Transformational leadership in an acquisition: A field study of employees

Transcription

1 The Leadership Quarterly 18 (2007) Transformational leadership in an acquisition: A field study of employees Louise A. Nemanich a,, Robert T. Keller b a School of Global Management and Leadership, Arizona State University, PO Box 37100, Phoenix AZ , USA b C. T. Bauer College of Business, University of Houston, 334 Melcher Hall, Houston, Texas , USA Abstract This field study of employees involved in a major acquisition integration addressed the relationships that leadership and climate had with subordinate acquisition acceptance, performance, and job satisfaction in an uncertain environment. Transformational leadership was positively related to acquisition acceptance, supervisor-rated performance, and job satisfaction (p <.01). Transformational leaders also impacted subordinate outcomes through the perceived climate they created for goal clarity and support for creative thinking. Both goal clarity and support for creative thinking partially mediated the relationship between transformational leadership and job satisfaction (p <.01; N = 447). Goal clarity did not mediate the relationship between transformational leadership and performance, but was positively related to performance (p <.05; N = 344). Support for creative thinking fully mediated the relationship between transformational leadership and acquisition acceptance (p <.01; N = 432). Implications for future research and for managers engaged in acquisition integrations are discussed Elsevier Inc. All rights reserved. Keywords: Transformational leadership; Mergers and acquisitions; Goal clarity; Creativity 1. Introduction Successfully managing employee attitudes and performance during an acquisition integration is vital because employees can choose to: leave the firm or remain, hold onto critical operational knowledge or share it, and raise their productivity level or become distracted by disruptions (Buono & Bowditch, 1989; Marks & Mirvis, 1992; Ranft & Lord, 2000). Acquisition integrations call upon employees to accelerate their productivity to manage routine job responsibilities, plus take on the additional tasks necessary to transition from two organizations to one. Yet resistance to changes in routines and uncertainty about responsibilities can reduce job performance at a critical time. Cultural changes and degradation of status can cause social identity issues that lead to frustration and anger about the acquisition and reduced job satisfaction (Hambrick & Cannella, 1993; Hogg & Terry, 2000). Unless managed successfully, these employee-level outcomes of an acquisition can manifest themselves at the firm level as an exodus of talent, tardiness, absenteeism, lower productivity, reduced customer satisfaction, less innovation, and ultimately reduced economic benefits from the acquisition (Buono & Bowditch, 1989, Ernst & Vitt, 2000; Hambrick & Cannella, 1993; Nygaard & Dahlstrom, 2002; Schweiger & DeNisi, 1991). Corresponding author. address: Louise.Nemanich@asu.edu (L.A. Nemanich) /$ - see front matter 2006 Elsevier Inc. All rights reserved. doi: /j.leaqua

2 50 L.A. Nemanich, R.T. Keller / The Leadership Quarterly 18 (2007) Bass (1985) set out a model of situational antecedents for transformational leadership and reiterated the importance of contextual antecedents in later work (Bass, 1998; Bass, Avolio, Jung, & Berson, 2003). In this model, transformational leadership is particularly effective in environments characterized by change, uncertainty, and distress, such as acquisition integrations (Bass, 1998; House & Aditya, 1997; Waldman, Ramirez, House, & Puranam, 2001; Yukl & Howell, 1999). Transformational leaders are essentially change agents; they visualize a future different than the status quo and inspire subordinates to work with them to achieve that new future (Vera & Crossan, 2004). In an acquisition integration, transformational leadership behaviors play a critical role in helping subordinates to accept that a bright future lies in an integration of the two organizations. By inspiring employees to work toward that future, moreover, they motivate employees to maintain their job satisfaction and performance despite the uncertainty and anxiety of the integration process. Our study makes an important empirical contribution to the transformational leadership literature because few studies have been conducted outside of stable environments to investigate the effects of transformational leaders on employees during organizational change (Bass, Avolio, Jung & Berson, 2003). The present study addresses this gap by answering this research question: In an acquisition integration, does transformational leadership have a beneficial relationship with employee acquisition acceptance, performance, and job satisfaction? Another gap in the transformational leadership literature is that it offers few insights into the mechanisms by which these effects take place (House & Aditya, 1997; Yukl, 2002). Some recent research has focused on understanding better how transformational leaders affect subordinate outcomes by looking at mediating mechanisms (Avolio, Zhu, Koh, & Bhatia, 2004; Jung & Avolio, 2000; MacKensie, Podsakoff, & Rich, 2001; Pillai, Schriesheim, & Williams, 1999; Zhu, Chew, & Spangler, 2005). In one study, transformational leaders indirectly affected subordinate safety behaviors through the type of climate they created (Barling, Loughlin, & Kelloway, 2002). We extend this work on climate as a mediating mechanism between transformational leadership and employee outcomes by analyzing the indirect use of transformational leadership behaviors during periods of rapid change to create climates emphasizing goal clarity and support for creative thinking. These climates have the potential to further mitigate the effects of uncertainty and change during acquisition integrations. By creating a climate emphasizing goal clarity, transformational leaders can alleviate ambiguity and thereby help employees to achieve their objectives. Transformational leaders also may promote a climate of creativity, thereby opening employees' minds to new ways of thinking and enabling them to better understand the need for change. Hence, our second research question is: Does transformational leadership also have an indirect relationship with employee acquisition acceptance, performance, and job satisfaction through the mediating mechanism of creating a climate that individuals perceive as providing goal clarity and supporting creative thinking? 2. Transformational leadership in acquisition integrations Transformational leadership is one of a class of theories known as neocharismatic leadership theories (House & Aditya, 1997). Transformational leadership is defined as a relationship between a leader and followers based on a set of leader behaviors perceived by subordinates as exhibiting idealized influence, motivational inspiration, intellectual stimulation, and individual consideration (Bass, 1985; Waldman, Javidan, & Varella, 2004). Extensive research evidence shows that transformational leadership improves subordinate satisfaction with the leader and subordinate perceptions of leader effectiveness (Howell & Shamir, 2005; Judge & Piccolo, 2004; Lowe, Kroeck, & Sivasubramaniam, 1996). A smaller body of work has studied transformational leadership and its relationship with variables such as subordinate job satisfaction, often with mixed results and smaller effect sizes (Judge & Bono, 2000; Podsakoff, MacKensie, & Bommer, 1996) Transformational leadership in the context of organizational change Transformational leadership theory postulates a contextual dependence with the emergence and effectiveness of transformational leadership being stronger in situations of crisis or uncertainty, such as acquisition integrations (Bass, 1990; Shamir, House, & Arthur, 1993; Yukl & Howell, 1999). Transformational leaders help subordinates to unlearn past routines, develop creative solutions to ambiguous problems, and respond appropriately to new environments (Bass, 1985; Bass et al., 2003; Vera & Crossan, 2004). There has been limited empirical transformational leadership research in an acquisition integration, or similar contexts with high uncertainty. Waldman et al. (2001) found that

3 L.A. Nemanich, R.T. Keller / The Leadership Quarterly 18 (2007) charismatic leadership by the CEO was positively related to firm performance in an environment of uncertainty, but not in more stable situations. Pillai & Meindl (1998) found that in crisis situations transformational leadership behavior tended not to emerge. In a study of employee attitudes toward a merger conducted several years after the integration process, charismatic leadership behaviors were positively related to employee satisfaction with the merger (Covin, Kolenko, Sightler, & Tudor, 1997) Direct effects of transformational leadership The first part of our model (Fig. 1) focuses on the direct relationships between transformational leadership behaviors and subordinate outcomes. Individuals enter an acquisition integration with their own cognitive frames, based upon their previous experience, that organize meaning, motivation, and cause effect relationships (Drazin, Glynn, & Kazanjian, 1999). They share commonalities in these frames with coworkers from the same prior employer, but not with coworkers from the other partner firm. Employees' pre-existing cognitive frames also may overlap very little and may even conflict with their leaders' frames representing the new reality of the merged firm. This cognitive dissonance, augmented by the uncertainty and disruptive change of the integration process, increases the influence and importance of leader behaviors (Nahavandi, 1993). Transformational leaders use idealized influence to empower followers, thereby raising their tolerance for uncertainty and their ability to adapt to new, changing conditions (Bass, 1998). The powerful communication skills associated with idealized influence can be used in public and private meetings with subordinates to augment leaders' abilities to help employees understand the benefits of acquisition (Maitlis, 2005). Through intellectual stimulation, leaders encourage subordinates to question the universality of previous cognitive frames, opening the door for new frames to develop (Vera & Crossan, 2004). By considering each subordinate as an individual, transformational leaders can provide support through the change process by facilitating social reconstruction to bring more uniform interpretations to people with separate experience bases (Bass, 1998; Gioia & Chittipeddi, 1991) Acquisition acceptance Acquisition acceptance is the extent to which an employee has a positive attitude about the acquisition and perceives it as beneficial. Acquisition acceptance is an important organizational outcome because employees who are more satisfied with an organization's change strategy, such as an acquisition, are likely to adjust more quickly to the new culture and strategies that the firm is trying to implement. Resistance to the acquisition is an underlying cause of unproductive behaviors ranging from reduced productivity to sabotage (Marks & Mirvis, 1992). In order for employees to be satisfied with the acquisition, they need to be able to understand the purpose for the acquisition and believe that this purpose is worthy of the disruption the acquisition causes. They need to be able to relate to the new vision for the future of the combined firm, and find it to be a desirable objective (Bass, 1998). The skills of a transformational leader facilitate this process by elucidating the vision and socially constructing common frames that are the groundwork for consensus building on a mutual sense of purpose. Leaders' skills at communicating the vision to employees and inspiring support for it provide the basis for building employee understanding of the business case for change. In support of these arguments, Covin et al. (1997) found that charismatic leadership was positively related to acquisition acceptance for two subsets of their respondents, managerial employees, and target firm employees. Hence, Fig. 1. Leadership in acquisition integration model.

4 52 L.A. Nemanich, R.T. Keller / The Leadership Quarterly 18 (2007) H1a. In an acquisition context, transformational leadership will be positively related to acquisition acceptance Performance Job performance, broadly speaking, encompasses both what employees do and how they do it. It is an assessment of the extent of an employee's accomplishments of the goals established by the organization and the acceptability of the employee's interpersonal behaviors relative to the norms of the organization. Transformational leaders inspire followers to perform beyond normal expectations. For example, transformational leadership positively affected the performance of teams in a longitudinal study (Keller, 2006) and of subordinates in a cross-sectional study (Whittington, Goodwin, & Murray, 2004). Dvir, Eden, Avolio, & Shamir (2002) found that transformational leadership had an indirect impact through a layer in the hierarchy on the performance of followers in an Israeli military field experiment. We propose that transformational leadership will have positive effects on individual performance during acquisition integration. By providing a vivid, comprehensible vision of the future for the merged firm, transformational leaders can motivate subordinates to help achieve the vision through their accomplishments. By building employees' identification with the new organization, transformational leaders motivate employees to engage in extra effort for the common good of the organization (Bono & Judge, 2003; Shamir et al., 1993). By setting an example of individualized consideration, transformational leaders model positive interpersonal behaviors that reduce conflict and, thereby, enhance productivity in the workplace. Therefore, H1b. In an acquisition context, transformational leadership will be positively related to subordinate performance Job satisfaction Transformational leaders are thought to enhance the job satisfaction of their subordinates by making employees feel special (through individualized consideration) and by making employees feel that they are called to a higher purpose (through idealized influence and inspiration motivation) (Bass, 1985). Prior research on the relationship between transformational leadership and job satisfaction, however, has yielded inconsistent results. While some studies have shown a significant positive relationship between transformational leadership and job satisfaction (Berson & Linton, 2005; Bono & Judge, 2003; Podsakoff et al., 1996), other studies have not found a significant relationship (Judge & Bono, 2000). In an acquisition situation, we expect transformational leadership to have a positive effect on job satisfaction. Through idealized influence, followers are motivated to adopt the leader's enthusiasm for conforming to the changes of the newly merged firm. Individualized consideration makes employees feel that they are valued and that their need to understand and resolve their personal uncertainties about the integration is respected (Bass, 1998). Transformational leaders make subordinates feel as if they are called to work toward a valuable purpose, such as building a new, larger firm (Bass, 1985). The effect of these transformational behaviors is that employees will be more satisfied with their jobs because they believe that they are doing important work for leaders who value their contributions. Thus, H1c. In an acquisition context, transformational leadership will be positively related to job satisfaction Mediation effects Bass (1985) posited an organic organization climate as a situational antecedent of transformational leadership that we have formulated as an intervening or mediating variable in the acquisition context. Hence, the second part of our theoretical model (Fig. 1) hypothesizes that, in addition to the direct effect that transformational leadership behaviors have on employees, transformational leaders also indirectly influence subordinate outcomes through the perceived climate that they create. This is a partial mediation model in which transformational leadership directly affects subordinate outcomes and also acts through an intervening climate variable (Howell, Dorfman, & Kerr, 1986). We adopt the cognitive schema approach that views climate as a microlevel property of the individual and defines it as individuals' cognitive representation of their environment (Anderson & West, 1998; Ashforth, 1985; Ragazzoni, Baiardi, Zotti, Anderson, & West, 2002). Leaders can influence climate formation, first, by holding a set of assumptions themselves and, then, by communicating them, engaging in symbolism, and inspiring consistent behaviors among their followers (Ashforth, 1985; Schein, 1997). In this paper, we focus on two climate characteristics

5 L.A. Nemanich, R.T. Keller / The Leadership Quarterly 18 (2007) that facilitate the resolution of uncertainty and adaptation to change during an acquisition integration: a climate of goal clarity, and a climate that supports the adoption of new ideas and new ways of doing things Goal clarity The unstable task processes, organization structures, and individual roles that characterize an acquisition integration lead to a lack of goal clarity (Ranft & Lord, 2000). By communicating their own motivation and enthusiasm for the new vision, leaders instill in their followers a motivation to change toward that vision and enthusiasm for attaining it (Conger & Kanungo, 1987; Struckman & Yammarino, 2003). Subordinate motivation to achieve the leader's vision provides the impetus to build a social construction of the goals that are critical to vision attainment. These behaviors lead to a climate that emphasizes team efforts to develop clearly understandable goals. The vision of a transformational leader is an image of an idealized, future state for the overarching organization that is substantially discrepant from the status quo (Conger & Kanungo, 1987; Rafferty & Griffin, 2004). Although a positive vision of future opportunity and growth is important for building a convincing case for change to justify an acquisition, employees' personal acceptance of the benefits of the acquisition integration are colored by their central uncertainty issue; that is, what does this integration mean to me? Therefore, employees in a context of high uncertainty, such as an acquisition, benefit from the ability to connect the transformational leader's distal organizational vision to proximate individual roles and goals. By creating a climate of goal clarity, transformational leaders assist subordinates to visualize how the new world of the integrated organization will help them meet their personal goals. Through this mechanism of connecting an organizational vision to clear goals, transformational leaders are able to build a clearer, positive perspective for employees of the benefits of the acquisition, thereby increasing acceptance. Hence, H2a. In an acquisition context, a climate of goal clarity will mediate the relationship between transformational leadership behaviors and subordinate acquisition acceptance. Employee performance is evaluated by comparing employee accomplishments and behaviors with standards established by the leader or by the organization. Thus, performance should be enhanced in contexts where mutual understanding leads to a high level of clarity and congruence between the perceptions of the leader and subordinates on the standards to be achieved. With an unambiguous understanding of goals, performance assessments by leaders will be more clear-cut. If subordinates have a clear understanding of their own individual objectives they will be more likely to achieve them. A recent meta-analytic study offers some support for these arguments with a finding of a negative relationship between role ambiguity, which includes an element of goal ambiguity, and employee performance (Tubre & Collins, 2000). Thus, H2b. In an acquisition context, a climate of goal clarity will mediate the relationship between transformational leadership behaviors and subordinate performance. Transformational leaders can positively influence subordinate job satisfaction by creating a vision for the new firm that makes employees feel that they are called to a higher purpose. They can further enhance employee job satisfaction by creating a climate emphasizing goal clarity. Goal clarity leads to higher job satisfaction because employees have a greater sense of task self-efficacy when faced with clear personal goals than with ambiguous goals (Arvey & Dewhirst, 1976; Dvir et al., 2002). A climate of goal clarity plays a mediating role between transformational leadership and job satisfaction by translating the leader's idealized vision into individual task objectives, thereby providing an additional source of job satisfaction. Thus, H2c. In an acquisition context, a climate of goal clarity will mediate the relationship between transformational leadership behaviors and subordinate job satisfaction Support for creative thinking A creative climate is a psychological perception that the work environment includes organizational and supervisory encouragement of new ideas and new ways of approaching work (Amabile, Conti, Coon, Lazenby, & Heron, 1996). A creative climate has influence beyond supporting ideas for new products in R&D; it also facilitates renewal in processes, such as when an accounting department revolutionizes its operations by converting its systems to enterprisewide software, e.g. SAP. Leaders can develop such a climate in their work group by establishing work processes that

6 encourage employees to be open to new ideas and by involving them in the decision-making process (Amabile, Schatzel, Moneta, & Kramer, 2004; Shalley & Gilson, 2004; Carr, Schmidt, Ford, & DeShon, 2003). Transformational leaders build a creative climate by processes such as questioning the status quo, stimulating followers to question the critical assumptions in their previous cognitive frames, and suggesting new ways of looking at work processes (Bass, 1985; Elkins & Keller, 2003; Vera & Crossan, 2004). Supporting this thinking, transformational leadership was found to be positively related to a creative culture and to follower creativity; it was also more effective when projects required more innovative thinking (Jung, Chow, & Wu, 2003; Keller, 1992; Shin & Zhou, 2003).

7 L.A. Nemanich, R.T. Keller / The Leadership Quarterly 18 (2007) The host firm limited our sampling frame to minimize disruption to their organization. In negotiating the selected respondents, our goals were to achieve a sampling frame that included employees located throughout the United States and engaged in a broad range of tasks generally representative of the organization. The resulting sampling frame included sales, research and development, customer service, and finance departments (N=919). Customer service and finance employees were located in a divisional headquarters office, research employees were located in a research laboratory campus, and sales employees were located throughout all 50 states. This sampling frame provided an attractive opportunity to test our model because the variety of settings and job content offered considerable variance in leadership style and climate. The integration involved extensive changes to job roles and/or job content for all employees in the sampling frame. Two instruments were used in the study: an employee survey and a supervisory performance rating form. Questionnaire data were collected from employees five months after the acquisition occurred. Employee survey instruments collected data on perceptions of the following: his/her immediate supervisor's transformational leadership behaviors, the emphasis on goal clarity and support for creative thinking in the work context, and his/her personal feelings of acquisition acceptance and job satisfaction. These procedures yielded 453 responses for a 52% response rate. After list-wise deletion of respondents with missing data for the independent measures, there were 447 useable surveys that comprised the sample for this study. The respondents were 61% male, 58% college degreed, and 58% with over ten years of tenure. Employees previously employed by the target firm comprised 65% of the respondent group. Supervisors were asked to evaluate each of their employee's performance using a separate instrument one month after the employee questionnaire data were collected. We requested these data a few weeks after supervisors were required to prepare the same performance assessments on a formal basis for compensation review. Performance data were requested for 447 employees, and supervisors returned completed data for 344, a 77% response rate by supervisors, and the sample size for performance data analyses. The resulting large sample size reduced the statistical validity threat of low power. Power analyses indicated power >.99 to detect a medium effect size (.13) for all three dependent variables. The models for acquisition acceptance and job satisfaction had a.80 power to detect an effect size of.04, while the performance model had a.80 power to detect an effect size of Sample context During seven on-site visits before and during data collection, an author engaged employees from the CEO to entry level new employees in unstructured discussions about the acquisition integration. The employee survey instruments also enabled us to collect written qualitative data through an open-ended invitation to provide comments. Many employee comments about the effectiveness of the overall integration suggested that the process had been managed fairly well: [acquiring firm] did everything possible to relieve the apprehension of the [target firm] employees, customers have not been inconvenienced by product shortages or delays, and the two companies have come together quite well. Other employees, however, were less positive: we are not able to perform at as high a level as before the merger and the acquisition has been the worst experience. Nevertheless, the two strongest themes that emerged were that the integration entailed extensive, disruptive change, and that this change created a challenging context of uncertainty. The extent of the integration involved massive changes in all parts of the organization. Based on discussions with managers, initial staff reductions were estimated to have been 20% 30%. Surviving employees experienced changed workloads saying, the current workload with the company is too much, and our employees are feeling the strain. Numerous offices (including the acquiring division headquarters office and the target firm R&D lab) and manufacturing plants were closed, requiring in some cases acquiring firm and in other cases target firm employees to relocate. A single value system, compensation system, and set of work processes were selected for the newly combined firm, which required employees (in some cases those from the target firm and in others those from the acquiring firm) to adjust to a new culture, accept changes in benefit programs, and learn new competencies. Some employees did not welcome these changes, [the firm other than the one I previously worked for] seems to dominate in the corporate culture and it's not a step in the right direction, and during a time when employees are asked to do more, don't pick that time to take away a very valued employee benefit. Integration of the product lines and customer bases necessitated that all employees learn new customers and products. Work practices at both previous firms were highly dependent upon software applications for customer service, sales, budgeting, and accounting. Development of new systems lagged the integration of jobs, hampering employee effectiveness: the fact that our IT systems do not

8 56 L.A. Nemanich, R.T. Keller / The Leadership Quarterly 18 (2007) speak to one another makes my job very difficult. One employee painted a vivid visualization of an overall sense of disruptive change with the comment, The organization feels like a lot of running with our hair on fire and lack of processes. At the time of our survey, most remaining employees had been offered future employment. Uncertainty remained, however, about such issues as which work practices would be used, what quality of relationship would be formed with new colleagues, and what future strategies would be pursued. Employees commented on the general sense of uncertainty by writing, We have gone through months of not knowing what to do about some situations. In looking for more specific sources of uncertainty and how they impact employee perceptions, we found that ambiguity about future firm strategy left employees with a lack of clarity about priorities for decision-making. An employee pointed up one such challenge by stating, There are two different mindsets; [acquirer] is volume over profits and [target] is profits over volume. Also, with so many new positions created and so many tasks reassigned, individuals perceived uncertainty about what exactly they should be doing. For example, employees wrote, My position is new, so the job functions and goals have not been clearly stated, and There are too many No.1 priorities Measures Transformational leadership Transformational leadership was measured using all twenty items (encompassing the dimensions of idealized influence, inspirational motivation, intellectual stimulation, and individualized consideration) from the Multifactor Leadership Questionnaire (MLQ) (Bass & Avolio, 2000), which is the most widely used measure of transformational leadership (Judge & Bono, 2000). The psychometric properties of this set of measures have been extensively validated in numerous large sample studies in business, government, and the military (Bass & Avolio, 2000; Tejeda, Scandura, & Pillai, 2001). Consistent with other researchers, we found the four dimensions to be highly correlated (bivariate correlations ranged from.71 to.86), representing one higher order factor of transformational leadership (Bass & Avolio, 2000; Bono & Judge, 2003). A factor analysis of the twenty items yielded only two factors with eigenvalues greater than 1.0 (11.0 and 1.0), and a single factor accounted for 55% of the variance. Therefore, our analyses utilize only the higher order transformational leadership factor composed of the average of all 20 items. A five-point Likert response scale was used for this and all other attitudinal measures. Cronbach's alpha was The high reliability for the MLQ scale is consistent with large sample results by Bono & Judge (2003) (N=954, alpha for 20-item transformational leadership =.94) and Bass & Avolio (2000) (N = 2154, alpha for each separate 4-item dimension ranges from.86 to.90). We also checked our sample means for each dimension of transformational leadership against those of Bass & Avolio's (2000) group of samples (N = 2154) and found no significant differences (p>.999) Climate Employees were asked to share their perceptions of what it is like to work in your team. Team was defined as the employee's supervisor and all of the people who report to him or her. Emphasis on goal clarity was measured using a three-item scale that had been previously developed and tested with a sample of 51 teams (Edmondson, 1999). Cronbach's alpha in that study was 0.82 and discriminant validity was confirmed with a multi-trait multi-method analysis. A sample item is The team has invested plenty of time to clarify our goals. Cronbach's alpha in the present study also was Individuals also were asked about the degree of support for creative thinking they perceived in their work environment using a new four-item scale developed for this study (see Table 1). Items were developed based on a review of the literature. Face validity was tested through a process of review first with an academic colleague experienced in the field of organizational learning and then with a focus group comprised of MBA students (N = 35). The measures were further refined in two pilot tests with employees of the host firm. (Pilot 1 respondents worked in supply chain management in the division impacted by the acquisition (N=103) and Pilot 2 respondents worked in a division unaffected by the acquisition (N = 80); pilot data were not incorporated in the analyses reported herein). Two scale items assessed to what extent employees felt that they were personally encouraged to do things in new ways (both incremental change and experimentation). Two items assessed employees' broader perceptions that the organization rewarded new ideas and took action to implement them. A sample item is I am encouraged to experiment with new ways to get my work done. Cronbach's alpha was 0.71.

9 L.A. Nemanich, R.T. Keller / The Leadership Quarterly 18 (2007) Table 1 Scale items and principal component analysis item loadings for new measures Acquisition acceptance A majority of the employees have come to accept this acquisition as necessary and worthwhile..74 My department has been strengthened by this acquisition..79 The policies and procedures of the combined company are improved compared to my previous company..72 The combined company has more to offer our customers compared to my previous company..69 Support for creative thinking This organization rewards creative thinking..70 I am encouraged to experiment with new ways to get my work done..70 I am encouraged to make incremental adjustments to improve the way we do things here..72 New ideas frequently get converted into new products, services, or processes in this organization Control variables Change often has a greater effect on members of the target firm than members of the acquiring firm in an acquisition (Nahavandi, 1993). Therefore, acquisition acceptance is expected to be higher for subordinates who were previously employed in the acquiring firm than those from the target firm. Heritage, the firm that an employee previously worked for, was dummy-coded as 1 for target company employees and 0 for the acquiring firm. Organizational tenure may also affect an employee's response to an acquisition because it reflects the length of time that the employee has invested in the previous organizational culture. Tenure data were collected in interval form and coded 1 (5 years or less), 2 (6 10 years), 3 (11 15 years), 4 (16 25 years), and 5 (26 years or more). We also controlled for individual differences in education, gender, functional department, and organizational position. Education data were collected in interval form and coded 1 (high school), 2 (some college), 3 (bachelor degree) and 4 (graduate degree). Gender was dummy coded as 1 for female and 0 for male. Functional departments were coded as dummy variables with finance as the reference group. Supervisory position was coded as 1 if the person had supervisory responsibilities and 0 otherwise Dependent variables Overall job satisfaction was measured using a three-item scale designed to correspond with Hackman & Oldham's (1975) JCM scale for global job satisfaction (McMurtrey, Grover, Teng, & Lightner, 2002). It was previously tested on a sample of 226 respondents from 32 firms and had a Cronbach's alpha of.74 (McMurtrey et al., 2002). A sample item is Generally speaking, I am very satisfied with my job. Cronbach's alpha in this study was Acquisition acceptance assessed employees' perceptions that the acquisition was a positive strategic move for the firm. The scale utilized two of the three items in a scale (Cronbach's alpha >.80) developed by Covin et al. (1997) to assess post-merger employee satisfaction in a sample of over 2000 employees. We added two items asking respondents to compare the combined company to their previous company (see Table 1). This revised scale was tested and refined in Pilot 1 (N=103). A sample item is A majority of employees have come to accept this acquisition as necessary and worthwhile. Cronbach's alpha was Subordinate performance was rated by direct supervisors and measured with two items that assessed accomplishment of objectives and acceptability of interpersonal behaviors. These measures correspond closely with the criteria the host firm used for formal employee performance assessment. Supervisors were asked to rate the overall performance of each employee relative to his/her peers with respect to achievement of objectives, and rate the overall performance of each employee relative to his/her peers with respect to interpersonal behaviors. Supervisors were guided that ratings should be similar to what you submitted for the annual salary administration program, which they had just completed. This guidance was intended to reduce leniency bias because the organization has limits in the salary administration program that prevent all employees from being evaluated as equally excellent performers. Cronbach's alpha was Analysis The host firm did not permit us to collect data from employees in multiple waves, which presents potential concerns for statistical validity due to common method variance. We did, however, acquire performance data from a separate source. There are few effective alternatives for gathering perceptual data other than from the respondent (Podsakoff & Organ, 1986). Therefore, we took additional design steps recommended in the literature to minimize the possibility of common method bias, including using different types of scale anchors and formats, using some reverse-scored items,

10 58 L.A. Nemanich, R.T. Keller / The Leadership Quarterly 18 (2007) inter-mixing items from different constructs, and interspersing dependent variables with independent variables (Hinkin, 1995; Podsakoff, MacKensie, Lee, & Podsakoff, 2003). We also tested for the presence of a common factor using a Harmon single factor test (Podsakoff et al., 2003). The results yielded six factors with eigenvalues greater than 1, and the first factor explained only 36% of the variance. The failure of a single factor to emerge and having less than a majority of the variance explained by the first factor suggest that common method bias is not a serious problem in these data (Podsakoff & Organ, 1986). To make this analysis more compelling, we also controlled statistically for common method bias by using a measure of employee affect as a control variable (Podsakoff et al., 2003). In the case of acquisition acceptance, we used job satisfaction as a control. For the other two dependent variables, we used acquisition acceptance as a control. This is a very conservative test because the affect control variable is expected to be correlated with the dependent variable and, therefore, its use partials out variance due to substantive relationships, as well as statistical bias. Although our research design is at the individual level (not aggregated), we performed statistical tests on our context variables to confirm that we were measuring climate perceptions. We had 88 teams with at least two respondents. Using two randomly selected raters per group, we tested for agreement among raters [ICC (1)] and for the reliability of average ratings [ICC (2)] using a two-way random effects consistency model (James, 1982; Klein & Kozlowski, 2000; Nichols, 1998). The results for goal clarity were.60 for ICC (1) and.75 for ICC (2) (p<.001). The same tests using four raters (55 teams) yielded an ICC (1) of.41 and an ICC (2) of.73 (p<.001). The results for support for creativity were an ICC (1) of.51 and an ICC (2) of.67 for two raters (p <.001), and an ICC (1) of.38 and an ICC (2) of.70 for four raters (p <.001). All of these results indicate both agreement among team members and acceptable reliability of average ratings, suggesting that we are measuring climate variables. We also tested for consistency within teams and differences among teams in ratings of the leader's transformational behaviors. The aggregation analysis results were an ICC (1) of.59 and an ICC (2) of.75 for two raters (p <.001), and an ICC (1) of.52 and an ICC (2) of.81 for four raters (p <.001). An ANOVA analysis indicated that transformational leadership ratings differed significantly among teams (F=1.54, p<.01, df=103, 327). Confirmatory factor analysis (AMOS 5.0) was used to test the measurement model for construct and discriminant validity of all perceptual variables (transformational leadership, goal clarity, support for creative thinking, job satisfaction, and acquisition acceptance). The results indicated a good fit for the proposed five factor structure (χ 2 =298.9/df 125, RMSEA =.06, CFI=.95). A four factor model (combining the dependent variables) and a three factor model (combining the mediating variables) had significantly poorer fit statistics (χ 2 = 552.9/df 129, RMSEA=.09, CFI=.88 and χ 2 =735.6/df 132, RMSEA=.10, CFI=.84, respectively), suggesting strong discriminant validity to the constructs. We also determined by analyzing modification indices that no significant improvements to model fit could be made by reassigning any measure to load on a different construct, supporting the construct validity of these measures. In order to demonstrate a mediation effect, Baron & Kenny (1986) stated that four relationship patterns must be supported: (1) transformational leadership must be correlated with the dependent variable; (2) transformational leadership must be correlated with the mediator; (3) the mediator must be correlated with the dependent variable; and (4) transformational leadership's relationship to the dependent variable must become non-significant in the presence of the mediator for full mediation or become at least weaker for partial mediation. Recent research has suggested alternatives to this method on the grounds that Baron and Kenny's procedure is limited in statistical power and too conservative with respect to avoiding Type I errors (MacKinnon, Lockwood, Hoffman, West, & Sheets, 2002). We believe, however, that our sample size is large enough to permit use of the more conservative test. A series of hierarchical regressions was used to assess these relationships. First, transformational leadership was regressed on the mediators, goal clarity and support for creative thinking. Then, hierarchical regressions were conducted for each subordinate dependent variable, whereby control variables were entered in Step1, followed by transformational leadership in Step 2, and finally the mediators were entered in Step Results Means, standard deviations, and correlations for all variables are shown in Table 2. All variance inflation factors in the regression analysis were less than 1.6, indicating that multicollinearity is not a serious concern (Hair, Anderson, Tatham, & Black, 1998). Table 3 reports the results of the regression analyses for the dependent variables. Transformational leadership had the hypothesized positive relationships (Step 2) with each of the individual outcomes (acquisition acceptance, performance, and job satisfaction) in the absence of the mediating variables, supporting

11 L.A. Nemanich, R.T. Keller / The Leadership Quarterly 18 (2007) Table 2 Means, standard deviations, and correlations Variable Mean SD Heritage Tenure Education Gender Supervisory position R&D Sales Customer Service Intellectual Charisma Motivation Consideration Transformational Goal clarity Creative thinking Acquisition acceptance a Performance b Job satisfaction c Note: N=447. a N=432. b N=344. c N=447. p<.05. p<.01.

12 60 L.A. Nemanich, R.T. Keller / The Leadership Quarterly 18 (2007) Table 3 Hierarchical regression results for acquisition acceptance, performance and job satisfaction Variable Acquisition acceptance a Performance b Job satisfaction c b SE β b SE β b SE β Step 1 Heritage Tenure Education Gender Supervisory position R&D Sales Customer service Job satisfaction Acquisition acceptance R Adjusted R Step 2 Heritage Tenure Education Gender Supervisory position R&D Sales Customer Service Job satisfaction Acquisition acceptance Transformational leadership R Adjusted R ΔR Step 3 Heritage Tenure Education Gender Supervisory position R&D Sales Customer Service Job satisfaction Acquisition acceptance Transformational leadership Goal clarity Creative thinking R Adjusted R ΔR Note: a N=432 b N=344 c N=447. Unstandardized (b) and standardized (β) regression coefficients are shown. p<.05. p<.01. Hypotheses 1a, 1b, and 1c. Although not shown in tabular form, transformational leadership also had significant positive relationships in regressions with each of the mediating variables, goal clarity (β =.40, p <.01) and support for creative thinking (β =.53, p <.01). Together, these relationships meet the first two requirements for mediation. In the model of acquisition acceptance, the control variable of employment heritage was a significant predictor. The affect control, job satisfaction, also had a significant relationship. Transformational leadership was a significant predictor of acquisition acceptance in Step 2. Supporting this result, acquisition acceptance differed significantly among teams in an ANOVA analysis (F = 1.45, p <.01, df = 103, 314). Support for creative thinking had the

13 L.A. Nemanich, R.T. Keller / The Leadership Quarterly 18 (2007) expected positive relationship with acquisition acceptance. Furthermore, when support for creative thinking was added to the model, transformational leadership became insignificant, indicating full mediation in support of Hypothesis H3a. Goal clarity, however, did not have a significant relationship with acquisition acceptance after controlling for transformational leadership. Therefore, Hypothesis H2a was not supported. Transformational leadership explained 1% of the variance in acquisition acceptance, and total variance explained by the model was 23%. In the model of performance, a supervisory role and working in the R&D function were significant predictors. Supervisors tended to receive higher performance ratings than individual contributors, and R&D employees tended to receive lower performance ratings than finance employees. Transformational leadership was a significant predictor of subordinate performance in Step 2. An ANOVA analysis indicated significant performance differences among teams (F=2.44, p<.001, df=90, 251). Goal clarity was a significant positive predictor of performance after controlling for transformational leadership behaviors. Transformational leadership's regression coefficient was not reduced; therefore, goal clarity did not mediate the relationship and Hypothesis H2b was not supported. Transformational leadership explained 6% of the variance in performance, and total variance explained by the performance model was 17%. In the model of job satisfaction, education, working in the sales function, and the affect control variable, acquisition satisfaction, were significant predictors. Employees with less education were more satisfied, and sales employees were more satisfied than finance employees. Goal clarity and support for creative thinking had the expected positive relationships with job satisfaction after controlling for transformational leadership, meeting the third requirement for mediation. Transformational leadership continued to have a significant, though reduced, relationship with job satisfaction when these variables were added to the model, indicating partial mediation (53%). Thus, the mediation hypotheses for job satisfaction (Hypotheses H2c and H3b) were supported for partial mediation. Transformational leadership explained 10% of the variance, while the total variance explained by the model was 26%. Although the dimensions of transformational leadership were highly correlated, we conducted additional ad hoc analyses of the relationships of each dimension with each of the subordinate outcomes. When analyzed separately, each dimension of transformational leadership was positively and significantly related to the subordinate outcomes in the same way as overall transformational leadership. To determine which dimension was most highly related to each outcome, we replaced transformational leadership in Step 2 of each regression analysis in Table 3 with the four dimensions of transformational leadership: intellectual stimulation, idealized influence, inspirational motivation, and individualized consideration (Table 4). These dimensions were added stepwise to prevent multicollinearity problems by minimizing redundancy. Variance inflation factors in these analyses were less than 2.5 and condition indices were less than 15, indicating acceptably low levels of multicollinearity. The results showed that idealized influence had the strongest relationship with acquisition acceptance (β =.14, p <.01 in Step 2 and fully mediated in the final model). Individualized consideration had the strongest relationship with performance (β=.33, p<.001 in the full model). Motivation had the strongest relationship with job satisfaction (β=.22, p<.01 in Step 2, partially mediated to β=.14, p<.05 in Step 3) and consideration added additional information (β=.16, p<.05 in Step 2 and fully mediated in Step 3). Acquisition integrations present complex environments to gather perceptual data, because subgroups of employees, such as those from the target firm or in an individual department may experience the environment differently than the average employee population. In another set of ad hoc analyses, therefore, we tested for interaction effects between department and heritage (Table 5). In predicting acquisition acceptance, the interaction of heritage and department was significant (p<.01, R 2 =.03). Former target firm employees were more satisfied with the acquisition than former acquiring firm employees, except in the R&D department where the situation was reversed. In predicting job satisfaction, there was no significant interaction between heritage and department. In predicting performance, the interaction was again significant (p<.01, R 2 =.03). Target firm employees were generally rated as higher performing, except in the sales department where the situation was reversed. 5. Discussion This study makes three important contributions to the transformational leadership literature with respect to context, outcome variables, and process understanding. First, we provide much needed insight into the relationships between transformational leadership behaviors and employee attitudes and performance in the demanding context of an acquisition integration. We provide qualitative evidence supporting the prevailing employee perceptions of an