Much of the empirical organizational commitment

Transcription

1 Copyright econtent Management Pty Ltd. Journal of Management & Organization (2012) 18(2): The mediating effect of leader member exchange on the relationship between Theory X and Y management styles and affective commitment: A multilevel analysis FARUK ŞAHIN The Faculty of Economics and Administrative Sciences, The Department of Business Administration, Nigde University, Nigde, Turkey Abstract The purpose of the present study was to investigate the relationship between McGregor s Theory X and Y management styles and affective commitment through the mediating mechanism of the leader member exchange (LMX). Adopting a multilevel perspective to explain the complex relations among variables, data were collected from 56 supervisors and 173 subordinates from yacht building companies in Turkey. The results indicated that the Theory Y management style related positively to affective commitment and LMX. In addition, the results indicated that LMX partially mediated the relationship between the Theory Y management style and affective commitment. However, the Theory X management style had no relationships with either LMX or affective commitment. Implications for future research and practice are discussed. Keywords: Theory X and Y management styles, affective commitment, leader member exchange Much of the empirical organizational commitment research has focused on affective commitment (Wasti, 2003). This emphasis on affective commitment has been mostly because of its association with desirable outcomes such as lower turnover, fewer intentions to quit and organizational citizenship behavior (Allen & Meyer, 1996; Meyer, Stanley, Herscovich, & Topolnytsky, 2002; Wasti, 2008). Although there are a variety of antecedents for affective commitment (Meyer & Allen, 1991, 1997), one of the most common is a high-quality relationship with one s supervisor (e.g., Eisenberger et al., 2010; Gerstner & Day, 1997). The quality of the relationship between a supervisor and subordinate can be described in terms of the leader member exchange (LMX) theory (Dienesch & Liden, 1986; Gerstner & Day, 1997; Liden, Sparrowe, & Wayne, 1997). LMX theory argues that leaders develop unique relationships with different subordinates and that the quality of these relationships is a determinant of how each subordinate will be treated (Graen & Uhl-Bien, 1995). Considering the benefits for individuals and organizations, it seems important to take action in order to strengthen subordinates affective commitment to the organization. Therefore, efforts should be based on understanding the determinants of affective commitment and the exchange relationships. Although research has indicated that supervisors can influence the development of subordinates affective commitment (e.g., Cohen, 1992; Wayne et al., 2009), the influence of supervisors management styles on affective commitment remains unclear. One of the research directions from the style approach is that by McGregor (1960), who developed Theory X and Y but which has seldom been empirically tested (Kopelman, Prottas, & Davis, 2008; Kopelman, Prottas, & Falk, 2010). McGregor (1960) proposed two sets of assumptions about human motivation that a manager can hold. These are Theory X assumptions that people basically dislike work, need direction and avoid responsibility and Theory Y assumptions that people like work, are creative and accept responsibility. Although McGregor proposed Theory X and Y over 50 years ago, these theories have contributed to management and leadership thinking and practice for many years (Heil, Bennis, & Stephens, 2000). Scholars have continued to discuss and debate since their introduction in One general criticism is the fact that McGregor s work was simple and undeveloped (Reddin, 1969) and that it did not take the impact and role of environmental factors into consideration (Thomas & Bennis, 1972). Some scholars have asserted that the assumptions underlying Theory Y are too general and need to be more specific (e.g., Bobic & Davis, 2003; Morse & Lorsch, 1970). Volume 18, Issue 2, March 2012 JOURNAL OF MANAGEMENT & ORGANIZATION 159

2 Faruk Şahin econtent Management Pty Ltd Despite these criticisms, McGregor s work has left a lasting mark on organizational management and leadership studies (Heil et al., 2000). Miner (2003) reported a recent peer review that ranked McGregor s Theory X and Y second in terms of recognition and in 33rd place in importance out of 73 organizational behavior theories, as rated by organizational behavior scholars. Furthermore, the Theory X and Y managerial assumptions can cause autocratic (work-focused) and democratic (people-focused) leadership behaviors, respectively (McGregor, 1960, 1967). McGregor s work, which includes basic principles from which to develop positive management styles, such as participative and empowering management styles (Kopelman et al., 2008, 2010), is still referred to commonly in the fields of management and leadership studies. However, Kopelman et al. (2008) stated that it has not been possible to test McGregor s Theory X and Y because of the lack of published, valid and reliable measures. Therefore, a study testing McGregor s Theory X and Y with field data would be an important contribution. According to McGregor (1960, 1966), the management styles of supervisors range from Theory X to Theory Y principles. Similarly, LMX relationships exist on a continuum, ranging from high to low quality (Graen & Uhl-Bien, 1995). Since supervisors establish different relationships with different subordinates, Kopelman et al. (2008) proposed that supervisors may apply the Theory Y management style with respect to in-group members who have high-quality relationships with them and the Theory X management style with respect to out-group members who have low-quality relationships with them. Although it is possible that supervisors management styles influence the quality of the LMX relationship, which subsequently affects desirable outcomes such as subordinates affective commitment, the empirical linkages among the Theory X and Y management styles, LMX relationship and affective commitment are relatively unexplored to date. This study was designed to examine these relationships using samples of Turkish managers and their subordinates. The present study makes several contributions to research in this area. First, this study provides one of the few empirical tests of McGregor s Theory X and Y management styles. Second, the results of this study are of importance not only to Turkish and global organizations operating in Turkey, but also to the organizations operating in other countries. Therefore, it is important to understand the consequences of the Theory X and Y management styles in a work setting. To advance the understanding of how the Theory X and Y management styles matter, in the present study several questions were formulated. Specifically, do the Theory X and Y managerial assumptions and beliefs that a manager holds influence LMX relationships? Do these assumptions also influence individual outcomes such as subordinates affective commitment towards the organization? Further, are those linkages mediated (indirect) or unique (direct)? All of these questions (as illustrated in Figure 1) are addressed in the literature review. To summarize, the main objective of the present study was to examine the relationships among McGregor s Theory X and Y management styles, LMX and affective commitment in order to evaluate the extent to which LMX mediates the influence of the Theory X and Y management styles on affective commitment. THEORETICAL BACKGROUND AND HYPOTHESES Theory X and Y management styles and affective commitment McGregor (1960, 1966) classified managers attitudes or belief systems, which he called assumptions, as Theory X and Y. Theory X and Theory Y describe two very different attitudes towards workforce motivation. Managers with Theory X attitudes hold that subordinates dislike work, avoid work if they can and show little ambition. According to this theory, subordinates must be closely supervised in order to make them efficient workers. Managers with Theory X mindsets tend to have a negative, pessimistic view of subordinates and display more coercive, autocratic leadership styles using external means of controls, such as threats and punishment (McGregor, 1960, 1966; McGregor & Cutcher- Gershenfeld, 2006). By contrast, Theory Y assumes that subordinates are willing and wanting to work and to achieve their maximum outputs; thus, they do not need to be closely supervised. Managers with 160 JOURNAL OF MANAGEMENT & ORGANIZATION Volume 18, Issue 2, March 2012

3 econtent Management Pty Ltd The mediating effect of the leader member exchange on the relationship Theory Y attitudes tend to have positive, optimistic views of subordinates and display more participative leadership styles using internal motivation and rewards (McGregor, 1960, 1966; McGregor & Cutcher-Gershenfeld, 2006). Overall, the Theory Y frame of mind looks at the human being as working for the organization and attaches value to the person. One of the aspects of the Group Level Theory Y Management Style Theory X Management Style three-component model of organizational commitment proposed by Meyer and Allen (1991) is affective commitment. This reflects the extent to which subordinates are emotionally attached to, identify with and are involved in the organization (Meyer & Allen, 1991; Mowday, Steers, & Porter, 1979; O Reilly & Chatman, 1986). By contrast, continuance commitment is based on subordinates perceptions of the costs associated with leaving the organization (Meyer & Allen, 1997; Reichers, 1985). Normative commitment is based on subordinates feelings of obligation to remain with the organization (Meyer & Allen, 1991, 1997). Compared with continuance or normative commitment, much of the empirical work has examined the relation between attitudinal correlates of affective commitment (Meyer & Allen, 1997; Wasti, 2008). Affective commitment has been shown to be strongly related to desirable individual and organizational outcomes (e.g., Mathieu & Zajac, 1990; Meyer et al., 2002; Vandenberghe, Stinglhamber, Bentein, & Delhaise, 2001; Wasti, 2003). Research has indicated that subordinates develop feelings of affective commitment if they perceive there to be organizational support and justice (Meyer & Allen, 1991) and if they see the organization as a place where they feel they are important (Meyer & Allen, 1997). LMX Individual Level FIGURE 1: THE MEDIATING ROLE OF LMX ON THEORY X AND Y MANAGEMENT STYLE AND AFFECTIVE COMMITMENT Affective Commitment In the eyes of subordinates, the supervisor represents the organization (Eisenberger, Stinglhamber, Vandenberghe, Sucharski, & Rhoades, 2002; Schyns, Paul, Mohr, & Blank, 2005). Supervisors have formal authority over subordinates. They convey and execute the organizational goals and, thus, hold a position that connects their subordinates to the overall organization. Subordinates affective commitment to the organization should matter as a consequence of supervisors management styles, since work experience such as job scope, support, reward and justice have been found to impact the development of affective commitment more than have the structural features of the organization or personal characteristics of subordinates (Rhoades, Eisenberger, & Armeli, 2001; Wasti, 2008). Social exchange theory can also explain the possible relationships between supervisors Theory X and Y management styles and subordinates affective commitment to the organization. According to this theory, social exchange tends to engender feelings of personal obligation, gratitude and trust (Blau, 1967; Gouldner, 1960). Subordinates develop relationships based on the implicit expectation of Volume 18, Issue 2, March 2012 JOURNAL OF MANAGEMENT & ORGANIZATION 161

4 Faruk Şahin econtent Management Pty Ltd reciprocally beneficial exchanges based on trust and felt obligation (e.g., Gouldner, 1960; Rhoades & Eisenberger, 2002). For example, giving autonomy and responsibility for work to a subordinate in an organization entails a close relationship and mutual trust, and the subordinate feels an obligation to reciprocate, for example by developing feelings of affective commitment and, thus, performing well. When subordinates and the supervisor trade benefits in a two-way process, exchange relations will be established (Blau, 1967; Gouldner, 1960). The concept of affective commitment has not been empirically investigated as a consequence of McGregor s management styles (Kopelman et al., 2008); therefore, one of the aims of the present study is to yield information about the relationship between supervisors Theory X and Y management styles and subordinates affective commitment to the organization. Based on the aforementioned explanations, the existence of Theory X and Y management styles is likely to be related to affective commitment. Specifically, subordinates who are supervised by Theory Y managers should be expected to have a higher level of affective commitment than do subordinates who are supervised by Theory X managers. Hypothesis 1a: Theory Y management style is positively related to affective commitment. Hypothesis 1b: Theory X management style is negatively related to affective commitment. Theory X and Y management styles and LMX There is general acceptance of the notion that effective leadership consists in part of good relationships between leaders and followers (e.g., Bass, 1990). According to LMX theory, supervisors expend their resources unequally among subordinates and they develop and maintain separate relationships with different subordinates, giving rise to in-groups and out-groups (Dansereau, Graen, & Haga, 1975; Graen & Cashman, 1975; Graen & Graen, 2005; Graen & Uhl-Bien, 1995). The in-group includes subordinates with strong social ties to their supervisor in a supportive relationship characterized by high mutual positive affect, trust, respect, loyalty and influence (e.g., Dienesch & Liden, 1986; Liden & Maslyn, 1998; Uhl-Bien & Maslyn, 2003). High-quality LMX relationships afford autonomy and valued resources for in-group members (Graen & Scandura, 1987; Yukl & Van Fleet, 1990). In-group members interact frequently with their supervisors (Yrle, Hartman, & Galle, 2002) and receive more assistance, encouragement and concern from their supervisors (Graen & Uhl- Bien, 1995; Sparrowe & Liden, 1997; Uhl-Bien & Maslyn, 2003; Wayne, Shore, & Liden, 1997). Relationships that are of low quality involve low levels of mutual influence from the supervisor as an out-group exchange (Graen & Scandura, 1987; Uhl-Bien & Maslyn, 2003; Yukl & Van Fleet, 1990). The out-group includes subordinates with few or no social ties to their supervisor in a strictly task-centered relationship characterized by low exchange and top-down influence (e.g., Dienesch & Liden, 1986; Graen & Scandura, 1987; Liden & Maslyn, 1998). Members in the out-group receive less access to the supervisor (Yrle et al., 2002), fewer resources and less information and they are motivated to do what is formally required by their job descriptions (Gerstner & Day, 1997; Graen & Uhl-Bien, 1995; Sparrowe & Liden, 1997; Uhl-Bien & Maslyn, 2003; Wayne et al., 1997). Leader member exchange relationships exist on a continuum, ranging from high to low quality (Graen & Graen, 2005; Graen & Uhl-Bien, 1995). These relationships affect the types of power and influence tactics supervisors use (e.g., Gerstner & Day, 1997; Graen & Uhl-Bien, 1995; Liden et al., 1997; Uhl-Bien & Maslyn, 2003). With regard to the source of power (French & Raven, 1968), supervisors primarily use expert, referent and reward power to influence members of the in-group, whereas they mostly use reward as well as legitimate and coercive power for out-group members (Liden et al., 1997). Empirical research has demonstrated that LMX has been linked to desired outcomes such as task performance, satisfaction, turnover and organizational commitment (Gerstner & Day, 1997). Regarding supervisor behaviors, the distinction between high- and low-quality exchange relationships is similar to that between Theory Y and Theory X managers, respectively. According to McGregor (1960, 1966), supervisors can apply either Theory X or Theory Y principles and that 162 JOURNAL OF MANAGEMENT & ORGANIZATION Volume 18, Issue 2, March 2012

5 econtent Management Pty Ltd The mediating effect of the leader member exchange on the relationship this does not prevent them from being classified as Theory X or Y managers. Management styles that supervisors apply range from Theory X principles, in which subordinates are coerced and controlled to do what is formally required by their job descriptions, to Theory Y principles, which are characterized by trust, respect and obligation. When a manager applies Theory Y principles, subordinates receive autonomy and responsibility for work, more opportunities to identify problems and find creative solutions to them (McGregor, 1960, 1966; McGregor & Cutcher-Gershenfeld, 2006). As a result, subordinates might perceive that the supervisor provides intangible and tangible resources, which leads to high-quality exchange relationships (Liden et al., 1997). By contrast, Theory X managers emphasize the close supervision of subordinates and the chain of command and motivate subordinates using extrinsic rewards (McGregor, 1960, 1966; McGregor & Cutcher-Gershenfeld, 2006). Therefore, subordinates that are supervised by Theory X managers might perceive there to be a social-emotional distance between their supervisor and themselves, which leads to formal, impersonal and low- quality exchange relationships (Graen & Uhl-Bien, 1995; Liden et al., 1997). As noted, both literatures have dealt with supervisor subordinate relationships. Furthermore, there has been a call for the investigation of the relationship between the Theory X and Y management styles and LMX. Thus, when the two literatures are examined conjointly, a relationship between the Theory X and Y management styles and LMX is expected (Kopelman et al., 2008). Specifically, subordinates that are supervised by managers with Theory Y attitudes should be expected to experience higher quality LMX relationships. Thus: Hypothesis 2a: The Theory Y management style is positively related to LMX. Hypothesis 2b: The Theory X management style is negatively related to LMX. LMX and affective commitment The relation between LMX and affective commitment is well established in the mainstream literature (Wayne et al., 2009). Previous research has suggested that subordinates who have high-quality relationships with their supervisors are more likely to be committed to their organizations (e.g., Gerstner & Day, 1997; Schriesheim, Castro, & Cogliser, 1999). Furthermore, Schyns et al. (2005) showed that LMX is positively related to subordinates affective commitment across different samples in different cultures. With regard to Turkish studies, Cevrioğ lu (2007) demonstrated the link between LMX quality and affective commitment. In sum, the quality of the exchange relationships between subordinates and supervisors is supposed to be an effective factor that influences the development of affective commitment among subordinates (Cohen, 1992; Eisenberger et al., 2010; Wayne et al., 2009). Hypothesis 3: LMX is positively related to affective commitment. The mediating effects of LMX Although the relationship between McGregor s Theory X and Y and affective commitment has not been empirically investigated, such a linkage might be evidenced by the mechanisms of the exchange relationship. The supervisor is an organizational member who is responsible for managing his or her subordinates performances. The supervisor is seen by his or her subordinates as a representative of the organization (Eisenberger et al., 2002; Schyns et al., 2005). Therefore, subordinates interpret the supervisor s actions as a formal functioning of the organization because the supervisor has formal authority over them. The management styles that the supervisor applies in a work setting involve different attitudes towards workforce motivation (McGregor, 1960, 1966) and this provides subordinates with different work-related experiences (Rhoades et al., 2001; Wasti, 2008). When subordinates receive such messages, each subordinate generates distinct perceptions, and this perception may be generalized to the organization, which subsequently determines the subordinate s attitude and behavior (i.e., commitment, loyalty and performance) in reciprocation to the organization (Rhoades et al., 2001; Wayne et al., 1997). For example, when a supervisor applies the Theory Y management style, subordinates might perceive that he or she provides intangible and Volume 18, Issue 2, March 2012 JOURNAL OF MANAGEMENT & ORGANIZATION 163

6 Faruk Şahin econtent Management Pty Ltd tangible resources, which leads to high-quality exchange relationships (Liden et al., 1997). Subordinates in high-quality exchange relationships can communicate with their supervisors frequently and have their supervisors support, encouragement and consideration (Sparrowe & Liden, 1997; Wayne et al., 1997). In return, subordinates repay these benefits by displaying desirable behaviors, such as affective commitment to the organization (Cohen, 1992; Wayne et al., 2009). Conversely, when a supervisor applies the Theory X management style, subordinates might perceive a strictly task-centered relationship characterized by low exchange and top-down influence, which leads to low-quality exchange relationships (Graen & Scandura, 1987; Liden & Maslyn, 1998). These relationships determine subordinates actions and behaviors in accordance with the prescriptions and proscriptions of the employment contract (Sparrowe & Liden, 1997; Wayne et al., 1997). Consequently, from this perspective, the Theory X and Y management styles will first influence subordinates perceptions of exchange relationships and subsequently further modify or change their affective commitment to the organization. Hence, in the present study, it is proposed that LMX mediates the relationships between the supervisor s management style and the subordinate s affective commitment to the organization. Hypothesis 4: LMX mediates the relationship between Theory X and Y management style and affective commitment. METHOD Sample and procedure Data were collected from 16 mid-sized yacht building companies located in Muğla in the western part of Turkey. The size of companies ranged from 39 to 63 employees. Drawing from the employee lists provided by the HR departments of the companies, most Turkish forms of the questionnaires were hand-delivered with stamped, selfaddressed envelopes for return, although in some cases the questionnaires were mailed via the postal service. A cover letter explained the purposes of the study and guaranteed anonymity of responses. Participation in the study was voluntary. Two sets of questionnaires were sent: one to subordinates and the other to their immediate supervisors. Data on LMX and affective commitment were obtained by the questionnaires directed at subordinates. Data on management style were obtained through the questionnaires directed at their immediate supervisors. The questionnaires of each employee and his or her immediate supervisor were matched using identity numbers. A total of 225 subordinate questionnaires and 75 supervisor questionnaires were distributed. Altogether, 196 completed questionnaires were returned by subordinates and 64 were collected from supervisors. Twenty-three subordinates questionnaires and eight supervisors questionnaires were excluded from the analysis because of missing data or incorrect marking. Finally, 173 subordinates questionnaires and 56 supervisors questionnaires were included for analysis, yielding a 76.8% response rate for subordinates and a 74.6% response rate for supervisors. Of the 173 subordinates, 29% were female and 71% were male. Of those who reported their education levels, 48.8% held a high school or a college degree, 40.2% held a bachelors degree and the remaining 11% held a graduate school degree. The majority of subordinates was aged 30 and below (59.3%), followed by years (22.5%), years (14.7%) and above 41 years (3.5%). Subordinates reported an average length of tenure within their organizations of 4.79 years (SD = 2.16). In the sample of 56 supervisors, 22% were female and 78% were male. Of those who reported their education levels, 13.9% held a high school or a college degree, 62.7% held a bachelors degree and the remaining 23.4% held a graduate school degree. The majority of supervisors was aged between 36 and 40 years (52.5%), followed by years (33.2%), above 41 years (12.3%) and below 30 years (2%). Supervisors reported an average length of tenure within their organizations of 8.21 years (SD = 5.44). Each manager supervised three employees on average. Measures All the main constructs included in the study were assessed with self-reported measures. Since 164 JOURNAL OF MANAGEMENT & ORGANIZATION Volume 18, Issue 2, March 2012

7 econtent Management Pty Ltd The mediating effect of the leader member exchange on the relationship the Theory X/Y scale was originally in English, two-way translations were performed by two bilinguals with English and Turkish proficiencies to ensure an equivalency of meaning (Brislin, 1980) and the scale was then piloted with a group of supervisors from the yacht building companies. The specific measures used in the present study are outlined below. McGregor s Theory X and Y The Theory X/Y scale included eight items that were adapted from the questionnaire developed by Kopelman et al. (2008). Four items measured the attitudes and assumptions of Theory Y and Theory X supervisors. A sample item for the Theory Y scale was For most people, work is as natural as play or recreation and a sample item for the Theory X scale was Most employees can t be trusted. Participants rated items on a five-point Likert-type scale ranging from 1 (strongly disagree) to 5 (strongly agree). A higher score indicated that the supervisor applied either the Theory Y or the Theory X management style in a work setting. Before testing the hypothesis, a series of exploratory factor analyses were conducted on the eight items of the Theory X/Y scale. This analysis yielded a two-factor solution that accounted for 70.24% of variation. All eight items had high factor loadings on the theoretical factor to which they were primarily attached. The two factors clearly represented the two postulated sets of attitudes and assumptions of supervisors (i.e., the Theory X and Y management styles). Factor 1 (Theory Y) accounted for 55.99% of the item variance and Factor 2 (Theory X) accounted for 14.24% of the item variance. The two factors had good reliabilities (0.86 for Theory Y and 0.84 for Theory X). Additionally, confirmatory factor analyses indicated that the two-factor model (χ 2 (N = 56) = with 19 df, p < 0.001, CFI = 0.90, GFI = 0.77, NNFI = 0.85 and RMSEA = 0.093) provided better fit compared to the one-factor model (χ 2 (N = 56) = with 20 df, p < 0.001, CFI = 0.78, GFI = 0.66, NNFI = 0.70 and RMSEA = 0.29; Browne & Cudeck, 1993; Hoyle, 1995). Leader member exchange Leader member exchange was measured using a seven-item questionnaire developed by Scandura and Graen (1984). The items asked subordinates the extent to which they had high-quality exchanges with their supervisor. An example of an item was Your supervisor understand(s) your job problems and needs. Participants rated items on a five-point Likert-type scale ranging from 1 (strongly disagree) to 5 (strongly agree). Higher scores indicated that the participant perceived a greater degree of higher quality relationships. Previous research has suggested that the scale has good psychometric properties and can be used as a single-factor measure (e.g., Özutku, Ağca, & Cevrioğlu, 2008). To confirm the scale s factor structure, a confirmatory factor analysis was applied. All the items significantly loaded on a single-factor (p < 0.05), and fit indexes provided evidence of acceptable fit: χ 2 (N = 173) = with 14 df, CFI = 0.95, GFI = 0.90, NNFI = 0.95 and RMSEA = (Browne & Cudeck, 1993; Hoyle, 1995). The scale s alpha reliability coefficient in this study was Affective commitment The subordinate s affective commitment to the organization was measured using an eight-item scale developed by Wasti (2003). This scale included five items from the work of Meyer, Allen, and Smith (1993) and three collectivistic culturespecific items generated by Wasti (2003) through interviews with Turkish employees. A sample item was I really feel as if this organization s problems are my own. Participants rated items on a fivepoint Likert-type scale ranging from 1 (strongly disagree) to 5 (strongly agree). Higher scores were indicative of higher level of affective commitment. Previous research has provided validation evidence for this scale and has shown that it forms a singlefactor with high reliability (e.g., Özutku, 2008; Özkaya, Kocakoç, & Karaa, 2006). A confirmatory factor analysis indicated acceptable fit for a single-factor (p < 0.05): χ 2 (N = 173) = with 20 df, CFI = 0.90, GFI = 0.80, NNFI = 0.89 and RMSEA = (Browne & Cudeck, 1993; Hoyle, 1995). For the present study, the scale s alpha reliability coefficient was Analytical strategy Since the sample consisted of individuals (i.e., subordinates) who were nested within groups Volume 18, Issue 2, March 2012 JOURNAL OF MANAGEMENT & ORGANIZATION 165

8 Faruk Şahin econtent Management Pty Ltd (i.e., supervisors), and the data obtained on individuals were not fully independent, hierarchical linear modeling (HLM) techniques were used to test the multilevel models (e.g., Hitt, Beamish, Jackson, & Mathieu, 2007; Hofmann, 1997; Raudenbush & Bryk, 2002). HLM models cross-level relationships by specifying distinct level-1 (i.e., lower-level observations such as subordinates) and level-2 (i.e., higher-level units such as groups or teams) models. HLM has several advantages. First, lower-level variables can be explained as a function of both lower-level and higher-level characteristics. Second, it allows for testing effects within- and between-levels and, therefore, produces a more precise assessment of lower-level variables. Furthermore, HLM takes account of biases in standard errors and statistical tests resulting from the non-independence of the data (Hofmann, 1997; Krull & MacKinnon, 2001). At the beginning, a null or fully unconditional model with no independent variables at the individual-level or group-level was estimated to test the significance level of individual-level variables (Bliese, 2000). Afterwards, to address the aim of the study (i.e., whether LMX mediates the management style affective commitment relationship), the mediation procedure suggested by Baron and Kenny (1986) for mediational analysis and Krull and MacKinnon s (2001) multilevel equations were used, which required three separate analyses. RESULTS The descriptive statistics and correlations of variables at an individual level are presented in Table 1. An examination of correlations revealed significant associations among variables. However, the main interest of the present study was in the HLM analyses. Before conducting multilevel research analyses, there had to be systematic within- and between-group variance in individual-level dependent variables (Bliese, 2000; Hofmann, 1997). Accordingly, using HLM, a null model that did not include any independent variables was estimated to test the significance level of individual-level variables, affective commitment and LMX. The chi-square estimates for the amount of variation in the changes in affective commitment (χ 2 = ; df = 55; p < 0.001) and LMX (χ 2 = 4548; df = 55; p < 0.001) between groups were significant. The computed intraclass correlations ICC1 were 0.25 for affective commitment and 0.36 for LMX. Both of them were greater than 0.12, indicating that 75% of the variance in affective commitment and 64% of the variance in LMX existed within groups; and 25% of the variance in affective commitment and 36% of the variance in LMX existed between groups. These results, which indicated that appropriate variance in individual-level variables existed between groups, justified further the HLM analysis (Bryk & Raudenbush, 1992). Hypothesis 1a predicted that the Theory Y management style is positively associated with affective commitment. Model 1 in Table 2 shows that the Theory Y management style had significantly positive relationships with affective commitment (γ 02 = 0.50, p < 0.001). Therefore, Hypothesis 1a was supported. Hypothesis 1b predicted that the Theory X management style is negatively related to affective commitment. As shown in Model 1 of Table 2, the Theory X management style had no relationships with affective commitment (γ 01 = 0.11, p = 0.64), so Hypothesis 1b was not supported. Hypothesis 2a predicted that the Theory Y management style is positively associated with LMX. Model 4 in Table 2 shows that the Theory Y management style had significantly positive relationships with LMX (γ 02 = 0.41, p < 0.01). Therefore, Hypothesis 2a was supported. Hypothesis 2b predicted that the Theory X management style is negatively related TABLE 1: DESCRIPTIVE STATISTICS AND CORRELATIONS Variables M SD Theory X management style (0.84) 2. Theory Y management style ** (0.86) 3. LMX ** 0.46** (0.93) 4. Affective commitment ** 0.45** 0.59** (0.94) **p < The statistics in this table are based on the individual-level data (N = 173). Scale reliabilities (α) are displayed between parentheses on the diagonal. 166 JOURNAL OF MANAGEMENT & ORGANIZATION Volume 18, Issue 2, March 2012

9 econtent Management Pty Ltd The mediating effect of the leader member exchange on the relationship TABLE 2: HLM RESULTS Variables Affective commitment LMX Model 1 Model 2 Model 3 Model 3 Coefficient SE t Coefficient SE t Coefficient SE t Coefficient SE t Intercept γ *** *** *** *** Theory X γ Theory Y γ *** ** ** 02 LMX γ ** *** 10 R 2 Level **p < 0.01; ***p < Model 1 equations [Level 1: (affective commitment) ij = β 0j + r ij ; Level 2: β 0j = γ 00 + γ 01 (Theory Y) j + γ 02 (Theory X) j + u 0j ]; Model 2 equations [Level 1: (Affective commitment) ij = β 0j + β 1j (LMX) ij + r ij ; Level 2: β 0j = γ 00 + u 0j ; β 1j = γ 10 ]; Model 3 equations [Level 1: (Affective commitment) ij = β 0j + β 1j (LMX) ij + r ij ; Level 2: β 0j = γ 00 + γ 01 (Theory Y) j + γ 02 (Theory X) j + u 0j ; β 1j = γ 10 ]; Model 4 equations [Level 1: (LMX) = β + r ; Level 2: β = γ + γ (Theory Y) + γ (Theory X) + u ]; ij 0j ij 0j j 02 j 0j The indices i and j refer to subordinates and groups (i.e., supervisors), respectively; β 0j is the intercept for group j; r ij and u 0j are the level-1 and level-2 residuals, respectively. to LMX. As shown in Model 4 of Table 2, the Theory X management style had no relationships with LMX (γ 01 = 0.10, p = 0.69), so Hypothesis 2b was not supported. Hypothesis 3 predicted that LMX is positively associated with affective commitment. Model 2 in Table 2 shows that LMX had significantly positive relationships with affective commitment (γ 10 = 0.30, p < 0.01). Therefore, Hypothesis 3 was supported. Hypothesis 4 predicted that LMX mediates the relationship between the Theory X and Y management styles and affective commitment. Before testing the mediation effect of LMX, three criteria should be met. First, the Theory X and Y management styles must be significantly related to affective commitment (Hypotheses 1a and 1b). Second, the Theory X and Y management styles must be significantly related to LMX (Hypotheses 2a and 2b). Third, LMX must be significantly related to affective commitment (Hypothesis 3). All three of these preconditions were supported only for testing the mediation effect of LMX on the relationship between the Theory Y management style and affective commitment because the results showed that the Theory X management style had no relationships with both LMX and affective commitment (Hypothesis 1b and Hypothesis 2b). Table 2 also shows the effects of the Theory Y management style and LMX on affective commitment. Models 1 3 show the results of a set of tests with affective commitment as the dependent variable. Comparing Model 1 with Model 3, the significant relationship between the Theory Y management style and affective commitment declined substantially when LMX was added to the equation (from γ 02 = 0.50, p < to γ 02 = 0.29, p < 0.01). These results indicated that LMX partially mediated the relationship between the Theory Y management style and affective commitment. Thus, Hypothesis 4 was partially supported. DISCUSSION The present study examined the potential mediating effect of LMX on the relationship between the Theory X and Y management styles and affective commitment in a multilevel research Volume 18, Issue 2, March 2012 JOURNAL OF MANAGEMENT & ORGANIZATION 167

10 Faruk Şahin econtent Management Pty Ltd design. The results indicated that the Theory Y management style is significantly and positively associated with LMX and affective commitment, adding to the understanding of the determinants of affective commitment and exchange relationships. However, the Theory X management style had no relationships with both LMX and affective commitment. The results further indicated that LMX partially mediated the relationship between the Theory Y management style and affective commitment. In other words, when a supervisor applies the Theory Y management style, subordinates might perceive that he or she provides intangible and tangible resources, which leads to high-quality exchange relationships (Liden et al., 1997). In return, subordinates repay these benefits by displaying desirable behaviors, such as affective commitment to the organization (Cohen, 1992; Wayne et al., 2009). This finding empirically supports the general proposition in social exchange theory. When an employer treats employees well, this engenders an obligation on the part of employees to reciprocate the favorable treatment (e.g., Blau, 1967; Gouldner, 1960). Moreover, this finding provides a possible explanation as to how the Theory Y management style influences subordinates affective commitment to the organization. The results did not verify that LMX mediated the relationship between the Theory X management style and affective commitment. This was because of the absence of significant relationships between the Theory X management style and both LMX and affective commitment. Perhaps, one explanation for why the Theory X management style is not associated with LMX and affective commitment is that subordinates engage in different reciprocation efforts depending on the exchange partner (Settoon, Bennett, & Liden, 1996). Although the supervisor is seen by subordinates as a representative of the organization (Eisenberger et al., 2002; Schyns et al., 2005), his or her identity is separate from that of the organization. Specifically, in a hierarchically structured organization, subordinates that are under the direct control of an immediate supervisor may not generalize their perceptions of their supervisor s management style to the overall organization. Furthermore, exchange relationships among subordinates who report to the same supervisor can be positively associated with work attitudes (Sherony & Green, 2002), even if they receive fewer resources and less information and are motivated to do what is formally required by their job descriptions (Gerstner & Day, 1997; Graen & Uhl-Bien, 1995; Sparrowe & Liden, 1997; Wayne et al., 1997). Another explanation for the absence of significant relationships among the Theory X management style, LMX and affective commitment is that posited relationships may be influenced by the requirements and demands of employees in different employment arrangements. The effective management of employees requires different leadership styles that are consistent with the characteristics of employment mode (Liu, Lepak, Takeuchi, & Sims, 2003). Organizations may use contract workers who are employed for specific, and usually, temporary jobs. The relationship between these employees and the organization depends on the contract terms and tends to be short-term-oriented (e.g., Rousseau, 1995). Organizations expect the work to be carried out without concern about the degree of emotional attachment or commitment from employees (Tsui, Pearce, Porter, & Hite, 1995); therefore, the relationship between employees and the organization usually entails the characteristics of a work-focused relationship. To be effective at managing this kind of employee, supervisors apply a directive leadership style that coincides with the Theory X management style (Liu et al., 2003). Theory X and Y assumptions and beliefs about human motivation exist on a continuum leading managers or leaders to create, exhibit and support practices and behaviors that can result in autocratic/directive and democratic/participative leadership styles (McGregor, 1960, 1967). Managers or leaders may apply some principles from each theory but that does not necessarily mean that they should be classified as one or the other (McGregor, 1960, 1966). Furthermore, McGregor (1966) stated that different circumstances require different leadership characteristics (p. 72). Given that the Theory X management 168 JOURNAL OF MANAGEMENT & ORGANIZATION Volume 18, Issue 2, March 2012

11 econtent Management Pty Ltd The mediating effect of the leader member exchange on the relationship style had no significant impact on LMX and affective commitment in the present study, it is worth investigating the efficacy of the Theory X management style under varying conditions (Kopelman et al., 2008). For example, would a Theory X manager be more successful in a command-and-control type of environment? Would a manager with a Theory Y mindset be more effective in organic structures? Additional research is needed to test these questions. The present study indicated that Turkish managers rated higher on the Theory Y management style than they did the Theory X management style. This finding is consistent with previous studies that have examined leadership style in the Turkish context. For example, Kabasakal and Bodur (2002) found that team-oriented, charismatic, participative, humane, autonomous and self-protected were the most preferred leadership attributes. Another study showed that participative style was found to be one of the most preferred leadership styles (Aycan & Pasa, 2003). Since the present study took place in a non-western country that has societal values different from those in the West, cultural context can provide an explanation for the preferences of management styles. Individualism and collectivism cultural dimensions reflect the degree of emphasis on independence versus interdependence with one s group (Hofstede, 1980). In individualist societies, people emphasize their personal goals, interests and values over those of the society and they are expected to maintain relationships as long as they are in line with personal preferences and cost/benefit analyses rather than with social norms (Triandis, 1995). By contrast, in collectivist cultures people value membership of a group and relating oneself to the group (Hofstede, 1980; Triandis, 1995). The meta-analysis by Oyserman, Coon, and Kemmelmeier (2002) indicated that collectivist beliefs include a sense of duty and obligation towards the group. In an organizational setting, the individualism/collectivism dimension reflects the extent to which people rely on social relationships to perform organizational work. Studies have shown that Turkish culture has high collectivism (Hofstede, 1980; House, Hanges, Javidan, Dorfman, & Gupta, 2004). Considering that Theory X and Y assumptions and beliefs about human motivation cause two different managerial practices and behaviors (McGregor, 1960, 1967), the individualism/collectivism cultural dimension may question the prediction for the Theory X and Y management styles relative to desired outcomes. As Kopelman et al. (2008) noted, boundary conditions may moderate the efficacy of management styles. Specifically, additional research should examine whether Theory Y managers perform better in collectivist cultures than they do in individualist cultures. Implications for research and practice This study offers several implications that might be helpful to managers and organizations in facilitating organizational and individual outcomes. In short, the significance of this research is threefold. First, evidence that Theory Y managerial assum ptions are conceptually and empirically related to exchange relationships in predicting an individual-level outcome reinforces the benefits of studying McGregor s work. Furthermore, Theory X and Y managerial assumptions and beliefs about human motivation may have special relevance to leadership behaviors. For example, McGregor (1967) postulated that Theory X and Y mindsets cause autocratic/directive and democratic/participative leadership behaviors, respectively. Thus, future research on leadership and management styles may find that the inclusion of McGregor s (1960, 1966) Theory X and Y improves the predictions of effectiveness. Second, the present study showed the empirical linkages between the Theory Y management style and the theoretically related outcome through the application of a multilevel research design. A manager with a Theory Y mindset enhances positive perceptions on the subordinate side, which leads to high-quality exchange relationships (Liden et al., 1997) and favorable attitudes and behaviors, such as affective commitment to the organization (Cohen, 1992; Wayne et al., 2009). The present study also verified that subordinates perceptions of exchange relationships with their supervisors partially mediate the Volume 18, Issue 2, March 2012 JOURNAL OF MANAGEMENT & ORGANIZATION 169

12 Faruk Şahin econtent Management Pty Ltd relationship between the Theory Y management style and affective commitment. Third, the present study yielded initial evidence that the Theory X/Y scale adapted from the work of Kopelman et al. (2008) has good psychometric properties. The two-factor measure can be useful for the assessment of managerial assumptions and attitudes about human nature. McGregor (1960, 1966) pointed out that Theory X and Theory Y managerial assumptions/ attitudes do not imply that they are unrelated; a manager or leader can have elements of both the Theory X and Theory Y mindset, which have different impacts on work environments. However, both management styles are linked to the achievement of the goals or objectives of an organization (McGregor, 1960). The process by which a manager motivates subordinates and the types of goals set are the main differences between these two management styles (McGregor, 1960, 1966; McGregor & Cutcher-Gershenfeld, 2006). McGregor (1960, 1966) advocated that managers should examine their assumptions and beliefs about human nature and that they should comprehend how these assumptions lead to their management practices and styles. As Heil et al. (2000) stated, the construct of Theory X and Theory Y (McGregor, 1960, 1966) were formulated to help managers examine their assumptions and beliefs. It is generally accepted that the change from a Theory X to Theory Y mindset is a difficult process. This process requires an extensive appraisal and intervening of the patterns of behavior in each style. Specifically, if organizations have adopted an organic structure with participative involvement-oriented cultures and empowered self-managed teams and managers serving as coaches rather than traditional mechanistic structures with control-oriented cultures, they need to realign the human resources management system and place emphasis on developing a Theory Y management style (Kopelman et al., 2008, 2010). Limitations Although the present study was conducted using multiple sources and even though it applied HLM analysis to investigate the relationships among variables, several limitations remain. Since a crosssectional design was used in the present study, it precluded the interpretation of the direction of causality among the considered variables. Future research that employs a longitudinal research design should examine the causal relationships found here. Second, the sample used in this study may raise questions about the generalizability of the findings. Therefore, the results found here should not be generalized until the findings have been replicated in other samples of interest as well as across nationalities and cultures. Third, since ordinal scales were used in this study and ordinal measurements describe an order to the data but not the relative size or degree of the variables measured, conclusions from such statistical data should be interpreted with a caution. Fourth, affective commitment was measured using an eight-item scale developed by Wasti (2003). Since this scale was composed of five items from the work of Meyer et al. (1993) and three collectivistic culture-specific items generated by Wasti (2003), additional work is needed to explore further the psychometric properties of the scale. Specifically, future research efforts investigating the convergent and discriminant validity of the scale through associations with similar and dissimilar constructs will be an important contribution. Finally, in this study LMX was used as a mediating factor and affective commitment was used as a predictable outcome. Nevertheless, other plausible mediators may influence the relationship between the Theory X and Y management styles and affective commitment (e.g., psychological contract, psychological climate). Future research could choose to test the other possible factors that are of central concern in studies of leadership effectiveness. CONCLUSION McGregor s (1960, 1966) Theory X and Y assumptions about human motivation that managers or leaders can hold lead them to create and demonstrate two general approaches to leadership behaviors, namely work-focused and peoplefocused styles. Based on the literature review, it was hypothesized that the Theory X and Y management styles influence the quality of exchange 170 JOURNAL OF MANAGEMENT & ORGANIZATION Volume 18, Issue 2, March 2012