. Hence, hypothesis 1 is supported by the sample.

Hypothesis 2 predicts that the fit between TMT communication network decentralization and perceived environmental uncertainty is positively related to performance. The hypothesis was tested using Models 1c and 2c as well. Interestingly, both coefficients for communication network decentralization are significant (p < 0.01 in Model 1c and p < 0.1 in Model 2c) but negative, the opposite of what is predicted in hypothesis 2. Hence hypothesis 2 is not supported. However, these counter-intuitive results are interesting and are discussed below.

DISCUSSION

Three findings emerged from this study. First, it is found that the more communication that occurs among top managers (as measured by communication network density), the greater the organization's performance. This result is in line with the argument that communication improves group performance through information sharing and through promotion of group identity and cohesion (Eisenberg and Witten, 1987). However, this finding is also in sharp contrast with the results reported in Smith el al. (1994), in which TMT communication was found to be negatively related to firm ROI and sales growth. While Smith el al. (1994) attribute the negative effect of intra-TMT communication to possible conflicts among team members, recent theoretical and empirical evidence indicates that conflict itself is a multi-dimensional construct that can have mixed influences on team performance (Amason, 1996).

Furthermore, this study found that the influence of TMT communication networks on organizational performance is contingent upon the level of environmental uncertainty lacing the firm. The better the tit is between intra-team communication density and environmental uncertainty, the better the firm's performance. This result suggests that to achieve the same level of performance in environments with various levels of uncertainty, TMTs need to maintain different levels of communication density. That is, there is no universally useful communication network design for organizations operating in environments with different levels of uncertainty.

The third finding addresses the least understood issue in TMT processes--the concentration of communication. While theorists note that intra-TMT communication attributes such as frequency, formality, and forms (Gargiulo and Benassi, 2000) at the individual level influence the performance of TMTs, research had not yet empirically investigated how the centralization patterns of TMT communication affect firm performance. While traditional contingency theory (e.g., Burns and Stalker, 1961) suggests that in uncertain environments, organizations need to decentralize information processing and decision-making, the results suggest the contrary. It was found that decentralized communication patterns had negative performance implications in more uncertain environments.

While opposite to the prediction made in this study, the negative interactive effect between network decentralization and environmental uncertainty on organization performance is plausible when the level of decision-making is considered. Baum and Wally (2003) argue that centralization of decision-making plays different roles in influencing organizational performance at different levels of managerial hierarchy. At the very top level where strategic decisions are made, centralization of decision-making is necessary to speed up decision-making and to reduce political behaviors (Eisenhardt, 1989). In contrast, at the lower level of operational decision-making, decentralization would be beneficial for motivating employees by engaging them in decision-making (Sims, 1996). Baum and Wally (2003) demonstrate that decentralization at the operational level and centralization at the strategic level increases speed in decision-making, which is positively related to firm performance in uncertain environments.

This contingent influence of centralization of decision-making on organizational performance is reasonable considering the nature of strategic decision-making. Strategy issues tend to be ambiguous, risky, and extend across multiple functional and environmental dimensions. Frequent coordination and rapid strategic decision-making requires strong leadership that can artfully integrate various sources of information and draw consensus among stakeholders with conflicting interests and goals. Concentrated communication reflects a strong leader in control who facilitates strategic decision-making. In contrast, operational decisions tend to be well-structured, repetitive, and less influential on overall organizations. Optimized decisions are more achievable when the decision alternatives and goals are easy to identify and evaluate. Under such circumstances, decentralization motivates employees to implement decisions through participative decision-making, while not neutralizing the quality of decisions. (2)

There are several limitations to this study that provide avenues for future research. First, data on the frequency of contact between managers were derived from the recollections of the managers. This raises validity concerns associated with retrospective data collection from managers such as accuracy and "critical incident" bias. Although this study followed Huber and Power's (1985) recommendations such as encouraging the managers in the sample examine their records of contacts (i.e., minutes of meeting, phone records, fax records, etc.) whenever possible to provide accurate information, this still remains a concern. Future research could enhance the validity of these measures by utilizing real time data collection methods. Furthermore, this study used subjective measures of organizational performance. Future research can increase the generalizability of the results if the relationships discussed in this paper were also found to influence other "objective" measures of performance such as earnings and stock market returns. Finally, due to the intensity of the data collection methodology, the firm-level sample size was only 32. This greatly limited the other variables that could be controlled for in the analyses. Future research using a larger sample size could consider more control variables such as CEO tenure and duality.

References

Alexander, J. W. and W. A. Randolph. 1985. "The Fit between Technology and Structure as a Predictor of Performance in Nursing Subunits." Academy of Management Journal 28: 844-859.

Amason, A. C. 1996. "Distinguishing the Effects of Functional and Dysfunctional Conflict on Strategic Decision Making: Resolving a Paradox for Top Management Teams." Academy of Management Journal 39: 123-148.

Athanassiou, N. and D. Nigh. 1999. "The Impact of Company Internationalization on Top Management Team Advice Networks: A Tacit Knowledge Perspective." Strategic Management Journal 19(1): 83-92.

Arendt, L. A., R. L. Priem, and H. A. Ndofor. 2005. "A CEO-adviser Model of Strategic Decision Making." Journal of Management 31: 680-699.

Barrick, M. R., B. H. Bradley, A. L. Kristof-Brown, and A. E. Colbert. 2007. "The Moderating Role of Top Management Team Interdependence: Implications for Real Teams and Working Groups." Academy of Management Journal 50(3): 544-557.

Balkundi, P and D. A. Harrison. 2006. "Ties, Leaders, and Time in Teams: Strong Inference About Network Structure's Effects on Team Viability and Performance." Academy of Management Journal 49(1): 49-68.

Baum, J. R. and S. Wally. 2003. "Strategic Decision Speed and Firm performance." Strategic Management Journal 24(1): 1107-1129

Blau, P. M. 1977. Inequality and Heterogeneity. Glencoe, IL: Free Press.

Bourgeois, L.J. 1985. "Strategic Goals, Perceived Uncertainty, and Economic Performance in Volatile Environments." Academy of Management Journal 28(11): 548-573.

Borgatti, S. P., M. G. Everett, and L. C. Freeman. 2002. UCINET for Windows: Software for Social Network Analysis. Analytic Technologies, Harvard.

Boyd, B., G. G. Dess, and A. Rasheed. 1993. "Divergence between Archival and Perceptual Measures of the Environment: Causes and Consequences." Academy of Management Review 1993, 18(2): 204-226

Burns, T. and G. Stalker. 1961. The Management of Innovation. Tavistock, London.

Burt, R. S. 1992. Structural Holes: The Social Structure of Competition. Cambridge, MA: Harvard University Press.

Cannella, A. A., J. H. Park, and H. O. Lee. 2008. "Top Management Team Functional Background Diversity and Firm Performance Examining the Roles of Team Member Collocation and Environmental Uncertainty." Academy of Management Journal 51: 768-784.

Carmeli, A. 2008. "Top Management Team Behavioral Integration and the Performance of Service Organizations." Group and Organization Management 33(6): 712-735.

Coleman, J. S. 1988. "Social Capital in the Creation of Human Capital." American Journal of Sociology 94: 95-120.

Dennis, A., J. Valacich, and R. Fuller. 2008. "Media, Tasks, and communication Processes: A Theory of Media Synchronicity." MIS Quarterly 32(3):575-600.

Dess, G. G. and R. B. Robinson. 1984. "Measuring Organization Performance in the Absence of Objective Measures: The Case of the Privately-held Firm and the Conglomerate Business Unit." Strategic Management Journal 5: 265-273.

--and D. W. Beard. 1984. "Dimensions of Organizational Task Environments." Administrative Science Quarterly 29: 52-73.

DeWine, S. 2001. The Consultant's Craft: Improving Organizational Communication (2nd edition). Boston: Bedford/St. Martin's.

Downs, C. W. and M. D. Hazen. 1977. "A Factor Analytic Study of Communication Satisfaction." Journal of Business Communication 14 (Spring): 63-74.

Doz, Y. L. and M. Kosonen. 2007. "The New Deal at the Top." Harvard Business Review 86(June): 1-7.

Duncan, R. B. 1972. "Characteristics of Organizational Environments and Perceived Environmental Uncertainty." Administrative Science Review 17: 313-327.

Eisenberg, E. M. and M. G. Witten. 1987. "Reconsidering Openness in Organizational Communication." Academy of Management Review 12(3):418-427.

Eisenhardt, K. M. 1989. "Making Fast Strategic Decisions in High-velocity Environments." Academy of Management Journal 32: 543-575.

Euske, N. A. and K. H. Roberts. 1987. "Evolving Perspectives in Organization Theory: Communication Implications." R. M. Jablin, L. L. Putman, K. H. Roberts, and L. W. Porter. Handbook of Organizational Communication 41-69 CA: Sage Publications.

Falcione, R. L. 1974. "Communication Climate and Satisfaction with Immediate Supervision." Journal of Applied Communication Research 2(Winter): 13-20.

Finkelstein, S., D. C. Hambrick, and A. Cannella. 2009. Strategic Leadership: Theory and Research on Executives, Top Management Teams, and Boards. Oxford, UK: Oxford University Press.

Fredrickson, J. W. 1984. "The Comprehensiveness of Decision Making Processes: Extension, Observations, Future Directions." Academy of Management Journal 27: 445-466.

Galbraith, J. 1973. Designing Complex Organizations. Reading, MA: Addison-Wesley

Garguilo, M. and M. Benassi. 2000. "Trapped in Your Own Net? Network Cohesion, Structural Holes, and the Adaptations of Social Capital." Organization Science 11(2): 183-197.

Goll, I. and A. A. Rasheed, 2005. "The Relationships between Top Management Demographic Characteristics, Rational Decision Making, Environmental Munificence, and Firm Performance." Organization Studies 26(7): 999-1023.

Granovetter, M. 1973. "The Strength of Weak Ties." American Journal of Sociology 78: 1360-1380.

Guzzo, R. A. and G. P. Shea. 1992. "Group Performance and Intergroup Relations." In M. D. Dunnette and L. M. Hough (Eds.), Handbook of Industrial and Organizational Psychology (2nd ed.), 3: 269-314. Palo.

Hansen, M. 1999. "The Search-transfer Problem: The Role of Weak Ties in Sharing Knowledge across Organization Subunits." Administrative Science Quarterly 44: 82-111.

Hinds, P. and M. Motensen. 2005. "Understanding Conflict in Geographically Distributed Teams: The Moderating Effects of Shared Identity, Shared Context, and Spontaneous Communication." Organization Science 16(3): 290-307

Huber, G. and D. Power. 1985. "Retrospective Reports of Strategic-level Managers: Guidelines for Increasing their Accuracy. Strategic Management Journal 6(2): 171-180.

Ibarra, H. 1993. "Network Centrality, Power, and Innovation Involvement: Determinants of Technical and Administrative Role." Academy of Management Journal 36(3): 471-502.

Johnson, J. D. 1990. "Effects of Communicative Factors on Participation in Innovations." Journal of Business Communication 27: 7-24.

Johnson, J. D., W. A. Donohue, C. K. Atkin, and S. Johnson. 2001. "Communication, Involvement, and Perceived Innovativeness--Tests of a Model with Two Contrasting Innovations." Group and Organization Management 26(1): 24-52.

Krackhardt, D. and J. R. Hanson. 1993. "Informal Networks: The Company behind the Chart." Harvard Business Review 71 (4): 104-111.

Larson, J. R., C. Christensen, T. M. Franz, and A. S. Abbott. 1998. "Diagnosing Groups: The Pooling, Management, and Impact of Shared and Unshared Case Information in Team-based Medical Decision Making." Journal of Personality and Social Psychology 75: 93-108.

Li, J. and D. C. Hambrick. 2005. "Factional Groups: A New Vantage on Demographic Faultlines, Conflict, and Disintegration in Work Teams." Academy of Management Journal 48: 794-813.

McGrath, J. E. 1984. Groups: Interaction and Performance. Englewood Cliffs, NJ: Prentice-Hall.

Michalisin, M. D., S. J. Karau, and C. Tangpong. 2004. "Top Management Team Cohesion and Superior Industry Returns--An Empirical Study of the Resource-Based View." Group and Organization Management 29(1): 125-140.

Michel, J. and D. C. Hambrick. 1992. "Diversification Posture and Top Management Team Characteristics." Academy of Management Journal 35: 9-37.

Miller, D. 1992. "Environment Fit versus Internal Fit." Organization Science 3: 159-178.

Miles, R. E. and C. C. Snow. 1978. Organizational Strategy, Structure, and Process. New York: McGraw-Hill Book Co.

Milliken, F. J. 1987. "Three Types of Perceived Uncertainty about the Environment: State, Effect, and Response Uncertainty." Academy of Management Review 12: 133-143.

Mintzberg, H. 1973. The Nature of Managerial Work. Englewood Cliffs, NJ Prentice-Hall.

Molm, L. D. 1994. "Dependence and Risk: Transforming and Structure Of Social Exchange." Social Psychology Quarterly 57: 163-176.

Monge, P. R. and M. E. Eisenberg. 1987. "Emergent Communication Network." in R. M. Jablin, L. L. Putman, K. H. Roberts, and L. W. Porter Handbook of Organizational Communication. CA: Sage Publications. pp. 304-342.

Nunnally, J. C. (1978). Psychometric Theorv (2nd ed.). New York: McGraw-Hill.

Paldam, M. 2001. "Social Capital: One or Many? Definition and Measurement." In S. Sayer Issues in New Political Economy. Oxford, UK: Blackwell Publishers Ltd. pp. 117-141.

Pfeffer, J. and G. R. Salancik. 1978. The External Control of Organizations. New York: Harper and Row.

Reagans, R. and E. W. Zuckerman. 2001. "Networks, Diversity, and Productivity: The Social Capital of Corporate R&D Teams." Organization Science 12: 502-517.

Shaw, M. E. 1964. "Communication Networks." In L. Berkowitz (Ed.) Advances in Experimental Social Psychology. New York: Academic Press. Vol 1. pp. 111-147.

Simons, T. L. and R. S. Peterson. 2000. "Task Conflict and Relationship Conflict in Top Management Teams: The Pivotal Role of Intra-group Trust." Journal of Applied Psychology 85: 102-111

Sims, H. P. and C. C. Manz. 1996. Business Without Bosses. Wiley: New York.

Smith, K. G., K. A. Smith, J. D. Olian, D. P. O'Bannon, and J. A. Scully. 1994. "Top Management Team Demography and Process: The Role of Social Integration and Communication." Administrative Science Quarterly 39: 412-438

Stephenson, K. and M. Zelen. 1989. "Rethinking Centrality: Methods and Applications." Social Networks 11: 1-37

Stevenson, W. B. and M. C. Gilly. 1991. "Information Processing and Problem Solving: The Migration of Problems through Formal Positions and Networks of Ties." Academy of Management Journal 34: 918-928

Suh, K. S. 1999. "Impact of Communication Medium on Task Performance and Satisfaction: An Examination of Media-Richness Theory." Information and Management 35(5): 295-312.

Tekleab, A. G., N. R. Quigley, and P. E. Tesluk. 2009. "A Longitudinal Study of Team Conflict, Conflict Management, Cohesion, and Team Effectiveness." Gwup and Organization Management 34(2): 170-205.

Tushman, M. L. 1979. "Work Characteristics and Sub-Unit Communication Structure: A Contingency Analysis." Administrative Science Quarterly 24(1): 82-98.

Thompson, J. D. 1967. Organizations in Action. New York: McGraw-Hill

Venkatraman, N. 1989. "The Concept of Fit in Strategy Research: Toward Verbal and Statistical Correspondence." Academy of Management Review 14: 423-444.

Wasserman, S. and K. Faust. 1994. Social Network Analysis. New York: Cambridge University Press.

Weirsema, M. F. and K. A. Bantel. 1992. "Top Management Team Demography and Corporate Strategic Change." Academy of Management Journal 35: 91-121.

Yasai-Ardekani, M. and P. Nystrom. 1996. "Designs for Environmental Scanning Systems: Tests of a Contingency Theory." Management Science 42: 187-204.

Xin Liang

Assistant Professor

University of Minnesota Duluth

Hermann Achidi Ndofor

Assistant Profesor

Texas A&M University

Richard L. Priem

Professor

The University of Wisconsin--Milwaukee

Joseph C. Picken

Clinical Profesor

The University of Texas at Dallas

(1) We would like to thank Drs. Paul Nystrom, Edward Levitas, Vincent Barken Phd seminar attendees at the University of Wisconsin-Milwaukee, and especially two anonymous JMI reviewers and JMI editor Dr Bienvenido Cortes for their valuable comments on earlier versions of this manuscript. However; all errors are our own.

(2) It is possible that the relationships identified in this study are also subject to the influence of CEO power. The results of post hoc tests show that after removing CEO centrality variable as a control variable from the regression models, the influence of environmental uncertainty on the relationship between TMT communication network characteristics and performance largely stay unchanged, except for the contingency relationship related to OVERALL PERFORMANCE and TMT communication network density. These results seem to provide some support for the influence of CEO power on TMT information processing, although to a moderate extent.
Table 1
Means, Standard Deviations, and Correlation Coefficients

                                   Mean     St. D.      1

1. TMT Tenure Heterogeneity       3.40      2.3
2. CEO Centrality                 0.10      0.02       0.02
3. TMT Functional Heterogeneity   0.56      0.21       0.03
4. Overall Performance            3.29      0.68      -0.13
5. Comparative Performance        3.46      0.81       0.17
6. Environmental Uncertainty      2.42      0.54       0.22
7. TMT Communication Network
   Density                        4.48      1.33       0.02
8. TMT Communication Network
   Decentralization               6780.30   4644.33   -0.08

                                        2             3        4

1. TMT Tenure Heterogeneity
2. CEO Centrality
3. TMT Functional Heterogeneity   -0.11
4. Overall Performance             0.35 ([dagger])   -0.02
5. Comparative Performance         0.17              -0.29     0.63 **
6. Environmental Uncertainty      -0.14               0.12     0.13
7. TMT Communication Network
   Density                        -0.47 **           -0.23     0.20
8. TMT Communication Network
   Decentralization               -0.50 **            0.06    -0.15

                                    5           6             7

1. TMT Tenure Heterogeneity
2. CEO Centrality
3. TMT Functional Heterogeneity
4. Overall Performance
5. Comparative Performance
6. Environmental Uncertainty       0.05
7. TMT Communication Network
   Density                         0.20    -0.03
8. TMT Communication Network
   Decentralization               -0.18    -0.35 ([dagger])   0.29

([dagger]): p < 0.1, *: p < 0.05, **: p < 0.01. All tests are
two-tailed.

Table 2
Hypothesis Testing and Results of Regression Analysis

                                          Overall Performance

                                  Model 1a  Model 1b  Model 1c

Constant                           2.33 *    0.96 **      1.92 **
TMT Tenure Heterogeneity          -0.04     -0.07        -0.05
CEO centrality                    11.12 **  23.80 **     14.49 **
TMT Functional Heterogeneity       0.06      0.47        -0.03
TMT Comm. Network Density                    0.38 *
TMT Comm. Network Centralization             0.11
Environmental Uncertainty (EU)               0.23
Fit between Network Density and
Environmental Uncertainty                                 0.32 *
Fit between decentralization and
Environmental Uncertainty                                -0.46 **
Adjusted R2                        0.05      0.28         0.38
Adjusted R2 change                           0.23 *       0.21 * (n)
F                                  0.23      2.99 *       5.29 **
N                                 32        32           29

                                       Comparative Performance

                                  Model 2a   Model 2b  Model 2c

Constant                           4.08 **    3.58 **   3.83 **
TMT Tenure Heterogeneity          -0.08      -0.09     -0.09
CEO centrality                    -0.73       2.54     -0.05
TMT Functional Heterogeneity      -0.241      0.22      0.19
TMT Comm. Network Density                     0.30 *
TMT Comm. Network Centralization             -0.13
Environmental Uncertainty (EU)               -0.01
Fit between Network Density and
Environmental Uncertainty                               0.37 *
Fit between decentralization and
Environmental Uncertainty                              -0.31 ([dagger])
Adjusted R2                        0.00       0.09      0.12
Adjusted R2 change                            0.09      0.12 *
F                                  0.88       2.00      2.89 ([dagger])
N                                 29         29        29

([dagger]) p < 0.1, * p < 0.05, ** p < 0.01. All tests are two-tailed.

Note: Adjusted R square change and significance test are between
model 1a and model 1a based on 29 observations.
COPYRIGHT 2010 Pittsburg State University - Department of Economics
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2010 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Liang, Xin; Ndofor, Hermann Achidi; Priem, Richard; Picken, Joseph C.
Publication:Journal of Managerial Issues
Article Type:Brief article
Geographic Code:1U2NY
Date:Dec 22, 2010
Words:8474
Previous Article:Investor reactions to substitution-based outsourcing agreements.
Next Article:Incumbents adaptation to competence-destroying change: role of prior experience and knowledge sourcing.
Topics:

Terms of use | Privacy policy | Copyright © 2020 Farlex, Inc. | Feedback | For webmasters